Systematic approach to the separation of mono- and hydroxycarboxylic acids in environmental samples by ion chromatography and capillary electrophoresis

Strategies for method development and optimization of conditions in ion chromatography (IC) and capillary electrophoresis CE are proposed. The approaches are based on well established theoretical models for analyte elution and rely on the inspection of effective capacity factor versus specific gradient ramp slope curves for IC and effective mobility versus pH curves for CE. Commonly used IC eluent systems, such as sodium hydroxide and sodium berate, and the CE electrolyte 3,5-dinitrobenzoic acid were investigated. A standard aqueous solution containing formate, acetate, alpha-hydroxyacetate (glycolate), propionate, alpha-hydroxypropionate (lactate), butyrate and beta-hydroxybutyrate anions was used to test the analytical conditions indicated by the models. Among the IC eluent systems, berate presented a better separation performance than hydroxide. However, a complete resolution of all components in the standard mixture was not accomplished experimentally at the chosen gradient ramp slope of 0.05 mM/ml. The analyte pairs acetate/lactate and glycolate/butyrate, whose effective capacity factors differ by approximately 0.2 units, co-eluted at 0.17 and 0.24 resolution, respectively, and the retention time of the last eluting analyte was relatively long (20 min). Nevertheless, the IC method provided the best overall limit of detection (LOD; 0.016-0.082 mg/l). Under the optimized CE conditions, all seven components in the standard mixture were resolved satisfactorily in less than 7 min. The analyte pair beta-hydroxybutyrate/butyrate presented the worst resolution, 0.45, and a difference in effective mobility of 1.6%. The CE methodology provided the best column efficiency, roughly a ten-fold improvement in terms of number of plates per meter over the IC method, but the limit of detection was comparatively poorer (0.050-0.36 mg/l). Both proposed IC and CE methodologies can be applied to the analysis of mono-and hydroxycarboxylic acids in samples of environmental interest, providing complementary information. The choice of the most appropriate method is a compromise between chemical composition of the sample and concentration level of the analytes under investigation. For instance, the classical co-elution of acetate/lactate that occurs in IC columns, can be solved by CE. But the CE method may lack sensitivity for these analytes, which compromises the analysis of certain real samples. The LOD for acetate and lactate as determined by IC is 0.036 and 0.082 mg/l, respectively, while for CE they increase to 0.11 and 0.20 mg/l. Under optimum conditions, the separation and identification of mono-and hydroxycarboxylic acids in an atmospheric particulate matter sample is illustrated by both techniques. (C) 1998 Elsevier Science B.V.