Doctoral thesis (Dissertations and theses)
Enhanced screening methods for pesticides in food based on travelling-wave ion-mobility-high-resolution mass spectrometry
Goscinny, Séverine
2017
 

Files


Full Text
THESIS_Severine GOSCINNY_ULG.pdf
Publisher postprint (8.02 MB)
Request a copy

All documents in ORBi are protected by a user license.

Send to



Details



Keywords :
Ion mobility mass spectrometry; Pesticides; drift time; collision cross section; calibration; travveling wave ion mobility
Abstract :
[en] The work presented in this thesis finds its origin in the need to develop ad hoc and novel analytical strategies to screen efficiently targeted pesticide residues in the sector of food control. Our work intended not only to monitor an increased number of pesticides but also to elaborate a new frame based on Ion Mobility-Mass Spectrometry (IMS-MS) for a confident and reliable use of screening pesticide data. The main part of this thesis was devoted to an investigation of using ion mobility, hyphenated to ultra-high performance liquid chromatography and Mass Spectrometry (UHPLC-IMS-MS), as a new dimension of separation for small polar organic molecules in complex biological samples without involving exhaustive sample preparation. The first objective of this study was dedicated to the development and optimization of UHPLC-IMS-MS using traveling wave ion mobility spectrometry (TWIMS). We have identified nitrogen as suitable gas for the mobility separation of multi-class pesticides and optimized the different T-wave parameters for a wide range of molecular weight pesticides. We observed for 100 pesticides tested that drift times gave a higher level of selectivity to the overall method as no interfering compound resulted at same retention time, drift time and measured mass from matrix blank samples. In addition, once the pesticide has been identified using its retention time, exact mass and drift time, the resulting cleaned mass spectrum facilitates the chemical structure elucidation. Key findings from this research show that drift times measurements are reproducible and independent from matrices tested and compound concentration leading to the proposition of using this parameter as a new identification point (IP). We have demonstrated that drift times can be used as an additional IP to increase confidence in the results without extending the analysis time or changing the sample preparation procedure. Based on reproducible data, we were the first group to report and recommend a 2% variation of drift times as quality criteria for identification purposes. In a few recent published papers, we noticed that research groups developing similar analytical strategies have extensively adopted the 2% proposed criterion. Obviously drift times are dependent of instrumental conditions. Hence, to apply the proposed approach with automated and wide scale screening methods further developments in processing software have been made which also allowed to convert drift times in calculated Collisional Cross Section (CCS) values by means of TWIMS calibration. The second aim of this thesis was dedicated to a comprehensive study of the reliability of CCS values derived from drift times for pesticides screening over a long period of time (3 years) with different ion mobility parameters. The results proved that CCS values are very consistent, as the measured values do not differ more than 1% from the in-house reference data library. Different calibration mixtures were also tested and results show that the type of the compound used for calibration does not influence CCS measurement. However, the results emphasized the importance of the first low m/z mobility calibration point (i.e. below 200 m/z) to guarantee full independence from instrument parameters and calibrant. The robustness of the CCS calibration was then tested by changing deliberately the IMS parameters and compare the measured CCSs with the reference CCSs. Three set of experiments were done by changing the IM velocity (static mode = fixed value and dynamic mode = velocity ramping) and also IM gas pressure. The first set of parameters gave CCS values within 2% error from the reference CCS throughout the pesticides mass range, while for the last 2 sets of parameters, deviations beyond 2% error away from the reference CCSs are clearly marked under the m/z 232 (first CCS calibration point). From these results, we clearly see that CCS values generated with the same calibration protocol are reproducible even for target masses under the first CCS calibration point, whereas robustness is guaranteed within the strict limits of the calibration range. To extend the calibration range to a larger window of molecular weights (in the lower and upper range) a combination of acetaminophene-polyalanine and a mixture of small molecules containing also polyalanine were tested. No significant deviations were observed between the two new calibration mixtures as all the pesticide results were within the 2% error. This observation demonstrates that the nature of the reference compounds used for calibration does not influence the outcome of the estimated CCS values for low molecular weight pesticides. Finally, an intercomparison study was designed to evaluate the relative capabilities of producing CCS values with a simple protocol for screening purposes. Four laboratories in three different countries (UK, USA and Belgium) were involved in the test. For this comparison, only monoisotopic ions were selected and 6 out of 40 pesticides selected had their masses under the first CCS calibration point (m/z 232). When comparing the percentage errors of each laboratory values to the calculated consensus value (mean of the CCS from the 4 instruments), two laboratories presented a few values outside the 2% error limits. Interestingly, when each laboratory used their in house reference CCS to calculate the percentage errors for a spiked mandarin extract, the results are well within the 2% CCS error window. These observations highlighted that the direct use of CCS values reported from the literature has to be treated with great care when applying the proposed approach. After this extensive study on CCS values, the third goal was to assess the possible added value of using CCS criterion when performing a UHPLC-IM-(HR)MS screening method of targeted pesticides. The approach has been tested in the framework of a proficiency test (EURL-PT-FV-16). The PT material was a sweet pepper homogenate containing 13 pesticides measured by LC in positive mode. Two additional pesticides can be added to this list, primicarb-desmethyl and tebuconazole, because they were in the sample but were not evaluated in the PT results by the organizers since their concentrations were below the minimum required reported limit of 0.01 mg kg-1. In total, we were expecting 15 detections by using our screening method. The approach proposed here is unconventional for screening and consists on one hand (scenario A), on large mass accuracy (20 ppm), large retention time window (0.5 min) and large CCS error (10%) in order to capture, in a first step, a maximum of detected compounds. 26 tentative candidates were then identified. In this scenario, when closing down first the mass accuracy window from 20 ppm to 5 ppm, 8 candidates were discarded. However, a wrong decision has been taken for two of them, (2 false negatives), requiring to trigger additional investigations. In the second step of scenario A, by selecting the 2% error for CCS, five additional residues did not meet the criterion and were ruling out, giving in total 13 detected compounds, actually present in the PT sample and 2 questionable data. In scenario B, when we directly apply a narrower CCS error percentage (from 10 to 2%) without filtering on mass accuracy we gain directly all the 15 compounds present in the PT sample without any false negatives and discard the 11 false positives in one step. These findings are indeed promising and require obviously confirmation at large scale within the pesticide laboratory end-users network equipped with IMS-MS and willing to apply the proposed screening strategy. Interlaboratory studies and PTs within the pesticide community are required to consolidate our findings during this thesis work. Regarding the treatment of data and software, considerable improvements have been made in terms of automatic extraction of mobility data and having a fully automated approach incorporating CCS values as a screening parameter has never been tested before this thesis. Nonetheless, there is room for improvements in terms of downsizing the acquisition file size which should enable faster processing and reporting time along with easier data storage. While the results from this thesis demonstrated that nitrogen was the most suited gas, Howdle et al. (1) have reported that the use of binary gas mixtures results in excellent selectivity enhancements over single gas composition for IMS separations and most importantly they demonstrated that this selectivity can be tuned by altering the binary gas composition. One can only barely imagine the impact of such results for IMS applications. Being able to choose the best gas mixture and/or develop gradient with variable gas composition during the mobility separation would most probably achieve better mobility resolution customized for specific applications or to create new opportunities for multi-class screening (e.g. pesticides, mycotoxins, food contact materials, food additives…). This new approach will trigger new research in the field of instrumental development to ensure control over the composition of gas delivered in the IMS cell and also how to adapt or create TWIMS calibration for these significant changes. The application space for ion mobility is vast, not only as a means of the specific pioneering application we developed here but in the richness of these measurements can provide in structural information of ions in many research fields including fundamental research. Conceptualization, design and construction of contemporary instruments were growing fast over the past few years. Emerging IM strategies developed by manufacturers based on proof-of-concept published papers encompass Overtone Mobility Spectrometry (OMS); confinement and mobility selective release like Trapped Ion Mobility Spectrometry (TIMS), or Cyclic Drift Tube Ion Mobility Spectrometry (Cyclic DTIMS) and more recently Cyclic Traveling Wave Ion Mobility Spectrometry (Cyclic TWIMS). For example, in Cyclic TWIMS, one remarkable feature is the use of one meter single path transit ring which was placed orthogonally to the primary beam travelling wave path of the instrument. This novel multi-pass cyclic device gives access to higher resolving power. As the Resolution (CCS/CCS) in T-wave increases in proportion to the square root of the device length, the one meter single pass in the ring being 4 times the length of a standard Synapt G2-S, the Resolution is doubled by cycle. Giles et al. reported up to 6 passes showing an acceptable ion transmission with a signal drop of only 15% (2). These emerging high-resolution IMS instruments clearly open new possibilities of separation of structural isomers low molecular weight compounds. Based on this research project, we are convinced that ion mobility used as an additional dimension of separation in hyphenated techniques can play a role in addressing the needs to provide reliable screening data and helping to reduce the rate of false negative and positive results prior to confirmation by quantitative methods. IMS-MS has become a key technology to answer chemical food safety issues as demonstrated by the rapidly expanding literature using IMS-MS over recent years. In a broader view, the concept of CCS data used as an additional identification tool and the development of new CCS databases will contribute to strengthen the needs in gathering reliable information to help the analyst in making right decision in the field of the very challenging qualitative screening analysis of trace chemical residues in food.
Disciplines :
Chemistry
Author, co-author :
Goscinny, Séverine ;  Université de Liège - ULiège > Form. doct. sc. (chimie - Bologne)
Language :
English
Title :
Enhanced screening methods for pesticides in food based on travelling-wave ion-mobility-high-resolution mass spectrometry
Defense date :
03 July 2017
Number of pages :
168
Institution :
ULiège - Université de Liège
Degree :
Doctor of Philosophy in Sciences
Promotor :
Eppe, Gauthier  ;  Université de Liège - ULiège > Molecular Systems (MolSys)
President :
Leyh, Bernard  ;  Université de Liège - ULiège > Département de chimie (sciences)
Secretary :
Focant, Jean-François  ;  Université de Liège - ULiège > Molecular Systems (MolSys)
Jury member :
De Pauw, Edwin  ;  Université de Liège - ULiège > Département de chimie (sciences)
Honing, Maarten
Le Bizec, Bruno
Mol, Hans
Available on ORBi :
since 01 December 2017

Statistics


Number of views
173 (13 by ULiège)
Number of downloads
3 (3 by ULiège)

Bibliography


Similar publications



Contact ORBi