This article was published in R&D Magazine April 1998 It is reproduced here with generous permission of Cahners Publishing Company Please Click the logo at the left to visit their web site.

Olfactometry Options
Perfect Product Development, Quality

Laura Vandendorpe
Assistant Managing Editor

Electronic noses, GC/MS systems,
and human sensory panels provide varied
solutions for product-odor concerns.

Perfumes, polymers, processed foods, and manure – an unusual montage to say the least, unless you are an olfactometry researcher. These four application areas comprise a large percent of the market served by olfactometers and electronic noses.

Just as the application areas for artificial smelling systems are diverse, so are the technologies that provide these services. Begun in the early part of this century to test perfume, human sensory panels are still frequently employed by companies to test consumer responses to odors generated by clothing, automobiles, and other goods.

The TD, or tank detector, from AromaScan measures chemical odors. The company won an R&D 100 Award in 1995 for its AromaScanner electronic nose.

Industry no longer relies solely on subjective noses, however. Over the past decade, technology designed to refine the olfactometry process has truncated into two different tool groups. On one path of development lie electronic noses and GC/MS equipment specialized to detect chemical "fingerprints" that produce odors. On the other fork lie technologies that operate in conjunction with traditional human sensory panels, providing baseline odors and monitoring devices to quantify and qualify human panel results.

"Before companies decide what type of olfactometry tool to use, several considerations should be made," says Chuck McGinley, technical director of St. Croix Sensory, Stillwater, Minn., and a long-time consultant in this field. "Managers must decide what decisions they will need to make, what data will help make those decisions, and whether they must determine if a product is acceptable to consumers. Each of these decisions will demand different types of tools – electronic noses, GC/MS systems, or human panels."

Electronic nose systems function primarily as quality-control tools in industrial manufacturing applications, says John Warburton, director of sales and marketing for Neotonics Scientific Ltd., Takeley, U.K. "These systems are particularly useful to ensure that product smells remain consistent across batches," he says.

Electronic noses use almost the same sample delivery and transfer techniques as gas-chromatography systems, but do not include columns or mass spectrometers, Warburton says. Chemical detection in most electronic noses is performed using a series of sensors, predominantly comprised of conducting-polymer, metal-oxide, and infrared sensors.

"This kind of technology can be used around the clock to analyze odor quality and can be operated by technicians who have much less training than those who operate gas-chromatography systems," says Aaron Kramer, project manager for Nordic Sensor Technologies AB, Linköping, Sweden. Electronic nose systems are also more affordable than most GC/MS systems.

With a modular design and plug-in sensor modules, the eNose 5000 from Neotronics Scientific provides RH control and data analysis.

Despite these benefits, electronic nose systems are not designed to replace standard analysis systems. "When the electronic nose detects a problem or quality deviation, researchers will still need to perform traditional chemistry methods with GC/MS to learn what is causing the problem," Kramer says.

The term "electronic nose" is a bit of a misnomer and has found varied acceptance within the field. Some companies, including Nordic Sensor Technologies AB, are trying to abandon it because the technology does not simulate the exact function of a human nose. Others in the industry have generalized the term to pertain to all technologies that mimic the sense of smell through chemical fingerprinting, says Wayne Gagne, applications specialist with AromaScan, Hollis, N.H.

Most systems that can be categorized as electronic noses create chemical footprints by analyzing sample headspace. The gaseous release generated by odorous samples is collected from equally sized samples over a period of time, Gagne says. Once odors are analyzed by an electronic nose, the chemical footprints are placed in a library and used as a comparison for other samples.

"The machine doesn’t know what any odors are – you have to give it information to correlate chemical footprints with odors," Gagne says. "We are finding that these tools are particularly important for the polymer and food industries."

Relatively new technologies, electronic noses have only been on the market for about a decade. Because of this, companies have spent a great deal of time over recent years educating clients about the need for their products. Worldwide, about seven companies are active in the market, and about as many are working to develop the technology, Warburton says.

Some electronic noses can be combined with gas chromatographs to provide more exact descriptions of the chemicals that comprise specific odorants. Despite this enhanced capability, the instruments will not notice that odors are disagreeable to humans unless they are told which chemical footprints cause unpleasant smells.

Featuring a hybrid array of sensors, the NST emission analyzer from Nordic Sensor Technologies AB monitors the quality of a wide array of food and beverages.

More than 1,000 genes have been shown to produce receptor proteins that interpret odors in mammals, says Stuart Firestein, associate professor of biological sciences at Columbia Univ., New York, and an expert in human olfactory research. These olfactory genes account for almost 2% of the entire mammalian genetic code. With these receptors, humans can detect more than 10,000 odors. When these smells combine, they delete or accentuate the overall odor, creating unexpected results.

Because the process of interpreting odors is extremely complex, it is crucial to integrate human panels with analytical tools when researching new products. Until product odors are approved by human panels and chemically footprinted, electronic nose technologies cannot act as effective quality-control tools.

"It would be nice to have artificial noses recognize new odors immediately, but they just can’t do that. The human ability to smell is just too complex to be modeled at this point," Columbia’s Firestein says. "Humans can smell and have an opinion about things we’ve never smelled before, like new-car odor. We certainly haven’t been selected through evolution to sense that smell, yet we are still capable of it."

Engineers are currently working to equip electronic noses with this complicated human ability through improved sensors and neural nets. Despite this, human olfactory panels will not be replaced for a long time, Kramer says. Instead, the two techniques remain complementary methods.

Human olfactory panels are now augmented by a number of procedures and technologies that provide quantification and qualification of their results. Before opinions are registered, panelists are screened to eliminate the 10% of the population with hypersensitive or hyposensitive noses. Most companies provide panelists with day-long training courses to acquaint them with equipment and ensure that they are able to verbally communicate the types of smells they experience. Less than a dozen people are on each panel; their opinions on each smell are averaged to create a mean response.

Panels provide five variables regarding each sample: threshold, intensity, persistence, hedonic tone, and character descriptors. To determine the first three variables, researchers use olfactometers to quantify results.

Threshold value is the common denominator to most olfactory work, McGinley says. To determine this variable, each olfactometer introduces increasing concentrations of an odorous air sample into clean air until panelists detect a scent. For this test, three scent puffs are offered two with pure air and one with an odorous sample. Panelists are asked to pick the one with the scent. If the smell is too diluted to be noticed, the sample concentration is increased until panelists detect it.

From these readings, a dilution threshold is calculated in "odor units." For example, if it takes 200 parts of clean air to dilute the sample to threshold, it has 200 odor units. This measurement provides a comparison tool for researchers testing odor-reduction technology in a variety of fields.

"Livestock provides a unique application in that each farm has its own particular odor related to its feed stock and water purity," says Larry Jacobson, associate professor in the biosystems and agricultural engineering department at the Univ. of Minnesota, St. Paul. "We use olfactometers to do comparative work and determine if techniques for reducing manure odor are effective or not," Jacobson says.

To determine intensity, the second variable, olfactometers introduce an odorous gas as a baseline against which other odors are compared. "The human nose cannot do an objective and precise job to evaluate odor intensity, so we pair it with technology that releases n-butanol," says Richard Gilbert, president of Tecnovir Intl., Quebec, Canada.

Using n-butanol as a reference odor, the portable Tecnodor from Tecnovir International helps sensory panels determine odor intensity.

A reference-odorant gas, n-butanol provides a standard against which the intensity of other gases can be scaled. Tecnovir’s product provides users with varying levels of n-butanol, the sample of odorous gas, and pure air. From these, panelists determine when the intensities of n-butanol and the sample match. The equipment determines a parts-per-million n-butanol equivalency for the sample.

"Allowing panelists to breath pure air before each test keeps their noses from adjusting to the smell of the ambient sample," Gilbert says. From the panelists’ classification of whether the sample is more or less intense than the n-butanol, the instrument varies the amount of n-butanol that is released. Within about five attempts, the olfactometer tabulates the n-butanol equivalent.

Persistence, the third variable in olfactometry testing, is determined by measuring the intensity of the odor in relation to its concentration. The intensity of some odors will be more persistent and have a greater "hang time" in the air. Persistence is derived from both the odor units and the atmospheric conditions at the sample site.

Hedonic Tone indicates how well consumers accept odors. Purely subjective, this variable is graded on a numerical scale and is frequently combined with the intensity rating.

Character descriptors offer a general assessment of sample odor as compared to common smells like coffee or rotten eggs. This is often coupled with hedonic tone to determine consumer acceptance of product odors.

McGinley estimates that there are less than a half-dozen companies that have created olfactometers designed to assist human sensory panels. Despite the importance of creating an objective platform for these tools, a worldwide standard has not been created.

Since the 1980’s, several olfactometry standards have been submitted by countries including France, Germany, the Netherlands, and the United States. These standards apply to equipment and procedures used with human sensory panels and not electronic noses.

In 1998, a European standard encompassing 18 countries is expected to be released, according to Michael McGinley, industrial hygienist with St. Croix Sensory. The standard follows ISO guidelines and is also expected to be adopted by Australia. An American revision of the current ASTM standard (E679-91) has been under development since 1991 and is expected to be released soon.

The AC'SCENT International Olfactometer from St. Croix Sensory mixes pure air with odorous samples to help determine odor threshold.

Both sets of standards detail factors like the presentation flow rate and the quality of the dilution gas used. The European standard also provides quality-assurance performance criteria for laboratory equipment and panelists.

Globally, the European standard is the favorite contender to become the baseline for olfactometry testing, Michael McGinley says. Still, he says, the American standard has several important attributes that could qualify it to become the global standard. A final decision will not be made until both plans are evaluated around the world.

Laura Vandendorpe
Assistant Managing Editor
R&D Magazine
Cahners Publishing Company
Editorial, Executive, and Sales Office:
1350 E. Touhy Ave.
Des Plaines, IL   60018
847-635-8800
http://www.rdmag.com
Circulation Office:
8773 South Ridgeline Blvd.
Highlands Ranch, CO  80126-2329

St. Croix Sensory again wishes to thank Cahners Publishing Company for allowing us to reproduce this article here.  This article is the sole property of Cahners Publishing and R&D Magazine.  Any duplication or reprinting may only be done with their permission.