RESUMO
α-In2Se3 semiconductor crystals realize artificial synapses by tuning in-plane and out-of-plane ferroelectricity with diverse avenues of electrical and optical pulses. While the electrically induced ferroelectricity of α-In2Se3 shows synaptic memory operation, the optically assisted synaptic plasticity in α-In2Se3 has also been preferred for polarization flipping enhancement. Here, the synaptic memory behavior of α-In2Se3 is demonstrated by applying electrical gate voltages under white light. As a result, the induced internal electric field is identified at a polarization flipped conductance channel in α-In2Se3/hexagonal boron nitride (hBN) heterostructure ferroelectric field effect transistors (FeFETs) under white light and discuss the contribution of this built-in electric field on synapse characterization. The biased dipoles in α-In2Se3 toward potentiation polarization direction by an enhanced internal built-in electric field under illumination of white light lead to improvement of linearity for long-term depression curves with proper electric spikes. Consequently, upon applying appropriate electric spikes to α-In2Se3/hBN FeFETs with illuminating white light, the recognition accuracy values significantly through the artificial learning simulation is elevated for discriminating hand-written digit number images.
RESUMO
In this work, we develop a gate-tunable gas sensor based on a MoS2/hBN heterostructure field effect transistor. Through experimental measurements and numerical simulations, we systematically reveal a principle that relates the concentration of the target gas and sensing signals (ΔI/I0) as a function of gate bias. Because a linear relationship between ΔI/I0 and the gas concentration guarantees reliable sensor operation, the optimal gate bias condition for linearity was investigated. Taking NO2 and NH3 as target molecules, it is clarified that the bias condition greatly depends on the electron accepting/donating nature of the gas. The effects of the bandgap and polarity of the transition metal dichalcogenides (TMDC) channel are also discussed. In order to achieve linearly increasing signals that are stable with respect to the gas concentration, a sufficiently large VBG within VBG > 0 is required. We expect this work will shed light on a way to precisely design reliable semiconducting gas sensors based on the characteristics of TMDC and target gas molecules.
