Modeling of the kinetics of higher alcohol and ester production based on CO2 emission with a view to control of beer flavor by temperature and top pressure

Citation
M. Titica et al., Modeling of the kinetics of higher alcohol and ester production based on CO2 emission with a view to control of beer flavor by temperature and top pressure, J AM SOC BR, 58(4), 2000, pp. 167-174
Citations number
25
Categorie Soggetti
Food Science/Nutrition
Journal title
JOURNAL OF THE AMERICAN SOCIETY OF BREWING CHEMISTS
ISSN journal
03610470 → ACNP
Volume
58
Issue
4
Year of publication
2000
Pages
167 - 174
Database
ISI
SICI code
0361-0470(2000)58:4<167:MOTKOH>2.0.ZU;2-5
Abstract
This article presents a model of beer fermentation in terms of flavor compo nent content. It describes the production of two higher alcohols (isoamyl a lcohol and phenyl ethanol) and three esters (isoamyl acetate, ethyl acetate , and ethyl hexanoate) selected according to their organoleptic threshold v alues in beer, as well as the effects of operating conditions on their form ation. A step-by-step modeling approach was used, which exploited prior bio logical information coupled with analysis of experimental data. These data resulted from a two-level full-factorial experimental design for three fact ors: temperature (10-16 degreesC), top pressure (50-800 mbar), and yeast in oculum (0.5 x 10(7)-2 x 10(7) cells per milliliter). Experiments were carri ed out in a 15-L, reactor. Because CO2 emission, which represented the only on-line measurement in this process, is directly associated with primary y east activity, the aroma dynamics were predicted from those for CO2, with y ield coefficients depending only on operating conditions (temperature, diss olved CO2, yeast inoculum size). Model parameters were identified from expe rimental data. An initial experimental validation gives a successful predic tion of most of the aroma compounds when compared with their measured value s. The model quantifies the effects of operating conditions on the producti on kinetics and final concentrations of aroma compounds. These results sugg est that this model could be successfully used for both prediction and cont rol of some aroma compounds under industrial conditions.