Electronic noses and tongues are being increasingly used in the food and beverage industries. First proposed for the detection and classification of odors in 1982, they have now reached the point where they are even visible to the ordinary consumer via inclusion in such television programs as CSI: Crime Scene Investigation, where forensic scientists are using them to help solve crimes.
E-noses and e-tongues are being used in the laboratory to develop new foods, beverages, and pharmaceuticals and in production plants to monitor the quality of products being produced. They help scientists develop new ingredients to improve the aroma and flavor of products, as well as ingredients to mask undesirable aromas. They are of value in determining and monitoring the shelf life of products and in ensuring that packaging materials do not contribute off-odors or off-flavors to products. In addition to determining shelf life, they are being employed to gauge the best time to harvest fruits and vegetables and the optimal length of time to age beer to develop the desirable aroma. They are also being used in medical diagnostics and other applications.
One additional area that is being pursued is food safety, namely, the detection of pathogens in foods and beverages. This is a relatively new area of application of e-noses and e-tongues, and faces a number of hurdles.
How E-Noses Work
E-noses and e-tongues are designed to mimic the way the human nose and tongue respond to odors and tastes. E-noses detect and measure volatile organic compounds (VOCs) whether they have an odor or not, whereas e-tongues test liquid samples or solids dissolved in liquids. Both types of instrument generate results correlated with human evaluation via chemometrics software.
Four basic types of gas sensors are used in commercial e-noses: quartz crystal microbalances (QCMs), metal oxide sensors, conducting polymer sensors, and surface acoustic wave (SAW) sensors. The sensors detect specific chemicals present in the headspace gases and produce changes in electrical resistance or, in the case of QCMs and SAWs, oscillation of the crystals.
Research is being conducted at various universities and companies around the world to develop new and improved sensors, as well as smaller, portable instruments. Some work is being done in the area of nanotechnology.
--- PAGE BREAK ---
The French company Alpha M.O.S. (www.alpha-mos.com) —whose U.S. headquarters is located in Hanover, Md.—was the first company to enter the market with an e-nose and is the largest supplier of e-noses and e-tongues. The company introduced its e-nose in 1993 and its e-tongue in 2000. Since those introductions, the company has made several improvements to optimize the instruments’ performance.
The company has developed several new headspace-sampling techniques, including autosamplers, solid-phase micro extraction (SPME), solid-phase dynamic extraction (SPDE), thermo-desorption devices, and also blenders to homogenize liquid or solid samples. In addition, a specific set of e-tongue sensors particularly sensitive to the five basic tastes has been developed for food and beverage applications.
Alpha M.O.S. offers several models of e-noses and e-tongues using various technologies for use in quality control and product development. The Fox and Gemini e-noses use metal oxide sensors, the Kronos e-nose uses fingerprint mass spectrometry, and the Prometheus combines metal oxide sensors and fingerprint mass spectrometry. The Heracles e-nose is based on ultra-fast gas chromatography (analysis in less than 60 sec). The Astree e-tongue uses chemically modified field effect transistor (CHEMFET) liquid sensors. All models deliver results and decision tools within minutes.
Electronic Sensor Technology (www.estcal.com), Newbury Park, Calif., developed its zNose® electronic sensor instrument in 1995. It uses a SAW sensor and ultra-fast gas chromatography to analyze and identify VOCs in less than 10 sec with parts-per-billion sensitivity. Models 4200 and 4300 are portable real-time analyzers that can detect and analyze all types of vapors and identify traces of organic, biological, and chemical compounds quickly and accurately. Model 7100 can identify vapors as low as parts-per-trillion concentration and amounts less than 50 picograms.
The New Zealand company Syft Technologies Ltd. (www.syft.com) —whose U.S. headquarters is located in Pittsburgh, Pa.—has introduced the Voice200® instrument, which instantaneously identifies and quantifies VOCs in complex gas mixtures in real-time via selective ion flow tube mass spectrometry (SIFT-MS). Because of its high sensitivity and multiple reagent ion technology, the instrument is said to be ideal for flavor and aroma detection, flavor release analysis, and online production process control. Samples obtained by direct-headspace, in-mouth, and retro-nasal techniques can be analyzed. The instrument can detect and measure VOCs within seconds, with no sample preparation required, at low parts per trillion levels. It instantly detects off-odors and fingerprints living organisms, such as those causing food safety and quality issues.
Application to Pathogens
I asked several e-nose researchers their views bout applying e-noses to detection of pathogens.
Kumar Mallikarjunan ([email protected]), Associate Professor, Biological Systems Engineering, Virginia Polytechnic Institute & State University, Blacksburg, is skeptical about using the e-nose for detecting food pathogens quantitatively.
--- PAGE BREAK ---
He is using four types of sensors to determine which will work for specific applications. They are nonselective, like the human nose. Humans can tell if a smell exceeds a certain threshold, he said, but can’t tell quantitatively what the level in parts per million of a specific volatile chemical is. The same applies to pathogens or spoilage, he said. Nevertheless, the e-nose can be used for screening as a first line of defense, giving a yes/no answer.
During spoilage, he said, microbes grow and produce metabolites, which cause off-odors. But some microorganisms, such as Escherichia coli, don’t give off an odor. However, he added, we can detect them by understanding how their metabolites are being released. He has been studying use of e-noses to detect Salmonella on chicken and plans to expand his work to other pathogens later.
Suranjan Panigrahi ([email protected]), Professor, Bio-imaging and Sensing Center, Agricultural and Biosystems Engineering, North Dakota State University, Fargo, has been working on use of e-noses for detecting Salmonella contamination in packaged meat. It’s a complex problem, he said, especially with respect to the dynamics of the biological systems involved—bacteria, a biological system, growing on meat, another biological system. In food applications, sampling is important, he said. If you sample a food product at location A and the pathogen is at location B, you will not find it unless you grind the sample.x To avoid this, he is focusing on developing portable intelligent sensors that can determine spoilage and pathogen contamination on intact packaged food products. He is using growth of Salmonella on packaged beef cuts as his model and an integrated sensing system that he terms an intelligent artificial nose (IAN) with various types of sensors. Every sensor technology has its advantages and disadvantages, he said, so he is combining the advantages of one technology with advantages of others.
Murat O. Balaban ([email protected]), Professor and Director of Fishery Industrial Technology Center in Kodiak, University of Alaska, Fairbanks, said that the use of electronic noses to detect food spoilage has been well demonstrated, but detection of pathogens is inherently more difficult. There may not be enough volatiles generated to reveal the pathogen’s presence, or not enough time or number of organisms—it doesn’t require a whole lot of microorganisms to do damage, he said. There might not be enough volatiles, or they might be masked by other volatiles.
Detection of pathogens by e-noses is an exciting area, he said, but it requires more research. To be able to detect them, we need generation of unique volatiles to provide a unique fingerprint, and they must be distinguishable from the background. Generally, spoilage comes first and generates a lot of volatiles. Most sensors are nonspecific, responding to all volatiles. That’s where the difference arises with pathogens. Evidence may be small and masked. If the pathogen generates a unique chemical and we concentrate on that, he concluded, it may be a better approach.
--- PAGE BREAK ---
Research on improvements in sensor technology around the world will lead to improved e-noses and e-tongues and additional applications in the food and beverage industries—and, most likely, more applications in food quality and safety.
by Neil H.
Mermelstein, a Fellow of IFT, is Editor Emeritus of Food Technology