人工神經網路(ANN)的運算，被認為是大量資料運算最節能、最省時的架構之一。然而，關鍵部分之一的光突觸元件仍被限制在作權重更新的學習訓練，這可能限制了其在人工神經系統中的應用。含有感光特性的電阻式記憶體(RRAM)結合塔米結構，利用塔米結構的光子侷限效應集成一個具有選擇特定波長的感光電阻式記憶體，這些結構具有電場誘導的雙極性電阻變化和記憶行為，其結合了特定顏色的識別以及學習與再學習的記憶特性，大大的降低製程的複雜性及元件厚度，受限於商用LED波長普遍落在480nm、525nm和550nm，所以選擇這三種波長做為實驗中元件的主要波長，由實驗可知，元件對應相同波段的光源具有多階阻態的RRAM特性，並作為仿視神經元的集成元件，並利用MNIST作為數字辨識的資料庫作元件權重對資料進行更新，可對應三種元件的權重分布作不同顏色所辨識出的優異結果。

Artificial neural networks (ANN) are considered one of the most energy-efficient and time-saving architectures for large-scale data computations. However, the photonic synapse, remains limited to weight updates during learning and training, potentially restricting its applications in artificial neural systems. Non-Volatile Resistive Random-Access Memory (RRAM) with photonic properties combined with Tamm plasma (TP) utilize TP's photon localization effect to integrate photonic resistive memories capable of selecting specific wavelengths. These structures exhibit electrically induced bipolar resistive switching and storage behavior, combining characteristics of color filters and RRAM. This integration significantly reduces process complexity and device thickness. Considering that commercial LED wavelengths typically fall between 480nm, 525nm, and 550nm, these three wavelengths were chosen as the primary wavelengths for experimental components. Experimental results demonstrate that these devices exhibit multi-level resistive switching characteristics corresponding to light sources within the same wavelength band. These integrated components simulate visual neurons and utilize MNIST as a database for digit recognition, updating component weights based on data. Importantly, the weights distributed among the three components correspond to different colors, resulting in excellent recognition outcomes.

[1] Xitong Hong, Yulong Huang, Qianlei Tian, Sen Zhang, Chang Liu, Liming Wang, Kai Zhang, Jia Sun, Lei Liao, Xuming Zou, “Two‐dimensional perovskite‐gated AlGaN/GaN high‐electron‐mobility‐transistor for neuromorphic vision sensor”, Advanced Science 9.27, 2202019 (2022).
[2] Dandan Hao, Junyao Zhang, Shilei Dai, Jianhua Zhang, Jia Huang, “Perovskite/organic semiconductor-based photonic synaptic transistor for artificial visual system”, ACS applied materials & interfaces 12.35, 39487-39495 (2020).
[3] Wang, Shyh. “Principles of distributed feedback and distributed Bragg-reflector lasers”, IEEE Journal of Quantum Electronics 10.4, 413-427 (1974).
[4] Barnes, William L., Alain Dereux, and Thomas W. Ebbesen, "Surface plasmon subwavelength optics", nature 424.6950, 824-830 (2003).
[5] M. Kaliteevski, I. Iorsh, S. Brand, R.A. Abram, J. Chamberlain, A. Kavokin, I. Shelykh, “Tamm plasmon-polaritons: possible electromagnetic states at the interface of a metal and a dielectric Bragg mirror”, Phys. Rev. B, 76 (2007).
[6] M.E. Sasin, R.P. Seisyan, M.A. Kalitteevski, S. Brand, R.A. Abram, J.M. Chamberlain, A.Y. Egorov, A.P. Vasil’ev, V.S. Mikhrin, A.V. Kavokin, “Tamm plasmon polaritons: slow and spatially compact light”, Appl. Phys. Lett., 251112 92 (2008),.
[7] I. Tamm, “On the possible bound states of electrons on a crystal surface”, Phys. Z. Sowjetunion, 1, 733-735 (1932).
[8] Chinmaya Kar a, Shuvendu Jena b, Dinesh V. Udupa b, K. Divakar Rao, “Tamm plasmon polariton in planar structures: A brief overview and applications”, Optics and Laser Technology 159, 108928 (2023).
[9] Saad Ullah, Jiaming Wang, Peixin Yang, Linlin Liu, Shi-E. Yang, Tianyu Xia, Haizhong Guo, Yongsheng Chen, “All-inorganic CsPbBr3 perovskite: a promising choice for photovoltaics”, Materials Advances 2.2, 646-683 (2021).
[10] Crosby, Glenn A., James N. Demas, “Measurement of photoluminescence quantum yields. Review”, The Journal of Physical Chemistry 75.8, 991-1024 (1971).
[11] Ielmini, Daniele, “Brain-inspired computing with resistive switching memory (RRAM): Devices, synapses and neural networks”, Microelectronic Engineering 190, 44-53 (2018).
[12] Van Gog, Heleen, Marijn A. Van Huis, “Structural and electronic properties of Frenkel and Schottky defects at the MgO {100} surface: spin polarization, mid-band gap states, and charge trapping at vacancy sites”, The Journal of Physical Chemistry C 123.23, 14408-14420 (2019).
[13] Ting-Chang Chang, Kuan-Chang Chang, Tsung-Ming Tsai, Tian-Jian Chu, Simon M. Sze, “Resistance random access memory”, Materials Today 19.5, 254-264 (2016).
[14] Furqan Zahoor, Tun Zainal Azni Zulkifli, Farooq Ahmad Khanday, “Resistive Random Access Memory (RRAM): an Overview of Materials, Switching Mechanism, Performance, Multilevel Cell (mlc) Storage, Modeling, and Applications”, Nanoscale research letters 15, 1-26 (2020).
[15] Wei Xiao a, Linbo Shan a, Haitao Zhang a, Yujun Fu a, Yanfei Zhao a, Dongliang Yang a, Chaohui Jiao a, Guangzhi Sun b, Qi Wang, Deyan He , “High photosensitivity light-controlled planar ZnO artificial synapse for neuromorphic computing”, Nanoscale, 13, 2502-2510 (2021).
[16] YiLong Wang, Minghui Cao, Jing Bian, Qiang Li, Jie Su, “Flexible ZnO Nanosheet-Based Artificial Synapses Prepared by Low-Temperature Process for High Recognition Accuracy Neuromorphic Computing”, Adv. Funct. Mater., 32, 2209907 (2022).
[17] Ramesh Kumar Mohapatra, Banshidhar Majhi, and Sanjay Kumar Jena, “Classification Performance Analysis of MNIST Dataset utilizing a Multi-resolution Technique”, IEEE, 15, 1-5 (2015).
[18] C. Symonds, G. Lheureux, J. P. Hugonin, J. J. Greffet, J. Laverdant, G. Brucoli, A. Lemaitre, P. Senellart, J. Bellessa, “Confined Tamm Plasmon Lasers”, Nano Lett., 13, 3179−3184 (2013).
[19] Jizhong Song, Jinhang Li, Leimeng Xu, Jianhai Li, Fengjuan Zhang, Boning Han, Qingsong Shan, Haibo Zeng, “Room-Temperature Triple-Ligand Surface Engineering Synergistically Boosts Ink Stability, Recombination Dynamics, and Charge Injection toward EQE-11.6% Perovskite QLEDs”, Adv. Mater., 30, 1800764 (2018).
[20] Te Jui Yen1, Albert chin1, Vladimir Gritsenko, “High performance All nonmetal Sinx Resistive Random Access Memory with Strong process Dependence”, Scientific Report, 10.1, 2807 (2020).
[21] Ye Wu1, Yi Wei1, Yong Huang, Fei Cao, Dejian Yu, Xiaoming Li, “Haibo ZengCapping CsPbBr3 with ZnO to improve performance and stability of perovskite memristors”, Nano Research, 10(5), 1584–1594 (2017).
[22] Shuang Gao, Gang Liu, Huali Yang, Chao Hu, Qilai Chen, Guodong Gong, Wuhong Xue, Xiaohui Yi, Jie Shang, Run-Wei Li, “An Oxide Schottky Junction Artificial Optoelectronic Synapse”, ACS Nano, 13, 2634−2642 (2019).
[23] Duygu Kuzum, Rakesh G. D. Jeyasingh, Byoungil Lee, H.-S. Philip Wong, “Nanoelectronic Programmable Synapses Based on Phase Change Materials for Brain-Inspired Computing”, Nano Lett., 12, 2179–2186 (2012).