Heterosynaptic Memtransistors Based on Switching Operation Mechanism Using Designed Organic/Inorganic Heterostructures for Neuromorphic Electronics
Memtransistors using low-dimensional semiconductors represent a promising gate-tunable heterosynaptic architecture for neuromorphic computing. However, active layers of these devices have not yet been artificially designed or controlled. In this study, gate-pulse-tunable heterosynaptic neuromodulation is achieved using memtransistors with organic semiconductor tris(4-carbazoyl-9-ylphenyl)amine (TCTA)/MoS<sub>2</sub> heterostructures designed via energy-band engineering and bottom-contact architecture. Memristive switching is realized through distinctive low- and high-conduction states with a switching ratio of 10<sup>2</sup>, modulated by gate pulses. As the gate voltage (V<sub>G</sub>) decreases from +30 to -30 V, the memristive hysteresis for the bottom contact TCTA/MoS<sub>2</sub> FET without post-treatment and an h-BN insulating layer appears at V<sub>G</sub> = -15 V and broadens with an increasing switching ratio. Intriguingly, as V<sub>G</sub> becomes increasingly negative (V<sub>G</sub> < -15 V), trap-related space-charge-limited conduction becomes dominant. Non-volatile heterosynaptic behavior is mimicked by drain pulses and modulated by gate-pulse polarities. Applying gate-pulse only, analogous responses are observed in synaptic modulation with time constants of 100 ms for potentiation and 60 ms for depression. The design of multi-functional memory and realization of gate-pulse-tunable memtransistors using nanoscale TCTA/MoS<sub>2</sub> can promote energy-efficient, tunable, and reliable heterosynaptic neuromorphic electronics.
Authors
- Dayeong Kwon
- Taek Joon Kim (ORCID: https://orcid.org/0000-0001-5135-7542)
- Sang wook Song
- Sang‐Hun Lee (ORCID: https://orcid.org/0000-0001-7553-8188)
- Jinsoo Joo
- Hye Lim Jeong
Institutions
- Korea University (KR)
- Korea University (JP)
Publication Details
- Journal
- Advanced Science
- Published
- 2026-01-05
- DOI
- https://doi.org/10.1002/advs.202517149
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Research Foundation
- Sungkyunkwan University