Exploring the Limits of Mycotoxin Analysis

In the paper “Exploring the Limits of UHPLC–MS/MS with Polarity Switching Towards the Quantification of 931 mycotoxins and Other Secondary Metabolites in Cereal-based Food”, what was the scientific and regulatory rationale for developing an 11-min ultrahigh-performance liquid chromatography tandem mass spectrometry (UHPLC–MS/MS) method capable of quantifying 931 mycotoxins and secondary metabolites in cereal-based foods?¹

From a scientific perspective, it was particularly interesting to evaluate how data quality was affected when the mass spectrometer available in our laboratory was pushed to its practical limits. The work began with monitoring data obtained from the routinely used method in our laboratory, which required two 20.5-min high performance liquid chromatography (HPLC) runs for each sample, performed in positive and negative ionization modes.² The method was then progressively accelerated in a stepwise manner through the transition to UHPLC and the reduction of the run time to 11 min. Ultimately, a UHPLC–MS/MS method was developed and validated that combined an exceptionally broad analyte scope—931 mycotoxins and secondary metabolites—with a very short analysis time of only 11 min in a single injection.¹

The underlying idea was straightforward: as the run time decreases, data quality is expected to deteriorate, and our aim was to determine when this deterioration becomes significant enough to compromise reliable quantification.

At the same time, there was a strong regulatory driver behind the study. In food safety, especially within the EU framework, laboratories are increasingly required to monitor a growing number of regulated and emerging mycotoxins within a single sample, while still meeting strict performance criteria defined by guidelines such as SANTE or Eurachem.^3,4 However, although multi-residue “mega-methods” are well established for pesticides, multi-class LC–MS/MS methods covering different contaminant groups are still less widely accepted in routine regulatory practice.

The rationale of our work was therefore to bridge this gap by developing a highly efficient UHPLC–MS/MS method that can simultaneously quantify a very large number of contaminants while remaining compliant with regulatory validation criteria and suitable for routine use. In particular, it was aimed to demonstrate that even with a highly condensed 11‑min run and a large scope, it is still possible to maintain the level of data quality required for quantitative analysis, rather than limiting such approaches to screening purposes.

In this sense, the study shows that highly efficient, large-scope, multi-class UHPLC–MS/MS methods are practically applicable tools for high-throughput routine monitoring and even emergency response scenarios, while still fulfilling regulatory expectations.

What analytical advantages did fast polarity switching provide compared with conventional dual-injection workflows in LC–MS/MS mycotoxin analysis?

One of the central aspects of this work was the use of fast polarity switching to increase analytical speed. Although this feature has been available for some time, our review of the literature indicated that its full potential has not yet been widely exploited in large-scope targeted food analysis, where most methods still rely on separate injections for positive and negative ionization modes.^5,6 By integrating both polarities into a single run, we were able to effectively halve the measurement time, thereby significantly increasing throughput while reducing solvent consumption and instrument usage. However, this approach comes at the cost of increasing the number of concurrent transitions for each cycle and introducing additional switching time, which makes data acquisition more challenging and requires careful optimization—and that was exactly what we wanted to investigate.

The researchers compared UHPLC versus HPLC and short versus long chromatographic gradients. From a food analysis perspective, what are the main trade-offs between analytical speed, chromatographic resolution, and quantitative performance?

From a food analysis perspective, the main trade-off of using UHPLC and shortened gradients is that chromatographic resolution, analytical speed, and throughput increase substantially, but the extremely high number of concurrent multiple reaction onitoring (MRM) transitions places much greater demands on MS/MS data acquisition. The combination of narrow peak widths and very high MRM concurrency reduced the available dwell time and, consequently, the number of data points acquired per peak.¹⁰ This became particularly critical at low concentration levels, where reduced dwell time limited ion collection and fewer data points for each peak led to poorer peak definition and more distorted peak shapes, ultimately compromising precision and quantitative performance.

Why did the researchers validate only 52 representative analytes out of the 931 monitored compounds, and how does this strategy support practical method validation in large multi-analyte chromatography workflows?

As the goal of this study was not to revalidate the entire method, which had already been comprehensively validated in previous work,^7,8.9 but to evaluate whether specific changes in data acquisition—such as polarity switching, shortened run times, and the transition to UHPLC—compromise data quality, the decision to validate only 52 analytes out of the 931 monitored compounds was deliberate and scientifically justified. Furthermore, the selected 52 analytes represent a broad range of physicochemical properties, retention times, ionization behaviors, and regulatory relevance, making them a representative and sufficiently challenging subset. Given the enormous workload that full validation of all analytes would require, this strategy provides a realistic and efficient way to assess method performance while allowing meaningful conclusions to be extrapolated to the entire dataset.

The study reported increased limits of quantification (LOQs) at lower spiking levels when analysis speed was increased. What chromatographic or mass spectrometric factors could explain this reduction in repeatability and sensitivity?

Before answering this question, I would like to clarify that LOQs in this study were determined according to Eurachem guidelines⁴ and were therefore based on precision data at low concentration levels.

The increase in LOQs observed at higher analysis speed can be explained mainly by a combination of chromatographic and mass spectrometric factors. From the chromatographic side, use of UHPLC and shortened gradients resulted in narrower chromatographic peaks, which reduced the available detection time across each peak. From the mass spectrometric side, the very large number of analytes monitored within a short run resulted in a high number of concurrent MRM transitions, which reduced the dwell time available per transition. Therefore, in the cycles with highest concurrency when the instrument reached the minimum dwell time threshold, the manually set cycle time was automatically extended, which further reduced the number of data points acquired across each chromatographic peak. Fewer data points per peak made peak integration less robust, while very short dwell time reduced the number of ions acquired per transition. Furthermore, at low concentration levels, this makes ion counting statistics increasingly important, leading to lower signal-to-noise (S/N) ratios, and less reproducible peak areas. As a result, both quantitative precision and thus sensitivity deteriorate, which is reflected in increased LOQs.^11,12 Importantly, despite this reduction in sensitivity, the method¹ still achieved LOQs suitable for regulated mycotoxins, demonstrating that it remains fit for purpose for routine applications.

How does matrix complexity in cereal-based foods such as muesli influence matrix effects during UHPLC–MS/MS analysis, and what strategies can analysts use to reduce ion suppression or enhancement?

Muesli contains a wide range of co-extracted compounds, including sugars, and lipids, which can interfere with ionization and lead to signal suppression or enhancement. Several strategies are available to correct for signal suppression or enhancement (SSE), including post-column infusion, isotope-labelled internal standards, post extraction spiking, and matrix-matched calibration.¹³ Ideal correction strategies such as isotope-labelled internal standards or matrix-matched calibration are not readily applicable when working with such a large number of analytes, as labelled analogues are unavailable or prohibitively expensive for the full analyte panel, and truly blank muesli matrices are difficult, if not impossible, to obtain. Therefore, in this study, matrix effects were evaluated by post-extraction spiking combined with the systematic assessment of multiple muesli samples. This approach enabled a realistic assessment of matrix behavior while maintaining feasibility for large-scope methods.

Importantly, matrix effects were assessed using five different muesli samples, allowing the evaluation of relative matrix effects, and thereby accounting for variability between different product types. In addition, lot-to-lot variation was incorporated into the estimation of expanded measurement uncertainty, reflecting one of the major contributors to variability in LC–MS/MS analysis.¹⁴ This strategy provides a more realistic estimation of method performance under real-world conditions, where matrix variability is unavoidable.

The authors estimated that replacing a dual-injection workflow with a single-injection method could reduce laboratory CO₂ emissions by approximately 22 tons annually. How can chromatographic method optimization contribute to sustainability and greener analytical chemistry in food testing laboratories?

Greener analytical chemistry is often associated with the development of entirely new solvents, materials, or instrument platforms; however, such solutions might not always be feasible for routine laboratories due to cost and infrastructure. In contrast, our study demonstrates that substantial environmental benefits can also be achieved through the optimization of established LC–MS/MS methods. By reducing the analysis time from two injections (41 min total) to a single 11-min run, we were able to substantially decrease solvent consumption, energy use, and total instrument operating time. In addition, simply switching from HPLC to UHPLC further reduced solvent consumption. Based on our estimates, this translates to a substantial reduction of CO₂ emissions in analytical laboratory per year, clearly demonstrating how optimizing LC–MS/MS methods can contribute to ongoing green transition without requiring major infrastructure changes. While the present study includes an estimate of the potential reduction in CO₂ emissions, a more detailed assessment focused specifically on overall CO₂ emissions has been carried out and will be reported separately.

Are you involved in any other projects using chromatography in food analysis?

I am currently involved in several chromatography-based projects in food analysis, particularly in collaboration with the Technical University of Denmark. There, I work closely with both the National Reference Laboratory for veterinary drugs and the European Reference Laboratory for processing contaminants. In these projects, I applied gained knowledge to ensure compliance with regulatory requirements to develop robust and practical LC–MS/MS methods that support food safety monitoring across the EU.

References

1. Kenjeric, L.; Sulyok, M.; Bueschl, C.; Malachova, A.; Krska, R. Exploring the Limits of UHPLC–MS/MS with Polarity Switching towards the Quantification of 931 Mycotoxins and Other Secondary Metabolites in Cereal-Based Food. J Food Compos Anal 2026, 151, 108908. DOI: 10.1016/j.jfca.2026.108908
2. Sulyok, M.; Suman, M.; Krska, R. Quantification of 700 Mycotoxins and Other Secondary Metabolites of Fungi and Plants in Grain Products. npj Sci Food 2024, 8, 49. DOI: 10.1038/s41538-024-00294-7
3. European Commission. Analytical Quality Control and Method Validation Procedures for Pesticide Residues Analysis in Food and Feed; SANTE 11312/2021 Rev. 2; European Commission: Brussels, Belgium, 2024.
4. Cantwell, H., Ed. Eurachem Guide: The Fitness for Purpose of Analytical Methods—A Laboratory Guide to Method Validation and Related Topics; 3rd ed.; Eurachem, 2025.
5. Bessaire, T.; Savoy, M.-C.; Ernest, M.; et al. Enhanced Surveillance of >1100 Pesticides and Natural Toxins in Food: Harnessing the Capabilities of LC–HRMS for Reliable Identification and Quantification. Foods 2024, 13, 3040. DOI: 10.3390/foods13193040
6. Zhang, D.; Sun, B.; Luan, Y.; et al. Simultaneous Broad-Spectrum Detection of Over 50 Mycotoxins in Maize Silage via Liquid Chromatography–Tandem Mass Spectrometry: Method Development and Application. Food Control 2025, 172, 111143. DOI: 10.1016/j.foodcont.2025.111143
7. Schamann, A.; Caprio, A.; Krpan, M.; et al. Mycotoxin Migration in Jam Samples: A Study on Natural Contamination and Experimental Inoculation. World Mycotoxin J 2025, Advance Online Publication. DOI: 10.1163/18750796-bja10028
8. Kenjeric, L.; Sulyok, M.; Malachová, A.; et al. Extension and Interlaboratory Comparison of an LC–MS/MS Multiclass Method for the Determination of 15 Different Classes of Veterinary Drug Residues in Milk and Poultry Feed. Food Chem 2024, 449, 138834. DOI: 10.1016/j.foodchem.2024.138834
9. Steiner, D.; Sulyok, M.; Malachova, A.; Mueller, A.; Krska, R. Realizing the Simultaneous Liquid Chromatography–Tandem Mass Spectrometry-Based Quantification of >1200 Biotoxins, Pesticides, and Veterinary Drugs in Complex Feed. J Chromatogr A 2020, 1629, 461502. DOI: 10.1016/j.chroma.2020.461502
10. Stahl-Zeng, J.; Lange, V.; Ossola, R.; et al. High Sensitivity Detection of Plasma Proteins by Multiple Reaction Monitoring of N-Glycosites. Mol Cell Proteomics 2007, 6 (10), 1809–1817. DOI: 10.1074/mcp.M700132-MCP200
11. Lee, H. Y.; Kim, E. G.; Jung, H. R.; et al. Refinements of LC-MS/MS Spectral Counting Statistics Improve Quantification of Low-Abundance Proteins. Sci Rep 2019, 9, 13653. DOI: 10.1038/s41598-019-49665-1
12. Walorczyk, S. Development of a Multi-Residue Method for the Determination of Pesticides in Cereals and Dry Animal Feed Using Gas Chromatography–Tandem Quadrupole Mass Spectrometry. II. Improvement and Extension to New Analytes. J Chromatogr A 2008, 1208 (1–2), 202–214. DOI: 10.1016/j.chroma.2008.08.068
13. Williams, M. L.; Olomukoro, A. A.; Emmons, R. V.; Godage, N. H.; Gionfriddo, E. Matrix Effects Demystified: Strategies for Resolving Challenges in Analytical Separations of Complex Samples. J Sep Sci 2023, 46, 2300571. DOI: 10.1002/jssc.202300571
14. Stadler, D.; Sulyok, M.; Schuhmacher, R.; Berthiller, F.; Krska, R. The Contribution of Lot-to-Lot Variation to the Measurement Uncertainty of an LC–MS-Based Multi-Mycotoxin Assay. Anal Bioanal Chem 2018, 410 (18), 4409–4418. DOI: 10.1007/s00216-018-1096-5