Event-Trigger Dual-Signal-Mode Flexible Piezotronic Bipolar Junction Transistor With Machine Learning for AI-Sound Recognition

Emad Iranmanesh*, Jucai Zhai, Congwei Liao, Zihao Liang, Yong Zhao, Hang Zhou, Shengdong Zhang, Charalampos Doumanidis, Anand Prakash Dwivedi, Kai Wang*

*Corresponding author for this work

Research output: Contribution to journalArticle

1 Scopus citations

Abstract

This work reports on a piezotronic n-p-n bipolar heterojunction transistor (PBJT) intended for AI-sound recognition with low power consumption footprint. An organic polymer (Poly (3-hexylthiophene)) as Base is sandwiched between two inorganic piezoelectric semiconductor ZnO layers as Emitter and Collector, respectively to form a vertically-stacked heterogeneous n-p-n BJT, which differs from the traditional BJT in that it can provide either current or voltage as an output signal to favor application’s preference. It can be regarded as two back-to-back auto-biased piezotronic heterojunction diodes where Base-Collector diode is reversely biased and Base-Emitter one is in a forward bias (or vice versa) upon sound wave stimuli. In addition, polarized charges generated at Emitter and Collector sides redistribute and as a result, give a significant enhancement in the output voltage. With machine learning AI, low frequency (LF) sound detection, mapping, and recognition have been demonstrated by this device.
Original languageEnglish
Pages (from-to)440-443
JournalIEEE Electron Device Letters
Volume44
Issue number3
DOIs
StatePublished - 1 Mar 2023

Fingerprint

Dive into the research topics of 'Event-Trigger Dual-Signal-Mode Flexible Piezotronic Bipolar Junction Transistor With Machine Learning for AI-Sound Recognition'. Together they form a unique fingerprint.

Cite this