A van der Waals Synaptic Transistor Based on Ferroelectric Hf0.5Zr0.5O2 and 2D Tungsten Disulfide,Advanced Electronic Materials

当前位置： X-MOL 学术 › Adv. Electron. Mater. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

A van der Waals Synaptic Transistor Based on Ferroelectric Hf0.5Zr0.5O2 and 2D Tungsten Disulfide
Advanced Electronic Materials ( IF 5.3 ) Pub Date : 2020-05-08 , DOI: 10.1002/aelm.202000057
Li Chen _{1,

2} , Lin Wang _{1,

2} , Yue Peng ₃ , Xuewei Feng _{1,

2} , Soumya Sarkar ₄ , Sifan Li _{1,

2} , Bochang Li _{1,

2} , Liang Liu ₅ , Kaizhen Han ₁ , Xiao Gong ₁ , Jingsheng Chen ₅ , Yan Liu ₃ , Genquan Han ₃ , Kah‐Wee Ang _{1,

2}

Affiliation

Neuromorphic computing on the hardware level is promising for performing ever‐increasing data‐centric tasks owing to its superiority to conventional von Neumann architecture in terms of energy efficiency and learning ability. One key aspect to its implementation is the development of artificial synapses that can effectively emulate the multiple functionalities exhibited by their biological counterparts. Here, building on an inorganic ferroelectric gate stack integrated with a 2D layered semiconductor (WS₂), a new type of ferroelectricity‐based synaptic transistor that differs from those relying on interface traps or floating gate configuration is reported. By virtue of a 6 nm thick ferroelectric hafnium zirconium oxide by atomic layer deposition and postannealing treatment, the device shows a channel resistance change ratio above 10⁵ corresponding to opposite ferroelectric polarization direction. Furthermore, by applying electrical stimulus to the gate, it demonstrates good capability to mimic various synaptic behaviors including long‐term potentiation, long‐term depression, spike‐amplitude‐dependent plasticity, and spike‐rate‐dependent plasticity. Given the inherent compatibility of the ferroelectric gate stack with existing fabrication technology, and the reliability of ferroelectricity engineering, this work paves the way toward practical implementation of synaptic devices in neuromorphic circuits.

中文翻译：

基于铁电Hf0.5Zr0.5O2和二维二硫化钨的范德华突触晶体管

由于在能量效率和学习能力方面优于传统的冯·诺依曼体系结构，因此硬件级的神经形态计算有望执行日益增长的以数据为中心的任务。其实现的一个关键方面是人工突触的发展，该突触可以有效地模仿其生物学对应物所展现的多种功能。在此，建立在与2D层状半导体（WS ₂），据报道，一种新型的基于铁电的突触晶体管不同于依赖于界面陷阱或浮栅配置的突触晶体管。通过原子层沉积和后退火处理，形成了厚度为6 nm的铁电氧化ha锆，该器件的沟道电阻变化比大于10 ⁵对应于相反的铁电极化方向。此外，通过在门上施加电刺激，它表现出了良好的模仿各种突触行为的能力，包括长期增强，长期抑制，尖峰幅度相关的可塑性和尖峰速率相关的可塑性。考虑到铁电栅极叠层与现有制造技术的固有兼容性以及铁电工程的可靠性，这项工作为在神经形态电路中实际实现突触设备铺平了道路。

更新日期：2020-05-08

点击分享查看原文

点击收藏

阅读更多本刊最新论文本刊介绍/投稿指南11