Osiris" optimization software was used for optimizing separations with nonaqueous or partially aqueous mobile phases, consisting of 2, 3 or 4 solvents. For nonaqueous eluents used in Nonaqueous Reversed Phase Liquid Chromatography (NARP-LC), classical transfer rules cannot be used to determine isoeluent mobile phase compositions that are generally the starting point of most of the optimization procedures. The ability of this software to determine isoeluent compositions under NARP conditions is demonstrated. This determination is performed by modeling the solute retention from different sets of two gradient runs, each set being carried out for each binary mobile phase studied. The application of retention models to nonaqueous mobile phases is discussed in terms of the accuracy of the predicted retention times. In addition, it is shown that these sets of gradient runs allow selection of the best composition space for the optimum composition search without requiring additional runs. It is demonstrated that such a selection procedure greatly reduces the number of preliminary experiments.

No abstract

Abstract. An analytical procedure has been developed for determining aliphatic alcohol ethoxylates (AAEs) in raw wastewater. AAEs were isolated by solid-phase extraction on a styrenedivinylbenzene disk. The maximum volume that was processed without plugging of the disk was 20 mL. Then AAEs were desorbed with methanol and, after evaporation to dryness, the residue was analyzed by reversed-phase liquid chromatography coupled to mass spectrometry via atmospheric pressure interface. Two desorption temperatures (ambient and 100 °C), two organic modifiers of the mobile phase (methanol and acetonitrile) and two atmospheric pressure interfaces (electrospray and atmospheric pressure chemical ionization) were tested. The lowest coefficients of variation for replicate measurements were obtained by desorbing AAEs from the extraction disk at 100°C and by using methanol-based mobile phase with atmospheric pressure chemical ionization source for extract analysis. Under these conditions, all the analytes (ranging in size from 10 to 16 alkyl carbons and from 1 to 25 ethoxylate groups) were recovered at more than 70 % and the limit of quantification (signal-to-noise ratio = 10) ranged between 0.05 and 4 µg L -1 .