Scientist discusses development of artificial synapses that mimic human brain function for next-gen AI chips

The Emerging Investigator Series by the journal Materials Horizons features outstanding work by young researchers in the field of materials science. In the latest Editorial, Dr. Eunho Lee, an Assistant Professor of Chemical and Biomolecular Engineering at Seoul National University of Science and Technology, Republic of Korea, where he leads the Functional Semiconductors and Devices Lab, discusses and elaborates upon his Emerging Investigator Series research paper titled “Improving ion uptake in artificial synapses through facilitated diffusion mechanisms.”

Dr. Lee explains, “Our research shows a simple way to make the next wave of AI hardware more efficient by improving electrolyte-based organic transistors, which are soft, low-voltage devices that process signals with ions as well as electrons. A long-standing bottleneck has been doping efficiency: how effectively ions can enter and leave the polymer channel to switch the device. We addressed this by engineering the polymer’s side chains so they actively attract and guide ions, like molecular ‘handles’ and ‘lanes,’ leading to faster and deeper ion uptake.”

The materials developed by Dr. Lee and his team point to practical uses in both AI hardware and bio-based interfaces. On the AI side, electrolyte-based organic transistors with diffusion-driven doping can function as analog synapses for ultra-low-power co-processors in wearables, cameras, and IoT nodes, enabling always-on sensing, in-sensor preprocessing, and adaptive learning with minimal energy. The same design rules also support hybrid integration with CMOS so that compact analog memory arrays reduce data movement and latency.

On the biology side, the soft and ion-friendly operation is well suited to skin and tissue environments, suggesting stable bioelectronic interfaces for closed-loop therapies and electrochemical biosensors that not only detect biomarkers but also classify patterns locally. Beyond health, these devices could support environmental monitors for water quality and point-of-care diagnostics that combine electrochemical readout with small on-board learning. In short, controlling ion motion at the molecular level enables safer, lower-voltage systems that learn at the edge.

“In the longer term, the ability to control ion motion in soft semiconductors could reshape how and where we run AI, and how electronics touch everyday life. In five years, we may see wearable and home devices that learn locally at very low voltage, which means longer battery life, less heat, and better privacy because raw data stays on the device. Health and well-being products could move beyond simple sensing toward adaptive analysis that filters noise, recognizes patterns, and personalizes feedback in real time,” remarks Dr. Lee.

Within a decade, the same materials principles could underpin soft human–machine interfaces and environmental monitors that operate reliably in wet or biological conditions, enabling continuous water quality tracking in communities and safer biointerfaces for rehabilitation and assistive technologies. Because these polymers are solution processable, there is also a path toward scalable manufacturing and lower cost, which can broaden their access.

Overall, the novel design rule established by the SeoulTech team led by Dr. Lee, namely tailoring side chains to steer ionic transport, offers a general strategy for energy-efficient, adaptive hardware that complements digital silicon. If it is possible to maintain stability and manufacturability while scaling up, this approach could reduce the energy footprint of AI, keep sensitive data local, and make intelligent systems more comfortable and more widely available.