离子迁移光谱法
重复性
化学
质谱法
色谱法
环境分析
分辨率(逻辑)
分析化学(期刊)
环境化学
计算机科学
人工智能
作者
Lauren Mullin,Karl J. Jobst,Robert A. Dilorenzo,Robert S. Plumb,Eric J. Reiner,Leo W.Y. Yeung,Ingrid Ericson Jogsten
标识
DOI:10.1016/j.aca.2020.05.052
摘要
Dust analysis provides a means to assess the degree of exposure of humans in an indoor environment to various contaminant classes such as flame retardants, pesticides and others. There is increasing interest in non-targeted acquisitions using high resolution mass spectrometry (HRMS) to better capture the contaminant profile. However, these studies are confronted with the challenge of assessing confidence in proposed identifications, particularly when authentic standards are not available. Here, we demonstrate the analysis of dust extracts representing various indoor environments (industrial e-waste processing and domestic) for high-abundance environmental contaminants using a data-independent LC-HRMS approach, incorporating ion mobility spectrometry (IMS) to provide additional characterization capability for the complex samples. Twenty-nine xenobiotic compound identifications were made based on both targeted and non-targeted processing approaches using accurate mass precursor and product ion measurement combined with an ion mobility derived collision-cross section (TWCCSN2) determination. Characterization of the repeatability of TWCCSN2 value measurements and their average relative error to compared authentic standards of 0.38% were consistent with various published studies and represent a robust measurement property. TWCCSN2 values were particularly useful in cases where confirmation after the initial dust analysis was performed using a different chromatographic method, due to the gas-phase measurement being unaffected by such changes. Observed compound TWCCSN2 values were then compared to predicted CCSN2 values obtained using two different machine-learning based predictive techniques. Results from one of the predictive programs indicates a promising avenue for use of these models for supporting compound identification in non-targeted analyses.
科研通智能强力驱动
Strongly Powered by AbleSci AI