Evaluating Anthropogenic Origin of Unidentified Volatile Chemicals in the River Rhine

Abstract

Surface water of rivers like the Rhine is a highly relevant environmental and an important source of the Dutch drinking water. To improve protection of the environment and drinking water supply, it is important to have a continuous overview of the chemical composition of the river. Such an overview may be obtained with contemporary, untargeted analytical platforms like gas chromatography-mass spectrometry. Interpretation of such untargeted data is however challenged by the presence of many chemicals of natural origin. We developed a novel approach to screen for anthropogenic chemicals using non-parametric tests on the time trends of yet unidentified chemicals. The approach uses PARAFAC2 to extract unknown components present in GC–MS data and provides an assessment of whether such components may be anthropogenic. This significantly reduces screening efforts required by human laboratory staff. In total, out of twelve suspect unknown components, eleven were classified as anthropogenic, providing compelling evidence that studying unknown components can be highly valuable for regulatory bodies. This approach filters out many naturally occurring compounds, leaving more resources available for wet-lab identification of suspected anthropogenic chemicals.