Label-free non-invasive detection of salivary diabetes biomarkers via chiral Au nanocubes capillary surface-enhanced Raman scattering

Chiral D Au NCs capillary SERS for label-free non-invasive saliva D-Glc assay. • Sensor achieves 0.30 nM LOD, covering clinical saliva glucose ranges. • Assay finishes in 10 min with 50 μL sample, high accuracy and anti-interference. • Assay shows correlation with GOD-POD gold std. (R 2  = 0.9313) and 0.947 AUC. • It offers a promising tool for early diabetes screening and clinical translation. Early, non-invasive screening and risk prediction of diabetes are crucial for effective prevention and control, as well as for delaying the onset and progression of complications. In this study, we developed a capillary surface-enhanced Raman scattering (SERS) sensor based on chiral gold nanocubes (D Au NCs) for the non-invasive, label-free, and rapid detection of d -glucose (D-Glc) in saliva. By leveraging the chiral recognition capability and strong SERS activity of D Au NCs, the sensor specifically identifies D-Glc and significantly amplifies its characteristic Raman signals, enabling quantitative analysis in complex saliva matrices without labeling. The sensor demonstrates outstanding analytical performance, covering the clinically relevant range of salivary glucose concentrations with a detection limit as low as 0.30 nM. Based on a capillary detection platform, the assay is simple and rapid, completed within 10 min using only a 50 μL micro-sample, while demonstrating strong anti-interference capability and high detection accuracy. Clinical saliva testing showed strong correlation with the glucose oxidase-peroxidase (GOD-POD) gold standard (R 2  = 0.9313), enabling reliable discrimination between healthy individuals and diabetic patients. Receiver operating characteristic (ROC) analysis yielded an area under the curve (AUC) of 0.947, confirming its high efficacy in diabetes risk screening and strong diagnostic performance. Consequently, the D Au NCs-based capillary SERS sensor provides a promising strategy for early, non-invasive diabetes screening and prevention, possessing broad potential for clinical translation and application.