Wearable surface-enhanced Raman spectroscopy (SERS) sensors can detect analytes in sweat containing interfering substances, but they face challenges in integrating sampling, sensing, and detection into a single system for remote in situ analysis. To address this, this paper proposes a flexible "hydrogel tentacle" optical fiber (HTOF) SERS sensor, enabling remote in situ detection of analytes in sweat. The sensor uses a hydrogel with excellent water absorption as the flexible SERS substrate, which is in situ crosslinked with Ca2+ at the optical fiber tip to assemble the flexible hydrogel tentacle (HT). The light transmitted through the fiber is coupled into the HT, enabling direct sampling of analytes. The sensor's performance was evaluated using 4-mercaptopyridine (4-Mpy), achieving an enhancement factor (EF) of 9.44 × 1010 and a limit of detection (LOD) of 8.69 × 10-11 M. After 40 days of storage, the sensor maintained 95.24% of its SERS activity, and the relative standard deviation (RSD) between different batches was as low as 2.64%. With the excellent bending and stretching properties of the HT, the sensor can be applied in wearable human devices. When combined with a one-dimensional convolutional neural network (1D-CNN) machine learning model, it can achieve semiquantitative recognition of uric acid, creatinine, and urea with an accuracy of up to 91.05%. The proposed sensor shows promising potential for applications in kidney disease assessment and health monitoring.
Shi et al. (Fri,) studied this question.