News
Prof. Daeshik Kang’s Team Develops "Sensor Capable of Reading Breaths and Capturing Subtle Body Signals"
POSTECH Researchers Develop Ultra-Sensitive Sensor to Capture Subtle Body Signals, Including Breath
When a person inhales, the skin on the bridge of the nose moves inward extremely subtly—a change so minute that conventional sensors can barely detect it. Recently, however, a POSTECH research team developed an ultra-sensitive sensor capable of precisely capturing these nearly imperceptible movements. By accurately reading subtle body signals like respiration, this sensor is expected to open new possibilities for wearable devices and robotics technology.
Imperceptible stretches and movements continuously occur on the surface of the human body and structures. These include vital signs such as heartbeats, respiration, and pulse, as well as extremely subtle forces or vibrations. Consequently, highly precise sensors that measure micro-movements have been considered core components in health monitoring devices, robotics, and human-machine interface technologies. However, conventional sensors faced limitations: while they respond well to larger deformations, their sensitivity drops significantly when measuring movements smaller than the width of a human hair.
To overcome this limitation, a research team led by Professor Daeshik Kang and Dr. Jieun Park from the Department of Mechanical Engineering at POSTECH, together with Dr. Minho Kim and Dr. Taewee Kim from Ajou University, developed a breakthrough sensor. The team created an ultra-thin, flexible sensor made of nanoscale gold (one-billionth of a meter). By precisely controlling "nano-cracks"—microscopic fractures on the gold surface—they drastically enhanced its sensitivity.
The key to this sensor lies in the "depth" of the cracks in the gold. Just as a deeper crack in a glass window makes it shatter under smaller impacts, deeper cracks in the sensor's gold electrode cause significantly larger changes in electrical signals, even in response to minute deformations.
However, if cracks become too deep, electrical connectivity is severed, rendering the sensor unusable. To solve this, the research team inserted a "semi-cured polyimide" intermediate layer beneath the electrode. Cured only halfway at a low temperature (200℃), this layer reduces resistance during the gold growth process, allowing cracks to penetrate deeper without causing structural disconnection. Simply put, they created a structure that does not break easily yet remains extremely sensitive to subtle changes.
Thanks to this design, the sensor achieved a gauge factor of 100,000 at a strain of just 2%. This value represents a sensitivity more than 50 times higher than that of conventional gold-based sensors. The research team attached this sensor to the bridge of a human nose to perform a breathing experiment. When breathing, the internal pressure of the nose changes, causing the skin on the bridge to move ever so slightly—a deformation virtually impossible to measure with conventional sensors. However, the team's sensor vividly detected this movement, and the measured signals precisely matched actual breathing patterns.
Because the sensor is extremely thin, flexible, and adheres seamlessly to the skin, it causes minimal discomfort when worn. This highlights its potential to replace traditional, bulky breathing monitoring equipment like masks or belts with a new paradigm of body signal detection technology.
Professor Daeshik Kang, who led the study, stated, "By precisely controlling the micro-growth of gold, we have realized a sensor that reacts to extremely subtle movements while operating stably. It has high potential for application in diverse fields, including wearable devices and soft robotics."
Meanwhile, this research was recently published online in the prestigious international academic journal Advanced Functional Materials. The study was supported by the National Research Foundation of Korea (NRF) through the Individual Basic Research Program, Basic Research Laboratory Program, and Ph.D. Fellowships.

