AI-based peak determination for pesticide analysis

Importance of the Topic

Water contamination by pesticides poses a serious risk to human health and ecosystems. Reliable detection and quantification of multiple pesticide residues in water samples are essential for monitoring environmental quality and ensuring regulatory compliance. Advanced chromatographic techniques coupled with improved data processing strategies can enhance analytical objectivity and throughput.

Objectives and Study Overview

This study evaluates an AI-driven peak integration algorithm, Peakintelligence for GCMS, in the simultaneous multi-component analysis of 140 pesticides in water. It compares the performance of the AI approach against a conventional integration method in terms of accuracy, repeatability and processing efficiency.

Methods and Instrumentation

A series of mixed pesticide standards were prepared at concentrations ranging from 0.003 to 0.5 mg/L, using internal standards (anthracene-d10, 9-bromoanthracene, chrysene-d12) to correct for variability. Chromatographic separation and detection were performed under splitless injection with helium carrier gas, operating in selected ion monitoring mode.

Used Instrumentation

• Gas Chromatograph–Mass Spectrometer: GCMS-QP2050
• Autosampler: AOC-30i
• Column: SH-I-5Sil MS (30 m × 0.25 mm ID × 0.25 µm)
• Software: LabSolutions Insight with Peakintelligence for GCMS

Key Results and Discussion

• Sensitivity and Linearity: Achieved reliable detection at 0.005 mg/L with calibration curves showing excellent linearity.
• Repeatability: Area ratio repeatability for 139 compounds averaged 2.21 %RSD (n=5), all under 5 %RSD.
• Integration Accuracy: Peakintelligence delivered consistent integration for low-level peaks and closely eluting compounds, whereas conventional algorithms showed misintegration in challenging regions.
• Processing Efficiency: AI-based integration reduced manual correction time and eliminated operator-dependent variability.

Benefits and Practical Applications

The AI-driven integration approach enhances analytical objectivity and reproducibility across laboratories. Time savings in data processing accelerate routine water quality monitoring, while robust peak detection supports regulatory compliance and risk assessment strategies.

Future Trends and Potential Applications

Further expansion of AI-based algorithms may include adaptation to liquid chromatography-mass spectrometry workflows and real-time on-site monitoring. Integration with high-throughput hydrogen carrier systems and advanced vacuum technologies will bolster sensitivity and sample throughput in environmental analytics.

Conclusion

Peakintelligence for GCMS, combined with the GCMS-QP2050 platform, offers a powerful solution for accurate, efficient and operator-independent quantitation of pesticides in water. This approach supports enhanced environmental surveillance and contributes to safer water resource management.

References

1. Mayser J.P., Kraft V., Weber W. AI-based peak determination for pesticide analysis. Shimadzu Europa GmbH; 2024.