Catching the Scent

David Pacchioli
September 01, 1999

At a public lecture last fall, Mary Chisholm passed out samples of androsterone, a hormone in the testosterone family. "It's a rather interesting compound," she explains. "To some people it smells like musty camphor. To others it is entirely revolting. A third group smells nothing at all." Her audience that night, Chisholm remembers, ran happily true to form: "A third of them smelled camphor; a third were entirely revolted; and a third smelled nothing at all. So it was an impressive demonstration of how complex the subject of olfaction is, and how much we don't yet know."

red white and black molecule
Robert Minard

An androsterone molecule.

Chisholm, an associate professor of chemistry at Penn State Erie, has spent much of her career trying to understand what she calls the chemistry-physiology link in the mechanism of olfaction. Her particular forte is chasing down the individual chemical compounds that add up to a given aroma. Last year, she made international news when she and a trio of undergraduate students —Matt Wilson, Nicole Taylor, and Aisha Mitchell—identified a previously undetected component in the odor of lime. That compound, 7-methoxycoumarin, has a woodsy, balsamic scent, she reports. Without it, lime just wouldn't be lime.

To find this particular fragrance, Chisholm and her students used their standard technique, called gas chromatography-olfactometry (GCO), "which allows us to separate the aroma components of any food, describe them one by one as they emerge from a column, and measrure their intensity." Actually, however, GCO is the second step. "First," Chisholm says, "we smell."

And not for sentimental reasons. "We use the nose as a detector," Chisholm explains, because many of these compounds are in concentrations of a few parts per billion or lower. No instrument can detect reliably at that low level."

In truth, some noses can't either. When it comes to assembling an aroma panel, "heavy smokers have poor response." Chisholm says. "So do people with chronic sinus trouble. But any reasonably healthy person is just fine. It's just a matter of being trained."

The rookies start with a standard mix containing compounds. "If they're able to detect all six and get the intensity reasonably well," Chisholm says, "then we start working on complex compounds—lime oils, for instance. We work on reproducibility, and on improving our descriptors."

Describing odors, she says, "is difficult, and until you start doing it you don't realize how difficult. Standardization is an issue, of course, but another interesting problem arises because a lot of describing odors has to do with recall and memory. People come up with odd descriptors based on associations."

A citrus aroma wheel, she notes, helps to standardize the descriptors generated by different sniffers, so that perceptions can be matched with uniform language. On the wheel, odors are grouped and arranged in tiers, and labeled with common comparative terms. At the first tier, a student might distinguish between woody and floral, Chisholm says. At the next, "the better ones can get down to distinguishing lavender from roses. But it doesn't really matter how accurate the descriptors are, just so they're consistent."

Chisholm, who has worked for years identifying aromas present in the wines produced around Lake Erie, got interested in limes while back in her native England, where she spent a sabbatical at the troubleshooting lab of Cadbury Schweppes, famous as makers of chocolates and ginger ale. "Since I don't much care for chocolate," she says, she gravitated toward citrus extracts, about which she found almost nothing describing aroma in the chemical literature. Studying lime oils, she says, "is much easier for my students than studying wine. Fermentation is a complicated, messy business."

Beyond its potential importance to soft-drink makers, Chisholm's recent work raises some basic olfactory questions. "There is no single component in lime odor that smells like what we recognize as lime," she points out. "So how do its component odors add up to that distinctive smell? What are the chemical interactions involved?" Or, to put it bluntly, Why do things smell the way they do? "These are questions we simply can't yet answer," she says.

Researchers have been investigating links between odor and molecular structure for some time, "but the vital link has not yet been found," Chisholm says. One clue, she suggests, may lie in the phenomenon of chirality, a common quirk of nature where two isomers (or versions) of the same molecule differ only in that their structures are exactly reversed. "They're mirror images, all but identical, " she says, and yet in some cases they exhibit very different odors. Limonene, for example, is a major component of all citrus oils, and a chiral molecule. "One of its isomers, naturally, smells fresh and lemony," she says. "The other smells like diesel oil.

"It's a good thing for us that second one is not present in lime."

Mary Chisholm, Ph.D., is associate professor of chemistry at Penn State Erie, the Behrend College, H32 Hammermill, Station Road, Erie, PA 16563; 814-898-6412, or

Last Updated September 01, 1999