Benchmarking of the quantification approaches for the non-targeted screening of micropollutants and their transformation products in groundwater

Abstract

A wide range of micropollutants can be monitored with non-targeted screening; however, the quantification of the newly discovered compounds is challenging. Transformation products (TPs) are especially problematic because analytical standards are rarely available. Here, we compared three quantification approaches for non-target compounds that do not require the availability of analytical standards. The comparison is based on a unique set of concentration data for 341 compounds, mainly pesticides, pharmaceuticals, and their TPs in 31 groundwater samples from Switzerland. The best accuracy was observed with the predicted ionization efficiency-based quantification, the mean error of concentration prediction for the groundwater samples was a factor of 1.8, and all of the 74 micropollutants detected in the groundwater were quantified with an error less than a factor of 10. The quantification of TPs with the parent compounds had significantly lower accuracy (mean error of a factor of 3.8) and could only be applied to a fraction of the detected compounds, while the mean performance (mean error of a factor of 3.2) of the closest eluting standard approach was similar to the parent compound approach.

Use of Passive and Grab Sampling and High-Resolution Mass Spectrometry for Non-Targeted Analysis of Emerging Contaminants and Their Semi-Quantification in Water

“…Several different approaches have been tested to overcome the lack of analytical standards and evaluate the most accurate ones for the semi-quantification [ 34 ]. Two strategies, based on the assumptions that TPs have the same response as the parent compounds and that the internal standard eluting closest to the compound of interest will have a similar response factor, appeared to be less reliable when comparing to the approach based on the prediction of the ionization efficiency of the compounds in the ESI source.…”

Machine Learning for Absolute Quantification of Unidentified Compounds in Non-Targeted LC/HRMS

“…However, it has been found that ionization efficiencies are correlated between different instruments [ 40 ]. This has previously enabled training a single ionization efficiency model and transferring the predictions to different instruments via correlation [ 15 ]. Therefore, the transferability of the model can be further investigated in the future, alongside validation engaging several instruments and laboratories.…”

“…To overcome these limitations, machine learning approaches for predicting the ionization efficiency of the detected compounds have recently been developed [ 10 , 11 , 12 , 13 ]. The predicted ionization efficiencies can be further used to quantify the tentatively identified compounds if analytical standards are lacking [ 11 , 14 , 15 ]. These approaches rely on descriptors deduced from the structure of the compound and, therefore, at minimum a tentatively known structure of the detected compound is required.…”

Spotlight on mass spectrometric non‐target screening analysis: Advanced data processing methods recently communicated for extracting, prioritizing and quantifying features

“…It should be noted that deciding on the size and composition of a sufficient ISTD set for intensity normalization is difficult, as the compounds in the environmental sample are initially unknown to the investigator and a large structural variety can be expected. Moreover, the factors influencing the RT and the ionization efficiency are to some extent different 70,71 …”