Flavor chemistry of beer : Part II : Flavor and threshold of 239 aroma volatiles

The flavor thresholds of the 239 compounds examined varied over a range of 2 x 1012, and it is therefore not sur­ prising that on 0 casion, compounds of high chemical pu rity (99% +) were organolepticall y impu re or even misflavored. Proof of organoleptic purity requires repeated purification until a constant flavor and a constant threshold are obtained. In each homol ogous series, the compounds containing 5 10, and usually 7 10 carbon atoms show the strongest flavor. The purified fusel alcohols were found to have pleasant, aromatic flavors. orne sulfu r compounds show threshold values which are several orders of magnitude below our present limits of detection by analysis. The flavor and threshold of a given compound can often be predicted quite reliably from the data presented. The significant flavor constituents of fresh beer are, in the approximate order of their importan e, ethanol, the hop bitter substances, carbon dioxide, the banana esters, the apple esters, and the fusel alcohols. In beers which are off-flavored (sulfury, diacetyl-like, or stale) the corresponding flavor compounds dimethyl sul fide + hydrogen sulfide, diacetyl, and 2-trans-nonenal can play a much greater role than any of the normal flavor constituents. INTRODUCTION Part I of this series dealt with the literature of beer flavor chemistry, the measurement of flavor strength, and the phenomenon of flavor interaction. The purpose of the present section is to report the flavor and thresh­ old of the 239 compounds studied to date, and to discuss the observed trends whkh may be applied to predict the strength and type of flavor of other compounds. haracteristics of the Mexican and U.S. beers used in the study were reported in Part I. Future papers in this series will add new or revised thresholds as they become available, and will expand the study of interactions until all compounds which can be flavor-active in beer have been included. The author is Director of Research and Development, The Stroh Brewery Company, Detroit, Michigan. Part I of this paper appeared in the previous issue of Tech­ nical Quarterly. By Morton C. Meilgaard SINTESIS Los limites de deteccion de sabor de 239 compuestos ex­ aminados variaron en el rango de 2 x 1012, y por 10 tanto no es sorprendente que en ocasiones, los compuestos de alta pu reza quimica (99%+) resultaron impuros organoJepti­ camente 0 aUn con sabores erroneos. La prueba de la pureza organoleptica requiere de purilicaciones repetidas hasta obtener un sabor y un limite de deteccion constantes. En cada serie hom610ga, los compuestos que contienen de 5 a 10 y generalmente de 7 a 10 atomos de carbono, muestran el sabor mas fuerte. Se encontro que los alcoholes de fusel purificados tienen sabores aromaticos agradables. Alguno compuestos de azufr meustran val ores de limite de detec­ cion que eswn varios 6rdenes de magnitud abajo de nuestros actuales Iimites de deteccio n por anali is. Se puede a me­ nudo p redecir con algo de seguridad el sabor y el limite de deteccion de un compuesto dado a partir de los datos presentados. Los constituyentes mas significativos del sabor de la cer­ veza fresca son, en orden aproximado de importancia, etanol, substancias amargantes del lupulo, dioxido de carbono, esteres on olor y sabor a pliltano, esteres con olor y abor a manzana y los alcoholes d fusel. En cervezas con sabor a des ompuesto (sulfu rosa, con sabor a diacetilo 0 pasada) 10 correspondientes compuestos de sabor, sulfuro de dimetilo + sulfuro de hidrog no, dia­ cetilo y 2-trans nonenal pu den jugar un papel mucho mayor que cualquiera de los constituyentes normales del sabor. Purity of Flavor Compounds The thresholds found in the present study varied by a factor of 2 x 1012• It is therefore not difficult to under­ stand that a flavor compound may be chemically very pure (99% +) and yet be organoleptically very impure or even misftavored. Examples of this type were phenyl­ ethanol which on purification increased its threshold from 35 ppm to 125 ppm, and tryptophol for which the threshold was raised from 50 ppm to higher and higher values until the compound be came flavorless at its limit of solubility of over 400 ppm. Other compounds for which the threshold increased from 3 to 40 times on purification were phenylethyl acetate, isoamyl alcohol, n-propanol, ethanol, methanol, butyric acid, octanal and isobutyraldehyde. The compound 1-octen-3-ol is re­ ported as being responsible for the flavor of many mush­ rooms (7) although Hoffmann(4) mentions that the odor is caused by some cyclic acetals (1,3-dioxalans) formed on decomposition. We found the compound tasted ex-