State Key Laboratory of Analytical Chemistry for Life Science, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210023, China.
J Am Chem Soc. 2024 Oct 2;146(39):27022-27029. doi: 10.1021/jacs.4c08833. Epub 2024 Sep 18.
Chemical synapse completes the signaling through neurotransmitter-mediated ion flux, the emulation of which has been a long-standing obstacle in neuromorphic exploration. Here, we report metal-organic framework (MOF) nanofluidic synapses in which conjugated MOFs with abundant ionic storage sites underlie the ionic hysteresis and simultaneously serve as catalase mimetics that sensitively respond to neurotransmitter glutamate (Glu). Various neurosynaptic patterns with adaptable weights are realized via Glu-mediated chemical/ionic coupling. In particular, nonlinear Hebbian and anti-Hebbian learning in millisecond time ranges are achieved, akin to those of chemical synapses. Reversible biochemical in-memory encoding via enzymatic Glu clearance is also accomplished. Such results are prerequisites for highly bionic electrolytic computers.
化学突触通过神经递质介导的离子流完成信号传递，模拟这种离子流一直是神经形态探索中的一个长期障碍。在这里，我们报告了金属有机骨架（MOF）纳米流突触，其中具有丰富离子存储位点的共轭 MOF 构成了离子滞后现象，同时还作为过氧化氢酶模拟物，对神经递质谷氨酸（Glu）敏感。通过 Glu 介导的化学/离子偶联，实现了具有自适应权重的各种神经突触模式。特别是，实现了毫秒级的非线性赫布和反赫布学习，类似于化学突触。通过酶促 Glu 清除实现了可逆的生化内存编码。这些结果是高度仿生电解质计算机的前提条件。
