Distinguishing Engineered TiO2 Nanomaterials from Natural Ti Nanomaterials in Soil Using spICP-TOFMS and Machine Learning
Bland et al.
Environmental Science & Technology, 2022
DOI: 10.1021/acs.est.1c02950
Engineered nanomaterials (ENMs) are increasingly released into the environment, but difficult to distinguish from naturally occurring nanomaterials (NNMs) that contain similar elements. In this study, researchers from Carnegie Mellon University set out to differentiate specific Ti-based ENMs from background NNMs in different soil samples.
For the analysis of nanoparticles, TOFWERK’s icpTOF R was used, which can measure most elements in the periodic table simultaneously on a single particle. This provides a rich dataset consisting of multielemental composition of thousands of individual nanoparticles in just a few minutes and allows the determination of the specific elemental fingerprint and mass distribution for each particle. Machine learning models were developed for this study to identify specific elements and mass differences to classify ENMs and NNMs. Such models can learn ratios of similar associated elements to differentiate ENMs and NNMs at a much deeper level than through observation alone. The results were successful, with the conclusion that the machine learning model best distinguished NNMs from ENMs when detectable differences in Ti-mass distribution existed between them.
