Northwestern Engineers Develop Printed Artificial Neurons Capable Of Two-Way Communication With Living Brain Cells

Northwestern engineers develop flexible, printed neurons that mimic brain signals, offering a breakthrough for AI efficiency and neuroprosthetics.

By: AXL Media

Published: Apr 15, 2026, 7:32 AM EDT

Source: Information for this report was sourced from EurekAlert!

Bridging the Gap Between Silicon and Biology

Engineers at Northwestern University have achieved a significant breakthrough in neurotechnology by developing printed artificial neurons that can effectively "talk" to living biological cells. Unlike traditional silicon-based transistors that remain rigid and fixed, these new devices are fabricated from flexible, soft materials designed to mimic the three-dimensional, dynamic nature of the human brain. Led by Professor Mark C. Hersam, the research team successfully demonstrated that these artificial neurons could produce electrical pulses indistinguishable from natural signals. This bi-directional compatibility opens new doors for brain-machine interfaces, potentially allowing for more seamless integration of medical implants with the human nervous system.

Aerosol Jet Printing of Neuromorphic Hardware

The fabrication process utilizes advanced aerosol jet printing to deposit specialized electronic inks onto flexible polymer substrates. These inks are composed of graphene, which functions as a conductor, and molybdenum disulfide nanosheets that act as semiconductors. While previous attempts to use such materials often failed due to the interference of stabilizing polymers, the Northwestern team leveraged these perceived imperfections to their advantage. By partially decomposing the polymer through electrical currents, they created conductive filaments that allow the device to produce complex, non-linear electrical responses. This method produces hardware that is not only low-cost and flexible but also significantly more environmentally friendly due to its additive manufacturing nature.

The Energy Efficiency Crisis of Modern AI

A primary motivation behind this research is the unsustainable power consumption of modern artificial intelligence. Current AI models require massive data centers that consume gigawatts of electricity and vast amounts of water for cooling. Professor Hersam noted that the human brain is approximately five orders of magnitude more energy-efficient than a digital computer, primarily because it uses diverse, specialized neurons rather than billions of identical transistors. By creating artificial neurons that mimic this biological efficiency, the researchers aim to build hardware capable of handling massive data processing tasks with only a fraction of the energy required by today’s silicon-...