Decoupling the effects of heating and flaming on chemical and sensory changes during flambé cooking

Anecdotal and literature accounts on the combustion, or flaming, of high-proof spirits during food preparation (flambé) have reported ethanol losses, browning and sensory changes of the final product. In this work, the effects of heating and combustion were decoupled by preparing model flambé (heated-and-ignited) systems alongside similar systems which were heated but not ignited. In a simple flambé systems consisting only of vodka, we observed a 24.7% loss of ethanol in the heated-and-ignited treatments and a 34.7% loss in the heated-not-ignited systems. In a model caramel sauce containing butter, sugar, and vodka, no significant difference in ethanol loss was observed between the ignited (13.2%) and not-ignited (14.1%) treatments. In both systems, the majority of ethanol loss was due to heating rather than combustion. No significant difference was observed in Hunter lightness (L) values between heated-and-ignited and heated-not-ignited treatments for the butter–sugar–vodka system, suggesting that the effect of flambé on browning was minimal. Maximum surface and flame temperatures were then measured in the vodka system by thermocouples. While maximum flame temperatures up to 532 °C were observed during 15 s of flaming, the maximum temperature at 1 cm above the pan surface was 67 °C, below temperatures typically required for significant Maillard or caramelization reactions on this time scale. In triangle tests using the vodka system, panelists were able to discriminate heated-and-ignited from the unheated control and, in one experiment, from the heated-not-ignited sample, even when treated samples were reconstituted with water and ethanol lost. However, for the butter–sugar–vodka system, the majority of panelists could not discriminate between ignited and not-ignited treatments.