Green leaf volatile (GLV) mixtures, commercial orange flavors, and commercial strawberry flavors were applied to beverage bases in which concentrations of citric acid as well as a sweetener (sucrose or aspartame/acesulfame-K) were varied. Sensory profiling showed that flavor-specific fruity character increased as perceptible sweetness increased, independent of whether the sweetness resulted from sucrose (a change from 9 to 12 Brix) or aspartame/acesulfame-K (a change from 0.2 to 0.4 Brix). Sweetness was affected only by the tastants in the base and not by the flavors, although flavor-specific interactions between sweetener type and sweetener level occurred. Flavor release from the sucrose bases was compared to flavor release from bases containing aspartame/acesulfame-K by static headspace measurements and by MS-Nose measurements using an artificial throat. These measurements showed greater flavor volatility from bases having low Brix (fewer soluble solids). This negative Brix effect was also evident in the sensory data for perception of some GLV green notes. The headspace data could not support a positive Brix effect, the typical salting out, which would correspond to the observed perceptual enhancement of fruity notes.

Pyrazine formation was studied as a model for carbohydrate fragmentation in the Maillard reaction. Thus 1-13 C-glucose, 2-13 C-glucose and 1-13 C-fructose were reacted with asparagine in 1,2-propanediol, and the volatile products isolated by steam distillation and extraction. The product mixture was analyzed by GC, NMR and GC-MS, and consisted mainly of dimethyl-, monomethyl, and to a lesser extent trimethylpyrazines. Although product yields and ratios for the reaction of asparagine with glucose differed to some extent from the same reaction with fructose, the 13 C-distribution of the resulting pyrazines was not much different. The 13 C-distribution in pyrazines originating from 2-13 C-glucose was quite different from the 13 C-distribution in pyrazines which were formed from glucose and fructose labeled at the C-1 position. The 13 C-incorporation in the pyrazines obtained from all three labeled hexoses was in agreement with retro-aldolization of the intermediate deoxyglucosones as the main cleavage mechanism (scheme 5). Both 1-and 3-deoxyglucosone appear to play an approximately equally important role in the formation of the methylated pyrazines.The formation of many flavor substances in the Maillard reaction is initiated by carbohydrate cleavage. The resulting mono-and dicarbonyl products may react further to form pyrazines, thiazoles, carbocyclic compounds, etc., depending on conditions and other reactants (7-5). Asparagine appears to be a particularly good amino acid in the formation of pyrazines (5,6), and was used here to study pyrazine formation and the underlying carbohydrate cleavage mechanism.