A fast and efficient method for the analysis of α-dicarbonyl compounds in aqueous solutions: Development and application.

Among the highly oxygenated species formed in situ in the atmosphere, α-dicarbonyl compounds are the most reactive species, thus contributing to the formation of secondary organic aerosols that affect both air quality and climate. They are ubiquitous in the atmosphere and are easily transferred to the atmospheric aqueous phase due to their high solubility. In addition, α-dicarbonyl compounds are toxic compounds found in food in biochemistry studies as they can be produced endogenously through various pathways and exogenously through the Maillard reaction. In this work, we take advantage of the high reactivity of α-dicarbonyl compounds in alkaline solutions (intramolecular Cannizzaro reaction) to develop an analytical method based on high performance ion chromatography. This fast and efficient method is suitable for glyoxal, methylglyoxal and phenylglyoxal which are detected as glycolate, lactate and mandelate anions respectively, with 100% conversion at pH > 12 and room temperature for exposure times to hydroxide ranging from 5 min to 4 h. Diacetyl is detected as 2,4-dihydroxy-2,4-dimethyl-5-oxohexanoate due to a base-catalysed aldol reaction that occurs before the Cannizzaro reaction. The analytical method is successfully applied to monitor glyoxal consumption during aqueous phase HO∙-oxidation, an atmospherically relevant reaction using concentrations that can be observed in fog and cloud water. The method also reveals potential analytical artifacts that can occur in the use of ion chromatography for α-hydroxy carboxylates measurements in complex matrices due to α-dicarbonyl conversion during the analysis time. An estimation of the artifact is given for each of the studied α-hydroxy carboxylates. Other polyfunctional and pH-sensitive compounds that are potentially present in environmental samples (such as nitrooxycarbonyls) can also be converted into α-hydroxy carboxylates and/or nitrite ions within the HPIC run. This shows the need for complementary analytical measurements when complex matrices are studied.