DOES ELECTRONIC NOSE ANALYSIS FIT FOR THE PETFOOD INDUSTRY?
*F. Cheli, A. Campagnoli, V. Bontempo
Department of Veterinary Sciences and Technologies for Food Safety, University of Milan,
Via Trentacoste, 2, 20134, Milan, Italy
*Corresponding author. E-mail: email@example.com
Prof. Federica Cheli, BSc, PhD, full professor in Animal Nutrition, Department of Veterinary
Sciences and Technologies for Food Safety, University of Milan, Milan, I. Head of the
Laboratory of Animal Feed of the Department (SINAL accreditation N? 0636).
Anna Campagnoli, MdV, PhD, research investigator in animal nutrition, Department of
Veterinary Sciences and Technologies for Food Safety, University of Milan, Milan, I.
Valentino Bontempo, DVM, PhD, full professor in Animal Nutrition, Department of
Veterinary Sciences and Technologies for Food Safety, University of Milan, Milan, I.
The increased focus and interest on pets' health and welfare make it essential that the petfood
complies to specifications ensuring good nutrition as well as prevention and treatment of cat
and dog diseases (Bontempo, 2005). The general pressure for safe and high quality petfood
and company policies for enhancing internal quality assurance programme have accelerated
the need in analytical control of petfood. In this context, there is an increasing need for rapid
new technologies and new applications for existing technologies for a more comprehensive
screening of petfood. The most recent technique based on the use of electronic nose may
represent a promising analytical approach by providing quantisation of quality and safety in
real time with the objectivity of an instrumental response (Cheli et al., 2007a). In the early
1990s commercial instruments became available on the market and immediate use in the
food industry became apparent, both as tools for rapid screening and quality control and as a
support for decision-making in the area of product quality.
Electronic nose - Odours are made of one up to several thousands chemical components generally light, small, polar and often hydrophobic. At the end of the 1980s, Gardner
introduced the term “electronic nose” for the first time and Todd and Persaud introduced the
concept of “artificial olfaction”. Gardner and Bartlett’s 1994 definition of an e-nose is
instructive: “An e-nose is an instrument which comprises an array of electronic chemical sensors with partial specificity and an appropriate pattern recognition system, capable of
recognizing simple or complex odours.” The electronic nose does not distinguish each
volatile substance, but express the global odour of a product. An electronic nose is composed
of an array of sensors, each sensor in the array is sensitive to a different range of chemicals, a
data pre-processor and a pattern recognition system (Figure 1). Sensor array formats interact
with different volatile molecules and provide an electronic signal that can be utilised
effectively as a fingerprint of the volatile molecules associated to the product (Gardner and
Bartlett, 1999; Zang, 2003). In order to enable rapid interpretation of the volatile pattern, the
data collected must be combined with a chemometric analysis based on multivariate analysis,
mainly exploratory and predictive methods, or artificial neural network (Jurs et al., 2000; Tothill and Magan, 2003). Once that the electronic nose has been trained and the appropriate data analysis method has been developed, the response can be obtained in real time.
Therefore compared with traditional odour analysis methods (sensory panel and gas chromatography), the advantages of the electronic nose are high sensitivity, easy sample preparation, user-friendliness, non-destructive operation, fast detection, lower cost and objectivity.
Electronic nose applications in petfood analysis – Currently the main applications of the
electronic nose technology are in the food industry with the aim of monitor freshness, onset of microbial spoilage or bioprocesses of food, determination of food authenticity (Zang, 2003). We can take advantage from electronic nose application in the food industry and foresee future analytical challenges in the petfood industry. A specific interesting application comes from researches carried on in order to evaluate the capability of the electronic nose in the characterization of animal protein sources in petfood. The electronic nose was able to detect a clear difference in volatile profile of petfood in the presence of proteins of different sources using PCA analysis. Satisfactory discrimination between presence and absence of animal proteins in commercial dog dry food was obtained. However, when high
concentrationS of different animal proteins co-occurr a non completely satisfactory discrimination and recognition of the protein source was achieved (Campagnoli et al., 2004, Cheli et al., 2007b). In view of these results, it could be suggested that application of the electronic nose in the petfood industry can provide an interesting approach for quality control and qualitative protein source characterization. However, in order to achieve a practical application, matrix interference must be solved and more defined statistical methods are necessary.
Another promising field of application of electronic nose in the petfood industry is in the analysis of fat quality. In petfood manufacturing, edible fat is included in the formulation as
an energy supplement, as well as a palatability enhancer. However, during petfood processing and storage, the added fat may be susceptible to oxidation (Lin et al., 1998). Oxidation of lipids is one common and frequently undesirable chemical change that can impact on flavour, aroma and nutritional quality and can compromise the odour and the palatability of the petfood. The selection of an optimum test is difficult due to the complexity of the chemical processes involved. In this context, the electronic nose may represent a promising analytical approach as it can provide an electronic signal that can be utilised effectively as a fingerprint of the off-flavour volatile molecules associated to fat spoilage. Results from application of electronic nose as fat quality control tool is reported in the food industry in order to evaluate raw material origin, quality compliance, storage time influence, and to detect the presence of contaminants (Yang et al., 2000; Guadarrama et al., 2001; Shen et al., 2001). Preliminary results on commercial dry dog food, containing 10-20% of fat content, analyzed by an electronic nose, indicate a significant correlation of the electronic nose response and the content of FFA (personal communication). These results suggest that the electronic nose may have a potential application in the evaluation of fat quality and control of petfood manufacturing process in order to guarantee petfood with a high palatability and nutritive value.
Mycotoxins still play a role and are a major concern for most petfood manufacturers internationally (Boermans and Leung, 2007). Mycotoxin contamination in pet food poses a serious health threat to pets. Cereal grains and nuts are used as ingredients in commercial pet food for companion animals. Cereal by-products may be diverted to animal feed even though they can contain mycotoxins at concentrations greater than raw cereals due to processing. Several mycotoxin outbreaks in commercial pet food have been reported in the past few years (Stenske et al., 2006). The underlined hypothesis, for the potential use of electronic nose in order to evaluate mould spoilage, is that the growth and the biochemical pattern of mycotoxin producing fungi induces nutritional losses, organoleptic deterioration, formation
of mycotoxins and off-flavours and therefore changes in the volatile compound composition
(Olssen et al., 2002). Volatiles can be used as taxonomic markers of mycotoxigenic and non-
mycotoxigenic fungi species (Magan and Evans, 2000; Sahgal et al., 2007). Applications of
electronic nose for rapid detection of fungal contamination are well documented (see for
review Cheli et al., 2008). Moreover, multivariate analysis for the extraction of additional
information from electronic nose data and evaluation of association of fungal content with
mycotoxins give promising results on the capability of this technique as a tool and model to
detect some mycotoxin class and partially quantify its level (Olssen et al., 2002; Tognon et
al., 2005; Cheli et al., 2007c; Dell’Orto et al., 2007).
A further promising application of electronic nose in the petfood industry is in the quality
control of packaging in order to evaluate packaging materials which could pose a quality
problem and support packaging choice. Results from applications in the food industry
indicate that discrimination of packaging materials by an electronic nose is possible
suggesting that this analytical approach could replace other more expensive and sophisticated
analytical techniques (Werlein, 2001; Van Deventer and Mallikarjunan, 2002).
Interesting applications of electronic nose alone or associated with electronic tongue could be
foreseen in the evaluation and standardisation of flavour types, prediction of flavour shelf life, ensuring correct level of flavour added to different formulation, evaluating palatability of raw
material and complete feeds when ingredients or additives with low acceptability, according
to peculiarity and preferences of different animal species, are incorporated.
Electronic nose may represent a promising “fit-to-purpose” analytical method to be routinely
used in the petfood industry. It is rapid, user-friendly, adaptable and therefore useful for real
time monitoring and control of petfood and industrial processes and for decision-making in the area of product quality. Every application needs a specific approach in order to optimize
reproducibility, calibration, significance, data interpretation, data base construction and
quality control transferability. Current major restrictions are high matrix dependence, lack of appropriate calibration material, need for increasing sensitivity, repeatability and optimization of the analysis and developing of suitable multivariate methods in order to develop robust models for quality control. The exploitation of electronic nose technology and development of prediction models in order to evaluate appropriate storage, manufacturing and packaging conditions, quantity and trace undesirable components are examples of electronic nose application which should improve rapidly. The future challenge of the electronic nose analysis is to develop an holistic approach in order to consider data from electronic nose and other artificial senses collectively (Huang et al., 2007; Figure 2). Multi sensor data fusion is an emerging technology to fuse data from multiple sensors in order to create so-called "sensing technology" designed to mimic human or animal sensing behaviour for the purposes of gathering information, increasing the amount of information extracted from a sample and enhancing accurate estimation of feed/food quality.
Petfood analysis: the general pressure for quality control and company
policies for enhancing internal quality assurance programme may take
K advantage from the use of real time analytical methods. E
Electronic nose: the electronic nose may represent a promising “fit-to-Y
purpose” analytical method for rapid monitoring and control of quality C
O and manufacturing processes, due to its high sensitivity, easy sample
preparation, user-friendliness, non-destructive operation, fast detection, C
E lower cost and objectivity.
Applications of the electronic nose for the petfood industry: the T
S exploitation of electronic nose technology and development of prediction
models should improve rapidly in order to evaluate petfood quality,
appropriate storage, manufacturing and packaging conditions, quantity and
trace undesirable components.
Figure 1. Electronic nose: the processing process
Samplesin Samplesin Data interpretation:Data interpretation:SmellSmellResponsedataResponsedatagaseousphasegaseousphasepattern recognitionpattern recognitionmeasurementmeasurement
SAMPLING AREASSAMPLING AREASSAMPLING AREASResults:Results:
ELECTRONICSELECTRONICSELECTRONICS-QUANTIFICATION-QUANTIFICATION-Control and referencing-Control and referencing-Control and referencing
Signal= Signal= Signal= f (identity, f (identity, f (identity,
Figure 2. The future challenge of artificial senses: multisensory data fusion for characterization of
food quality and safety. (modified from Huang et al., 2007).
Artificialhead Artificialhead featurefeatureFeatureFeatureSmellSmellSmellSmell
FeatureFeatureTaste Taste Taste Taste ArtificialmouthArtificialmouthNeuralNeuralextractionextractionsensorssensorssensorssensorsfeaturefeaturenetwork network
SimplefeatureSimplefeatureFeatureFeatureForce Force Force Force
1) Boermans, HJ, Leung, MCK, 2007. Mycotoxins and the pet food industry: Toxicological
evidence and risk assessment. Int J Food Micr 119: 95–102.
2) Bontempo, V, 2005. Nutrition and Health of Dogs and Cats: Evolution of Petfood. Vet
Res Comm 29(Suppl. 2): 45–50.
3) Campagnoli, A, Pinotti, L, Tognon, G, Cheli, F, Baldi, A, Dell'Orto, V, 2004. Potential
application of electronic nose in processed animal proteins (PAP) detection in feedstuffs.
BASE 8: 253-255.
4) Cheli, F, Campagnoli, A, Pinotti, L, Maggioni, L, Savoini, G, Dell’Orto, V, 2007a.
Testing feed quality: the “artificial senses”. FEED INTERNATIONAL: May/June, 24-26. 5) Cheli, F, Campagnoli, A, Pinotti, L, D’Ambrosio, F, Crotti, A, 2007b. Electronic nose
application in animal protein source characterisation in petfood Proceedings of LXI
Annual Meeting of the Italian Society for Veterinary Sciences –Salsomaggiore Terme –
Parma, Italy, September 26-29: 431-432..
6) Cheli, F, Pinotti, L, Campagnoli, A, Fusi, E, Rebucci, R, Baldi, A. 2008. Mycotoxin
analysis, mycotoxin-producing fungi assays and mycotoxin toxicity bioassays in food
mycotoxin monitoring and surveillance. Ital J Food Sci 20: in press.
7) Cheli, F, Savoini, G, Campagnoli, A, Dell’Orto, V, 2007c. Electronic nose as fit-for-
purpose approach applied to aflatoxin determination in Zea mais. Miraglia M, Brera C
(Ed.). Rapporti ISTISAN 07/37, Roma, I: 253-256.
8) Dell’Orto, V, Savoini, G, Nichilo, A, Campagnoli, A, Cheli, F. 2007. Electronic nose
combined with chemometric techniques for evaluation of deoxinivalenol contamination
in Triticum durum. Miraglia M, Brera C (Ed.). Rapporti ISTISAN 07/37, Roma, I: 207-
9) Gardner, JW, Bartlett, PN, 1994. A brief history of electronic noses. Sens Actuators B
Chem 18: 211-220.
10) Gardner, JW, Bartlett, PN, 1999. Electronic noses: Principles and applications. Oxford,
Oxford University Press.
11) Guadarrama, A, Rodriguez-Mendez, ML, Sanz, C, Rios, JL, de Saja, JA, 2001. Electronic
nose based on conducting polymers for the quality control of the oil aroma –
discrimination of quality, variety of olive and geographic origin. Anal Chim Acta 432:
12) Werlein, H.D, 2001. Discrimination of chocolates and packaging materials by an
electronic nose. Eur Food Res Technol 212 :529–533.
13) Jurs, PC, Bakken, GA, Mcclelland, HE, 2000. Computational methods for the analysis of
chemical sensor array data from volatile analytes. Chem Rev 100: 2649-2678.
14) Huang, Y, Lan, Y, Hoffmann, WC, Lacey, RE, 2007. Multisensor data fusion for high
quality data analysis and processing in measurement and instrumentation. J Bionic Eng 4:
15) Lin, S, Hsieh, F, Huff, HE, 1998. Effects of lipids and processing conditions on lipid
oxidation of extruded dry pet food during storage. Anim Feed Sci Technol 71: 283–294. 16) Magan, N, Evans, P, 2000. Volatiles as an indicator of fungal activity and differentiation
between species, and the potential use of electronic nose technology for early detection of
grain spoilage. J Stored Prod 36, 319-340.
17) Olssen, J, Borjesson, T, Lundstedt, T, Schnuerer, J, 2002. Detection and quantification of
ochratoxin and deoxinivalenol in barley grain by GC-MS and electronic nose. Int J Food
18) Sahgal, N, Needeman, R, Cabanes, FJ, Magan, N, 2007. Potential for detection and
discrimination between mycotoxigenic and non-toxigenic moulds using volatile
production pattern: A review. Food Addit Contam 24:1161-1170.