An automated sample pretreatment system for liquid–liquid extraction and its application in the analysis of four steroid hormones in human plasma

Liquid–liquid extraction (LLE) is commonly employed in sample preparation, but faces limitations in automation due to solvent leakage during pipetting, interference from conductivity-based liquid level detection, stringent mixing requirements, and dependence on centrifugation, all of which constrain throughput. A fully automated, centrifugation-free LLE system was developed, incorporating a pipetting unit, high-speed vortex mixing, nitrogen evaporation, a positive-pressure module, and a custom-designed 24-well phase separation filter plate. The system was optimized for the analysis of four steroid hormones—aldosterone, cortisone, cortisol, and 11-deoxycorticosterone. Validation followed CLSI C62-A guidelines and included a comparison with manual centrifugation- and column-based LLE techniques. The automated method demonstrated correlation coefficients exceeding 0.99 for all analytes, with sensitivity comparable to manual protocols (LOD: 10–20 pg/mL). Spike recoveries ranged from 85.9 % to 111.8 %, and intra-/inter-day precision was within 10 %. Matrix effects remained under 20 %. In real plasma samples, results obtained using the automated method were consistent with those from manual centrifugation within ±15 %. The automated system reduced batch preparation time by over 50 % compared to manual processes while maintaining accuracy and reproducibility. This study introduces an innovative, efficient, centrifugation-free automated LLE platform for steroid hormone analysis in plasma. The system overcomes the limitations of manual extraction, supports high-throughput workflows, and demonstrates substantial potential for clinical and pharmaceutical applications. • A fully automated, centrifugation-free LLE platform enables quantification of four steroid hormones in plasma. • A custom-designed 24-well phase separation plate purifies emulsions and isolates aqueous and organic phases. • The results are comparable to manual LLE and shows high-throughput potential for clinical and pharmaceutical applications.