分子光谱成像对玉米粉中微塑料的深度无损表征

目的 采用不同的光谱成像方法对玉米粉中的微塑料进行原位无损表征，并通过化学计量学优化其成像效果。方法 本研究尝试利用光热共振红外光谱成像和共聚焦拉曼光谱成像两种分子光谱技术，结合化学计量学方法对玉米粉中微塑料的存在进行深入表征。结果 采用光热共振红外成像 方法能够快速获取样品中污染物存在信息，但数据完整性和可视化方面受到一定限制。通过对拉曼光谱图像数据降解，应用独立成分分析可以得到玉米粉中微塑料清晰的分布轮廓。此外，光热红外成像能够在低浓度水平上得到污染物的分布图像，但检出率不高（10%），具有快捷检测特点。拉曼光谱成像受到扫描速度的影响，在低浓度污染样品中检出较少，但在中浓度污染样品得到了30%的检出率，以及中得到了较好的可视化效果。结论 通过分子光谱成像技术串联数据挖掘方法可以实现对粉状食品中微塑料的可视化检测，为食品中微塑料的无损便捷检测提供方法参考。

In-depth nondestructive characterization of microplastics in corn flour by molecular spectral imaging

Objective To perform the in situ nondestructive characteristics of the micro -plastic in corn flour using different spectral imaging methods, and optimize its imaging effect through chemical measurement. Methods Two molecular spectral techniques, photothermal resonance infrared spectral imaging and confocal Raman spectral imaging, combined with chemometric methods were used to characterize the presence of microplastics in corn flour. Results The photothermal resonance infrared imaging method could obtain the presence of pollutants rapidly, but its data integrity and visualization were limited. By degrading the Raman spectral image and conducting independent component analysis, a clear distribution of microplastics in corn flour could be obtained. Moreover, photothermal resonance infrared imaging could detect pollutants at low concentration levels with a low detection rate of 10%, but with a fast detection speed. Affected by the scanning speed, Raman spectral imaging exhibited poor detection in low-concentration samples, but provided a better visualization in mid-concentration samples with the detection rate of 30%. Conclusion The visual detection of microplastics in powdered food can be realized by molecular spectral imaging coupled with data mining method, which can also provide a method reference for non-destructive detection of microplastics in food.