AI Peak Integration for MassHunter software automates manual peak integration during the data analysis process in GC/SQ

Importance of Topic

Peak integration in single quadrupole GC/MS is fundamental to quantitative analysis but remains manual, time-consuming, and susceptible to user variability.
Automating this step with artificial intelligence ensures consistent baselines, higher throughput, and improved data quality.

Objectives and Study Overview

The primary goal was to develop and evaluate an AI-driven peak integration solution integrated into Agilent MassHunter Quant.
A custom machine learning model was trained using a standardized workflow and real-world phthalate data to replace manual corrections.

Methodology and Instrumentation

Equipment and Software:

• Agilent single quadrupole GC/MS instruments
• MassHunter Workstation version 10.2 with Agile2 default integrator
• Cloud-based AI Peak Integration database add-on

Model Development:

• Data from 1,247 consumer product samples targeting 14 phthalates and one internal standard
• Continuous learning by logging manual integration events to refine AI predictions
• Use of SIM/Scan acquisition and randomized data to minimize bias

Main Results and Discussion

Throughput:

• Manual integration: 60–120 seconds per chromatogram under ideal conditions
• AI integration: average cloud processing time ≈30 seconds
• Batch comparison: 100 samples in <25 minutes vs 2 hours manually (4× speed improvement)

Accuracy and Precision:

• Improved Critical Success Index with additional training data
• Higher precision on uneven baselines and co-eluted high-molecular-weight phthalates (e.g., DINP, DIDP)
• Reduced integration bias compared to default parameter-less integrator

Benefits and Practical Applications

The AI Peak Integration tool delivers:

• Reproducible and consistent analytical results across operators
• Substantial reduction in chemist review time
• Ease-of-use facilitating adoption by new and experienced analysts
• Accelerated sample turnaround supporting higher throughput laboratories

Future Trends and Opportunities

Expanding AI integration to other compound classes:

• Regulated analytes in oil, gas, and agriculture (pesticides)
• Environmental semi-volatile and short-chain chlorinated paraffins

Additional AI applications may include automated calibration curve generation and predictive maintenance of chromatographic systems.

Conclusion

Integrating machine learning into GC/MS peak integration transforms a manual bottleneck into an automated, reliable process.
Laboratory efficiency, data consistency, and overall productivity are markedly enhanced by AI Peak Integration.

Reference

1 Phthalates: The Everywhere Chemical. NIEHS. 2023.
2 Hauser, R.; Calafat, A. M. Phthalates and Cumulative Risk Assessment: The Tasks Ahead. Environ. Health Perspect. 2016, 124(6), A104–A105.
3 Edwards, S. C.; Arbuckle, T. E.; Clayton, B. E.; et al. Phthalate and Novel Plasticizer Concentrations in U.S. Fast Food Packaging. J. Expo. Sci. Environ. Epidemiol. 2021;31(4):550–562.