A rational strategy based on experimental designs to optimize parameters of a liquid chromatography-mass spectrometry analysis of complex matrices.

A high number of factors controlled by the experimenter has to be optimized to successfully separate, ionize and detect compounds when analyzing complex matrices by liquid chromatography hyphenated to high resolution mass spectrometry (LC-UV/MS). Key steps to manage such hyphenation are focused on desolvation and ionization processes. In this study, a design of experiments approach was used to optimize decisive parameters (i.e. nebulising, drying and sweep gas flow rates, ion transfer capillary voltage and temperature) for electrospray ionization and atmospheric pressure chemical ionization sources both in positive and negative modes. Central composite designs including 131 experiments each were built to cover rationally a sufficiently wide range of operating conditions. Each run was repeated three times to insure stable conditions of ionization and thus a satisfactory repeatability. Extracted ion chromatograms of twelve model oxygenated compounds were integrated and used as responses for experiment designs. Quadratic models for each standard allowed to take into account interactions between factors. Then responses were simultaneously maximized to achieve optimized factors. To illustrate the methodology relevance, optimal conditions were applied to a lignocellulosic biomass fast pyrolysis oil. Thanks to our high sensitivity method, a large number of molecular formulae was identified, as for instance in negative-ion mode electrospray with more than 5500 identified molecular formulae whereas analysis of the same sample by mass spectrometry without any prior chromatographic separation provided less than 2000 molecular formulae. In short, this study proposed a rational methodology to optimize ionization efficiency for LC-UV/MS analysis of complex mixtures.