Pitfalls and Inherent Biases in Liquid Handling Robotics: Investigations in Automation for SI Traceable Measurements.

Industries such as pharmaceuticals, petrochemicals, and clinical research are adopting automated sample preparation processes to streamline repetitive tasks and improve efficiency. As robotics underpins more complex methods of analysis, the evaluation and optimization of workflows and apparatus are vital. In this study, the accuracy and precision of an automated workflow were investigated using two liquid handling platforms with syringes and integrated balances, with the focus on their suitability for accurate multianalyte quantification. The methodology was developed by optimizing key parameters such as the syringe draw and dispense speeds. Additionally, the physical setup was evaluated as the syringe tip, vial sizes, and decapper tool all played a vital role in the performance of the system. The performance of robotic sample preparation was compared to traditional gravimetric manual preparation of amino acid blends, using double exact matching isotope dilution mass spectrometry (DEM-IDMS). After optimal conditions for the liquid-handling robotics were established, the accuracy of the DEM-IDMS method with automated sample preparation was assessed using NIST SRM 2389a amino acids in 0.1 mol/L HCl. The optimization process successfully reduced a systematic bias identified in the robotic workflow from 10% to <3.5%. Moreover, the measurement uncertainty achieved with the optimized automated method (≤5.3%, expanded at the 95% confidence interval for all amino acids) was comparable to that of manual preparation, while saving a full day of analyst time (1 h setup versus 7 h manually weighing). The research outputs highlighted are invaluable to improve the understanding of the advantages and remaining challenges in the use of automated sample preparation for high-accuracy analysis.