Development of Sensory Adapting Mimicking Neuromophic Phototransistor by two SKKU Research Teams
- 공과대학
- Hit4964
- 2020-10-14
Hong Sung-in (Professor Kim Sun-guk's Research Team) in the Department of New Material Engineering,
Dr. Choi Seung-hee (Professor Yoon Dae-ho's research team),
Development of Sensory Adapting Mimicking Neuromophic Phototransistor
- Photoreceptor compliance sensor using phase separation material by light
- Expected application of the new neuromorphic chip
Dr. Hong Sung-in (first author, professor Kim Sun-guk's research team) and Dr. Choi Seung-hee (co-author, research team of Professor Yoon Dae-ho) of the New Material Engineering Department announced that they have developed a neuromophic photo-transistor that mimics human sensory compliance through joint research.
Recently, research on five-sensitivity-enhancing electronic devices and materials linking humans and electronic devices has been actively conducted, and the role of smart interfaces that help them like artificial eyes, artificial skin, and artificial nose is proposed when the function of five-sensitivity is not working well. In particular, image sensors in charge of human eyes are manufactured using silicon, III-V materials.
However, conventional image sensors simply had the ability to distinguish by reading colors, so they could not imitate human photoreceptor conformations. Optical receptor conformance refers to placing a counterweight on the repetitive signals around them to conform to them and to read other signals. For example, when entering a dark place, the human eye adapts to the darkness to distinguish between different colors.
From the point of view of Neurological Engineering, which imitates the behavior of the nervous system, the current research trends are limited to synaptic behavior, requiring new neuromorphic devices such as receptors or nervous systems. The researchers predicted that the sensory adaptation photo-transistor would be an attractive neurologic device.
The research team developed a phototransistor based on the heterocompatibility of ferrovskite-electrometallic calcogen compounds, and developed a photo-sensor that mimics human photoreceptor conforming behavior by using the phase separation characteristic by light of ferrovskite.
The research team found that the dual complex of halide perovskite nanocrystalline structures and two-dimensional nanoparboard materials, in which phase separation occurs during light reactions, demonstrates optic neural network simulation behavior that can selectively adapt to RGB light sources. [See Figure 2]
Furthermore, materials with various band structures can be produced by adjusting the composition and composition ratio of halide ferrovskite based on the theory of density function in the first principle, and the change of band gap was identified by comparing materials that maximize phase separation effect. [See Figure 3]
Using this technology, it can develop an artificial intelligence image sensor that can detect or adapt only certain colors depending on the input of external light. Existing endoscopes suggest the possibility of developing intelligent image sensors that can more easily detect lesions that were difficult to distinguish due to red blood where bleeding occurred. [See Figure 4]