餐厨循环油的波谱性质及其FTIR识别技术的基础研究

Refined edible oils and gutter oils were analyzed by FTIR and the spectrums were recorded and analyzed by discriminant analysis(DA) and PCA methods in this study.The removal rules of components in gutter oil with various refine methods were also explored. Based on previous results,the whole absorption band of gutter oil was screened with aim to find specific band. Then,identification of the specific componment and its peak assignment will be performed in combination of Mass Spectrum analysis,Nuclear Magnetic Resonance,Infrared spectral analysis. According to selected representive references, the quantitive analysis model of gutter oil was built by using FTIR coupled with chemometrics method. In addition, we aim to build an analytical method to acchieve simultaneous determination of multi parameters in gutter oil. The predicted results were compared with those obtained from the present standard method to validate the model developed in term of accuracy and precision.

本研究拟通过PCA法及DA方法分析对比标准食用油脂及地沟油的红外图谱，结合不同工艺条件精炼对地沟油成分脱除规律的研究，从内源污染物中搜寻地沟油潜在的特征吸收峰，然后进行色谱分离，根据核磁、质谱、红外波谱解析结果确定地沟油特征吸收峰对应的特征组份。再选择代表性参比物质结合化学计量学知识建立地沟油的识别模型，建立油脂多组分的FTIR同时测定方法，与现行标准方法对照，对开发的模型有效性进行验证，最终实现地沟油的定性、定量、快速、准确识别。

目前我国餐厨循环油掺假问题严重，而大多数方法只适用于特定类型地沟油的检测判定，通用性不强，灵敏度不高，而且容易出现误判、漏判，识别准确性不高。本项目基于FTIR技术耦合化学计量学方法开发了一整套绿色、高效、快速、准确的食用油脂综合质量快速检验技术。. 项目收集了11大类、130种样品的精炼油（大豆、棕榈、菜籽、玉米、山茶、芝麻、花生、向日葵、橄榄、牛油、猪油）FTIR/ATR光谱及透射光谱，初步建立了精炼食用油红外光谱库。采用PLSR-LA、PCA等技术开发了食用油种类定性识别模型并成功用于未知样品。11类60个精炼食用油双盲验证识别率100%；115个二元精炼油互掺样品识别率高达96.4%，最低掺假浓度检出限0~8%，成功解决了精炼油以次充好、低值冒充高值油的掺假识别难题；精炼食用油、餐厨循环油二元互掺样品验证表明识别率97.8%。此外，以我国大宗、易掺假的山茶、芝麻、花生油为代表，建立了低值菜籽油、猪油、大豆油、精炼餐厨循环油掺假的定量检测模型，检测限低至0.5%~3%。. 对比分析精炼油IR光谱及餐厨循环油IR光谱，确定了TPC、TFA、AV、PV、PL、水分作为特征指标，用于判别食用油综合质量的指标，并对其红外特征吸收峰归属进行鉴定。研究首先开发了基于简单溶剂萃取耦合FTIR的水分、PL、AV、PV的快速分析模型，方法简便、快速、准确、重现性好。其次，采用FTIR/ATR结合波谱区间优化（iPLS、fiPLS、biPLS），分别构建了食用油中TPC、TFA、AV、PV、IV、PL、水分等指标的快速分析模型，各子模型预测准确度平均偏差0.5~3%。同时，项目集成各子模型构建了多指标同时分析鉴别食用油掺假的模型，实现了单次分析即可同时测定多指标，预测RSD在2.7%~3.2%之间，显示出所开发的红外模型稳定性好，具有较高的预测准确度和精密度。