Conductive polymers for energy efficient neuromorphic computing

Description:

Brain-inspired neuromorphic computing offers a promising solution to address the low efficiency and high energy consumption associated with traditional von Neumann architecture. Neuromorphic computing emulates the neural networks of human brains, integrating data processing and memory within neurons and synapses. It enables more adaptable data processing and enhanced energy efficiency, making it particularly advantageous for AI-related applications.

Supervisors:

Prof. dr. Christian Nijhuis (supervisor) - c.a.nijhuis@utwente.nl
Dr. Ivana Qianqi Lin (supervisor) - q.lin@utwente.nl
Charlotte Lin (daily supervisor) - q.lin-1@utwente.nl
Harilal Kulangara Babulal (daily supervisor) -
h.kulangarababulal@utwente.nl

We have identified promising conductive polymer candidates that show time-dependent switching exactly like our synapses do, the most important feature to reduce power consumption in data processing. In this project, we prepare simple junctions (eutectic gallium and indium, “EGaIn” technique) to test different polymer formulations and their electrical properties and neuromorphic behaviour, which then can be incorporated into neural networks. Depending on the interest, the focus can be on synthetic modification, fabrication or electrical characterization:

1. Deposition of polymers in cross-bar array device, can we scale fabrication?

2. How do electrical properties vary with device area? Can we improve switching speeds?

3. Screening of different conductive polymers to optimize chemical structures. What is the role of dopants? Can we use mixed polymers?

This project welcomes students from Chemistry, Physics, and Nano/Advanced/Biomedical Technology. We are open to suggestion on project tailored to your interest.