Ultra-Sensitive and Unlabeled SERS Nanosheets for Specific Identification of Glucose in Body Fluids

The detection of biomolecules utilizing surface-enhanced Raman spectroscopy (SERS) is highly trending in research, but the unlabeled and sensitive detection of saccharides in body fluids is challenging. Here, a novel SERS substrate (AgNS600) is fabricated by physical scratching of smooth silver sheets, which successfully achieves enhanced SERS fingerprinting by capturing glucose molecules through nanoscale grooves on the surface of the substrate to form multi-dimensional "hot spot" regions. The platform proves to be rapid, reliable, and reproducible, with a lower detection limit of 0.5 amol L-1 in deionized water. Further detection of multiple monosaccharides, polysaccharides, and mix samples, which are distinguished with machine learning and heat map, demonstrates the versatility and specificity of the platform. Importantly, real samples are explored and the results show that the platform is able to detect and recognize SERS signals of glucose in the blood, urine, tears, and perspiration, and non-destructive predict diabetes by analyzing the signal intensity of different characteristic peaks. In this work, it is the first non-chemical, unlabeled determination of glucose with high sensitivity via SERS, which provides a potential strategy for advancing the clinical development of SERS technology and the early diagnosis of glucose metabolic diseases. Ultra-sensitive, unlabeled silver sheets are designed for the specific identification of glucose in body fluids via surface-enhanced Raman spectroscopy (SERS). The detection limit in deionized water is as low as 0.5 amol L-1. Multiple mechanical learning algorithms are applied to analyze the SERS spectra and successfully predict the variation of glucose levels in the urine of clinical patients.image