The perceptual learning function
of a simulating human body is
very important for constructing a neural computing system and a brainlike
computer in the future. The sense of smell is an important part of
the human sensory nervous system. However, current gas sensors simply
convert gas concentrations into electrical signals and do not have
the same learning and memory function as synapses. To solve this problem,
we propose a new sensing idea to induce and activate the synaptic
properties of transistors by adjusting the oxygen vacancy in the active
layer. This sensor combines gas detection with synaptic memory and
learning and overcomes the disadvantage of the separation of synaptic
transistors and sensors, thus greatly reducing the cost of production.
This work combines the detection of N,N-dimethylformamide (DMF) gas with the synaptic mechanism of human
olfactory nerves. We successfully fabricated an InGdO nanofiber field-effect
transistor by electrostatic spinning and simulated the response of
human olfactory synapses to target gas by regulating the oxygen vacancy
of the InGdO nanofiber. The synaptic transistor response under different
concentrations of unmodulated pulses is tested, and the pavlovian
conditioned reflex experiment is simulated successfully. This work
provides a new idea of a gas sensor device, which is very important
for the development of high-performance gas sensors and bionic electronic
devices in the future.