双极扩散
神经形态工程学
逆变器
晶体管
逻辑门
计算机科学
电子线路
电压
电子工程
非易失性存储器
电气工程
跨导
波形
材料科学
和大门
光电子学
信号(编程语言)
纳米技术
作者
Shengyu Cong,Junxin Chen,Miao Xie,Ziyi Deng,Chaoyue Chen,Riping Liu,Jiayao Duan,Xiuyuan Zhu,Zhengke Li,Yuhua Cheng,Wei Huang,Iain McCulloch,Wan Yue
出处
期刊:Science Advances
[American Association for the Advancement of Science]
日期:2024-10-09
卷期号:10 (41): eadq9405-eadq9405
被引量:26
标识
DOI:10.1126/sciadv.adq9405
摘要
Organic electrochemical transistor (OECT)–based inverter introduces new prospects for energy-efficient brain-inspired artificial intelligence devices. Here, we report single-component OECT-based inverters by incorporating ambipolar p(gDPP-V). Notably, p(gDPP-V) shows state-of-the-art ambipolar OECT performances in both conventional (p/n-type mode transconductance of 29/25 S cm −1 ) and vertical (transconductance of 297.2/292.4 μS μm −2 under p/n operation) device architectures. Especially, the resulting highly stable vertical OECT–based inverter shows a high voltage gain of 105 V V −1 under a low driving voltage of 0.8 V. The inverter exhibits undiscovered voltage-regulated dual mode: volatile receptor and nonvolatile synapse. Moreover, applications of physiology signal recording and demonstrations of NAND/NOR logic circuits are investigated within the volatile feature, while neuromorphic simulations with a convolutional neural network and image memorizing capabilities are explored under the nonvolatile behavior. The ambipolar OECT–based inverter, capable of both volatile and nonvolatile operations, provides possibilities for the applications of reconfigurable complementary logic circuits in novel neuromorphic computing paradigms.
科研通智能强力驱动
Strongly Powered by AbleSci AI