人体呼气在线检测质谱仪应用于食用糖果风味鉴别研究

Study on the application of human breath on-line detection mass spectrometer to the flavor identification of edible candy

目的利用自主研发的单光子电离飞行时间质谱仪(SPI-MS2000)检测呼气中的痕量特征物质,建立一种在线呼气鉴别风味的分析方法。方法采用SPI-MS在线检测人体食用榴莲糖、薄荷糖、柠檬糖后呼气的成分变化,再利用皮尔逊相关系数(Pearson correlation coefficient,PCCs)、偏最小二乘法(partialleast square-discriminant analysis,PLS-DA)及层聚类分析(hierarchical cluster analysis,HCA)模型对数据进行分析处理,同步验证结果,并通过回归预测值与观测值、置换检验模型对分类方法进行评估。结果3类糖果的皮尔逊相关系数差异明显,与PLS-DA及HCA模型分析结果相互印证,方法模型分析未知样品的拟合优度达到99.7%。结论通过在线质谱识别人体食用糖果后的呼气,可以实现食品风味的在线快速鉴别。为SPI-MS 2000应用于食物在口腔中的残留及风味释放等领域提供重要的参考价值。

Objective To establish an online breath analysis method for flavor identification by using a self-developed single photon ionization time-of-flight mass spectrometer(SPI-MS 2000). Methods The changes in the composition of breath after human consumption of durian candy, mint candy, and lemon candy were detected on-line by SPI-MS. The data was analyzed and processed using the Pearson correlation coefficient(PCCs) model, partial least square(PLS-DA) model and hierarchical cluster analysis(HCA) to verify the results simultaneously, the classification method was evaluated by the regression predicted value and observed value and the replacement test model. Results The Pearson correlation coefficients of the three types of candies were significantly different, which was mutually confirmed with the analysis results of the PLS-DA and HCA models. The method model analysis of unknown samples had a goodness of fit of 99.7%. Conclusion The exhalation of human body after eating candy can be identified by online mass spectrometry. the food flavor can be quickly identified online. It provides an important reference value for SPI-MS 2000 in the application of food residue and flavor release in the oral cavity.