Antifouling superhydrophilic hydrogel-based electrochemical sensor for robust multi-analyte sweat analysis

Reliable sweat analysis in complex biological environments remains challenging due to biofouling-induced signal degradation and analyte interference. Herein, a superhydrophilic antifouling hydrogel-based electrochemical sensor is developed by integrating TiO₂-decorated MXene (MXene@TiO₂) into a borax-crosslinked PVA network for intelligent multi-analyte sweat analysis. The MXene@TiO₂ not only provides conductivity and electrocatalytic activity, but also enhances hydrophilicity of the sensing interface, thereby suppressing nonspecific protein adsorption. The sensor enables sensitive and selective detection of dopamine, uric acid, and ascorbic acid, with detection limits as low as 0.0491, 0.1187, and 0.2451 μM, respectively. Under protein-interference conditions containing 10 mg mL^-1 bovine serum albumin (BSA), the current attenuation for DA, UA, and AA was limited to 24.88%, 13.37%, and 12.65%, respectively, demonstrating resistance to biofouling-induced signal loss. Machine-learning-based classification and regression were further employed for analyte identification and concentration prediction, achieving a classification accuracy of 97.9% and an R² of 0.978 for concentration prediction. Furthermore, a wearable sensing platform incorporating wireless communication and mobile-based data processing enables wearable signal acquisition and intelligent analysis, while droplet-based validation yielded recoveries from 99.45% to 100.80%. By integrating superhydrophilic antifouling interface engineering with machine-learning-assisted signal interpretation, this work provides a robust strategy for intelligent multi-analyte sweat sensing and wearable health monitoring.