Large-scale fully printed “Lego Bricks” type wearable sweat sensor for physical activity monitoring

The study addresses the need for low-cost, high-throughput, and reliable wearable sweat sensors to support real-time, non-invasive monitoring of electrolytes linked to hydration and heat-related disorders. Traditional blood or saliva analyses are invasive, slow, and costly. Sweat contains key biomarkers such as Na⁺ (20–160 mM) and K⁺ (2–16 mM), where abnormalities relate to dehydration, cramps, heat stroke, and shock. Existing wearable ion-selective electrode (ISE)-based sensors face challenges in stability, repeatability, and scalable fabrication, and often do not evaluate electrolyte changes under external interventions (e.g., fluid supplementation). The research aims to develop a modular, fully printed, large-scale manufacturable “Lego Bricks” wearable sweat sensor integrating paper-based microfluidics and flexible ISEs to sensitively and stably monitor Na⁺ and K⁺ dynamics across body sites and during prolonged exercise with interventions.

Wearable sweat ion electrochemical biosensors are attractive for continuous, in situ monitoring due to their sensitivity and specificity. They typically combine (1) microfluidics for sweat capture/routing/storage, (2) all-solid-state ISEs, and (3) signal transmission. Prior work has focused on improving ISE stability and repeatability and multifunctional integration (simultaneous Na⁺, K⁺, glucose, lactate, uric acid). However, fewer studies analyze electrolyte alterations under external interference (e.g., rehydration strategies) or concurrently assess multiple ions’ correlations. Common fabrication methods for flexible devices (laser etching, 3D printing, inkjet printing) can be costly, complex, and material-restrictive, limiting high-throughput production. Paper-based microfluidics, since 2007, offer capillarity-driven, low-cost, scalable devices. Screen printing is similarly high-throughput and cost-effective for electrodes. These methods are promising for large-scale, homogeneous, integrated sweat sensing platforms.

• Device concept and fabrication: A modular “Lego Bricks” architecture was implemented by combining screen-printed flexible electrodes on PET with wax-printed paper-based microfluidic layers. Components (Na⁺ and K⁺ ISEs; microfluidic layers 3DM-P, 3DM-W, and adjusted 3DM-W) can be vertically stacked via double-sided adhesive to realize POCT or wearable configurations.
• Electrode fabrication (screen printing): Carbon and silver inks were printed on PET and cured at 150 °C for 15 min. Ag/AgCl reference was formed via NaCl treatment. Patterns were designed in AutoCAD.
• AuNPs ion–electron transducer: Electrodes were pretreated in 50 mM H₂SO₄ by CV (−0.4 to +1.0 V, 10 cycles) and AuNPs were electrodeposited from HAuCl₄/Na₂SO₄ using CV (−1.5 to +1.5 V, 50 mV/s) for 12 cycles. AuNPs improved hydrophobicity, double-layer capacitance, and electron transfer.
• Ion-selective membranes: Na⁺-selective membrane (ionophore X, Na-TPFB, PVC, DOS in THF) and K⁺-selective membrane (valinomycin, NaTPB, PVC, DOS in cyclohexane) were drop-cast on AuNP-modified working electrodes and dried overnight. Optimal membrane loadings were identified (Na⁺ ~0.1 μL/mm²; K⁺ ~0.12 μL/mm², per optimization). Reference electrode used PVB with NaCl, F127, and MWCNTs to stabilize potential.
• Paper-based microfluidic layers (wax printing): Hydrophilic/hydrophobic patterns were printed on Whatman paper with a Xerox ColorQube 8570 and baked at 100 °C for 2 min to create 2-layer 3D channels via wax penetration. Designs enabled sweat capture, routing, and storage by capillarity; layers were folded and vertically bonded with laser-cut double-sided adhesive.
• Device assembly: For POCT tests, Na⁺/K⁺ electrode arrays were paired with microfluidic layer 3DM-P. For on-body regional monitoring, 3DM-W was integrated with Na⁺/K⁺ ISEs. For dehydration monitoring during prolonged exercise, adjusted 3DM-W was integrated with Na⁺ ISE.
• Electrochemical characterization: CV and EIS were conducted in 5 mM [Fe(CN)₆]³⁻/⁴⁻. Open-circuit potential (OCP) was used for ion sensing. Contact angle measurements and SEM/TEM characterized AuNPs and membrane morphology. Mechanical bending tests (0–180°) assessed resistance stability; corrosion resistance and potential drift were evaluated in long-term tests; storage stability was assessed over 0–3 weeks at 4 °C.
• Calibration and selectivity: Hysteresis, forward/reverse calibration, reproducibility (multiple devices), and selectivity versus common ions were tested. Single-point calibration was applied before each measurement (Na⁺ in 10 mM NaCl; K⁺ in 2 mM KCl).
• Human subject testing: Three healthy volunteers (22–27 years, 25–30 °C ambient) performed indoor cycling at prescribed intensities. Sensors were affixed to cleaned forearm and back skin with waterproof medical tape, connected to a portable PalmSens workstation. Real-time on-body monitoring was conducted; sensors were calibrated pre- and post-test to generate concentration readouts from calibration curves. Ethical approval and informed consent were obtained.
• Ex-situ and gold-standard comparison: Sweat was collected during exercise for ex-situ testing via the POCT device (3DM-P + Na⁺/K⁺ ISEs) and ICP-OES after dilution in 0.1% HNO₃. On-body vs ex-situ agreement and ICP-OES validation were analyzed.
• Prolonged exercise with interventions: During ~150 min constant-rate cycling, dehydration onset was inferred from rising sweat Na⁺. Interventions compared electrolyte water vs pure water supplementation (with rest) and subsequent exercise resumption. Na⁺ dynamics and recovery were analyzed across subjects.

• AuNPs transducer performance: CV showed successful Au deposition; EIS indicated reduced charge-transfer resistance vs bare carbon; optimal electrodeposition at 12 cycles yielded highest electrochemical surface area (ECSA increased from 1.4×10⁻⁵ to 7.0×10⁻⁵ cm²) and hydrophobicity (contact angle ~141°). Excess cycles led to aggregation and decreased current.
• Sensor electrochemical metrics:
  • Na⁺ ISE: Hysteresis ~1 mV; reversible sensitivities ~43.1 and 44.3 mV/decade (forward/reverse). Single-point calibration used 10 mM NaCl; a reported sensitivity value of 542.9 mV/decade (as stated). High reproducibility across devices; good selectivity against common interferents.
  • K⁺ ISE: Hysteresis ~0.5 mV; reversible sensitivities ~102.1 and 107.2 mV/decade; single-point calibration sensitivity ~102.5 mV/decade (2 mM KCl set to 0 V). High reproducibility; good selectivity.
• Mechanical and stability: AuNP-modified electrode resistance remained ~155 Ω with RSD 0.52% under 0–180° bending. Potential drift over 3 h: Na⁺ ~0.15 mV/h; K⁺ ~0.05 mV/h. Sensors retained performance after storage up to 3 weeks at 4 °C (light-protected).
• On-body regional monitoring: Real-time on-body Na⁺/K⁺ readings overlapped with ex-situ measurements under constant exertion. Pearson correlations between on-body and ex-situ: Na⁺ r≈0.97; K⁺ r≈0.87. Back exhibited slightly higher Na⁺ and initial K⁺ than forearm, consistent with higher sweat gland density/sweat rate; both ions stabilized over time with similar trends across sites.
• Prolonged exercise and interventions: During ~150 min cycling, Na⁺ concentration rose around 90 min indicating dehydration (e.g., Subject 1: ~80→100 mM; Subject 2: ~67→80 mM). Electrolyte water plus brief rest reduced Na⁺ and returned to normal within ~40–50 min, whereas pure water plus rest decreased Na⁺ but did not return it to baseline, indicating electrolyte supplementation is more effective for rehydration under these conditions.
• Manufacturing and scalability: The “Screen+Wax” process enabled rapid, low-cost, high-throughput fabrication of flexible electrodes and paper microfluidics; modular “Lego Bricks” assembly supported customizable POCT and wearable configurations for large-scale sweat analysis.

The work demonstrates that a modular, fully printed wearable ISE platform can sensitively and stably monitor key sweat electrolytes while being manufacturable at scale and configurable to different applications. Hydrophobic AuNPs as solid-contact layers enhanced electron transfer, hydrophobicity, and capacitance, improving stability (low hysteresis and drift), reproducibility, and response characteristics of Na⁺ and K⁺ ISEs. Integration with wax-printed paper microfluidics provided passive sweat handling with minimal evaporation/contamination and comfortable skin conformity. The platform accurately tracked dynamic Na⁺ and K⁺ variations across body regions and during prolonged exercise; strong agreement between on-body and ex-situ measurements, and with ICP-OES validation, supports measurement fidelity. Application studies showed site-specific differences (back vs forearm) largely attributable to sweat rate and physiology, and demonstrated that electrolyte-containing fluids more effectively reverse dehydration-associated elevations in sweat Na⁺ than pure water. Collectively, the findings address the research goal of enabling cost-effective, large-scale, customizable, and reliable sweat electrolyte monitoring for exercise and health management.

This study introduces a fully printed, modular “Lego Bricks” wearable sweat ion sensor that combines screen-printed flexible ISEs (Na⁺, K⁺) with wax-printed 3D paper microfluidics for scalable, low-cost fabrication and customizable assembly. AuNPs solid-contact layers yield high hydrophobicity, enhanced electron transfer, and stable potentiometric performance. The sensors enable accurate POCT and real-time on-body monitoring, revealing consistent Na⁺ and K⁺ dynamics across body sites and demonstrating superior rehydration efficacy of electrolyte water over pure water during prolonged exercise. The platform is poised for large-scale population monitoring and integration into physical activity and health applications. Future directions include: integrating sweat rate sensing to quantify total electrolyte loss; expanding the analyte panel (e.g., glucose, lactate, uric acid, pH, dopamine, cortisol); generalizing solid-contact materials and membrane integration for higher-throughput fabrication; enhancing device architectures and microfluidics; and conducting larger cohort studies with rigorous statistical analyses.

• Sample size was small (three healthy volunteers) with limited demographics; larger, statistically powered studies are needed.
• Sweat rate was not measured; without sweat rate normalization, interpreting total electrolyte loss is limited.
• Sensor stretchability was not assessed (PET substrate seldom stretched); only bending deformation was evaluated.
• Reported Na⁺ sensitivity values varied across descriptions; single-point calibration was required before each test to address device-to-device offsets.
• Only Na⁺ and K⁺ were measured; broader biomarker panels and cross-analyte correlations were not explored here.
• Long-term wear, skin compatibility over extended durations, and environmental robustness (temperature/humidity variations) were not comprehensively evaluated.