Model of retention time and density of gradient peak capacity for improved LC-MS method optimization: Application to metabolomics

Abstract

A general and deterministic model is derived from the fundamentals of liquid chromatography to calculate retention time, peak width, peak capacity, and density of peak capacity in gradient liquid chromatography. The calculation of these chromatographic properties accounts for 1) the presence of initial (separation of the earliest eluters) and final (column wash) isocratic steps before and after the linear gradient, respectively, 2) the pre- (flow through needle and preheater tubes) and post-column (outlet and emitter tubes before MS detection) dispersion, 3) the compression of the chromatographic band, and 4) the retention of the organic modifier onto the RPLC column. The multiple and variable method parameters may include the column dimensions, particle size, flow rate, temperature, initial and final isocratic hold times, gradient time, gradient steepness, column conditioning/sample load time, and the pre- and post-column tube dimensions. The model enables the users to perform robust multi-dimensional optimization of UHPLC-MS methods and offers the possibility to predict the expected MS feature density for increased method performance. Method optimization can be further improved by matching the observed MS feature density (number of metabolites detected as function of time) to the predicted density of peak capacity. It is directly applied to the optimization of high-throughput RPLC separation methods specifically designed for large-scale urinary metabolic phenotyping.

Keywords: Density of peak capacity; LC-MS method optimization; Non-targeted metabolomics; Retention time and peak width prediction; Sample feature density; Urine metabolites.