Shen Jiang, Qiuyun Li, Guangrun Wu, Xuming Mu, Xiaotong Wang, Yunpeng Wang, Yanli Wu, Jing Wu, Yang Li
In the landscape of biomolecular detection, surface-enhanced Raman spectroscopy (SERS) confronts notable obstacles, particularly in the label-free detection of biomolecules, with glucose and other sugars presenting a quintessential challenge. This study heralds the development of a pioneering SERS substrate, ingeniously engineered through the self-assembly of nanoparticles of diverse sizes (Ag@AgNPs). This configuration strategically induces 'hot spots' within the interstices of nanoparticles, markedly amplifying the detection signal. Rigorous experimental investigations affirm the platform's rapidity, precision, and reproducibility, and the detection limit of this detection method is calculated to be 6.62 pM. Crucially, this methodology facilitates nondestructive glucose detection in simulated samples, including phosphate-buffered saline and urine. Integrating machine learning algorithms with simulated serum samples, the approach adeptly discriminates between hypoglycemic, normoglycemic, and hyperglycemic states. Moreover, the platform's versatility extends to the detection and differentiation of monosaccharides, disaccharides, and methylated glycosides, underscoring its universality and specificity. Comparative Raman spectroscopic analysis of various carbohydrate structures elucidates the unique SERS characteristics pertinent to these molecules. This research signifies a major advance in nonchemical, label-free glucose determination with enhanced sensitivity via SERS, laying a new foundation for its application in precision medicine and advancing structural analysis in the sugar domain.
