J Phys Chem Lett. 2025 Oct 15:11043-11052. doi: 10.1021/acs.jpclett.5c02780. Online ahead of print.

ABSTRACT

In the pursuit of high-performance and reliable next-generation neuromorphic computing systems, the precise modulation of material properties is of crucial significance for enabling advanced brain-inspired functionalities in optoelectronic devices, but still a challenge yet. This study demonstrates that the Zn/Sn ratio in kesterite chalcogenide (Cu2ZnSn(S,Se)4, CZTSSe) optoelectronic memristors play a pivotal regulatory role in fundamental resistive switching behaviors, as well as the synaptic function simulation and brain-like morphological applications thereof. For a Zn/Sn ratio of 1.1, the device’s On/Off ratio (i.e., memory window) can be increased by two-orders-of-magnitude, from 17 at a ratio of 0.7 to 1027, while the Set/Reset voltages significantly reduced to -0.38/0.34 V. Furthermore, under electrical and light (visible 655 nm to near-infrared 808 nm) stimuli, the device can more effectively emulate diverse biological synaptic behaviors, including long-term potentiation/depression (LTP/LTD), short-/long-term memory (STM/LTM), spike-intensity-dependent plasticity (SIDP), spike-duration-dependent plasticity (SDDP), spike-rate-dependent plasticity (SRDP), the “learning-forgetting-consolidation” process, and recapitulation of classical conditioning. Especially, by exploiting the device’s NIR-modulated multilevel conductance and its Zn/Sn-dependent On/Off ratio characteristics, high-security signal encryption/decryption and efficient image denoising have been achieved, respectively. These results further highlight the broad application potential of CZTSSe materials in the development of high-performance neuromorphic optoelectronic synaptic devices.

PMID:41091078 | DOI:10.1021/acs.jpclett.5c02780