Mimicking Pavlovian Conditioning with WSe2 Phototransistors

Neuromorphic engineering aims to bypass the energy and latency limitations of traditional Von Neumann architectures by emulating the biological efficiency of the mammalian brain. Two-dimensional materials, with their exceptional light-matter interaction, have emerged as prime candidates for next-generation optoelectronic artificial synapses. In this work, we demonstrate a robust optoelectronic synaptic device based on a mechanically exfoliated tungsten diselenide (WSe2) field-effect transistor. Unlike complex heterostructures, the proposed device exploits intrinsic defect-mediated charge trapping mechanisms to achieve neuromorphic functionalities. We report a reversible modulation of channel conductance, where optical stimuli induce potentiation via persistent photoconductivity and electrical gate pulses trigger depression. A key finding is the ability to selectively switch between Short-Term and Long-Term Plasticity simply by tuning the drain bias polarity, utilizing the asymmetry of Schottky contacts. To validate the device's potential for bioinspired computing, we successfully emulate Pavlovian associative learning at the hardware level. These results establish exfoliated WSe2 as a simple yet versatile platform for light-stimulated neuromorphic computing.