Automatic Optimization of Gradient Conditions by AI Algorithm for Impurity Analysis

Importance of the Topic

Automating the optimization of gradient conditions in liquid chromatography accelerates method development, reduces manual workload, and lowers dependency on specialist expertise. This approach addresses labor-intensive cycles of mobile phase preparation, scheduling, analysis, and data evaluation, enabling more efficient workflows across pharmaceutical and analytical laboratories.

Objectives and Study Overview

This study applies the AI-driven optimization module of LabSolutions MD to develop an LC method for separating montelukast and its related impurity (Imp1). The target criteria are a resolution greater than 3.0 between peaks and a maximum elution time below 15 minutes.

Used Instrumentation

• LC System: Shimadzu Nexera X3
• Column: Shim-pack Scepter Phenyl-120, 100 mm × 3.0 mm I.D., 1.9 µm
• Mobile Phase A: 0.15% formic acid in water
• Mobile Phase B: 0.1% formic acid in acetonitrile
• Column Temperature: 30 °C; Flow Rate: 0.5 mL/min; Injection Volume: 10 µL
• Detection: UV at 238 nm (SPD-M40, UHPLC cell)

Main Results and Discussion

The AI algorithm conducted initial analyses using five gradient profiles (5% B to 95% B over variable time windows). The initial resolution for montelukast and Imp1 was insufficient (Rs ~0.91). Two iterative correction analyses were performed, during which the algorithm adjusted gradient slopes and introduced an isocratic hold. In the second correction, the resolution exceeded 3.0 and the last peak eluted within 15 minutes, satisfying both criteria. The optimized gradient shortened run time and enhanced peak separation by dynamically refining conditions based on resolution and elution time feedback.

Benefits and Practical Applications

• Significant reduction in labor and time for method development
• Accessible to non-chromatography experts, lowering the learning curve
• Standardized approach that improves reproducibility across labs
• Integration with screening and robustness evaluation supports a full AQbD workflow

Future Trends and Potential Applications

Extending AI-driven optimization to diverse compound classes and complex matrices can further streamline LC method development. Integration with cloud-based data repositories and predictive modeling may enable knowledge-sharing networks. Future enhancements might include automated robustness testing and real-time adaptive control of chromatographic systems.

Conclusion

The AI-based gradient optimization feature of LabSolutions MD effectively automated the development of a rapid, high-resolution method for montelukast and its impurity. By achieving target resolution and run time in minimal iterations, this approach demonstrates the potential to transform LC method development, delivering faster results with reduced expert intervention.

References

• Shimadzu Technical Report C190-E309, Automatic Optimization of Gradient Conditions with LabSolutions MD
• Shimadzu Technical Report C190-E284, Efficient Method Development Based on Analytical Quality by Design with LabSolutions MD Software