A Linear Model for Measuring the Experimental Resolution, Minimum Detection Limit, and Accuracy of Common Analytical Techniques

Abstract

Objective: This study aims to establish methods based on linear models to measure the experimental resolution, minimum detection limit, and accuracy of flame atomic absorption spectrometry (FAAS), gas chromatography (GC), and liquid chromatography–mass spectrometry (LC–MS) and explore their applications.Methods: A series of equally diluted samples (copper, benzene, and renin) are measured. A linear regression analysis is performed between the detected concentrations and the expected concentrations. Under the optimal detection state, the minimum dilution gradient corresponding to the linear regression equation is defined as the experimental resolution. Meanwhile, the resolution of the measurement result using zero as the baseline is considered as the minimum detection limit of the detection system. The accuracy is calculated by using the deviations of the coefficient of determination, slope, and intercept. When the fluctuation range is less than 5%, the accuracy is considered acceptable.Results: The experimental resolutions of FAAS, GC, and LC–MS all reach 10% or better, which can be used for quantitative detection. The minimum detection limits of FAAS, GC, and LC–MS are 0.064, 0.001 μg/mL, and 0.088 ng/mL, respectively. The sensitivity of LC–MS is higher than that of the other two instruments according to the present data. When the coefficient of determination is greater than 0.98, the accuracy is further calculated. The accuracy results (relative errors) of FAAS, GC, and LC–MS are 1.015%, 1.430%, and 3.475%, respectively. FAAS is considered to be the most accurate.Conclusions: The method for measuring the experimental resolution, minimum detection limit, and accuracy based on linear models is simple and reasonable. The experimental performance is evaluated from multiple perspectives, which simplifies and improves the process and effect of the evaluation in order to suggest improvement strategies for analysis experiments and related equipment.