6782 J. Agric. Food Chem. 2003, 51, 6782−6790 Evolution of Chemical and Sensory Properties during Aging of Top-Fermented Beer BART VANDERHAEGEN,* HEDWIG NEVEN,STEFAN COGHE,KEVIN J. VERSTREPEN, HUBERT VERACHTERT, AND GUY DERDELINCKX Centre for Malting and Brewing Science, Katholieke Universiteit Leuven, Kasteelpark Arenberg 22, B-3001 Heverlee, Belgium The aging and consequent changes in flavor molecules of a top-fermented beer were studied. Different aging conditions were imposed on freshly bottled beer. After 6 months of aging, the concentration changes were recorded for acetate esters, ethyl esters, carbonyls, Maillard compounds, dioxolanes, and furanic ethers. For some flavor compounds, the changes with time of storage were monitored at different temperatures, either with CO2 or with air in the headspace of the bottles. For some molecules a relationship was determined between concentration changes and sensory evaluation results. A decrease in volatile esters was responsible for a reduced fruity flavor during aging. On the contrary, various carbonyl compounds, some ethyl esters, Maillard compounds, dioxolanes, and furanic ethers showed a marked increase, due to oxidative and nonoxidative reactions. A very high increase was found for furfural, 2-furanmethanol, and especially the furanic ether, 2-furfuryl ethyl ether (FEE). For FEE a flavor threshold in beer of 6 µg/L was determined. In the aged top-fermented beer, FEE concentrations multiple times the flavor threshold were observed. This was associated with the appearance of a typical solvent-like flavor. As the FEE concentration increased with time at an almost constant rate, with or without air in the headspace, FEE (and probably other furanic ethers) is proposed as a good candidate to evaluate the thermal stress imposed on beer. KEYWORDS: Top-fermented beer; beer aging; 2-furfuryl ethyl ether; furanic ethers; flavor stability INTRODUCTION of top-fermented beer. Using advanced purge and trap extraction techniques combined with gas chromatography and mass The limited flavor stability of beer is still one of the major spectrometry, the changes in the volatile fraction of beer were and less understood problems in the brewery industry. For beers, examined, as were the effects of temperature and oxygen. In there is a need to optimize production processes toward shelf addition, a relationship between the sensory evaluation of beer life. This is certainly necessary in a continuously competitive aging and the appearance of specific compounds was sought market. Most research on beer aging has been done on lager and discussed. beer, which represents the largest part of the beer market. Especially, lipid oxidation during brewing and the appearance of trans-2-nonenal in beer have received much attention (1- MATERIALS AND METHODS 3). The latter reaction is considered to be responsible for the Chemicals. The following substances with corresponding purity were development of papery and cardboard-like flavors during beer supplied by Sigma Aldrich Chemie GmbH (Munich, Germany): ethyl aging (4). However, most top-fermented beers do not show the acetate (99.9%), propyl acetate (99.9%), butyl acetate (99.7%), pentyl same type of aging characteristics as found for lager beer (5, acetate (99.7%), hexyl acetate (99.7%), heptyl acetate (98%), octyl 6). Top-fermented beers, with their typically higher fermentation acetate (99+%), isobutyl acetate (99.8%), isoamyl acetate (99.7%), ethyl temperatures (20-24 °C) and different yeast strains, usually propionate (99.7%), ethyl butyrate (99.7%), ethyl pentanoate (99.7%), + contain more esters and fusel alcohols, and they often have a ethyl hexanoate (99 %), ethyl heptanoate (99%), ethyl octanoate + + higher alcohol content (6-10% v/v) (7). Although very little (99 %), ethyl nonanoate (98 %), ethyl decanoate (99%), diethyl succinate (99.5%), ethyl phenylacetate (99+%), ethyl pyruvate (98%), research has been done on the aging of such beers, it is often ethyl lactate (98%), ethyl 3-methylbutyrate (99.7%), ethyl 2-methyl- mentioned that the development of Madeira- or port wine-like butyrate (99%), ethyl 2-methylpropionate (99%), acetaldehyde (99.5%), flavors is typical. The objective of this study was an in-depth hexanal (98%), octanal (99%), nonanal (95%), trans-2-nonenal (97%), investigation on the chemical mechanisms involved in the aging benzaldehyde (99.5%), phenylacetaldehyde (98+%), 3-methylbutanal (98%), 2-methylbutanal (95%), diacetyl (99.5%), 2,3-pentanedione * Author to whom correspondence should be addressed (e-mail (97%), 4-methylpentan-2-one (99%), 2-furfural (99%), 5-methyl-2- [email protected]; telephone +32-16321460; fax furfural (99%), 2-acetylfuran (99+%), 2-furanmethanol (99%), thiazole +32-16321576). (99+%), 2-heptanol (99%), and 2,2-diphenyl-1-picrylhydrazyl (95%). 10.1021/jf034631z CCC: $25.00 © 2003 American Chemical Society Published on Web 10/03/2003 Properties of Top-Fermented Beer J. Agric. Food Chem., Vol. 51, No. 23, 2003 6783 The furanic ether 2-furfuryl ethyl ether with a purity of 95% was Table 1. Aging Conditions Imposed on a Top-Fermented Beer purchased from Narchem Corp. (Chicago, IL). Beer Aging Conditions. A fresh pale top-fermented beer (7.5% v/v aging condition storage temp (°C) headspace contents alcohol) was obtained from a Belgian brewery and subjected to five I0CO2 different aging conditions. Initially, the dissolved oxygen concentration II 20 CO2 was 0.5 mg/L, and the headspace contained <0.1 mg of oxygen. The III 20 air beer was aged in bottles containing 250 ( 0.2 mL of beer and 10 ( IV 40 CO2 0.5 mL of headspace volume, filled with either CO2 or air. Air was V 40 air brought into the headspace by opening the bottles, flushing the headspace shortly with air, and capping it again. Beer samples with a Flavor Threshold of 2-Furfuryl Ethyl Ether. The flavor threshold CO2 headspace were stored at 0 ( 0.2, 20 ( 0.2, and 40 ( 0.2 °C, whereas beer samples with air in the headspace were stored at 20 ( of 2-furfuryl ethyl ether added to a neutral lager beer was determined 0.2 and 40 ( 0.2 °C. After 0, 12, 25, 51, 84, 119, and 187 days of according to EBC Analytica (12) method 13-9, using a trained panel aging, two samples for each storage condition were analyzed in of 15 members. Triangle tests were performed with concentrations of duplicate. 1.2, 2.3, 4.7, 9.4, 18.7, 37.2, 75.6, and 149.9 µg/L of 2-furfuryl ethyl ether. Beer Analysis. Total polyphenols were quantified according to the method of De Clerck and Jerumanis (8). RESULTS AND DISCUSSION Flavanoids were quantified according to the method of Delcour and Janssens de Varebeke (9). Sensory Analysis of Aged Beer. Table 1 summarizes the The reducing power of beer was measured using the method five different conditions used to study beer aging. described by Kaneda et al. (10). Degassed beer (0.2 mL) was added to For sensory analysis, in a first session, the panel was asked 2.8 mL of 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical solution to describe the characteristics of aroma and taste of each aged × -4 (1.86 10 mol/L). The reducing power was defined as the beer. On the basis of the results of the first degustation, six discoloration of the DPPH solution (relative decrease in absorbance) major aroma and taste descriptors were selected and used for after 5 min at 525 nm. further sensory evaluation. In a second session the panelists were Beer color was measured at 430 nm according to the method of asked to quantify their intensities. The mean of the scores for Seaton et al. (11). each flavor aspect is given in Figure 1. It appeared that after 6 Analysis of Volatile Compounds. Prior to analysis, beer was months, aroma and taste were strongly affected by the type of degassed by kieselguhr filtration. Then, 200 µL of internal standard ° (250 mg/L 2-heptanol) and 200 µL of a 10% antifoam solution Sigma storage condition. Samples stored at 40 C lost the initial fruity Aldrich Chemie GmbH (Munich, Germany) were added to 50 mL of estery flavor, and strong Madeira- or port-like flavors developed degassed beer. Five milliliters was then transferred into the Tekmar (Figure 1B). In these conditions, a significant solvent and a Dohrman 3000 (Emerson, St. Louis, MO) purge and trap concentrator minor papery flavor appeared. Samples stored at this temperature unit with a Vocarb 3000 trap (Supelco, Bellefonte, PA) in the following showed an increased pungency (Figure 1D). conditions: helium was the carrier gas, 10 min purge at 140 °C, 8 min For samples with air in the headspace, the aroma was dry purge at 140 °C, 6 min desorption at 250 °C, 10 min bake at 260 described as more pronounced caramel-like and red fruit-like ° C. The indicated temperatures are those of the adsorbing trap; the (Figure 1A). Oxidation caused the beer’s bitter taste to decrease ° beer sample temperature was kept at 20 C during purging. The and evolve toward sweet and warming flavors (Figure 1C). relatively high trap temperature of 140 °C during the purge and dry Changes in Beer Color, Polyphenol Concentration, and purge steps avoided saturation of the trap with ethanol. Before entering the GC, volatiles were focused using a cold trap with an MFA 815 Reducing Power. Figure 2A shows the increase of beer color control unit (ThermoFinnigan, San Jose, CA) in the following condi- during aging. An almost linear increase was observed for tions: initial temperature, -70 °C; final temperature, 200 °C. GC was samples stored at 40 °C in the absence of oxygen. The formation performed using a Fisons GC 8000 gas chromatograph equipped with of colored Maillard products in these storage conditions is the a Chrompack CP-WAX-52-CB column (length ) 50 m, i.d.