Non-enzymatic near-infrared up-conversion fluorescent microfluidic sensor integrated with smartphone platform for rapid sweat glucose detection.

Intermittent and non-invasive glucose monitoring offers remarkable benefits for personalized diabetes management. We present a novel smartphone-integrated near-infrared (NIR) fluorescent microfluidic sensor for efficient sweat collection and fast, reliable detection of glucose. The microfluidic sensor contains four microchambers, each equipped with a bilayer membrane: the upper is sensing layer consisting of polyurethane matrix and non-enzymatic glucose probe PY-POFs, while the lower is light-converting layer consisting of PMMA matrix and up-conversion nanoparticles (UCNPs). Upon 808-nm NIR excitation, the light-converting layer excited the sensing layer via upconversion luminescence, and the glucose-sensitive fluorescence peaking at 470 nm is captured with a smartphone. The sensor exhibits a rapid response time of 15 s, excellent selectivity and stability due to the non-enzymatic nature, as well as a low detection limit of 20 µM because of low-background interference. Moreover, the practical applicability of our device is verified by tracking glucose fluctuations in sweat samples from human volunteers under pre- and post-prandial conditions, with results well correlated to those acquired by commercial blood glucose meters. The prototype demonstrated holds significant promise as a reliable, portable, and user-friendly sensing platform, suitable for rapid intermittent point-of-care non-invasive metabolic monitoring.