Analysis of Radix Bupleuri (Chaihu) using Agilent 1200 Series LC Systems with Evaporative Light Scattering Detector

Importance of the topic

This study addresses the quality control of Radix Bupleuri (Chaihu), a widely used Traditional Chinese Medicine rich in saikosaponins. Reliable chromatographic fingerprints and quantification of bioactive saponins are critical for ensuring therapeutic consistency, detecting adulteration, and authenticating geographical origin. The development of fast, robust methods aids laboratories in meeting industry and regulatory requirements.

Objectives and study overview

The primary goal was to develop and compare conventional HPLC and Rapid Resolution LC (RRLC) methods for fingerprinting and quantifying six key saikosaponins in various Bupleurum species and preparations. Optimization of evaporative light scattering detector (ELSD) parameters and parallel UV diode array detection (DAD) was performed to maximize sensitivity and baseline stability. Samples from different origins were analyzed to reveal species-specific profiles.

Methodology and instrumentation

Sample preparation involved ultrasonic extraction of 0.5 g powdered root with methanol (2 % pyridine), centrifugation, concentration and dilution to 10 mL, followed by 0.22 µm filtration. The Agilent 1200 Series RRLC system was configured with:

• Binary Pump SL with vacuum degasser
• High-performance Autosampler SL
• Thermostatted Column Compartment SL
• Diode Array Detector SL (210 nm and 248 nm, micro flow cell)
• Evaporative Light Scattering Detector (40 °C, 55 psi, gain 7, 3 s filter)
• ZORBAX SB-C18 columns: 4.6 × 250 mm, 5 µm (conventional HPLC) and 3 × 50 mm, 1.8 µm (RRLC)

Conventional HPLC used a 60 min gradient (30 → 100 % acetonitrile), 1 mL/min, 15 µL injection. The RRLC method employed a 6–15 min gradient, 0.8 mL/min, 2 µL injection to accelerate throughput.

Main results and discussion

• ELSD provided flat baselines and detected non-UV-active saikosaponins without interference from coextracted impurities, while DAD captured chromophoric compounds at low wavelengths.
• RRLC reduced analysis time from 60 min to under 15 min, maintaining resolution and peak order comparable to conventional HPLC.
• Fingerprint profiles varied among Bupleurum species: typical North and South Chinese samples differed in late-eluting peaks; B. bicaule exhibited additional clusters not seen in pharmacopoeial species, underscoring the need for dual detection.
• Quantitative calibration using log-transformed ELSD responses achieved excellent linearity (r2>0.995) for six saikosaponins, with LODs of 8–25 µg/mL.

Benefits and practical applications

• Combined RRLC-ELSD/DAD offers a rapid, cost-effective approach for routine quality control and origin authentication of Radix Bupleuri.
• Dual detection ensures comprehensive monitoring of both chromophoric and non-chromophoric constituents.
• Shortened analysis times increase laboratory throughput and reduce solvent consumption and operational costs.

Future trends and applications

• Expansion of chromatographic fingerprint libraries for machine learning–based species classification and geographic traceability.
• Integration with LC-MS to structurally identify unknown peak clusters and minor constituents.
• Further miniaturization and automation of sample preparation and data processing for high-throughput screening in herbal quality control.

Conclusion

The optimized Agilent 1200 Series RRLC method with ELSD and DAD provides robust fingerprints and reliable quantification of saikosaponins in various Bupleurum samples. This approach significantly improves analysis speed, baseline stability, and detection coverage, thereby enhancing the quality assurance of Traditional Chinese Medicine products.

Reference

1. China Pharmacopeia Commission, China Pharmacopeia, 2005.
2. Xiao R. et al., Chinese Journal of Pharmaceutical Analysis, 26(2), 162–167, 2006.
3. Zhang H. et al., Chinese Journal of Pharmaceutical Analysis, 27(8), 1150–1153, 2007.
4. Liau B.-C. et al., Journal of Pharmaceutical and Biomedical Analysis, 43, 1174–1178, 2007.