Abstract
Despite the recent advancements in sensor detection of biomarkers in sweat, there is no sensor capable of long-term detection in constricted or load-bearing applications where other flexible plastic sensors might cause discomfort. In this work, a carbon nanotube-based fabric sensor capable of real-time detection of sodium in human sweat is presented. The sensor consists of an electrospun nylon-6 base functionalized with multi-walled carbon nanotubes for enhanced conductivity and p-tert-calix[4]arene for enhanced sodium selectivity. Monitoring sweat is a compelling choice to gain insight into a person's hydration at a molecular level, as sweat is rich in physiological and metabolic information that can be obtained non-invasively. Body surface sweat is approximately 99% water but also contains over 40 other compounds, most notably the presence of extracellular ions: sodium and chloride. Sodium has been shown to have a notable correlation to body fluid balance and is lost in ample quantities during exercise, making it a promising candidate for the detection of dehydration (i.e., hypernatremia) and overhydration (i.e., hyponatremia). Other compounds commonly found in surface sweat include intracellular ions, metabolites, hormones, small proteins, and peptides. This rich composition and its correspondence to blood chemistry can be the key to accessing the body's biomolecular (health) state through non-invasive monitoring and diagnostics.
When monitoring biomarkers in sweat, the sampling step has the greatest impact on accuracy. Traditional methods for monitoring sweat sodium concentration involve collecting a sweat sample and performing chemical analysis in two separate steps. Typical sweat collection procedures include whole body washdown, sweat collection patches, arm bags, and Macroducts® which require some combination of tedious procedures, trained professionals, and large equipment for analysis. These methods are often limited by insufficient sample volumes, non-negligible sample evaporation, possible analyte degradation between sampling and analysis all of which strongly impacts the reliability and sensitivity of the measurement.
To circumvent the problems caused by separating sampling and analysis, sweat sampling can also be completed using wearable sensors. Through in-situ sampling and analysis, wearable sweat sensors can perform the necessary real-time measurements with freshly generated sweat rather than a mixture of new and old sweat. Functional absorbent materials (e.g., paper, nonwoven, cellulosic, or hydrogels) are typically introduced between the skin and the sensing component; although as shown in this work, they can also be introduced as sensing components during sampling. The advantages of functional absorbent sensing materials are their low cost, multiple functions, real-time and continuous sensing, and most importantly the enhanced breathability of sweat glands. As such, these wearable sensors avoid the issue of altering the sweat composition underneath because they remove the old sweat accumulation as new sweat wicks into the material and washes away old sweat. This also allows physiological monitoring for longer periods of time than traditional epidermal microfluidic devices. Additionally, functional absorbent materials can also be integrated into fabrics with inherent moisture wicking properties to bring sweat to a sensing area.
This work presents the characterization and on-body testing of a previously developed functional absorbent material sensor made of a multi-walled carbon nanotube (MWCNT) functionalized nylon-6 nanocomposite. The wearable-fabric sensor works by measuring resistance of the nanocomposite material as sodium binds to the sensor fabric. The functionalization of the nanocomposite material was verified by FTIR; while XRD spectra show that the electrospun nylon-6 is made of -form crystals and the resulting nanocomposite is intercalated/exfoliated with an increase in crystalline size. The optimized MWCNT/nylon-6 nanocomposite sensor has a MWCNT loading close to the percolation threshold at approximately 1-2 wt%. Experimental results demonstrated that the sensor is selective to sodium in sweat at a sensitivity of 10 mM sodium. When worn on the body (attached to human skin), the sensor is capable of real-time monitoring of sodium concentration in sweat in the physiologically relevant range of 10-110 mM with up to 95% accuracy. Additionally, the obtained signal can be sent out via Bluetooth so that changes in hydration status can be detected in real-time.
Overall, this carbon nanotube-based fabric sensor can be integrated into "smarter" clothing to read an array of biomarkers in sweat. This could lead to better understanding of health and disease processes (e.g., analysis of common diabetic neurological complications, diagnosis of cystic fibrosis (CF), sweat monitoring for advanced prosthetic limb applications and bedridden patients, and athlete performance tracking) resulting in better treatments and health outcomes for all patients.
Figure 1