Biofluids are ideal sample sources for wearable in situ surface-enhanced Raman scattering (IS-SERS) sensors due to their noninvasive collection. However, the limited spatial reach of conventional hot spots (HSs), coupled with the fluidic and biocomplex nature of biofluids, means that only a small portion of target analytes can be effectively captured inside the HSs. To overcome this, we propose a metal@MOF particle-in-cavity (MMPIC) detection model. This architecture enhances the cascade electric field, expanding and concentrating HSs within and around the MOF dielectric. The integration of conical nanocavities with nanoporous MOFs enables effective analyte confinement and enrichment within the MOF matrix as well, ensuring colocalization with HSs in the same microregion. Additionally, the molecular sieving and graded refractive index properties of the MMPIC structure provide strong resistance to interference from both biofluids and their components. Together, these features improve both the sensitivity and robustness of the model. As a proof of concept, a microfluidic patch and a smart mask were developed based on the MMPIC model, enabling precise quantification of biomarkers-such as pH, glucose, ammonia, and 4-ethylbenzaldehyde-down to 1 ppb in real human sweat and exhaled breath. This work introduces a universal wearable IS-SERS detection model and validates its applicability across diverse real-world scenarios, offering valuable guidance for future wearable in situ sensing technologies.
Li et al. (Tue,) studied this question.