Type of emulsifying agent impacts aroma release.

Our sense of smell is nothing short of exquisite. Engineers can only dream of designing a machine that could match it in terms of its range and sensitivity. Our ability to detect certain aromas in the sub-part per billion or even sub-part per trillion range is the reason many similar products do not have the same aroma. Some estimates suggest we can distinguish around 10,000 different smells.

In many foods and beverages, the fat phase occurs as part of an emulsion. Various emulsifying agents are commercially available, each with different physico-chemical properties. For this reason, it appears to some that the type of emulsifier might have a significant impact on the aroma release profile--the flavor perception--of foods and beverages.

Scientists at the University of Pennsylvania wanted to evaluate the influence of the type of emulsifier on the release of select aroma molecules in model oil-water emulsions. Their efforts indicate that the selection of an emulsifying agent can serve as an important parameter for controlling the release of aroma molecules in oil-water emulsions.

A novel instrumental design was developed for analyzing aroma molecules released from model emulsions. This system consisted of a model mouth that was directly interfaced to a flame ionization detector. Time-release profiles for several aroma compounds--aldehydes, esters and ketones--were obtained from oil-water emulsions, which had a median particle diameter of 0.5 micron. Researchers used three emulsifying agents: whey protein isolate, Tween 20 and gum arabic. Static headspace analysis with gas chromatography was also used to quantify the concentration of the free aroma compounds in the model emulsions.

Time-release profiles showed that the type of emulsifier had a significant effect on aroma release kinetics in the emulsions. The release from certain aroma-surfactant combinations, e.g., whey and hexanal, was particularly slow, suggesting that chemical binding occurred. This was also supported by static headspace analysis.

Further information. Devin Peterson, 215 Borland Laboratory, Department of Food Science, Pennsylvania State University, University Park, PA 16802; phone: [phone]; email: [e-mail].