Measurement of vapor-liquid equilibria of ß-damascenone and simulation of its behavior during a batch distillation

International audience

Inglés. Distilled spirits are essentially non-ideal mixtures of ethanol and water containing hundreds of highly diluted aroma compounds that will dictate their aroma typicity. Despite the prominent role of the distillation operation in building the aromatic quality of spirits, its mastery is hampered by a high dependency on empirical knowledge and the lack of a more scientific basis. Computer simulation is a valuable tool to achieve a better understanding of the behavior of the different components in a mixture during distillation processes, but the implementation of simulation modules for spirit distillation requires reliable vapor-liquid equilibria (VLE) data of the different components in the mixture, in addition to the choice of a suitable thermodynamic model. When VLE data is not available in the literature, experimental measurements might be needed, as is the case for ß-damascenone, an aroma compound that plays a key role in the aroma of several alcoholic beverages. This work focused on the experimental acquisition of new VLE data for ß-damascenone highly diluted in ethanol-water mixtures and the simulation of its behavior during batch multistage and simple distillations. The VLE measurements were conducted in a Gillespie-type recirculating still operating at adiabatic and isobaric conditions, and the experimental data was correlated by activity coefficient model NRTL to estimate binary interaction coefficients that are required for simulation purposes. Simulation modules were then built with ProSim® BatchColumn for two laboratory-scale distillations of model solutions of ß-damascenone in ethanol-water mixtures, and the results were compared with experimental data. This validation step allowed to verify that the NRTL coefficients accurately represent the behavior of the aroma compound in the simulation of a spirit distillation. This approach can be extended to other aroma compounds and the results can be applied for simulations of industrial-scale distillations.