Automatic Optimization of Gradient Conditions by AI Algorithmand Seamless Method Transfer

Significance of the Topic

Automating liquid chromatography (LC) method development addresses the substantial time and expertise required for repeated gradient optimization and analysis scheduling. By reducing human intervention, laboratories can significantly improve throughput, consistency, and accessibility of chromatographic methods across research, quality control, and industrial environments.

Objectives and Study Overview

This study demonstrates the automatic optimization of gradient conditions at multiple column oven temperatures using an AI algorithm in LabSolutions MD software. Seven small-molecule compounds served as a model mixture. The software consecutively optimized gradients at 30, 40, and 50 °C to meet a minimum resolution criterion of 1.5. A subsequent case examined seamless transfer of a UHPLC method to conventional HPLC while preserving separation patterns.

Methodology and Used Instrumentation

LabSolutions MD employs an AI-driven cycle of gradient condition search and correction analysis. Users input flow rate and column oven temperature; the software alternates improved gradient proposals and corrective analyses until resolution and last-peak elution time criteria are met. The model sample comprised hydrocortisone, furosemide, ketoprofen, naproxen, probenecid, diclofenac, and indomethacin. Gradient profiles spanned 30 % to 60 % organic modifier with a resolution threshold of 1.5.

Major Results and Discussion

At 30 °C, automatic optimization achieved baseline separation for all seven peaks (resolution ≥ 1.5). At 40 °C and 50 °C, critical peak pairs failed to meet the resolution requirement despite AI optimization, indicating 30 °C as the optimal oven temperature. For method transfer, LabSolutions MD recalculated gradient times, flow rates, and elution steps to adapt from a 100 mm×3.0 mm, 1.9 µm UHPLC column on a Nexera X3 system to a 150 mm×4.6 mm, 5 µm column on a Nexera-lite HPLC. Chromatograms before and after transfer displayed consistent separation patterns and maintained baseline resolution of challenging peak pairs.

Benefits and Practical Applications

• Reduces manual labor and expert intervention in LC method development.
• Ensures reproducible gradient optimization across temperature settings.
• Enables non-specialists to develop and transfer methods reliably.
• Simplifies parameter adjustments for UHPLC-to-HPLC method translation.

Future Trends and Potential Applications

• Integration of multi-objective AI to optimize selectivity, speed, and sensitivity simultaneously.
• Real-time feedback loops coupling online detectors and machine learning for adaptive gradient control.
• Cloud-based collaborative platforms for sharing and refining automated methods globally.
• Expansion to other separation modes, such as ion chromatography or chiral LC, using AI-driven workflows.

Conclusion

LabSolutions MD’s AI algorithm effectively automates gradient optimization across different column oven temperatures and supports reliable UHPLC-to-HPLC method transfer. This approach streamlines method development, reduces dependence on chromatographic expertise, and enhances reproducibility in diverse laboratory settings.

Used Instrumentation

• Nexera X3 UHPLC system
• Nexera-lite conventional HPLC system
• LabSolutions MD software with AI optimization module
• Shim-pack Scepter C18-120 columns (100 mm×3.0 mm, 1.9 µm and 150 mm×4.6 mm, 5 µm)
• SPD-M40 UV detector
• Standard detection cell