本研究聚焦於氧化鈮薄膜之質子型電化學憶阻器元件在仿神經運算行為中的寫入與抹除行為，利用濺鍍法生長氧化鈮薄膜，並製成氧化鈮薄膜憶阻器元件。此元件主要包含以鄰苯二甲酸氫鉀溶液作為質子的來源及氧化鈮薄膜作為反應層之電化學憶阻器。本研究利用外加電場，使溶液中之質子嵌入或嵌出氧化鈮薄膜並與之發生反應，進而改變氧化鈮薄膜之電阻值，實驗結果揭示氧化鈮之質子型電化學憶阻器元件在仿神經運算行為之電子神經突觸的應用潛力。
本實驗透過掃描式電子顯微鏡與原子力顯微鏡比較氧化鈮薄膜在質子嵌入前後，薄膜表面形貌與粗糙度之差異。利用穿透式電子顯微鏡觀察氧化鈮薄膜在質子嵌入後之微結構變化。也透過電性量測搭配XPS分析，比較質子嵌入前與嵌入後之氧化鈮薄膜憶阻器元件，觀察到溶液中之質子與氧化鈮之氧原子產生鍵結，同時部分鈮原子之價數從+5變為+4。最後運用光激螢光分析氧化鈮薄膜之缺陷對質子嵌入與電性量測的影響。

This study focuses on the write and erase behaviors of a protonic electrochemical memristor device based on niobium oxide thin films in neuromorphic computing applications. Niobium oxide thin films were grown using sputtering techniques, and memristor devices were fabricated from these films. The device primarily consists of an electrochemical memristor with potassium hydrogen phthalate (KHP) solution as the proton source and niobium oxide thin films as the reactive layer. By applying an external electric field, protons from the solution are inserted into or extracted from the niobium oxide thin films, altering their resistance. The experimental results reveal the potential of proton-type electrochemical niobium oxide memristors for applications as artificial synapses in neuromorphic computing.
In this experiment, the surface morphology and roughness of the niobium oxide thin films were compared before and after proton insertion using scanning electron microscopy (SEM) and atomic force microscopy (AFM). Transmission electron microscopy (TEM) was employed to observe the microstructural changes in the niobium oxide thin films after proton insertion. Additionally, electrical measurements combined with X-ray photoelectron spectroscopy (XPS) analysis were used to compare the niobium oxide memristor devices before and after proton insertion. It was observed that protons from the solution bonded with oxygen atoms in the niobium oxide, and some niobium atoms' oxidation states changed from +5 to +4. Finally, photoluminescence (PL) analysis was utilized to study the effects of defects in the niobium oxide thin films on proton insertion and electrical measurements.

[1] A. de Vries, "The growing energy footprint of artificial intelligence," Joule, vol. 7, no. 10, pp. 2191-2194, Oct 18 2023, doi: 10.1016/j.joule.2023.09.004.
[2] M. A. Hofman, "Evolution of the human brain: when bigger is better," Front Neuroanat, vol. 8, Mar 27 2014, doi: 10.3389/fnana.2014.00015.
[3] S. Li et al., "Magnetic skyrmions for unconventional computing," Materials Horizons, 10.1039/D0MH01603A vol. 8, no. 3, pp. 854-868, 2021, doi: 10.1039/D0MH01603A.
[4] S. Furber, "Large-scale neuromorphic computing systems," J Neural Eng, vol. 13, no. 5, Oct 2016, doi: 10.1088/1741-2560/13/5/051001.
[5] D. Markovic, A. Mizrahi, D. Querlioz, and J. Grollier, "Physics for neuromorphic computing," Nat Rev Phys, vol. 2, no. 9, pp. 499-510, Sep 2020, doi: 10.1038/s42254-020-0208-2.
[6] V. Pawlak, J. R. Wickens, A. Kirkwood, and J. N. Kerr, "Timing is not Everything: Neuromodulation Opens the STDP Gate," Frontiers in Synaptic Neuroscience, Review vol. 2, 2010-October-25 2010, doi: 10.3389/fnsyn.2010.00146.
[7] G. Q. Bi and M. M. Poo, "Synaptic modifications in cultured hippocampal neurons: Dependence on spike timing, synaptic strength, and postsynaptic cell type," J Neurosci, vol. 18, no. 24, pp. 10464-10472, Dec 15 1998, doi: 10.1523/jneurosci.18-24-10464.1998.
[8] Y. Z. Zhang, P. Huang, B. Gao, J. F. Kang, and H. Q. Wu, "Oxide-based filamentary RRAM for deep learning," J Phys D Appl Phys, vol. 54, no. 8, Feb 25 2021, doi: 10.1088/1361-6463/abc5e7.
[9] A. Sawa, "Resistive switching in transition metal oxides," Mater Today, vol. 11, no. 6, pp. 28-36, Jun 2008, doi: 10.1016/S1369-7021(08)70119-6.
[10] H. Kang et al., "Two- and three-terminal HfO-based multilevel resistive memories for neuromorphic analog synaptic elements," Neuromorph Comput En, vol. 1, no. 2, Dec 2021, doi: 10.1088/2634-4386/ac29ca.
[11] M. Le Gallo and A. Sebastian, "An overview of phase-change memory device physics," J Phys D Appl Phys, vol. 53, no. 21, May 20 2020, doi: 10.1088/1361-6463/ab7794.
[12] S. G. Sarwat, B. Kersting, T. Moraitis, V. P. Jonnalagadda, and A. Sebastian, "Phase-change memtransistive synapses for mixed-plasticity neural computations," Nat Nanotechnol, vol. 17, no. 5, pp. 507-+, May 2022, doi: 10.1038/s41565-022-01095-3.
[13] H. Abbas, J. Y. Li, and D. S. Ang, "Conductive Bridge Random Access Memory (CBRAM): Challenges and Opportunities for Memory and Neuromorphic Computing Applications," Micromachines-Basel, vol. 13, no. 5, May 2022, doi: 10.3390/mi13050725.
[14] H. L. Ma et al., "Coexistence of unipolar and bipolar modes in Ag/ZnO/Pt resistive switching memory with oxygen-vacancy and metal-Ag filaments," Chinese Phys B, vol. 25, no. 12, Dec 2016, doi: 10.1088/1674-1056/25/12/127303.
[15] D. Verma, T. C. Chen, B. Liu, and C. S. Lai, "BiOSe-based CBRAM integrated artificial synapse," Heliyon, vol. 9, no. 12, Dec 2023, doi: 10.1016/j.heliyon.2023.e22512.
[16] Q. F. Ou, B. S. Xiong, L. Yu, J. Wen, L. Wang, and Y. Tong, "In-Memory Logic Operations and Neuromorphic Computing in Non-Volatile Random Access Memory," Materials, vol. 13, no. 16, Aug 2020, doi: 10.3390/ma13163532.
[17] T. Mikolajick et al., "Next generation ferroelectric materials for semiconductor process integration and their applications," J Appl Phys, vol. 129, no. 10, Mar 14 2021, doi: 10.1063/5.0037617.
[18] M. Jerry et al., "Ferroelectric FET analog synapse for acceleration of deep neural network training," in 2017 IEEE international electron devices meeting (IEDM), 2017: IEEE, pp. 6.2. 1-6.2. 4.
[19] J. H. Baek et al., "Artificial synaptic devices based on biomimetic electrochemistry: A review," Mater Res Bull, vol. 176, Aug 2024, doi: 10.1016/j.materresbull.2024.112803.
[20] X. H. Yao et al., "Protonic solid-state electrochemical synapse for physical neural networks," Nat Commun, vol. 11, no. 1, Jun 19 2020, doi: 10.1038/s41467-020-16866-6.
[21] S. Y. Cheng et al., "Tuning electromagnetic absorption properties of transition metal oxides by hydrogenation with nascent hydrogen," Chem Eng J, vol. 417, Aug 1 2021, doi: 10.1016/j.cej.2020.127980.
[22] M. D. Pickett, G. Medeiros-Ribeiro, and R. S. Williams, "A scalable neuristor built with Mott memristors," Nat Mater, vol. 12, no. 2, pp. 114-117, Feb 2013, doi: 10.1038/Nmat3510.
[23] L. G. Gao, P. Y. Chen, and S. M. Yu, "NbO based oscillation neuron for neuromorphic computing," Appl Phys Lett, vol. 111, no. 10, Sep 4 2017, doi: 10.1063/1.4991917.
[24] Q. Duan, Z. Jing, K. Yang, R. Huang, and Y. Yang, "Oscillation neuron based on threshold switching characteristics of niobium oxide films," in 2019 IEEE International Workshop on Future Computing (IWOFC, 2019: IEEE, pp. 1-3.
[25] A. C. Kozen et al., "In Situ Hydrogen Plasma Exposure for Varying the Stoichiometry of Atomic Layer Deposited Niobium Oxide Films for Use in Neuromorphic Computing Applications," Acs Appl Mater Inter, vol. 12, no. 14, pp. 16639-16647, Apr 8 2020, doi: 10.1021/acsami.0c01279.
[26] https://www.metrohm.com/zh_tw.html (accessed 2024).
[27] https://www.bruker.com/zh/products-and-solutions/diffractometers-and-x-ray-microscopes/x-ray-diffractometers/d8-advance-family/d8-advance-eco.html (accessed 2024).
[28] https://www.jeol.com/products/scientific/xps_esca/JPS-9030.php (accessed 2024).
[29] https://ctrmost-cfc.ncku.edu.tw/p/404-1210-7290.php?Lang=zh-tw (accessed 2024).
[30] B. L. Sopori et al., "Hydrogen in silicon: A discussion of diffusion and passivation mechanisms," Sol Energ Mat Sol C, vol. 41-2, pp. 159-169, Jun 1996, doi: Doi 10.1016/0927-0248(95)00098-4.
[31] J. L. Benton, C. J. Doherty, S. D. Ferris, D. L. Flamm, L. C. Kimerling, and H. J. Leamy, "Hydrogen Passivation of Point-Defects in Silicon," Appl Phys Lett, vol. 36, no. 8, pp. 670-671, 1980, doi: Doi 10.1063/1.91619.
[32] X. F. Zhou, Z. C. Li, Y. Q. Wang, X. Sheng, and Z. J. Zhang, "Photoluminescence of amorphous niobium oxide films synthesized by solid-state reaction," Thin Solid Films, vol. 516, no. 12, pp. 4213-4216, Apr 30 2008, doi: 10.1016/j.tsf.2007.12.112.