Determination of low concentration glucose solution using Raman spectroscopy based on the internal standard method

• The field of glucose detection often overlooks the structural changes that glucose undergoes in solution, typically assuming that its structure remains consistent with that of solid glucose. However, the interaction between glucose molecules and water molecules in solution primarily occurs in the form of pyranose-structured cyclic hemiacetals, with the proportion of free linear glucose being only 0.002 %, a figure that can be considered negligible. By integrating theoretical simulations with experimental validation, this method establishes a robust theoretical foundation for the accurate measurement of low-concentration glucose solutions. It also clarifies the intrinsic relationship between the precise structures of solid and liquid glucose molecules. Furthermore, the observed characteristic peak at 1125 cm −1 aligns perfectly with literature reports, fundamentally confirming its validity as a Raman characteristic peak of glucose. Simultaneously, this approach mitigates the potential risk of accidental correlations in small sample data, thereby demonstrating the reliability of the 1125 cm −1 characteristic peak in predicting glucose concentrations. • In glucose detection, glucose in solution exhibits significantly different Raman spectroscopic characteristics compared to its solid state due to dynamic changes in molecular conformation. Laser Raman spectroscopy has emerged as a powerful tool for analyzing such complex systems, thanks to its unique advantages. This technology achieves high specificity in detection by monitoring the characteristic narrow peaks or structural vibration modes of glucose molecules. Notably, it is unaffected by fluctuations in pH and temperature, and it does not require sample pretreatment or destruction, thereby fulfilling the criteria for non-destructive testing. • The use of the internal standard method with water as the internal standard, without introducing additional substances for non-destructive measurement, can mitigate the influence of sample quantity on quantitative results. This approach eliminates accidental errors, simplifies the methodology, enhances linearity, improves the accuracy of analytical results, and achieves greater quantitative precision. Utilizing two quantitative analysis methods—peak relative intensity and relative area—the linear fitting coefficients were 0.992 and 0.993, respectively, demonstrating consistent resistance to interference and high linearity. The spectral area method optimizes the limit of detection (LOD) value by integrating multiple peak signals and suppressing noise, all while maintaining the minimum detectable concentration (MDC). • The measurement accuracy of this glucose solution is high, with a LOD value of 40.3 mg/dL representing the quantitative result of this study, and a MDC value of 12.5 mg/dL serving as the qualitative result. Publicly available research indicates that human blood glucose levels typically range from 70 to 200 mg/dL. Studies have demonstrated that the combination of the internal standard method and confocal Raman spectroscopy can enhance the detection limit of glucose solutions and facilitate accurate quantitative analysis of low-concentration glucose solutions. Recent studies conducted by teams from MIT and Samsung, employing Piraman spectroscopy technology, have for the first time directly captured glucose characteristic signals through in vivo detection in human skin, confirming their linear correlation with blood glucose concentration. This breakthrough has effectively resolved the long-standing debate regarding the capability of Raman systems to achieve in vivo glucose detection and has removed significant barriers to the clinical application of this technology.Raman spectroscopy offers distinct advantages in detecting complex biological systems due to its narrow bandwidth, low polymer specificity with minimal overlap, low environmental sensitivity (such as interference from temperature and humidity), and technical compatibility. The miniaturization of optical components and the optimization of machine learning algorithms are expected to overcome the limitations of volume and cost in Raman detection systems. With non-invasive detection and precise molecular recognition capabilities, these systems hold significant promise for applications in real-time blood glucose monitoring. Glucose is an essential substance for life and exists in a cyclic equilibrium in solution. Raman spectroscopy is utilized to differentiate substances based on their glucose concentration. This study employed solid-state density functional theory to simulate the Raman spectra of D(+)-glucose, D-glucose monohydrate, and two types of glucopyranose. Experimental spectra were obtained for D-glucose monohydrate and solutions of varying concentrations. The calculation results of various glucose configuration theories indicate that 1125 cm −1 is the characteristic peak of glucose. This finding aligns with the experimental results and literature presented in this article, addressing concerns regarding the validity of glucose characteristic peaks. The peak at 1640 cm −1 , attributed to water, serves as the normalized internal standard. The correlation coefficients obtained from linear regression analysis were all greater than 0.990, with limits of detection (LOD) of 48.1 mg/dL and 40.3 mg/dL, respectively. The minimum detectable concentration (MDC) for qualitative analysis is 12.5 mg/dL. The relative area method has been demonstrated to be a direct and effective approach for accurately quantifying low-concentration glucose solutions.