Separating Motion Artifacts from Biomarker Signals in Flexible MoS2 FET Cortisol Sensors: A Coupled Multiphysics Model and Finite-Element Analysis of Strain-Binding Cross-Talk and Mitigation Strategies
This paper presents a coupled multiphysics model and finite-element analysis to quantify and evaluate mitigation strategies for motion-induced artifacts in flexible MoS2 FET cortisol sensors, revealing that while twin-sensor subtraction and substrate thinning reduce strain-binding cross-talk, current designs struggle to consistently meet the 1 ng/mL accuracy target due to persistent Schottky contact effects and aptamer-dependent limitations.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
The human body is a constant source of chemical signals, and one of the most important is cortisol, a hormone released during stress. Measuring cortisol in sweat offers a way to monitor stress levels without needles or blood draws, potentially leading to wearable devices that track our well-being in real time. To do this, scientists are turning to sensors made from molybdenum disulfide, a material so thin it is only a few atoms deep. This extreme thinness makes the sensor incredibly sensitive to chemical changes, allowing it to detect tiny amounts of hormones. However, this same thinness creates a major problem: the material is also highly sensitive to physical stretching and bending. When a person moves their arm or walks, the sensor bends, and this physical strain changes the electrical signal just as much as the hormone does. For a device to be useful, it must be able to tell the difference between a signal caused by a stressful event and a signal caused by the simple act of moving a wrist.
Researchers at Kwara State University and Olabisi Onabanjo University in Nigeria set out to solve this specific puzzle. They built a detailed computer model to simulate how a flexible sensor made of molybdenum disulfide behaves when worn on the skin. Instead of building a physical device and testing it in a lab, they created a virtual chain of events that starts with a person moving their body and ends with the electrical reading on the sensor. They modeled three common movements: bending the wrist, bending the elbow, and walking. Their goal was to see exactly how much of the electrical signal comes from the cortisol and how much is just "noise" created by the motion. They found that for a standard sensor design, the movement of the wrist alone creates a false signal that is nearly half the size of the maximum hormone signal the device is supposed to measure. In fact, a single bend of the wrist could make the device think the user is under extreme stress when they are actually at rest.
The researchers discovered that the source of this confusion is not just the thin film itself, but the metal contacts that connect the film to the rest of the circuit. When the sensor bends, the metal contacts change their electrical properties even more dramatically than the sensing material does. About 70 percent of the false signal generated by movement comes from these contact points. This finding is crucial because it suggests that simply making the sensor thinner or adding a protective layer is not enough to fix the problem. In fact, the study showed that adding a stiff, rigid island under the sensor to try to keep it flat does not isolate the channel as expected. Under some loading conditions, the rigid island retains 18 to 109 percent of the baseline strain, meaning it offers little to no protection. Under other conditions, where the skin dictates the patch's shape, the island can actually cause the channel to experience up to 300 percent of the baseline strain. This contradicts a common idea in wearable electronics that stiffening a small area protects the sensitive part, showing that what works for stretching does not necessarily work for bending.
To separate the true hormone signal from the motion noise, the team tested several strategies. They found that simply filtering out fast changes in the signal, a technique often used for other types of wearable sensors, does not work here. This is because bending a joint is a slow, one-way motion that looks very similar to a slow change in hormone levels, so a filter cannot tell them apart. The most promising solution involves using two sensors side by side. One sensor is designed to detect cortisol, while the other is a "dummy" sensor that cannot detect the hormone but experiences the exact same physical movement. By comparing the two, the system can subtract the motion noise. However, this method only works if the two sensors are perfectly matched and if the main sensor is built on an extremely thin base. The study showed that by combining a very thin plastic base with a matched pair of sensors, the error could be reduced to a level that is close to the target for accurate measurement.
The results of this simulation provide a clear roadmap for building these devices, but they also highlight the difficulty of the task. The researchers emphasize that their findings are based on computer models, not physical experiments, so the numbers represent predictions rather than final proof. They found that for sensors placed on the back of the wrist, where bending is frequent and sharp, the motion error is so large that it is nearly impossible to get an accurate reading without using the twin-sensor method and an ultra-thin base. For the elbow, the situation is slightly better but still challenging. The study concludes that while the technology holds great promise for monitoring stress, the path to a reliable wearable device requires careful engineering of the sensor's physical structure, not just its chemical sensitivity. The work serves as a guide for future inventors, showing exactly where the pitfalls lie and which design choices are likely to succeed in separating the body's chemical signals from its physical movements.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.