中国是世界上受真菌毒素污染最严重的国家之一, 因污染造成的粮食及粮油产品经济损失巨大, 真菌毒素快速检测与防控迫在眉睫。 由于传统湿化学检测方法的时滞性、 复杂性、 高成本、 使用大量化学试剂等问题, 无法满足粮食生产、 流通及加工过程中快速实时检测的需要。 分子光谱是分子振动能级间或转动能级间跃迁产生的光谱, 反映了分子内部的结构信息, 可确定分子的转动惯量、 分子键长健强及离解能, 用于样本中化学组分及性质的检测。 粮食样本(真菌毒素)在激发光的激励作用下能级跃变产生的光通过光路系统被光电探测器接收, 光谱强度与被测物浓度在一定范围内符合Lambert-Beer定律, 可实现粮食真菌毒素的快速、 定量检测。 相比真菌毒素传统检测方法费时费力、 成本高、 大量使用化学试剂等问题, 光谱分析技术具有快速、 无损、 绿色等显著技术优势。 在分析粮食真菌毒素检测的重要性、 迫切性的基础上, 介绍了光谱分析的技术原理与理论基础, 近红外光谱是电偶极矩变化引起的振动光谱, 拉曼光谱是分子极化引起的振动光谱, 而荧光光谱反映具有长共轭结构的分子信息, 光谱成像在检测维度上由一维拓展到二维分布, 通过光谱解析和特征分析可进行真菌毒素快速准确检测。 进一步分析了近红外光谱、 拉曼光谱、 荧光光谱和光谱成像等技术在粮食真菌毒素检测的研究现状及发展动态, 指出了各技术的优势与存在的不足, 研究表明光谱分析技术受到越来越多的学者关注, 基于光谱分析技术的粮食真菌毒素的检测探索, 已成为食品安全检测领域的热点问题。 通过文献综述可以发现, 光谱分析技术为粮食中真菌毒素的快速筛查、 定性判别或高灵敏检测提供了新的途径, 但仍存在诸多尚需解决的问题, 在系统探讨光谱分析技术瓶颈问题的基础上, 展望了需进一步突破的研究方向, 特别是在检测的认可度、 检测精度和稳定性方面。 通过光谱理论解析明确粮食真菌毒素检测的可行性, 通过微观尺度提高粮食真菌毒素检测的稳定性, 通过多模态光谱信息融合提高粮食真菌毒素检测的精确性, 为粮食真菌毒素的光谱快速检测技术提供参考。

China is one of the countries with serious concerns about mycotoxin contamination of agricultural food and feed commodities in the world. Mycotoxin contamination leads to a substantial economic loss of grain and oil products coupled with public health hazards; hence, the rapid detection and control of mycotoxin are imminent. The traditional wet chemical detection methods cannot meet the needs of rapid and real-time detection in the process of grain production, supply, distribution, and processing. Even though classical techniques such as HPLC are accurate and sensitive, they have the disadvantage of being time-consuming, entail complex sample preparation, expensive and consume large volumes of chemical reagent. Molecular spectrum is the spectral response produced by the transition between the vibrational or rotational energy levels of molecules, which interprets the structural information of molecules. It can determine the rotary inertia, band length, bond strength, and dissociation energy of molecules, and can be used for the detection of chemical components and properties in samples. The light produced by the transition of the molecules of mycotoxin contaminated grain sample under the excited state is acquired by the photodetector through the optical path system. The spectral intensity and the concentration of the tested substance are underpinned by the Lambert-Beer law within a certain range, which can realize the rapid and quantitative detection of mycotoxin in grain. Compared with the traditional methods of fungal toxins detection, spectral analysis technologies have significant technical advantages of rapid, non-destructive and green. The importance and urgency of mycotoxin detection in grain were analyzed, and then the technical principle and theoretical basis of spectral analysis techniques employed for the detection were introduced. Near-infrared spectroscopy is the vibration caused by the change of electric dipole moment, Raman spectrum responses to the vibration caused by molecular polarization, while the fluorescence spectrum reflects the molecular information with long conjugated structure. Spectral imaging expands from one-dimensional to two-dimensional distribution in detection, and detects mycotoxin quickly and accurately by spectral and feature analysis. This work analyzed the research progress and development trend of different spectral analysis technology, and also exposed the advantages and disadvantages of each technique. The investigation revealed the increasing researchers focus on this research field, and the detection and exploration of grain mycotoxin based on spectral analysis technology, which has become a research hotspot of food safety. Through literature review, it can be found that spectral analysis technology provides a novel approach for rapid screening, qualitative identification, or high-sensitivity detection of mycotoxin in food, but there are still many problems that need to be solved. The applications along with major barriers and limitations of these spectral techniques are discussed, with emphasis on the development of recognition, accuracy and stability. Spectroscopic techniques have the potential to fulfill the need for mycotoxin detection. However, they still require enhancement of theory interpretation, detection scale and accuracy. We believe this review will be an effective guide for rapid detection of mycotoxin in the grain to provide a methodological reference.