Case Study

Sensory Analysis Tracks Flavor-Package Interactions

Packages are designed to protect food products. Consequently, a natural expectation is that the packaging material itself will not affect the product. This is, however, not always the case. Not only may some packaging materials contribute off-flavors to food products, they can even scalp flavor aromatics from the food. Fortunately, sensory analysis offers a tool to identify potential negative package interactions. 21st Sensory, a Bartlesville, OK-based sensory research firm, has both the testing facility and trained panelists available to handle such packaging studies.

Rounding up the resins
In one recent project, a water bottler was looking to switch suppliers for the HDPE resin it used to blow-mold its bottles. Prior to the switch, this bottler wanted to be sure the new polymer wouldn't transfer off-notes into the water.

"I think it's overlooked in the beverage area, in particular, and many water bottlers are unaware of potential interactions until they get a complaint," says Kathleen Rutledge, 21st Sensory president. "Some will start with a nice filtered water and end up with something that smells and tastes of plastic."

In blow-molding, the resins are heated and forced out through dies. The high heat level required has the potential to generate byproducts and/or decomposition products that can transfer to the water. Not all resins created equal, either. Each supplier will have slightly different formulations that may be better suited to certain applications.

To determine if switching resins would make a difference for the water bottler, 21st Sensory first collected several samples. The control water sample was the original resin bottles that had been blow-molded and filled by the bottler. The first test sample consisted of bottles blow-molded by the new resin supplier and filled at 21st Sensory's research center with water supplied by the bottler. The bottler provided two additional test samples. One featured the new resin blow-molded and filled by the bottler and the other consisted of random samples pulled from regular production. The last sample was a test control using bottles blow-molded by resin supplier and filled at the 21st Sensory lab with ultra pure filtered water.

Training key to success
To conduct the actual evaluation, the panel leader assigned a panel of eight highly trained panelists to the project. For this particular project, panelist sensitivity and resistance to "taste fatigue" were key considerations.

"Panelists say that water samples are the most demanding to evaluate," says Rutledge. "When you're just tasting water, the metallics and astringencies you're looking for are very subtle."

The panel first reviewed flavor and aroma intensity references using a range of standard reference solutions of salt and sucrose in water. The panel also used additional food-based references such as peanut butter, cola beverages, grape jelly, peas, chicken soup and soy sauce. The group then reviewed qualitative references which included:

Candle wax. Aromatic associated with birthday candles, paraffin.
Candle wax (Burnt). Aromatic associated with burnt candle wax, also a grilled meat fat, smoky association.
Dusty. Aromatics associated with dust.
Metallic. The aromatics associated with metals such as iron or copper.
Stale. The aromatics associated with water that has been allowed to sit in a sealed container without aeration.
Vinyl, new. Aromatic associated with newly unwrapped vinyl products.

Five of the samples were coded and the group familiarized themselves with the samples. They trial balloted the samples for both flavor and aroma and discussed the data for agreement and readiness to complete testing.

Off-flavor evaluation
To perform the actual evaluation, panelists received a list of sample codes in random order. In one session, panelists tested samples for aroma intensity and character. In the second session, they tested the flavor intensity and character of the samples. When all panelists had completed testing, the panel leader wrote all scores on the white board for discussion. The group agreed on a consensus score for the aroma and flavor intensity of each sample and agreed on the qualitative characteristics.

All of the samples exhibited very, very slight aroma and flavor characteristics. On a 0 to 15 scale with 0 being no measurable characteristic, to 15 being very strong, these samples were virtually free of any aroma (scores ranged from 0.4 to 0.9). In flavor, the samples were measured at very low intensities from 1.2 to 2.2.

The panel concluded that the test samples exhibited no difference when compared against control samples and random samples. With this information, the bottler decided to begin purchasing the new resin.

21st Sensory Inc. P.O. Box 3913, Bartlesville, OK 74006. Tel: 918-333-1011. Fax: 918-333-7773.