UPC2 Strategy for Scaling from Analytical to Preparative SFC Separations

Significance of the Topic

Supercritical fluid chromatography (SFC) offers high resolution and fast separations with carbon dioxide as the main mobile phase. Scaling methods reliably from analytical to preparative instruments is critical for efficient purification workflows. However, CO2 compressibility leads to changing mobile phase densities when flow rate, column dimensions, or particle size vary, complicating direct method transfer.

Objectives and Overview of the Study

This work establishes a systematic strategy based on density modulation to align average mobile phase densities across different SFC configurations. By matching density profiles rather than relying on empirical adjustments, methods developed on sub-2 μm analytical columns can be predictably transferred to preparative instruments using 5 μm stationary phases under isocratic and gradient conditions.

Methodology

A standard mixture of six compounds (caffeine, carbamazepine, uracil, hydrocortisone, prednisolone, sulfanilamide) was analyzed at 0.2 mg/mL on analytical systems and 3.75 mg/mL on preparative. Key parameters (flow rate, temperature, back-pressure) were varied to generate pressure and density profiles. Density simulations employed NIST REFPROP modeling of CO2/methanol mixtures and assumed linear pressure drops along the column. Adjusted ABPR settings compensated for density differences to equalize average density between configurations.

Instrumentation Used

• ACQUITY UPC2 system with PDA detection
• Waters Prep 100q SFC system with PDA detection
• ACQUITY UPC2 BEH 2-Ethylpyridine columns (1.7 μm, 2.1×150 mm; 3.0×50 mm)
• Viridis BEH 2-Ethylpyridine and Viridis OBD Prep columns (5 μm, 2.1×150 mm; 19×150 mm)
• LCMS Certified Max Recovery vials

Results and Discussion

Changing from 1.7 μm to 5 μm particles increased pressure drop and mobile phase density, altering retention, selectivity, and resolution. Density profile simulations revealed mismatched average densities of 0.88 g/mL (1.7 μm) vs. 0.80 g/mL (5 μm) under identical conditions. By raising ABPR for the 5 μm column, the average density was adjusted to match 0.89 g/mL, restoring comparable chromatograms.

Applying this approach to varying flow rates (0.48, 1.4, 4.0 mL/min) demonstrated that flow-induced density changes can also be compensated to maintain selectivity despite efficiency losses at higher velocities. Analytical methods were successfully scaled to preparative flow rates (~83 mL/min) on 19×150 mm columns with minimal alteration of retention order.

For gradient separations (2–10% methanol), dynamic density changes at the column inlet and outlet were simulated. Adjusting pressure at each flow condition achieved matched average densities, yielding analogous resolution and peak shape between analytical and preparative runs.

Benefits and Practical Applications

• Rapid, predictable transfer from analytical screening to preparative purification
• Reduced solvent consumption and waste disposal costs
• Applicability to both achiral and chiral separations, enabling high-throughput column and modifier screening

Future Trends and Applications

Integration of real-time density monitoring and automated pressure control may further streamline SFC scale-up. Advanced predictive models could extend this approach to novel stationary phases and solvent modifiers. Coupling density-matched SFC with mass spectrometry or fraction collection robots promises fully automated workflows for pharmaceutical, natural product, and fine chemical industries.

Conclusion

Density modulation offers a systematic, physics-based framework for scaling SFC methods across instruments and column formats. By aligning average mobile phase densities, retention and selectivity are preserved under isocratic and gradient conditions, supporting efficient method development and preparative separations.

References

1. A. Tarafder et al., J. Chromatogr. A 1238:132–145, 2012.
2. A. Tarafder et al., J. Chromatogr. A 1258:136–151, 2012.
3. E.W. Lemmon et al., NIST REFPROP, Version 9.1.
4. R. Span & W. Wagner, J. Phys. Chem. Ref. Data 25(6):1509–1596, 1996.
5. O. Kunz & W. Wagner, J. Chem. Eng. Data 57(11):3032–3091, 2012.