Optoelectronic heterostructure transistor based on perovskite-silicon for neuromorphic computing

The Von-Neumann computing has its own shortcomings, which can be overcome by certain extent by using neuromorphic computing, as it excels in parallel processing and self-adaptive learning by consuming lower amount of energy. Essential components of neuromorphic computing are synaptic devices that mimic biological synapses. The photonic neuromorphic chips can be an attractive alternative for the next generation of artificially intelligent systems as they require comparatively low power, having low crosstalk, and high density compared to conventional neuromorphic chips based on electronic synapses. In this regard, the present work discusses the design and analysis of heterojunction optoelectronic transistor based synaptic device. The design is considered to be carried out using the low cost materials that do not require complex design process. MAPbI3 (methyl ammonium lead iodide) as a perovskite and silicon are considered to be the material for one of the designs. Also, in order to overcome the instability issue of organic perovskite (MAPbI3), the inorganic perovskites are considered for design. And to obtain high responsivity, the multiple quantum well structure of same crystal geometry inorganic perovskite materials such as CsPbI3 (Cesium lead iodide) and CsPbBr3 (Cesium lead bromide) is being considered for quantum well and for quantum barrier, respectively in the photo absorber layer of the heterojunction optoelectronic synaptic transistor. Considerable improvement in the responsivity has been observed, which suggest that the proposed designs have the realistic potential to take up a part in the neuromorphic computing.