Hybrid nanotube electrodes developed for safer brain-machine interfaces

Brain–computer interfaces are technologies that enable direct communication between brain activity and external devices, enabling researchers to monitor and interpret brain signals in real time. These connections often involve arrays of tiny, hair-like electrodes called “microelectrodes” which are implanted within the brain to record or stimulate electrical activity.

For decades, microelectrodes have faced a challenge in balancing conductivity with tissue compatibility. Rigid metal or silicon-based electrodes enable stable signal recordings but often damage the delicate brain tissues, whereas softer polymer electrodes reduce harm but suffer from poor signal transmission.

Bridging this gap, a research team led by Associate Professor Jong G. Ok from the Department of Mechanical and Automotive Engineering, Seoul National University of Science and Technology, Seoul and Dr. Maesoon Im from Brain Science Institute, Korea Institute of Science and Technology (KIST) developed a microelectrode with three-dimensional “forests” of carbon nanotubes (CNTs) that efficiently conduct electricity like metals but also flex like soft tissue. Embedded in an elastic polymer base, the arrays are approximately 4,000 times softer than silicon and about 100 times softer than polyimide.

The findings of the study were made available online in Advanced Functional Materials on June 27, 2025.

To fabricate these arrays, the researchers used a multi-step process for vertical growth of CNTs and a proprietary polymer-CNT hybridization technique. The resultant arrays demonstrated stable insertion in brain tissues, enabling precise recording of the visual responses. The arrays also showed a marked reduction of inflammatory responses compared to tungsten microwires—making them a promising option for safer brain applications.

“By combining the vertically aligned CNTs with a flexible polymer, we achieved both high electrical performance and mechanical compliance within one device,” says lead author, Dr. Ok. “This dual capability enables a long-term, stable neural interface without harming the surrounding tissue.”

The in-vivo experiments in mice also validated the device’s capacity to record light-evoked responses from visual cortex neurons (visual center present at the back of the brain). Additionally, one-month implantation of CNT arrays showed lower activation of astrocytes and microglial cells (cells involved in immune response) than those observed in conventional electrodes—highlighting the superior long-term compatibility of CNT arrays.

The findings open possibilities for use in visual prosthetics, especially for patients with retinal degeneration or optic nerve damage. Additionally, the technology could also be extended to cortical implants for brain-machine interfaces and tools for studying visual processing in neuroscience.

“Refining this technology to read visual attention could unlock new opportunities in brain-assisted communication, Retinal Prosthesis (Bionic Eye) and immersive AR/VR experiences,” highlights Dr. Ok.

In the long run, the researchers aim to scale down the arrays for subcellular dimensions to record brain signals at higher resolutions. These advances could inspire next-generation bioelectronic devices, helping restore or enhance vision via direct brain connectivity.