Related Experiment Video
Updated: Jan 15, 2026
10:32
Design, Surface Treatment, Cellular Plating, and Culturing of Modular Neuronal Networks Composed of Functionally Inter-connected Circuits
Published on: April 15, 2015
8.9K
Thin-Film Silicon Nanosheet-Based Neuromorphic Assemblies
Rana Biswas1,2,3, Moneim Elshobaki1, Jeremy B Essner4
1Microelectronics Research Center and Department of Electrical and Computer Engineering, Iowa State University, Ames, Iowa 50011, United States.
ACS Applied Materials & Interfaces
|October 7, 2025
Summary
Silicon nanosheet assemblies exhibit brain-like neuromorphic behavior. These materials show potential for developing energy-efficient, next-generation computing networks inspired by the human brain.
Area of Science:
- Materials Science
- Neuroscience
- Computer Engineering
Background:
- Neuromorphic computing aims to mimic the brain's structure and function for energy-efficient computation.
- Silicon-based materials are extensively explored for electronic applications, but novel architectures are needed for advanced functionalities.
- Understanding charge dynamics in nanoscale materials is crucial for developing new computing paradigms.
Purpose of the Study:
- To demonstrate and characterize neuromorphic behavior in silicon nanosheet (Si-NS) assemblies.
- To investigate the underlying mechanisms responsible for the observed neuroplastic characteristics.
- To assess the potential of Si-NSs for future brain-inspired computing.
Main Methods:
- Synthesis of Cl- and H-passivated Si-NSs via topotactic deintercalation of CaSi2.
- Fabrication of Si-NS assemblies between gold source-drain electrodes on a silicon platform using photolithography and solution processing.
- Characterization of electrical properties, including response to spiking voltage inputs and analysis of current outputs using power-law decay models.
- Density Functional Theory (DFT) and electron-trapping simulations to elucidate charge dynamics.
Main Results:
- Si-NS assemblies exhibited neuromorphic behavior, with spiking voltage inputs eliciting decaying spiking current outputs (power-law decay, ~t^-β, β ≈ 0.7-1.1).
- A distinct 'learning' phase was observed at higher frequencies, analogous to synaptic potentiation.
- DFT and simulations indicated that Si-dangling bonds at NS surfaces are likely responsible for charge trapping and release, driving neuroplasticity.
Conclusions:
- Si-NS assemblies demonstrate promising neuromorphic characteristics, including plasticity and learning.
- The observed behavior is attributed to charge dynamics at Si-NS surfaces, particularly Si-dangling bonds.
- Si-NSs represent a viable material for developing energy-efficient, next-generation neuromorphic computing networks.
