Molecular-Layer-Engineered Artificial Synapses for Ultralow-Energy Computing

forthcoming

Although the rise of artificial intelligence technology greatly benefits our life by improving work productivity, the increasingly high energy consumption required to operate these software programs to cope with huge computational tasks becomes a serious issue. This is mainly due to inefficient data transport between separated memory and computing units (von Neumann bottleneck) in traditional computers. To minimize energy cost while maintaining computing power, this MSCA-PF research project - MoleSynap - aims to design brain-inspired electronic devices. Specifically, MoleSynap will fabricate artificial synapses by emulating biological synapses. These artificial synaptic devices will be designed in three-terminal transistor configuration (synaptic transistors), given their excellent regulation of charge carrier flow during device operation. Current organic polymer-based synaptic transistors suffer from the decreasing long-term device accuracy and high device-to-device variations, mainly due to nonperfect active-layers thin-films prepared by traditional solution-based wet methods. In this regard, MoleSynap will go beyond the current state-of-the-art on two different levels: (1) employ emerging two-dimensional coordination polymers and ionogel electrolytes as active-layer materials for improving synaptic characteristics, and (2) develop all-dry solvent-free thin-film deposition technology to fabricate active-layer thin-films with high precision and consistency. Overall, MoleSynap aims to significantly advance the field in developing organic synaptic devices for low-energy-cost computing. Moreover, MoleSynap will lay the strong foundation for building my own research group, and fulfill my long term scientific goal of bridging basic science (fundamental polymer physics) and applied science (polymer electronics devices) in polymer materials to promote environmental sustainability.