Formation of r-Dicarbonyl Compounds in Beer during Storage of Pilsner ADRIANA BRAVO,* ,† JULIO C. HERRERA, ‡ ERIKA SCHERER, † YON JU-NAM, † HEINRICH RÜBSAM, † JORGE MADRID, ‡ CARSTEN ZUFALL, † AND RAFAEL RANGEL-ALDAO †,§ Unidad de Investigacio ´n e Innovacio ´n, Empresas Polar, 4ta. Transv. de Los Cortijos de Lourdes, Caracas 1071, Venezuela, and Departamento de Quı ´mica and Departamento de Biologı ´a Celular, Universidad Simo ´n Bolı ´var, Apartado 89000, Caracas 1080A, Venezuela With the aim of determining the formation of R-dicarbonyl intermediates during beer aging on the shelf, R-dicarbonyls were identified and quantified after derivatization with 1,2-diaminobenze to generate quinoxalines. The sensory effects of R-dicarbonyls were evaluated by the quantification of key Strecker aldehydes and by GC-olfactometry (GCO)analysis of beer headspace using solid phase microextraction. Four R-dicarbonyls, reported here for the first time, were detected in fresh and aged beers, three were derived from the 2,3-enolization pathway of mono- and disaccharides, and the fourth was derived from the epimerization of 3-deoxy-2-hexosulose. Ten R-dicarbonyls were quantified during beer processing and during different periods of beer aging at 28 °C. The aging periods were from 15 to 105 days. During beer aging, 1-deoxydiuloses were produced and degraded, while 1,4- dideoxydiuloses were produced at the highest rates. The GCO analysis indicated that forced beer aging increased the amounts of furaneol, trans-2-nonenal, and phenylacetaldehyde. The blockage of R-dicarbonyls inhibited the accumulation of sensory-active aldehydes in the beer headspace. KEYWORDS: r-Dicarbonyl quantification; beer aging; 1,4-dideoxydiuloses; 1-deoxydiuloses; GC-olfac- tometry INTRODUCTION The most appreciated characteristic of beer by consumers is its fresh flavor. This appealing characteristic is progressively lost on the shelf due to flavor changes. These flavor changes deteriorate the beer flavor and are referred to as aging. A technological solution to this important practical problem, however, is not in sight because of the incomplete knowledge concerning the mechanisms involved in beer aging and the flavor-active compounds responsible for flavor deterioration. Until now, most of the approaches considered in the brewing industry have been oriented toward the elimination of oxygen uptake throughout the brewing process to eliminate trans-2- nonenal or its precursors from malt and wort (1). Nevertheless, despite all efforts concerning this subject, only a relatively small improvement in beer flavor stability has been attained (2). Reducing the oxygen content in packaged beer only gives a product of limited flavor stability (3). Therefore, other chemical mechanisms and/or pathways different from the oxidative pathway that also contribute to flavor formation should be considered, among them the Maillard reaction, which may have an important role in flavor stability. In our previous work, the formation of a group of R-dicar- bonyl intermediates, arising most likely from carbohydrate degradation, were determined in beer for the first time (4). In that study, aging took place at 28 °C, which can be considered a mild to low temperature for most model Maillard reaction studies. It was observed that the formation of 5-hydroxymeth- ylfurfural (5-HMF) from the dehydration of 3-deoxy-2-hexo- sulose (3-DH) correlated well with beer aging. However, it was not clear as to how R-dicarbonyl intermediates originating from 2,3-enolization, 1-deoxydiuloses, were involved as precursors of aroma-active compounds such as Strecker aldehydes and cyclic enolones. Our hypothesis was that these R-dicarbonyls may play an important role in beer flavor deterioration even at this mild condition since a significant increase in the concentra- tion of 1,4-DDP and 1,4-DDH was observed in beer stored at 28 °C, 1,4-DDP and 1,4-DDH being derived from the Strecker degradation of 1-DP and 1-DH in the presence of amino acids (5). An alternative mechanism for 1,4-DDH formation, the so- called “peeling off” mechanism, was proposed for the quasi- water-free thermolysis of maltose (6). In our previous work, the formation of 1-DH also was detected in beers, (4) but its concentration could not be estimated and neither could the * Corresponding author. Tel.: +58 212 202 3905; fax: +58 212 202 3065; e-mail: adriana.bravo@empresas-polar.com. † Unidad de Investigacio ´n e Innovacio ´n, Empresas Polar. ‡ Departamento de Quı ´mica, Universidad Simo ´n Bolı ´var. § Departamento de Biologı ´a Celular, Universidad Simo ´n Bolı ´var. 4134 J. Agric. Food Chem. 2008, 56, 4134–4144 10.1021/jf703696p CCC: $40.75  2008 American Chemical Society Published on Web 05/08/2008