The last two weeks have been going under the title “thesis-thesis-thesis” in our group. This spring one BSc student and two MSc students from our group have defended their thesis: Sara Khabazbashi, Helen Sepman, and Thomas Ledbetter. Both Sara and Thomas have dug into the analytical standard free quantification for pollutants. Sara specifically focussed on metabolites of PCBs while Thomas’ aimed to quantify the suspected compounds contributing to the extractable organic fluorine. Helen, on the contrary, was studying the processes occurring in the charged droplets, such as chemical reactions.
Sara’s project was carried out in collaboration with the group of Jana Weiss who kindly donated standards of PCB metabolites and was largely inspired by the inputs from Hans-Joachim Lehmler group at the University of Iowa. PCBs are legacy pollutants still present in our environment. Hundreds of PCB metabolites exist, where both the extent of chlorination as well as the structural modifications introduced during metabolism (hydroxylation, methoxylation, sulphonation, glucuronidation) vary largely. Standards are available almost only for the PCBs themselves. However, for understanding the environmental fate of the PCBs the metabolites are of crucial importance. Sara looked into the quantification of hydroxylated PCBs in particular. Namely, how the current ionization efficiency prediction model is performing on the PCB metabolites and how to improve it further to make accurate predictions for PCB metabolites. Interestingly, hydroxylated PCBs also possess high variability in ionization efficiency for structural isomers, such as ortho- and para-substituted compounds. In the end, she was able to quantify multiple hydroxylated PCBs in human blood plasma with an error below a factor of 3 (does not include the error from matrix effect). The results are highly promising for the quantification of hydroxylated PCBs. In the future, we hope to expand towards other functional groups present in PCB metabolites, currently, the availability of training data for sulphonation and glucuronidation is limited by the availability of analytical standards. Additionally, we noticed that due to the very wide range of isomeric compounds the separation and identification of PCB metabolites is very difficult. Here, we hope to employ our LC/cyclic-IMS/HRMS for improving the separation of different isomeres.
The project of Thomas’ was carried out in collaboration with the group of Jonathan Benskin, whose research focus is on closing the Florine mass balance. Part of the research question of Thomas’ MSc thesis was on quantifying the fluorinated compounds that do not have analytical standards available. For this Thomas incorporated 30+ fluorinated compounds into the ionization efficiency scale in electrospray negative mode and retrained a gradient boosted trees-based model for predicting the ionization efficiency values. The validation with spiked samples promised a mean prediction error below a factor of 2, which is a very amazing result for such complex samples. The predictions were thereafter used to quantify the suspects detected in marine mammal samples. Though some additional work on suspect identification and quantification is still needed, the preliminary results showed promise in this approach.
From September two MSc students, Emma Palm and Louise Malm, will start again in our group on developing better identification and quantification methods for non-targeted screening. Additionally, one student, Lisa Jonsson, will start implementing the non-targeted screening strategies we have developed in the industry.