Poster Abstracts

International Symposium on

Olfaction and Electronic Nose (ISOEN)

May 2-5, 2011

Rockefeller University 1230 York Avenue, New York, NY 10065

Chair Perena Gouma

SUNY at Stony Brook, USA

Engineering Conferences International 32 Broadway, Suite 314

New York, NY 10004, USA Phone: 1 - 212 - 514 - 6760, Fax: 1 - 212 - 514 - 6030

www.engconfintl.org – info@engconfintl.org

Exhaled breath analysis for the monitoring of elderly COPD patients health-state

Giorgio Pennazza, University 'Campus Bio-Medico di Roma', Italy

g.pennazza@unicampus.it Simone Scarlata, Italy Geriatrics Department, University 'Campus Bio-Medico di Roma', Italy

s.scarlata@unicampus.it Marco Santonico, Department of Electronic Engineering, University of Rome ‘Tor Vergata’, Italy

antonico@ing.uniroma2.it Domenica ChiurcoGeriatrics Department, University 'Campus Bio-Medico di Roma', Italy

d.chiurco@unicampus.it Arnaldo D'AmicoItaly, Department of Electronic Engineering, University of Rome ‘Tor Vergata’

damico@eln.uniroma2.it Raffaele Antonelli Incalzi, Geriatrics Department, University 'Campus Bio-Medico di Roma', Italy

r.antonelli@unicampus.it Health state needs to be frequently monitored in elderly people. The potential role of breath analysis as a summary index of health status deserves consideration. Breath analysis has shown classificatory and discriminative properties in respiratory diseases , , , but respiratory function and health status correlates of breath analysis have not been thoroughly assessed. Furthermore, there is insufficient information on breath analysis reproducibility. This pilot study assesses how effectively a gas sensors array can follow the evolution of elderly patients with COPD, the most common chronic respiratory disease. The main result refers to the reproducibility of breath analysis compared to spirometry: the values of the variance of gas sensor array and spirometry, calculated for each subject along three measurements (time period of three weeks), were comparable except for a larger spread for breath analysis. Gas sensors array patterns was significantly correlated with other heath status parameters, mainly with eosinophiles (a proxy for allergic state diagnosis), and Barthel index (a mobility and independence index). This highly interesting correlations deserve a larger confirmatory study.

Monitoring the halitosis with an electronic nose

Enrico Marchetti, Department of Surgical Sciences, University of L’Aquila, Italy enrico.marchetti@cc.univaq.it

Giorgio Pennazza, Faculty of Engineering, University 'Campus Bio-Medico di Roma, Italy ' g.pennazza@unicampus.it

Marco Santonico, Department of electronic Engineering, University of Rome "Tor Vergata", Italy santonico@ing.uniroma2.it

Rosamaria Capuano, Department of electronic Engineering, University of Rome "Tor Vergata", Italy capuano@ing.uniroma2.it

Stefano Mummolo, Department of Surgical Sciences, University of L’Aquila, Italy stefano.mummolo@cc.univaq.it

Giuseppe Marzo,Department of Surgical Sciences, University of L’Aquila, Italy giuseppe.marzo@cc.univaq.it

Corrado Di NataleDepartment of electronic Engineering, University of Rome "Tor Vergata", Italy dinatale@uniroma2.it

Halitosis condition can be revealed by the presence of peculiar Volatile Organic Compounds (VOCs) in the exhaled breath; besides, exhaled breath composition gives rise to a distinctive volatile fingerprint. These two aspect allow halitosis diagnosis following two approaches: a quantitative one, based on the measure of certain VOCs concentration by mean of selective sensors ; a qualitative analysis, based on the discrimination of a peculiar volatile fingerprint by mean of an array of metalloporphyrins coated quartz microbalances . This electronic nose is developed at the University of Rome Tor Vergata, and it has been already demonstrated to be sufficiently sensitive and selective to discriminate between different conditions of halitosis . With the respect to the past tests, this study is centered on the characterization of the time evolution of halitosis conditions in 15 individuals along a period of 4 weeks. In particular, the tests were aimed at studying the effect of two different treatments. Professional operators assessed oral odor, and their evaluation was used for classifier training. Results show that the electronic nose can identify the presence of oral malodor and the attenuation of the condition achieved by the application of the treatment.

Hand held numeric prototype for breath analyzing sensor

Aditya Shyam Ambre, State University of New York at Stony Brook J 2146 Cx, 700 Health Sciences Drive, Stony Brook, New York, 11790, United States of America

T: 631-796-1644, F: 631-632-8052, aditya67ambre@gmail.com Perena Gouma, State University of New York at Stony Brook

Milutin Stanacevic, State University of New York at Stony Brook

This paper describes a hand held numeric prototype device that will incorporate resistive sensor arrays for breath analysis diagnostics. The sensor array monitors three specific gases: isoprene, ammonia, and carbon dioxide. The challenge is to continuously track the sensor responses. The device has the advance peripheral interface controller with internal ADCs and LCD driver. An external DAC is used for controlling the amplification of the sensors to have a controlled output on LCD. Changes in the sensor’s resistance due to the presence of a gas are used in an inverting amplifier configuration for conversion from resistance to voltage to make it compatible with the ADC. The feedback resistance for OP-AMP is selected in such a way that the required feedback factor is achieved for which we have inserted resistances in parallel fashion. The DAC is used to control the flow of the current through the sensor that will in turn control the amplification. PIC provides the digital output to DAC via a PWM signal on serial data line. The feedback obtained maintains a stable configuration for the sensors as it gives the real time information. The amplified voltage is then given to the DACx port of the PIC. Once the results are obtained the program copies it into LCD registers for display purposes. The display is a 16*2 LCD display of the readings of all the three sensors. The LCD is configured with the help of three control lines. The PIC is programmed with the help of the programmer kit. The program takes care of the timing constraints with the help of the clock provided by the 8MHz clock circuit. Our design is unique in terms of the sensor array, the DAC controlled feedback mechanism, detailed display of the sensors.

An electronic nose distinguishes the exhaled breath condensates obtained by two different devices and two different breath patterns

Andreas Rembert Koczulla, Department for Internal Medicine, Division for Pulmonary Diseases, Philipps-

Universität Marburg, 35043 Marburg, Germany, koczulla@med.uni-marburg.de Sarah Noeske, Department for Internal Medicine, Division for Pulmonary Diseases, Philipps-Universität

Marburg, 35043 Marburg, Germany, noeske@med.uni-marburg.de Akira Hattesohl Germany, Department for Internal Medicine, Division for Pulmonary Diseases, Philipps-

Universität Marburg, 35043 Marburg, Germany, hattesohl@staff.uni-marburg.de Rudolf Joerres, Institute for Occupational, Social and Environmental Medicine, Ludwig-Maximilians-

Universität Munich, 80336 Munich, Germany, Rudolf.Joerres@med.uni-muenchen.de Timm Greulich, Department for Internal Medicine, Division for Pulmonary Diseases, Philipps-Universität

Marburg, 35043 Marburg, Germany, greulich@med.uni-marburg.de Claus Franzl Vogelmeier, Department for Internal Medicine, Division for Pulmonary Diseases, Philipps-

Universität Marburg, 35043 Marburg, Germany, vogelemeier@med.uni-marburg.de

Measurement of exhaled breath condensate (EBC) is a non invasive and simple method to investigate parameters of the epithelial lining fluid of the lung. Collection of EBC is easy to perform, inexpensive and reproducible1. EBC measurements can be performed with a variety of commercial available EBC collection devices like RTube? (Respiratory Research, Inc.,Charlottesville, VA, USA) and ECoScreen turbo? (VIASYS Healthcare, Germany). Advantages of the RTube? are portability, low weight and the simple way of using it. The ECoScreen turbo? is characterized by a constantly cooling system of approximately -5 °C. Exhaled respiratory air contains countless volatile organic compounds (VOCs). The composition of those VOCs varies dependent of different pulmonary diseases. With the customizable electronic nose Cyranose 320? (C-320) (Smith Detection, USA) it is possible to detect VOCs and distinguish between different VOC patterns2. We could clearly separate between tidal and forced breathing patterns MD:2.210 (LDA: 64 %) and the two different devices Rtube and EcoScreen MD: 2,392 (LDA: 74 %). Our approach shows that the device and the breathing pattern has influence on the detectable VOC patterns. Device and breathing pattern should be mentioned in experimental set ups.

An electronic nose based on hybrid MOS-SAW sensors for detection of different biomarkers of lung cancer

Di Wang, Zhejiang University, China,wd85@zju.edu.cn

Le Wang, Zhejiang University, China,wangle@zju.edu.cn Kai Yu, Zhejiang University, China,toshiba@zju.edu.cn

Yishan Wang, Zhejiang University, China,YishanWang@zju.edu.cn Ping Wang, Zhejiang University, China,cnpwang@zju.edu.cn

Jin Yu, Zhejiang University, China,dgnkyj@gmail.com In the application of electronic nose in detection of lung cancer, many biomarkers need to be detected, including VOCs of both low molecular weight and high molecular weight. But a single kind of gas sensors, such as metal oxide semiconductor (MOS) sensor and surface acoustic wave (SAW) sensor, can not be sensitive to all the biomarkers, so the electronic nose based on single kind of sensors can not obtain all usefull information and give a reliably detection result. In this paper, we proposed an electronic nose based on MOS and SAW gas sensors. The MOS gas sensors are used to detect the VOCs whose molecular weights are lower than 150, e.g. benzene, and the SAW sensors are adopted for the detection of VOCs, whose molecular weights are higher than 150, e.g. tridecane and tridecanone. This novel electronic nose provides a better precision in the diagnosis of lung cancer than the system adopting single detecting method we developed before.

Multi-model diagnosis method for lung cancer based on MOS-SAW breath detecting e-nose

Yishan Wang, Biosensor National Special Lab, China, wys513.1028@gmail.com Kai Yu, Biosensor National Special Lab, China, andrew.yukai@gmail.com

Di Wang, Biosensor National Special Lab, China, wd85@zju.edu.cn Le Wang, Biosensor National Special Lab, China, wangle686@msn.com

Jin Yu, Biosensor National Special Lab, China, dgnkj@gmail.com Ping Wang, Biosensor National Special Lab, China, cnpwang@zju.edu.cn

Some of the volatile organic compounds (VOCs) in the exhaled breath of lung cancer patients are considered as the biomarkers of lung cancer published in recent years. MOS-SAW breath detecting e-Nose based on this theory combines the metal oxide semiconductor (MOS) sensors and the SAW sensor, which compose the detecting units of the system. In this paper, we introduce a multi-model diagnosis method, which is used to process the signals of the MOS-SAW e-Nose, to establish the diagnosis models, and to detect unknown samples. We adopted four pattern recognition algorithms: principle components analysis (PCA), linear discriminant analysis (LDA), artificial neural network (ANN), and partial least squares (PLS). The diagnosis models were established through these four algorithms. And then the best model was selected to detect unknown samples. Here we established six models and used leave-one-out cross-validation to test these models.

Portable gas sensor for breath analysis

Marco Righettoni, ETH Zurich, Switzerland, righettoni@ptl.mavt.ethz.ch Antonio Tricoli, ETH Zurich, Switzerland, tricoli@ptl.mavt.ethz.ch Samuel Gass, ETH Zurich, Switzerland, gasss@student.ethz.ch

Sotiris E. Pratsinis, ETH Zurich, Switzerland, sotiris.pratsinis@ptl.mavt.ethz.ch The detection of acetone in the human breath is a promising method for the noninvasive diagnosis and monitoring of diabetes. Here, a portable and highly selective acetone-sensor was developed by direct flame synthesis and deposition of a chemo-resistive nanostructured SiO2-doped WO3 film on back-heated interdigitated Pt electrodes on an alumina substrate inside a T-shape chamber of similar size to a mobile telephone. This device enables real-time measurements of ultra low (≤ 20 ppb) acetone concentrations at realistic breath conditions (90% relative humidity). The sensing properties (selectivity, detection limit, sensor response and recovery times) were investigated as a function of operating temperature, relative humidity, interfering analyte concentration (ethanol) and inlet flow-rate. The optimized device resulted in short (10 - 20 s) total response and recovery times. Furthermore, the sensor response was robust against changes in the inlet volumetric flow-rate.

Continuous exhaled breath analysis on the ICU: Feasibility study

Lieuwe DJ Bos, Academic Medical Center, Amsterdam, Netherlands, l.d.bos@amc.uva.nl Marcus J Schultz, Academic Medical Center, Amsterdam, Netherlands,m.j.schultz@amc.uva.nl

Peter J Sterk, Academic Medical Center, Amsterdam, Netherlands,p.j.sterk@amc.uva.nl Summary: During admittance to the intensive care unit (ICU), critically ill patients frequently develop secondary infections and/or multiple organ failure, leading to increased morbidity and mortality. Discontinuous, infrequent diagnostic assessments in highly dynamic critically ill patients create a gap between the time of development of complications and the time of diagnosis. This gap can be closed by increasing the measurement frequency to a near continuum. Although very much needed, such tests are scarce. An eNose (Cyranose) has been used previously in intubated and mechanically patients, although only in a single assessment (1). Walsh et al. found continuous pH measurement in exhaled breath condensate on a neonatal ICU to be feasible (2), however expensive. We hypothesized that an electronic nose (eNose) provides a continuous read-out of (changes in) patterns of exhaled volatile organic compounds (breathprints) in intubated and mechanically ventilated ICU patients, allowing non-invasive monitoring of the pulmonary metabolic status of the patient. We conclude that continuous monitoring of exhaled breath by electronic nose is feasible in the intensive care setting. Our data suggest that within-patient changes in breathprints can be used to assess and/or predict the clinical course of the patients. Methods and Results: This was a preliminary observational study in 6 critically ill, intubated and mechanically ventilated patients (1 subarachnoid hemorrhage, 1 sepsis, 1 pneumonia, 3 cardiopulmonary surgery). Breathprints were collected by means of an eNose every 10 seconds for approximately 7 hours. The attending nurse and researcher observed each patient during the sample period. The sample size was too small for statistical analysis between patients, but varying fluctuations were observed within patients. The time to stabilization of the breathprint after the start of the measurement is approximately 7 minutes. Breathprints fluctuated considerably over time (SEM 1.18). However, typical changes could be detected: e.g., salbutamol inhalation, decreased static compliance and increased minute volumes all caused a rapid change in the breathprints (illustrated in figure).

Nanosensor array-based breath analyzer for disease diagnosis

Perena Gouma, Stony Brook University 314 Old Eng. Bldg, Stony Brook, NY, 11794-2275, USA

T: 631-632-4537, F: 631-632-8052, pgouma@notes.cc.sunysb.edu Arun Prasad, Stony Brook University

Milutin Stanacevic, Stony Brook University

A novel concept of a three-nanosensor array microsystem that may serve as a coarse handheld diagnostic tool breath analyzer to provide a first detection device for disease diagnosis is described in this work. The specification and performance of a simple metal oxide nanosensor operating between three distinct temperatures are discussed, focusing on the need for a non-invasive blood cholesterol monitor. The sensor array in this study involves MoO3-based nanostructured thin films operating at 420°C, at 450°C, and at 500°C, respectively. Variations in the film’s thickness appear to play a role in the actual sensor sensitivity, but not in the relative selectivity. The sensor operating at 420°C is a promising isoprene detector. These sensors behave electrically as resistors and therefore a specialized multi-channel instrumentation is required to obtain readouts. The electrical resistance of each of the sensors in array is being composed of a combination of two series resistances. Interfacing the sensor array to an integrated circuit for electrical readout and temperature control provides for a complete microsystem capable to capturing a single exhaled breath and analyzing it with respect to its relative content of isoprene, carbon dioxide, and ammonia gas. This inexpensive sensor technology may be used as a personalized medical diagnostics tool in the near future.

A supervised feature extraction method for GCMS data based on PLS: Application to the detection of adulterated olive oil

Cosmin Burian, Departament de Enginyeria Electrònica, Elèctrica I Automatica, Universitat Rovira i Virgili,

Tarragona, Spain, cosminburian@yahoo.com Jesus Brezmes, Departament de Enginyeria Electrònica, Elèctrica I Automatica, Universitat Rovira i Virgili,

Tarragona, Spain, jesus.brezmes@urv.cat Eugenio Martinelli, Department of Electronic Engineering, University of Rome Tor Vergata, Rome, Italy,

martinelli@ing.uniroma2.it Corrado Di Natale Department of Electronic Engineering, University of Rome Tor Vergata, Rome, Italy,

dinatale@uniroma2.it The high demand for extra virgin olive oil by consumers and its related high price has turned olive oil adulteration as one of the biggest agricultural frauds in Europe. The adulteration is done diluting the extra virgin oils with cheaper oils either from olives or other vegetable crops. The detection of adulteration is often complicated and more than one test is necessary to determine the olive oil authenticity and the identity of the adulterant. This is the case of hazelnut oil, one of the most diffused adulterant, whose physical and chemical features are very close to those of olive oil . In this work, a method to identify olive oil adulteration is presented. It is based on GC-MS analysis coupled with multivariate data analysis and a features selection step. A comparison of the results in term of identification performance of the adulterated samples is provided for different data analysis techniques such as Partial Least square Discriminant Analysis, SIMCA , KNN and LDA .

Portable electronic nose to discriminate artificial aged wine from barrel-aged wine

Jose Pedro Santos, Instituto Física Aplicada. Consejo Superior de Investigaciones Científicas, Spain josepe@ifa.cetef.csic.es

Juan Mariano, Cabellos Departamento de agroalimentación. IMIDRA, Spain Teresa Arroyo, Departamento de agroalimentación. IMIDRA, Spain

Maria Carmen Horrillo, Instituto Física Aplicada. Consejo Superior de Investigaciones Científicas, Spain

A novel portable electronic nose is presented as a fast tool to differentiate traditional aged wines from artificial aged wines. Aging method discrimination is important due to the increasing production in artificial

aged wines with oak chips, a practice recently approved by the European Commission, which has remarkable economic implications. Alternative methods to classical chemical analysis are of great interest. Few of such studies have been made about this subject and none with electronic noses, for example with

an electronic tongue. The instrument we have developed for testing the wines is a general purpose wireless portable electronic nose base on micromechanized tin oxide sensors. This system differentiates

correctly the wine samples obtained with the different aging methods

Using a multi-way analysis for the application of an electronic noses in wine quality control

N. Prieto, Research Academic Staff, Spain, nprietoc@fmc.uva.es J. A. De Saja, professor, Spain, sajasaez@fmc.uva.es

R. Leardi, professor, Italy, riclea@dictfa.unige.it M. L. Rodríguez-Méndez, professor, Spain, mluz@eis.uva.es

Summary: In this work a multi-way data treatment (Tucker3) method has been applied to optimize the performance of an electronic nose for assessing the ageing of red wines. Using this method the data obtained from GC-Ms and Uv-Vis have been correlated with e-nose results in order to increase the relevant chemical or physical information captured and to represent a sensory profile of the wines analysed. The data modelling has demostrated the capability of the e-nose to follow the wine ageing process and to indicate the variables which have a major role for the discrimination between wines according to their ageing (barrel or stainless steel tanks with addition of small oak wood pieces) and the geographic origin of the oak wood (american or french). Methods: Wine samples were prepared in stainless steel tanks (100 liters). Ageing was carried out following two different procedures, a first set of samples was aged in oak barrels using french or oak american oak wood. A second set samples were aged following an alternative method by adding pieces of oak wood (chips and staves from american or french origin) to stainless steel tanks. The analytical evaluation was performed using an array of MOX sensors selected according to the previous experience of our group. The GC-MS analysis were carried out with a Clarus 500 GC with a mass detector Clarus 560 Perkin Elmer and spectroscopy analysis using a UV-2101PC shimadzu. The data set was analysed in Matlab with a program developed by R. Leardi and the N-way Toolbox (R.Bro). Results: The analysis of raw data have showed some trends to discriminate between the wines aged in barrel and those were treated with oak chips (figure 1, 2). In orden to explain the influence on sensory characteristics on depth it was necessary to apply a more complex model For this purpose, the dataset was pretreated to remove the differences among the variables (j-scaling) and to remove the differences among the sampling times (jk-scaling). In order to choose the optimal Tucker3 model with the lowest number of factors and good fit a test routine was applied. A model [1 2 2] with a 69,8% of the explained variance for the j-scaling pretreatment and a model [2 2 2] (fig.3) with 39,7% of the explained variance for jk-scaling have been chosen. The model interpration is based on loadings plots. Fig. 4 shows that the system is able to detect the use of an alternative ageing system. The correlations found with the compounds quantified by GC-MS (see for instance the whiskeylactone and eugenol levels during wine ageing process) are registred in fig. 5, 6 shown that the result obtained with the e-nose are complementary to GC and UV-Visible.

Aroma analysis by GC/MS and electronic nose dedicated to Negroamaro and Primitivo typical Apulian wines

Maria Tufariello, Italy, IMM-CNR, tufariello@le.imm.cnr.it

Simonetta Capone, Italy, IMM-CNR, simona.capone@le.imm.cnr.itAlessandro Leone, Italy, IMM-CNR, alessandro.leone@le.imm.cnr.it

Luca Francioso, Italy, IMM-CNR, luca.francioso@le.imm.cnr.it Pietro Siciliano, Italy, IMM-CNR, pietro.siciliano@le.imm.cnr.it

Volatile composition of monovarietal young red wines made from Negroamaro and Primitivo cultivars, two autochthonous grape varieties of Vitis vinifera grown exclusively in Apulia region in southeast of Italy, was investigated. Volatile compounds were extracted with Solid Phase Extraction (SPE) method, and then analysed by Gas Chromatography–Mass Spectrometry (GC/MS). This data was used to support the wine analysis carried out on the same wine samples by an Electronic Nose (EN) based on chemoresistive gas microsensors. The EN device was also calibrated to detect some wine defects, i.e. some negative markers of quality when exceeding a concentration threshold. This activity is part of a Strategic Project of the research founding program of Apulia Region government: “Innovative biotechnological approaches to improve quality and safety of typical Apulian wines” – acronym INNOWINE.

Olive oil headspace characterization by a gas sensor array

Marco Santonico, Univrersity of Rome Tor Vergata, Italy, santonico@ing.uniroma2.it

Giacomo Gianni, Camera di Commercio di Firenze, Italy, giacomo.gianni@fi.camcom.it

Marzia Migliorini, Camera di Commercio di Firenze, Italy, marzia.migliorini@fi.camcom.it

Corrado Di Natale, Univrersity of Rome Tor Vergata, Italy, dinatale@uniroma2.it

Artificial olfactory systems are attractive tools for the appraisal of the quality of many different foodstuffs whose features can be assessed by the analysis of the volatile compounds profile. An interesting case, is the olive oil where the volatile compounds reflect the sensory properties and are modified in presence of defects. These properties make both sensorial analysis and gas chromatography the most used methods for quality controls. Although efficient, panels and GC are scarcely practice because of the subjectivity of the evaluation and the complexity of the measurement respectively. There are several examples of applications of electronic noses for olive oil defects identification; here we investigate the capability of such an instrument to detect a large range of extra-virgin olive oil defects. For the scope, olive oil samples have been artificially modified by adding a number of compounds known to be connected to the more frequent specific defects: fusty, musty, muddy and rancid. Results demonstrate the sensitivity of the electronic nose to the compounds characterizing the defects and then the capability of the instrument to identify the defects.

Detection of acetic acid in wine by means of an electronic nose

Jesus Lozano, Universidad de Extremadura, Spain, jesuslozano@unex.es Fernando Alvarez, Universidad de Extremadura, Spain, fafranco@unex.es

Jose Pedro Santos, Consejo Superior de Investigaciones Científicas, Spain, josepe@ifa.cetef.csic.es carmen Horrillo, Consejo Superior de Investigaciones Científicas, Spain, carmenhorrillo@ifa.cetef.csic.es Wine quality often is defined in incredibly diverse ways. Quality of a wine can be affected by positive and negative quality factors, such as off-odors. Quick and accurate identification of off-odors is advantageous to both winemakers and wine merchants. For the winemaker early remedial action often can correct the situation before the fault becomes serious or irreversible . Some of the most important off- odors producing compounds in wine are acetic acid, ethyl acetate, cork related and sulphur compounds. Vinegary wines typically are sharply acidic, with an irritating odor derived from the combined effects of acetic acid and ethyl acetate. Ideally, the acid concentration should not exceed 0.7g/liter. Traditionally, gas chromatography and mass spectroscopy are used to determine the composition of the wine samples. However, they are complicated and expensive methods, and also unsuitable for real-time, on-line, in situ applications, non-destructive sensing and portable instruments. The use of electronic noses, as a combination of sensors array and pattern recognition methods, has been demonstrated as a valid approach to creating a smart odor sensing system applied to wine aroma. A portable electronic nose based on metal oxide semiconductor thin-film sensors has been developed to detect acetic acid present in four types of wines. The wines analyzed are from the same cellar but are made with different varieties of grapes. Data analysis was performed by two pattern recognition methods: principal component analysis (PCA) and Artificial Neural Networks (ANN)

An ‘olfactory fatigue’ measurement method for Chinese liquors classification with a metal oxide gas sensor array

Huayao Li, Nanomaterial and Smart Sensor Research Laboratory, Department of Materials Science and

Engineering, Huazhong University of Science and Technology, China, lhyflanker@gmail.com Shunping Zhang Nanomaterial and Smart Sensor Research Laboratory, Department of Materials Science

and Engineering, Huazhong University of Science and Technology, China,pszhang001@gmail.com Changsheng Xie Nanomaterial and Smart Sensor Research Laboratory, Department of Materials Science

and Engineering, Huazhong University of Science and Technology, China,csxie@mail.hust.edu.cn China

Odors with the similar main components are hard to be discriminated with metal oxide (MOX) gas sensor arrays under frequently-used measurements (dynamic headspace test for example). ‘Olfactory Fatigue’ could be used to classify similar odors by sensory analysis. For example, in the similar odors classification of odors A and B with ‘Olfaction Fatigue’ method, not the odor A and B were tested under the background of air respectively, but the odor A (/B) was first long time smelled to be adapted, and then odor B (/A) was immediately tested by the analyst’s noses. Under this ‘Olfaction Fatigue’ testing process, the difference between odors A and B could be amplified. Could the ‘Olfaction Fatigue’ method be used in the E-nose measurements? The paper answered this question through Chinese liquors classification measurements with an array of six MOX gas sensors. Meanwhile, the mechanism of selectivity enhancement of MOX gas sensors under the ‘Olfaction Fatigue’ method was presented. Ten brands of Chinese liquors with different flavors were tested under ‘Olfaction Fatigue’ measurements. The results showed that all samples could be 100% correctly classified under the sample brands with ‘Olfaction Fatigue’ method. The results indicate the outstanding performances of E-noses with ‘Olfaction Fatigue’ method in many other applications of similar odors classification. Meanwhile, the ‘Olfaction Fatigue’ measurement process is very simple just like the dynamic headspace test.

A combined gas and liquid chemical sensors array for fuel adulteration detection

Nadja K. L. Wiziack1, Leonardo G. Paterno1, L.H.C. Mattoso2, Fernando J. Fonseca1 1Department of Electronic Systems Engineering, University of São Paulo, São Paulo, Brazil 2Embrapa Agricultural

Instrumentation, São Carlos, São Paulo, Brazil nadja@lme.usp.br

The simultaneous use of different sensor technologies is an efficient method to increase the sensor performance. In this work, a multisensor system combining electronic tongue (ET) and electronic nose (EN) was developed to achieve a better quality control of fuels. ET and EN were manufactured and used, both separately and simultaneously, to evaluate the quality of commercial brazilian fuels. Both ET and EN were composed of sensor arrays based on microelectrodes with different geometries and sensoactive materials. These sensors were designed, fabricated and combinatorially evaluated to choose those better discriminating fuels. The electrical impedance of each sensor in the combined multisensor system was measured when exposed to fuel samples in both liquid and vapor phases simultaneously. Multivariate data analysis was used to calibrate and validate the sensors, as well as to interpret the results. The combined system was shown to significantly improve the substance discrimination compared to isolated ET and EN

Novel sensors for the artificial mouth

Patrick MIELLE , INRA, UMR CSGA, France, Patrick.Mielle@dijon.inra.fr Jacques Maratray, INRA, UMR CSGA, France

Pierre Temple-Boyer, LAAS-CNRS Toulouse, France, temple@laas.fr Christian Salles, INRA, UMR CSGA, France

The ‘Artificial Mouth’, together with its ability mimic human mastication, to deconstruct food pieces, and analyse the released aromas, was presented at the previous ISOEN edition. Numerous improvements have been achieved from this point, mostly by the development of a liquid sampling head. This allows now low forces chewing and the implementation of liquid phase sensors for in-line analysis of the chemical extracted by the saliva. This is a very time-saving solution which enables getting rid of at-line sampling and analysis by time-consuming techniques such as HPLC. Similarly to human chewing, tasty compounds are released in saliva during the food piece mastication in the ‘Artificial Mouth’, and so, are available continuously. Glutamate is present in numerous food, as taste enhancer, has a nice and sought "umami" taste, specific receptors and different inter individual sensitivities, and is a fair marker of the release of tasty compounds. The three sensors (for pH, salt, or glutamate concentration) have the same size, so they are easily interchangeable. Up to now, only one kind of parameter may be analysed at a time by the different sensors. Nevertheless, combined electrodes may be developed in the future.

Use of an electronic tongue to detect geosmin in distilled water

Guilherme de Souza Braga, Universidade de São Paulo, Brazil, gbraga@lme.usp.br Leonardo Giordano Paterno, Universidade de São Paulo, Brazil, lpaterno@lme.usp.br Brazil

Fernando Josepetti Fonseca, Universidade de São Paulo, Brazil, fernando.fonseca@poli.usp.br An electronic tongue (ET) system consisting of conducting polymer sensors was employed to detect geosmin (GSM) in distilled water. GSM is a tainting compound and known to cause undesirable tastes and odours in water and aquaculture farming , . Diluted solutions of GSM were prepared in distilled water at different concentrations. The electrical response (capacitance) of the sensors was analyzed using principal analysis component (PCA). The data obtained were separated into different clusters indicating a good sensibility of the ET system to this compound in distilled water. For GSM concentrations higher than 300ng.L-1, the ET showed a signal saturation. So far, the detection limit of our system is 25ng.L-1. Nevertheless, close grouping between repeated tests indicated that the ET system response is reproducible.

Assessment of volatile sulfur compounds production by select oral bacteria with cysteine and methione as substrates measured by OralChromaTM

Nathanael Salako, Kuwait University, Kuwait, natsalako@hsc.edu.kw

Khalaf Al-Shammari, Kuwait University, Kuwait, Kalshammari@hsc.edu.kw Leeba Philip, Kuwait University, Kuwait, leeba@hsc.edu.kw

Introduction: Oral halitosis is commonly the result of microbial action mainly from the tongue coating and plaque. The three most common VSCs incriminated in oral malodor are H2S, CH3SH and dimethyl sulfide ((CH3)2S). VSCs are produced from the sulfur containing amino acids like cysteine, and methionine. Earlier studies have successfully used either Halimeter (portable sulfide monitor) or Gas Chromatograph (GC) to measure the VSCs produced by the bacteria. Halimeter cannot differentiate between the three VSCs and it is more sensitive to H2S than to CH3SH and is almost insensitive to (CH3)2S. The major shortcoming of GC is that it is difficult to use in the field or clinical setting. Recently however, portable GC instruments such as OralChromaTM (Abilit Cooperation, Japan)have been developed which can measure all three gases with equal high sensitivities. Objective: To assess the ability of common oral anaerobic bacteria isolated from Kuwaiti population to produce VSCs using L- cysteine, and L- methionine as substrates and to measure the VSCs produced using the OralChromaTM. Methods and Materials: Broth cultures of the common anaerobes isolated and identified, were centrifuged and pellets resuspended in phosphate buffer (pH 7.7) with an OD550 of 0.3. 100μl of this suspension and 870 μl of buffer were added in two sterile 15ml head space vials. Reaction was initiated by addition of 30μl of 33mM L- methionine, and L-cysteine respectively in each vial and incubated at 37oC for 90 minutes. 500μl of 3M phosphoric acid was added to tubes and was kept aside for 10 minutes. Production of VSCs were measured using OralChromaTM . Result: With L-cycsteine as substrate eight anaerobes produced high levels of VSCs of which H2S was the major component. When using L-methionine as the substrate, a different complement of VSCs producing anaerobes were identified with CH3SH the major constituent. Conclusion: The OralChromaTM produces a more comprehensive assessment of VSC production by the oral microflora. The use of both cysteine and methionine provide a more detailed assessment of VSCs profile by oral bacteria than the use of either of the substrates alone.

Fusion of potentiometric & voltammetric electronic tongue for classification of black tea taste based on theaflavins (TF) content

Nabarun Bhattacharyya, CDAC, Kolkata, India, nabarun.bhattacharya@cdackolkata.in Andrey Legin, St. Petersburg University, Russian Federation, andrew@kl13930.spb.edu

Irina Papieva, St. Petersburg University, Russian Federation, papieva_irina@mail.ru Subrata Sarkar, CDAC, Kolkata, India, subrata.sarkar@live.com

Dmitry Kirsanov , St. Petersburg University, Russian Federation, d.kirsanov@gmail.com Anna Kartsova, St. Petersburg University, Russian Federation, kartsova@gmail.com

Arunangshu Ghosh, Jadavpur University, India, Rajib Bandyopadhyay, Jadavpur University, India,rb@iee.jusl.ac.in

Black tea is an extensively consumed beverage worldwide with an expanding market. The final quality of black tea depends upon number of chemical compounds present in the tea. Out of these compounds, theaflavins (TF), which is responsible for astringency in black tea, plays an important role in determining the final taste of the finished black tea. The present paper reports our effort to correlate the theaflavins contents with the potentiometric and voltammetric electronic tongue data. Noble metal-based electrode array has been used for collecting data though voltammetric electronic tongue where as liquid filled membrane based electrodes have been used for potentiometric electronic tongue. Black tea samples with tea taster score and biochemical results have been collected from Tea Research Association, Tocklai, India for the analysis purpose. In this paper, voltammetric and potentiometric e-tongue responses are combined to demonstrate improvement of cluster formation among tea samples with different ranges of TF values.

Portable e-tongue based on multi-channel LAPS array with PVC membrane for rapid environment detection

Ha Da, Department of Biomedical Engeering,Zhejiang University, China, ilovenmg@163.com

Wang Ping, Department of Biomedical Engeering,Zhejiang University, China, cnpwang@zju.edu.cn

Yu Hui, Department of Biomedical Engeering,Zhejiang University, China, gogo019@gmail.com Hu Ning, Department of Biomedical Engeering,Zhejiang University, China,

zjubmehuning@gmail.com Wu Cheng Xiong, Department of Biomedical Engeering,Zhejiang University, Cocos (Keeling)

Islands,wucxiong@gmail.com Zhou Jie, Department of Biomedical Engeering,Zhejiang University, China,

zhoujiecomeon@163.com Dmitry Kirsanov d, Chemistry Department,Mendeleev Centre, Saint Petersburg University,

Russian Federation, kirsanov@gmail.com Andrey Legin, Chemistry Department,Mendeleev Centre, Saint Petersburg University, Russian

Federation, andrew@kl13930.spb.edu A new kind of portable e-Tongue based on multi-channel LAPS array with PVC membrane has been designed for the rapid detection of environment situation, especially the seawater. It has the great advantages of depositing membranes which are offered by Chemistry Department, Saint-Petersburg State University on the sensors artificially with convenience and efficiency. To detect various heavy metal ions (Pb2+, Cd2+, Zn2+) simultaneously, respective Polyvinyl Chloride(PVC) membrane could be prepared on the surface of the silicon-based sensor in different channel. For the pre-treatment of the membrane preparation, we have chosen a method of silanization to improve the adhesion between the organic and inorganic interfaces. It also serves as an improvement for preparing uniform and compact membrane on the plane structure. When it refers to the post-processing of the membrane, we have proposed to maintain the PVC structure in the solutions of high concentration. Firstly, it can extend the life time of membrane. Secondly, it is a useful way as maintenance of the activity, leading to the repeatability and consistency of the sensors.

Data fusion from voltammetric and potentiometric sensors to build a hybrid electronic tongue applied in classification of beers

Zouhair Haddi, Moulay Ismaïl University, Morocco, haddi.zouhair@gmail.com Benachir Bouchikhi, Moulay Ismaïl University, Morocco, benachir.bouchikhi@gmail.com

Juan Manuel Gutierrez, CINVESTAV, Mexico, jmgutier27@gmail.com Xavier Cetó, Universitat Autonoma de Barcelona, Spain,

xavidelafiguerosa@gmail.com Aitor Mimendia, Universitat Autonoma de Barcelona, Spain,

aitor.mimendia@gmail.com Manel del Valle, Universitat Autonoma de Barcelona, Spain,

manel.delvalle@uab.es This paper reports the use of an electronic tongue based on data fusion of two different sensor families to recognize three types of beer. The employed sensor array was formed by three modified graphite-epoxy voltammetric sensors1 plus six potentiometric sensors2. The sensors array coupled with feature extraction and pattern recognition methods3, namely Principal Component Analysis (PCA) and Discriminant Factor Analysis (DFA), were trained to classify the data clusters related to different beer types. PCA was used to visualize the different categories of taste profiles and DFA with leave-one-out cross validation approach permitted the qualitative classification. The aim of the work is to improve performance of existing electronic tongue systems, by exploiting the new approach of data fusion of different sensor families.

Discrimination of soils and assessment of some soil fertility parameters using an electronic tongue

Aitor Mimendia, Universitat Autonoma de Barcelona, Spain, aitor.mimendia@gmail.com

Juan Manuel Gutierrez, Universitat Autonoma de Barcelona Salgado, Spain, jmgutier27@gmail.com Josep Maria Alcañiz, Universitat Autonoma de Barcelona, Spain JoseMaria.Alcaniz@uab.es

Manel del Valle, Universitat Autonoma de Barcelona, Spain, manel.delvalle@uab.es Soil analysis is a crucial step in many areas in which this natural resource must be evaluated, as in soil fertility studies. The analytical approaches employed have to deal with two common difficulties that appear in environmental analysis: the strong variation due to natural heterogeneity and the complex problem of interactions between chemical species. Labour intensive and time-consuming procedures are thus normally employed to assess soil characteristics, or to identify soil type. In this communication, a new strategy to perform soil classification and/or characterization is reported, which is the coupling of chemical sensors with a pattern recognition method, what is known as an electronic tongue. Following this approach, the system proposed in this paper uses a sensor array formed by potentiometric sensors with generic cross response against several cations and anions, plus a pattern recognition method based on Artificial Neural Networks (ANN); the sensor-based system allows performing a simple laboratory procedure where the advanced data processing methodology permits to extract the meaningful information. In this way this work represents the first application and testing of an electronic tongue in soil analysis. Apart from the qualitative classification application, a quantitative analysis of certain chemical features related to soil fertility has also been attempted.

An impedancemetric electronic tongue for discrimination of adulteration process of ethanol fuel with water

Thiago Paixão, Universidade de São Paulo / Instituto de Química, Brazil,

trlcp@iq.usp.br Lígia Bueno, Universidade Federal do ABC / Centro de Ciências Naturais e Humanas,

Brazil,ligia_quim@yahoo.com.br The most widely used biofuel in Brazil is ethanol from sugarcane. One of the major frauds involving this product is undue adding of more water to ethanol fuel (7.4% water (v/v) is permitted by Brazilian legislation1) in order to increase its volume, which will result in tax evasion, as well as consumer harm. Furthermore, this counterfeit product causes damage to vehicles. For this reason, this work will reported our efforts to fabricate an impedancemetric electronic tongue to discriminated counterfeit ethanol fuel samples by addition of water towards to development of sensor that can be allocated into fuel tank of vehicles, resulting in a binary answer (good or bad fuel). To the best of our knowledge, no reports were found on the literature including this topic. Additionally, the search for new analytical methods for ethanol fuel analysis becomes necessary due to the addition of impurities to fuel, by counterfeiters, trying to difficult the fraud identification by traditional methods (such as hydrometers and conductivity meters). The capability of the proposed device to differentiate samples was evaluated for ethanol fuel with and without addition of water. In all analysed cases, a good separation between different samples was noticed in the score plots obtained from the principal component analysis (PCA) e using hierarchical cluster analysis (HCA) dendrogram.

Sensory evaluation and electronic tongue analysis for sweetener recognition in coke drinks

Dániel Szöllősi, Corvinus University of Budapest Somlói str. 14-16, Budapest, 1118, Hungary

T: +36 1 482 6023, F: +36 1 482 6361, daniel.szollosi@uni-corvinus.hu Zoltán Kovács, Corvinus University of Budapest László Sípos, Corvinus University of Budapest Zoltán Kókai, Corvinus University of Budapest

András Fekete, Corvinus University of Budapest

Consumption of beverages with low energy has an increasing role. Furthermore hydrolyzed starch products such as inverted syrup show a wide application in the beverage industry. Therefore the importance of methods which can monitor the usage of natural and artificial sweeteners is increasing. The task was to describe the relevant sensory attributes and to determine the applicability of the electronic tongue to discriminate the coke drink samples with different sweeteners. Furthermore the aim was to find relationship between the taste attributes and measurement results provided by electronic tongue. Four different coke drink sample was used for the experiments. There were two samples containing different kinds of artificial sweeteners (acesulfame-K, aspartame and Na cyclamate) and two samples contained sucrose or inverted syrup. Profile analysis was used for the sensory experiments by 21 trained panelists. Alpha Astree potentiometric Electronic Tongue was used for the electronic tongue tests. The sensory results were evaluated by one way analysis of variance and principle component analysis was used to evaluate the results of the electronic tongue tests. Panelists found significant differences between the samples in 15 cases from the 18 sensory attributes defined previously by the consensus group. The highest differences between the coke samples were observed in sweet taste and sweetener taste. The samples were definitely distinguished by the electronic tongue. The main difference was found between the samples made with natural and artificial sweeteners. However electronic tongue was able to distinguish samples containing different kind of artificial and different kind of natural sweeteners, as well. Taste attributes of coke drinks determined by sensory panel were predicted by partial least squares regression method based on the results of electronic tongue with close correlation and low prediction error. Correlation coefficients were 0.89 and 0.99 while RMSEP were 3.86 and 0.45 for astringent taste and sweet taste, respectively.

Development of taste sensing system using inorganic membrane

Yohichiro Kojima, Tomakomai National College of Technology, Japan, iorisiori1999@yahoo.co.jp Yuki Hasegawa, Saitama University, Japan, yuki@ees.saitama-u.ac.jp

Recent interest in the development of taste sensors has been motivated by their potential application in the food and beverage industries as well as environmental monitoring. We developed a novel taste sensor for liquid and verified its effectiveness using coffee. We fabricated an inorganic metal oxide membrane liquid sensor using the laser ablation method. The sensor shows a sufficient sensitivity for electrolyte solutions, while it shows a relatively low response for non-electrolyte solutions. We differentiated and identified five brands of commercially available coffee using the sensor.

Sensing characteristic of polyaniline/TiO2 nanocomposites

Han Yong Lee, Republic of Kyungpook National University, Korea, brood38@hanmail.net Joon Boo Yu, Republic of Kyungpook National University, Korea, jyacht@naver.com Jeung Soo Huh, Republic of Kyungpook National University, Korea, jshuh@knu.ac.kr

Summary Conducting polymers were found to be a better choice for gas-sensing material due to its good environmental and chemical stability, ease of synthesis, reversible response and shorter response time. However, Conducting polymer gas sensors have lower sensitivity than metal oxide gas sensor. There is a tremendous approach for the enhancement of the sensitivity of sensors by combining the organic materials with inorganic counterparts to form composites . We study about the conducting polymers combined with metal oxides. A polyaniline-titanium dioxide (PANi/TiO2) nanocomposite was prepared by interfacial polymerization at room temperature. The sensors were fabricated by dipping method. PANi/TiO2 nanocomposite was transferred onto an alumina substrate with a comb-type electrode of 0.5 mm gap between each other. Sensitivity measurements of PANi/TiO2 nanocomposite and pure aniline were conducted with benzene, ethyl benzene, toluene, styrene and xylene gas at 50 ppm. Methods and Results PANi/TiO2 nanocomposites were prepared by interfacial polymerization. We compared with PANi/TiO2 nanocomposite and pure aniline to know effective contribution of TiO2 in the nanocomposites sensor. Surface morphology of sensor was observed with scanning electron microscope (SEM) as shown Figure 1. They ware obviously shown with a mesh-like structure, in which the particle dimensions fall in a nanometer range. It is evident that there are many pores on the surface, which seem to contribute to the short response time and good reversibility of the sensors . Polyaniline/TiO2 sensor materials weres transfered by dipping method. The obtained powder was completely dissolved in the chloroform with an additional DBSA. The solution was transferred onto an alumina substrate with a comb-type electrode of 0.5 mm gap between each other. The sensing characteristics investigation was carried out by using a gas sampling system including the mass flow controller (MFC), temperature controller. The sensor characteristics were then tested in a testing chamber after flowing injection of the target VOC vapor. Nitrogen gas was used as a carrier gas to transport the target into the chamber. VOC vapors were introduced to the chamber followed by ambient air for approximately 3 minutes per cycle. The operating temperature was set at 25oC using a cooling bath (Figure 2). The sensitivity of Polyaniline/TiO2 and pure polyaniline gas sensors show similar trend, but Polyaniline/TiO2 gas sensors higher sensitivity than pure polyaniline sensors over two times. It appeared that the TiO2 components influenced the morphology of the nanocomposite, which led to the improvement of the sensor sensitivity.

A novel method for synthesis of 3-D Wo3-CuO nanogrids

Jusang Lee, SUNY at Stony Brook 314 Old Engineering Building, Stony Brook, New York, 11794, USA T: 631-632-8497, F: 631-632-8052, Jusang.Lee@stonybrook.edu

Perena Gouma, SUNY at Stony Brook

This study involves a novel method for synthesis of 3-D WO3/CuO nanogrids by the combined use of sol-gel, electrospinning and thermal oxidation processes, respectively. The electrospinning technique was employed to produce nanofiber using metal oxide-sol-gel and polymer solution mixture as precursors. Electrospurned non-woven nanofiberous mats provide the template guiding the growth of interconnected CuO and WO3 nanopoarticles. The mechanism of formation of unique nano-architectures was result of Cu diffusion upon thermal oxidation within each metallic core nanofiber on Cu-mesh substrate. Morphology and grain size of the products were investigated by scanning electron microscopy (SEM), transmission electron microscopy (TEM)and X-ray diffraction (XRD). The average size of the nanoparticles in the nanogrids is 20 nm. Considering the simplicity of the synthesis method and, such open and high surface areas of nanogrids might be a promising candidate photocatalyst and gas sensing applications.

Tunneling through surface barrier and oxygen in-diffusion in nanostructured SnO2 gas sensors.

Cesare Malagù, Univeristy of Ferrara, Italy,malagu@fe.infn.it The surface barrier of SnO2 was studied, for grains of the order of the depletion width, ?? the condition which makes the material properly nanostructured. The model allowed us to explain the different behavior of conductance in gas of two sets of sensors with grains having two well distinct characteristic radii (R< ? and R>??). In the first case (R<?), the overlapping of barriers can take place in the presence of a gas, an effect which strongly reduces the tunneling contribution to conductivity with respect to the thermionic one, at the same temperature. The behavior of conductance in the presence of reducing gases has been explained through the mechanism of barrier modulation through gas chemisorptions, assuming that the density of vacancies can only be modified by interstitial oxygen diffusion in and out of the nanograins. The necessity of adopting the electrochemical potential to describe equilibrium instead of the chemical one was discussed

TiO2 nanostructures for room temperature gas sensing

Daniel F. Rodríguez, Comisión Nacional de Energía Atómica, Argentina, drodrig@tandar.cnea.gov.ar Patricia Perillo, Comisión Nacional de Energía Atómica, Argentina, perillo@cnea.gov.ar

Carlos Rinaldi, CNEA, Argentina, rinaldi@cnea.gov.ar Alberto Lamagna, CNEA, Argentina, alamagna@cnea.gov.ar

Gas sensors TiO2 based were prepared by anodic oxidation method. Morphology of the TiO2 nanostructures was characterized by scanning electron microscopy (SEM), X-ray diffraction (XRD) and Raman spectroscopy. At room temperature, the sensors exhibited highly sensitive and fast response-recovery (less than 2 min) to NH3 gas of concentrations ranging from 50 to 200 ppm. Based on the experimental observations, a model of potential barrier to electronic conduction at the grain boundary for the TiO2 composite sensors was also discussed.

A gas sensor array for environmental air monitoring: A study case of application of artificial neural networks

Michele Penza, ENEA, Italy, michele.penza@enea.it

Domenico Suriano, ENEA, Italy, domenico.suriano@enea.it Gennaro Cassano, ENEA, Italy, gennaro.cassano@enea.it

Riccardo RossiItaly, ENEA, Italy, riccardo.rossi@enea.it Marco Alvisi, ENEA, Italy, marco.alvisi@enea.it

Valerio Pfister, ENEA, Italy, valerio.pfister@enea.it Livia Trizio, LENVIROS srl and University of Bari, Italy, liviatrizio@lenviros.it

Magda Brattoli, LENVIROS srl and University of Bari, Italy, m.brattoli@chimica.uniba.it Gianluigi De Gennaro, LENVIROS srl and University of Bari, Italy, giangi@chimica.uniba.it

An array of commercial gas sensors and nanotechnology sensors has been integrated to quantify gas concentration of air-pollutants. A variety of chemoresistive gas sensors, commercial (Figaro and Fis) and developed at ENEA laboratories (metal-modified carbon nanotubes) were tested to implement a database useful for applied artificial neural networks (ANNs). The ANN algorithm used is the common perceptron multi-layer feed-forward network based on error back-propagation. Electronic Noses based on various sensor arrays related to mammalian olfactory systems have been largely reported . Here, we reported on the perceptron-based ANNs applied to a large database of 4800 datapoints for environmental air monitoring. The ANNs performance has been individually assessed for any targeted gas. The response of the classifier has been measured for NO2, CO, CO2, SO2, and H2S gas. The NO2 characteristics exhibit that real concentrations and predicted concentrations are very close with a normalized mean square error (NMSE) in the test set as low as 6%.

Development of an electronic nose for environmental monitoring: detection of specific environmentally important gases at their odor detection threshold concentration

Licinia Dentoni, Politecnico di Milano, Italy,ldentoni@chem.polimi.it Laura Capelli, Politecnico di Milano, Italy Selena Sironi, Politecnico di Milano, Italy

Renato Del Rosso, Politecnico di Milano, Italy Paolo Centola, Politecnico di Milano, Italy Matteo Della Torre, Sacmi s.c.a.r.l., Italy Fabrizio Demattè, Sacmi s.c.a.r.l., Italy

The use of a sensor array is demonstrated to be an effective approach to evaluate hazardous odor (or gas) emissions from industrial sites. Therefore the possibility to use electronic noses for the prolonged survey of odor emissions from industrial sites is of particular interest for environmental monitoring purposes . At the Olfactometric Laboratory of the Politecnico di Milano, in collaboration with Sacmi Group, Imola, an innovative electronic nose for the continuous monitoring of environmental odours is being developed. The aim of this work is to show the laboratory tests conducted to evaluate the capability of the electronic nose to recognize some specific environmentally important gases at their odor detection threshold concentration. The laboratory studies up to now focused on ammonia and butyric acid, those being compounds that can typically be found in the emissions from waste treatment plants, that may cause health effects when they exceed a given concentration level. The laboratory tests proved the sensors to be sensitive towards the considered compounds and the system to be capable of discriminating between odorous or non-odorous air, with a detection limit comparable with the detection limit of human nose.

Cumulative measurement principle for the detection of small amounts of gaseous species

Andrea Geupel, University of Bayreuth, Germany, andrea.geupel@uni-bayreuth.de Gregor Beulertz, University of Bayreuth, Germany, gregor.beulertz@uni-bayreuth.de

David Kubinski,Ford Research and Advanced Engineering, United States dkubinsk@ford.com Jacobus Visser jvisser@ford.com,Ford Research and Advanced Engineering, United States

Ralf Moos, University of Bayreuth, Germany, Ralf.Moos@uni-bayreuth.de Summary: To control or monitor processes and the compliance with air quality regulation, highly selective and long-term stable gas sensors with a high accuracy in the low ppm detection range are needed. Gas sensor techniques detecting the actual concentration of analyte gas species often suffer from inaccuracy at low concentrations and signal instability. In contrast, gas sensors detecting the cumulative amount of the analyte species during the measurement period are beneficial for monitoring even very small amounts of analyte species. The idea of the cumulative detection principle is the accumulation of the gas molecules in the sensitive layer and therefore measuring the timely integrated concentration of the gas species (“the amount”). Hence, the sensitive layer of the cumulative sensor consists of a gas storage material, which is able to store the analyte molecules by a chemical bonding and thereby changes its material composition and its electrical properties. In this work, the principle and the benefits of the cumulative detection is described and measurement results of an integrating NOx sensor are shown to demonstrate the general integrating functionality of this cumulative measurement technique during exposure of small concentrations of analyte gas. Methods and Results: One possible sensor setup of a cumulative impedimetric gas sensor is shown in Fig. 1. The electrical transducer element, here interdigital electrodes, are arranged on a ceramic substrate and are covered with the sensitive analyte trap material, which is able to store chemically analyte molecules by forming a new material composite. As illustrated in Fig. 2., the exposure of the sensitive layer to the analyte gas results in a successive accumulation of the gas molecules in the sensitive layer. The material transformation results in a change in its complex impedance. Once the storage capacity gets saturated, the material needs to be regenerated to recover its functionality. In Fig. 3a., the ideal response of a cumulative sensor on several periods with different analyte concentrations is illustrated. Since the analyte molecules are accumulating on the sensor, the signal increases in the presence of the analyte gas. But it remains constant at 0 ppm, since the analyte amount stored on the sensor remains the same. The slope of the curve depends on the actual analyte concentration (double the concentration means double the amount). The characteristic line correlates the sensor signal with the amount of exposed analyte gas. An ideal integrating or cumulating sensor has a linear characteristic line, independent on the actual concentration, as shown in Fig. 3b. The integrating sensing principle was realized by the authors for the detection of NOx using commercial NOx trap materials, known from NOx storage catalysts (NSC), as sensitive layers . NSCs contain precious metals oxidizing NO to NO2 which is successive stored on oxides or carbonates (e.g. BaCO3) by forming nitrates. The measurement results, presented in Fig. 4., clearly proof the integrating properties of this total NOx sensor. At 375°C, the relative resistance change |ΔR/R0|, calculated from the complex impedance at 1 kHz is increasing linearly while 10 ppm or 5 ppm NO or NO2 are in a gas stream of 10 % O2, 5% CO2 and 2% H2O in N2, whereas the signal remains almost constant at 0 ppm NOx. The characteristic line (Fig. 4b), plotted from those measurement data, shows a linear correlation between |ΔR/R0| and the amount of NOx ANOx. Those results demonstrate that the cumulative or integrating measurement principle with an appropriate trap material is very well suited for the reliable detection of low analyte gas concentrations.

Electronic nose system combined with membrane interface probe for detection of VOCs in water

Junghwan Cho, University of Massachusetts Lowell, United States, Junghwan_cho@uml.edu

Zachary Howard, University of Massachusetts Lowell, United States, Zachary_Howard@student.uml.edu

Pradeep Kurup, University of Massachusetts Lowell, United States, Pradeep_Kurup@uml.edu

This paper describes a novel electronic nose system combined with a membrane interface probe (MIP) for detecting volatile organic compounds (VOCs) in water. The MIP is an in situ tool that allows the detection of certain VOCs in the soil via a pushed or driven probe . The MIP was combined with a sensor array consisting of four different tin-oxide gas sensors known as an electronic nose (e-nose). The designed e-nose system was calibrated in aqueous media spiked with benzene, toluene, ethylbenzene, and p-xylene (BTEX) at concentrations of 100, 250, and 500 ppm. Since the experiment was conducted utilizing five repetitions for each analyte, a data set of 60 measurements was prepared for principal components analysis (PCA). The results of the PCA showed that two principal components contain more than 99% variance information and each VOC is separable and detectable by the e-nose.

Electronic noses implementation on landfill site

Louis Vivola, ALPHA MOS, France, vivola@alpha-mos.com

frederic Clet, ALPHA MOS, France, frederic.clet@sita.fr

elena Senante, ALPHA MOS, France, elena.senante@suez.fr

Odours nuisance are an increasingly important consideration for composting and drying sites, tips and waste water purification plants, in relation to local inhabitants complaints. Odour measurement remains a complex and difficult task, with current methods focused on one-off measurements and delayed results. As an alternative, RQ Box Electronic Noses (Alpha MOS, France) offer the ability to continuously measure odour and volatile compounds directly on-site. The Electronic Nose is a well-established instrument in the laboratory setting or in Quality Control where it is used to perform Volatile Organic Compound (VOC) or aroma analysis. In order to offer continuous and real time monitoring of olfactory nuisances, specific sensor array devices have been developed. This wireless equipment called RQ Box (as Air Quality Box) includes several complementary detectors (metal oxide sensors and specific detectors for target gases detection) in order to measure odors on the field. Thanks to real time measurement ability and wireless communication, these devices can be installed in several areas of a same industrial site or in the neighborhood and generate an accurate map of olfactory nuisance. A network of 2 RQ Box electronic noses was installed on a waste storage site from Suez Environnement in order to monitor odour emissions, characterize the odour sources and improve the neighborhood comfort.

Hand-held device for monitoring dissolved organics in fresh and recycled water on PPB levels

Serge Zhuiykov, Senior Research Scientist, Australia, serge.zhuiykov@csiro.au Brett Sexton,Principle Research Scientist, Australia,

rett.sexton@csiro.au Pamela Hoobin, Research Project Officer, Australia, pamela.hoobin@csiro.au Michael Best, Senior Electronics Engineer, Australia, michael.best@csiro.au

Donavan Marney, Senior Research Scientist, Australia, donavan.marney@csiro.au New hand-held robust UV255 sensor for control of dissolved organic carbon (DOC) in water has been developed by CSIRO, CMSE1. Device is designed for compact operation and simplified circuitry. UV 255 sensor exhibited a good liner response at low concentration range of DOC (<5 ppm) and has shown the detection limit less than 10 ppb of DOC. The current design allows for immersion of the sensor into a solution. The sensor can also be modified for flow-through applications by allowing water to flow through side arms. New sensor discriminates readily between various types of potable and recycled water.

Pursuing contamination detection on aircraft cfrp surfaces by artificial olfaction techniques

Saverio De Vito, ENEA UTTP/MDB P.le E. Fermi,1, Portici, NAples, 80055, Italy

T: +39 0817723364, F: +39 0817723344, saverio.devito@enea.it Ettore Massera, ENEA UTTP/MDB

Grazia Fattoruso, ENEA UTTP/MDB Maria Lucia Miglietta, ENEA UTTP/MDB Girolamo Di Francia, ENEA UTTP/MDB

In the last decade, Aerospace industry has experimentally become to use Carbon Fiber Reinforced Polymer (CFRP) for secondary structure fabrication. CFRP structures can be easily bonded via adhesive assembly procedures. The lack of standardized procedures for CFRP adhesive bonds quality assessment has hence prevented their use in the rivetless assembly of aircraft main structures. Such rivetless assembly could easily lead to extremely lightweight aircraft with very significant savings in fuel usage and CO2 emissions for commercial air traffic. As such developing, the so called “green aircraft” is one of the main drivers of FP7 aerospace research programs and of the CleanSky JTI. Actually, it represents the most financed EU research initiative (800MEUR).Within this framework, the FP7-ENCOMB (Extensive NDT for Composite Bonds quality assessment) project aims to investigate and develop novel NDT (Non Destructive test) technologies for the assessment of both the adherends surface (pre-bond) and adhesive bonds quality. The ENCOMB project, lead by Fraunhofer Institute, brings together 14 of the top quality EU research institutions from 6 EU States, including ENEA. Several technologies will be investigated, adapted and developed in order to establish their suitability for CFRP adhesive bonds quality assessment. ENEA will be in charge of investigating the application of Artificial Olfaction technologies for detecting the contamination of adherends surfaces. More specifically, the selection and the development of sampling, sensing and data processing technologies will represent the main effort of the ENEA research group in the ENCOMB project. The sampling subsystem should be developed to maximize the uptake of volatile molecules from the CFRP surfaces. On the basis of the results of the first experimentation, carried out by the use of spectrometry and general purpose electronic nose (Bloodhound ST306), the architecture of an ad-hoc sampling system will be designed. Sensors selection will be carried out by investigating different technologies so it will probably lead to a hybrid sensor array; MOX, Polymer, EC, PID and IMS sensing technologies will be screened for their capability to detect main volatile compounds building up by contamination agents at significantly low concentration levels (hundreds of ppb). Finally, ad-hoc sensor data processing techniques will be designed and developed to address the complexity and the different performance of the selected sensor technologies with the aim to enhance the discrimination capability and the sensor drift resiliency. Operative requirements of such a tool include portability and quick availability of contamination assessment results.

Tin oxide nanowire sensors and their potential for selective detection of the toxic gases SO2 and H2S

Anton Koeck, deputy head of division, Austria, anton.koeck@ait.ac.at

Numerous applications ranging from industrial process control to personal safety systems and environmental monitoring have a strongly increasing demand for highly sensitive and selective gas sensors. Novel devices employing nanowires as sensing elements have been successfully developed for improving the performance of gas sensing devices1-3. Recently we have reported on SnO2-nanowire sensors for highly sensitive CO and CH4 detection4. We present nanosensors based on single crystalline SnO2-nanowires, which are very sensitive to the highly toxic gases SO2 and H2S. The nanowires are realized in a two-step atmospheric pressure synthesis process on Si-substrates4. We have found that the sensing performance of SnO2-nanowire sensors strongly depends on their operating temperature. The sensors start to detect the SO2 and H2S gases at temperatures above 200°C. A SO2 gas concentration of 26.8 ppm decreases the resistance of the nanosensor, above 350°C the resistance is increased up to a maximum signal of ~6% at 400°C. A H2S concentration of only 1.4 ppm increases the resistance, above 300°C the resistance is decreased up to a maximum of 30% at 400°C. This very high sensor signal demonstrates extraordinary sensitivity of the nanowire sensors with a resolution in the ppb range. The distinct temperature dependence provides the possibility for selective detection of the two toxic gases SO2 and H2S.

Microwave-hydrothermal synthesis and vibrational spectroscopy of nanostructured (Ni,Mn,Co)Sb2O6 compounds for chemical sensing

Anderson Dias, Federal University of Ouro Preto Department of Chemistry, ICEB II, Sala 67, Campus Morro do Cruzeiro, Ouro Preto, MG, 35400-000,

BRAZIL T: 55-31-35591716, F: 55-31-35591707, anderson_dias@iceb.ufop.br

Nowadays, the emission of CO2 has significantly contributed to the global warming. Apart from the strategies to reduce these emissions, the accurate determination of the concentration of environmental gases demands reliable gas sensor materials. Regarding gas sensor materials, transition-metal antimonates of general formula ASb2O6 emerge as potential compounds for gas sensing and monitoring. Besides, the production of nanomaterials with optimized chemical sensing properties could be associated with environmentally-friendly processes. Among the methods employed to synthesize nanostructured materials, the hydrothermal process associated with the use of microwave energy has become a promising route to the production of nanosized materials, offering advantages in comparison with conventional methods, such as low-energy consumption associated with faster precipitation kinetics. In this work, (Ni,Mn,Co)Sb2O6 nanostructured materials were produced by microwave-hydrothermal synthesis and the crystal structures and sensing properties were studied by electron microscopy, Raman scattering and impedance measurements. The results from high-resolution transmission electron microscopy and selected area electron diffraction showed that crystalline, nanostructured materials were produced, as also evidenced by X-ray diffraction after microwave synthesis. Vibrational spectroscopic results confirm the crystal structure, which were validated by group-theory calculations. The Raman spectra of (Ni,Mn,Co)Sb2O6 at room temperature showed a elevated number of bands, higher than those predicted by group-theory calculations. It was possible to discern 12 bands (although some of them, which are very weak and broad, could be combination modes), which could be attributed to defect-chemistry occurred during microwave processing. It is also worthy noticing the stretching modes of O-Sb-O at around 300 cm-1 split into several bands. The same feature occurs relatively to the other compounds, i.e., we can clearly identify groups of bands in the regions 60-700cm-1. The AC electrical characterization demonstrates the important role of the applied frequency on the gas sensing properties of (Ni,Mn,Co)Sb2O6compounds. Defect-chemistry detected by Raman scattering associated with impedance measurements showed the relationship between processing conditions and final properties.

Discrimination of body odor using odor sieving sensor system

Tadashi Takamizawa, U.S.E Co.,Ltd., Japan, takamizawat@use-ebisu.co.jp

Kazuki Miyagi, U.S.E Co.,Ltd., Japan, miyagik@use-ebisu.co.jp Hitoshi Miyauchi, U.S.E Co.,Ltd., Japan, miyauchih@use-ebisu.co.jp

Kenshi Hayashi , Kyushu University, Japan, hayashi@ed.kyushu-u.ac.jp Masahiro Imahashi, Kyushu University, Japan, imahashi@belab.ed.kyushu-u.ac.jp

The themes of our research are (i) to develop an experimental system which is based on the principle of the molecular sieve effect, to detect and discriminate features of dilute human body odor; and (ii) to study the characteristics of human body odor using the experimental system mentioned in (i). The aforementioned system is used to comfirm whether there are differences between human body odor features, but is not intended for medical use. The objective of the first experiment in our research was to distinguish human body odor types using T-shirts worn by examinees. In our second experiment based on the same principle, samples were five oil clear sheets used to wipe the faces of five examinees. The research results lead to a drastic change in our experimental system development. We have developed our experiment methodology under a consistent policy. In the present research, we have conducted experiments, especially focusing on the discrimination of human body odor features. A small number of examinees were examined for this purpose, and for each examinee experiments were carried out a number of times. Through these experiments, we detected statistically-significant differences between measurements except those for one combination of examinees. This result shows that the aforementioned system and principle enabled discrimination of body odor between examinees to a certain extent. Therefore, further development of this system and addition of a sieving method based on different features of human body order to the system will likely to enable more obvious discrimination of examinees through their body odor features.

Towards an analogue neuromorphic VLSI instrument for the sensing of complex odours

Muhammad Fazli Ab Aziz, Universiti Teknologi Malaysia, Skudai, Malaysia, fazli_zz@yahoo.com.sg

Fauzan Khairi Che Harun, Universiti Teknologi Malaysia, Skudai, Malaysia, fauzan@fke.utm.my

James A. Covington, United Kingdom School of Engineering, University of Warwick, Coventry, United Kingdom, J.A.Covington@warwick.ac.uk

Julian W. Gardner, United Kingdom School of Engineering, University of Warwick, Coventry, United Kingdom, J.W.Gardner@warwick.ac.uk

Almost all electronic nose instruments reported today employ pattern recognition algorithms written in software and run on digital processors, e.g. micro-processors, microcontrollers or FPGAs. Conversely, in this paper we describe the analogue VLSI implementation of an e-nose through the design of a neuromorphic olfactory chip. The neuromorphic chip contains 3 neural sections where each section contains 3 receptor neurons, 27 synapses, and 1 principal neuron. The modelling, design and fabrication of the chip have already been reported. Here a smart interface has been designed and characterised for the neuromorphic chip. The interface board contains all the control variables for the neuromorphic chip and 9 sensor response inputs. For test purposes, the board was designed so that the input and output responses can be controlled and monitored via a PC and using National Instrument Labview software. Thus we can demonstrate the functionality of the aVLSI neuromorphic chip, producing differing principal neuron firing patterns to real sensor response data. Further work is directed towards integrating 9 separate neuromorphic chips to create a large neuronal network to solve more complex olfactory problems.

Optimizing the operating temperature for an array of MOX sensors on an open sampling system

Marco Trincavelli, AASS Research Centre, Örebro University, Sweden

, marco.trincavelli@oru.se Alexander Vergara, BioCiruit Institute, University of California, San Diego, United States,

vergara@ucsd.edu Nikolai Rulkov, BioCiruit Institute, University of California, San Diego, United States, nrulkov@ucsd.edu

Jose S. Murguia, BioCiruit Institute, University of California, San Diego, United States, ondeleto@uaslp.mx

Achim Lilienthal , AASS Research Centre, achim@lilienthals.de Ramon Huerta, BioCiruit Institute, University of California, San Diego, United States,

ramon.huerta@gmail.com Chemo-resistive transduction is essential for capturing the spatio-temporal structure of chemical compounds dispersed in different environments. Due to gas dispersion mechanisms, namely diffusion, turbulence and advection, the sensors in an open sampling system, i.e. directly exposed to the environment to be monitored, are exposed to low concentrations of gases with many fluctuations making, as a consequence, the identification and monitoring of the gases even more complicated and challenging than in a controlled laboratory setting. Therefore, tuning the value of the operating temperature becomes crucial for successfully identifying and monitoring the pollutant gases, particularly in applications such as exploration of hazardous areas, air pollution monitoring, and search and rescue. In this study we demonstrate the benefit of optimizing the sensor’s operating temperature when the sensors are deployed in an open sampling system, i.e. directly exposed to the environment to be monitored.

Portable e-nose and multivariate data analysis to identify different kinds of drugs Zouhair Haddi, Sensor Electronic & Instrumentation Group, Physics Department, Moulay Ismaïl University,

Morocco, haddi.zouhair@gmail.com Aziz Amari,Sensor Electronic & Instrumentation Group, Physics Department, Moulay Ismaïl University,

Morocco, amariaziz2004@yahoo.fr Hassane Alami,Sensor Electronic & Instrumentation Group, Physics Department, Moulay

Ismaïl University, Morocco, alami@fs-umi.ac.ma Nezha El Bari, Biotechnology Agroalimentary and Biomedical Analysis Group, Biology Department,

Moulay Ismaïl University, Morocco, n_elbari@hotmail.com Eduard Llobet, MINOS, Electronic Engineering Department, Universitat Rovira i Virgili, Spain,

eduard.llobet@urv.net Benachir Bouchikhi, Sensor Electronic & Instrumentation Group, Physics Department, Faculty of

Sciences, , Moulay Ismaïl University, Morocco, benachir.bouchikhi@gmail.com

In this work, we present a portable electronic nose system equipped with an array of commercially available metal oxide gas sensors and suitable pattern recognition methods in order to classify several types of drugs. E-nose applications are mainly centered in food and beverage industries , environmental monitoring, medicine, etc.; however, this is the first time that a portable e-nose could perform successfully illegal drug identification. Five drugs, namely cannabis buds, cannabis plants, hashish, snuff tobacco and tobacco leaves were experimented. The portable e-nose was based on a simple dynamic headspace sampling, a PIC microcontroller for real time data acquisition and a laptop computer for data storage and processing . Principal Component Analysis (PCA), Multivariate Analysis of Variance (MANOVA) and Support Vector Machines (SVMs) permitted discrimination and identification of five clusters corresponding to the five drugs analyzed. This study gives further evidence that the portable e-nose could be useful for both health and security protection agencies.

Development of odor gas sensor using TiO2 nanostructures

Bong-Hwan Cho, Korea, Republic of Department of Materials Science and Metallurgical Engineering, Kyungpook National University, magic1111i@nate.com

Shao-Lin Zhang, Department of Sensor and Display Engineering, Kyungpook National University, China,slzhang@knu.ac.kr

Joon-Boo Yu, Korea, Republic of Department of Materials Science and Metallurgical Engineering, Kyungpook National University, jyacht@naver.com

Jeong-Ok Lim, Korea, Republic of Department of Materials Science and Metallurgical Engineering, Kyungpook National University, jolim@knu.ac.kr

Hyung-Gi Byun, Korea, Republic of Department of Materials Science and Metallurgical Engineering, Kyungpook National University, byun@kangwon.ac.kr

Jeung-Soo Huh, Korea, Republic of Department of Materials Science and Metallurgical Engineering, Kyungpook National University, jshuh@knu.ac.kr

Summary: Detection of odor gas in an easy, portable and economical method is urgently desired in industry and domestic field. Micro gas sensor array using metal oxide semiconductor is considered as one of the promising method to detect and classify odor gases . It is known that sensing responses are strongly dependent on the material structures and operating temperatures. In this study, TiO2 commercial powder (degussa, P-25), TiO2 nanorods and TiO2 nanosheets were used as sensing materials. Micro sensor substrates fabricated by MEMS technology were used. Sensing materials and working temperatures were controlled and sensors were fabricated under controlled conditions. Sensing properties of sensors were measured to odor gases such as toluene, acetaldehyde, hydrogen sulfide, methyl mercaptan, ammonia. Sensing results were analyzed and compared to each another. Methods and Results: TiO2 nanorods and TiO2 nanosheets were synthesized by hydrothermal and sonochemical method, respectively. Commercial TiO2 nanopowder and NaOH were used as starting materials. The effects of hydrothermal temperature and ultrasound power on the formation of TiO2 nanostructures were investigated. The formation mechanisms of TiO2 nanorods and nanosheets were discussed. Silicon-based sensor substrates with interdigital Pt electrodes and co-planar heaters were pre-prepared using MEMS process (Fig. 1). Sensing materials were coated on the silicon-based substrates by a dropping method. The sensors were subsequently soldered onto commercial sockets and hosted in a test chamber (Fig. 2). Synthetic dry air at a constant flow rate of 250 sccm was acted as carrier gas. Various kinds of odor gases, such as toluene, acetaldehyde, hydrogen sulfide, methyl mercaptan and ammonia, were used as target gases. The sensor signal was obtained from the resistance variation of sensors under air and odor gas exposure. The obtained data were compared and analyzed for TiO2 nanopowder sensor and nanorod sensor (Fig. 3.). The results showed that nanomarerial such as TiO2 has potential for odor gas detection.

Odour profile of different varieties of extra-virgin olive oil during deep-frying using an electronic nose and SPME-GC-FID.

Valeria Messina, CINSO, Argentina, ona_2003@hotmail.com

Andrea Biolatto, INTA, Argentina, abiolatto@hotmail.com Adriana descalzo, INTA, Argentina, descalzo@hotmail.com

Ana Sancho, INTA, Argentina, asancho@hotmail.com Noemi Walsoe de Reca, CINSO, Argentina, nwalsoe@hotmail.com

Gabriela Grigioni, INTA, Argentina, grigioni@gotmail.com Aroma, taste, colour and nutritive properties of extra virgin olive oil distinguish it from other edible vegetable oils. Extra virgin olive oil is considered to be stable oil but it is susceptible to oxidation, because of these reasons pleasant sensory characteristics of oil change to unpleasant ones . The aim of the performed work was to evaluate with an electronic nose changes in odour profile of Arauco and Arbequina varieties of extra-virgin olive oil during deep-frying complementing with conventional analysis. Samples were heated at 180°C during intervals of 60 min until 180 min. Volatile compounds were analysed with SPME-GC-FID. Arauco variety showed a positive correlation between doped SnO2 sensors, being at 120 min and 180 min the highest response for sensors. Arbequina showed a negative correlation between sensors. Arauco variety showed the highest production of volatile compounds at 60 min of deep frying. Arauco and Arbequina varieties showed different volatile production pattern and also could be clearly identified with the electronic nose, showing that electronic nose approach represents an alternative powerful tool to traditional methods of odour measurements.

Odour profile and colour characteristics of waxy brakedown paralysis process in garlic assessed by instrumental methods

Valeria Messina, CINSO, Argentina, ona_2003@hotmail.com

Gabriela Grigioni, ITA, Argentina, ggrigioni@speedy.com Monica Giñazu, UnCuyo, Argentina, guiñazu@citefa.gov.ar

Rosa baby, CINSO, Argentina, vmessina@citefa.gov.ar Noemi Walsoe de Reca, CINSO, Argentina, walsoe@citefa.gov.ar

Waxy breakdown is a physiological disorder, not due to infection by microorganisms, which affects garlic during latter stages of growth and is often associated with periods of high temperature near harvest. Early symptoms are: small, light yellow areas in the clove flesh that darken to yellow or amber with time . Finally the clove is translucent, sticky and waxy, but the outer dry skins are not usually affected. This paralysis affects odour and colour of garlic. Waxy breakdown was studied and separated in two groups according to damage: cloves totally brown color (group a) and cloves totally white color and wool aspect (group b). The odour profile of garlic was measured by an electronic nose comprising 18 semi-conductor oxide metallic sensors (MOS), coupled with a mass spectrometer system. Odour results were analyzed by Discriminant function analysis. Colour determination was done using a BYK-Gardner spectro-guide 45/0 gloss, with D65 illuminant and CIE Lab system. Data analysis of colour was analyzed by ANOVA. Clear separations between the odour profiles of cloves samples were observed. In relation to colour parameters, different behaviors were observed. The methodologies applied have shown the possibility to differentiate garlic sensory quality expressed by odour and colour characteristics.

Odors discrimination by olfactory epithelium biosensor

Qingjun Liu, Department of Biomedical Engineering, Zhejiang University, Hangzhou, 310027, P. R. China qjliu@zju.edu.cn

Ning Hu, Department of Biomedical Engineering, Zhejiang University, Hangzhou, 310027, P. R. China zjubmehuning@gmail.com

Weiwei Ye, Department of Biomedical Engineering, Zhejiang University, Hangzhou, 310027, P. R. China yeweiwei92@gmail.com

Fenni Zhang, Department of Biomedical Engineering, Zhejiang University, Hangzhou, 310027, P. R. China zhangyuxinhaha@163.com

Hua Wang, Department of Biomedical Engineering, Zhejiang University, Hangzhou, 310027, P. R. China biosensor@zju.edu.cn

Ping Wang, Department of Biomedical Engineering, Zhejiang University, Hangzhou, 310027, P. R. China cnpwang@zju.edu.cn

Humans are exploring the bionic biological olfaction to sense the various trace components of gas or liquid in many fields. For achieving the goal, we endeavor to establish a bioelectronic nose system for odor detection by combining intact bioactive function units with sensors. The bioelectronic nose is based on the olfactory epithelium of rat and microelectrode array (MEA) . The olfactory epithelium biosensor generates extracellular potentials in presence of odor, and presents obvious specificity under different odors condition. The odor response signals can be distinguished with each other effectively by signal sorting. On basis of bioactive MEA hybrid system and the improved signal processing analysis, the bioelectronic nose will realize odor discrimination by the specific feature of signals response to various odors.

Odour mapping under strong backgrounds with a metal oxide sensor array

Andrey Ziyatdinov, Department d'Enginyeria de Sistemes, Automà tica i Informà tica Industrial (ESAII), Universitat Politècnica de Cataluna, Pau Gargallo 5, 08208, Barcelona, Spain. CIBER-BBN in

Bioengineering, Biomaterials and Nanomedicine, Spain. c/ Pau Gargallo 5, Barcelona, Barcelona, 08028, Spain

T: +34 93 401 0773, F: +34 93 401 7045, andrey.ziyatdinov@upc.edu José MarÃa Blanco Calvo, Sergi Bermúdez i Badia, Miguel Lechón, School Laboratory for Synthetic

Perceptive, Emotive and Cognitive Systems (SPECS), Universitat Pompeu Fabra, Barcelona, Spain Santiago Marco, Institute for Bioengineering of Catalonia (IBEC), Baldiri Reixac, 13, 08028 Barcelona,

Spain. Electronics Department, Universitat de Barcelona, Martà i Franqués 1, 08028 Barcelona, Spain Paul F.M.J. Verschure, School Laboratory for Synthetic Perceptive, Emotive and Cognitive Systems

(SPECS), Universitat Pompeu Fabra, Barcelona, Spain Alexandre Perera, Department d'Enginyeria de Sistemes, Automà tica i Informà tica Industrial (ESAII),

Universitat Politècnica de Cataluna, Pau Gargallo 5, 08208, Barcelona, Spain. CIBER-BBN in Bioengineering, Biomaterials and Nanomedicine, Spain.

Summary

This text describes the data from an initial set of navigation experiments in the scope of the Bio-ICT European project NEUROCHEM. The acquisition system was composed of two segments, a robotic platform developed in SPECS at UPF and an embedded computer running a custom GNU/Linux distribution developed within the project by UPC (Fig. 1). The embedded computer held a Metal Oxide gas sensor array (TGS262010, TGS260010 and TGS2810 varieties) with a total autonomy of 1.5 hours. The system was placed in a wind tunnel facility in UPF in order to characterize the response of the metal oxide sensor array under the presence of one odour source jointly with a strong background. The compounds used were Ethanol (as background), Acetone and Ammonia at 5%, 11% and 20% dilution in water, respectively. These compounds were diffused in the wind tunnel with help of an ultrasound diffuser at two separate locations (Fig. 2). Four series of measurements were performed aiming to explore the capabilities of the sensor array in constructing the odour map (source 1) in presence of a strong background (source 2) under a controlled environment. Data pre-processing included correction of a certain time delay in sensor response in respect to the robot position in the tunnel and de-noising through a low-pass filter. A separation method based on Independent Component Analysis (ICA) was applied to the sensor data in order to decorrelate the signal form the two sources. ICA assumes a model of mixing, x = As (1) were the sources s=[s1, s2, …, sm]' are mutually independent random variables, and A is an unknown invertible mixing matrix. This algorithm finds a matrix W such that the output y = Wx (2) is a good estimate of the sources s. Pearson variant of ICA finds y through a Mutual Information minimization process [1]. Results of Pearson ICA are able to decorrelate the two odour sources as seen in Figure 3 and 4 for Ammonia and Ethanol, as seen split from fist and second ICA component, and in Figure 5 and 6 for Acetone and Ethanol. Ethanol can be considered as a very strong background as metal oxide sensors are very sensitive to this compound. This poster shows that a preprocessing based on Independent Component Analysis is able to discriminate two odour sources. Further work will include automatic determination of the number of components present in the tunnel and the application of the Neurochem platform in surge-and-cast behavioral models.

References [1] Karvanen and Koivunen, Blind separation methods based on Pearson system and its extensions, Signal Processing 82 (2002) 663-673 [2] Figures are available on the web: http://neurochem.sisbio.recerca.upc.edu/public/conferences/Ziyatdinov-ISOEN-2011-Poster-Abstract.pdf

A portable gas sensor system for environmental monitoring and malodours control: Data assessment of an experimental campaign

Michele Penza, ENEA, Italy, michele.penza@enea.it

Domenico Suriano, ENEA, Italy, domenico.suriano@enea.it Gennaro Cassano, ENEA, Italy, gennaro.cassano@enea.it

Riccardo Rossi, ENEA, Italy, riccardo.rossi@enea.it Marco Alvisi, ENEA, Italy, marco.alvisi@enea.it

Valerio Pfister, ENEA, Italy, valerio.pfister@enea.it Livia Trizio, LENVIROS srl and University of Bari, Italy, liviatrizio@lenviros.it

Magda Brattoli, LENVIROS srl and University of Bari, Italy, m.brattoli@chimica.uniba.it Martino Amodio, LENVIROS srl and University of Bari, Italy, m.amodio@chimica.uniba.it

Gianluigi De Gennaro, LENVIROS srl and University of Bari, Italy, giangi@chimica.uniba.it A portable sensor-system based on solid-state gas sensors has been designed and implemented as proof-of-concept for environmental air-monitoring applications, malodours olfactometric control and landfill gas monitoring. Commercial gas sensors and nanotechnology sensors are arranged in a configuration of array for multisensing and multiparameter devices. Wireless sensors at low-cost are integrated to implement a portable and mobile node, that can be used as early-detection system in a distributed sensor network. Various portable sensor-systems , , have been proposed for environmental air monitoring applications. Real-time and continuous monitoring of hazardous air-contaminants (NO2, CO, PAH, BTEX, etc.) has been performed in field measurements by comparison of chemical analyzers from environmental protection governmental agency (ARPA-Puglia). In addition, experimental campaigns of the integrated portable sensor-system have been realized for assessment of malodours emitted from an urban waste site. The results demonstrate that the sensor-system has a potential capacity for real-time measurements of air-pollutants, malodours from waste site, and control of landfill gas.

Temperature-centric evaluation of sensor transients

Tuba Ayhan, Istanbul Technical University, Turkey, ayhant@itu.edu.tr Kerem Muezzinoglu, University of California San Diego, United States,

kerem.muezzinoglu@gmail.com Alexander Vergara, University of California San Diego, United States,

vergara@ucsd.eduMustak Yalcin, Istanbul Technical University, Turkey, mustak.yalcin@itu.edu.tr

Controllable sensing conditions provide the means for diversifying sensor response and achieving better selectivity. Modulating the sensing layer temperature of metal-oxide sensors is a popular method for multiplexing the limited number of sensing elements that can be employed in a practical array. Time limitations in many applications, however, cannot tolerate an ad-hoc, one-size-fits-all modulation pattern. When the response pattern is itself non-stationary, as in the transient phase, a temperature program also becomes infeasible. We consider the problem of determining and tuning into a fixed optimum temperature in a sensor array. For this purpose, we present an empirical analysis of the temperature’s role on the performance of a metal-oxide gas sensor array in the identification of odorants along the response transient. We show that the optimal temperature in this sense depends heavily on the selection of (i) the set of candidate analytes, (ii) the time-window of the analysis, (iii) the feature extracted from the sensor response, and (iv) the computational identification method used.

Gas identification by dynamic measurements of SnO2 sensors

Juan Vorobioff , Comision Nacional de Energia Atomica , Argentina, vorobioff@cnea.gov.ar Carlos Rinaldi, Comision Nacional de Energia Atomica , Argentina,rinaldi@cnea.gov.ar

Daniel Rodriguez, Comision Nacional de Energia Atomica, Argentina,drodrig@tandar.cnea.gov.ar Alfredo Boselli , Comision Nacional de Energia Atomica , Argentina,boselli@cnea.gov.ar

Alberto Lamagna, Comision Nacional de Energia Atomica, Argentina,alamagna@cnea.gov.ar

Summary It is well known that the use of chambers with the sensors in the e-nose improves the measurements, due to a constant gas flow and the controlled temperature sensors . Normally, the chamber temperature is above room temperature due to the heat generated by the heater of sensors. Also, the chamber takes a long time to reach a stable equilibrium temperature and it depends on environmental conditions. Besides, the temperature variations modify the humidity producing variations in resistance measurements . In this work using a heater system that controls the temperature of the chamber, the desorption process on SnO2 sensor array was study . Also, it was fitted the data signal sensors using a two exponential decay functions in order to determine the desorbing constant process. These constants were used to classify and identify different alcohols and their concentrations. Methods and Results It was developed a teflon chamber sensor of 8 cm3 rectangular shape. A set of heaters were added to the chamber sensor, in order to improve the quality and the performance of the measurements. The temperature was fixed at 43 ° C. It was placed a temperature and humidity sensor. By increasing the temperature of the chamber, the humidity content was reduced and the time desorption of the SnO2 sensors was decreased. This fact resets the baseline faster, reduces the measurement and the drift time. A great improvement was achieved keeping stable chamber temperature and more repetitive measurements were obtained. A set of measurements of ethyl alcohol, isopropyl alcohol, at atmospheric pressure, subjeted to 14 s sampling and 80 s of purge, were done. Eight SnO2 sensors were analyzed and measurements were repeated four times, in all cases similar results were obtained. Due to these improvements it was possible to calculate the first desorption constants and it was plotted as a function of concentration (Fig. 1). Fig.2 shows the resistance logarithm of the sensors as a function of the time, for water vapor and different concentrations of ethyl alcohol. As can be seen the desorption lines are highlited. The constants values of desorption characterize unequivocally each compound and it was used as a method of identification using the PCA algorithm (see Fig 3). Also these constants could be used for training e-nose and determine the concentrations of the alcohol solutions. These measurements not depend on line base. It is not necessary to obtain a permanent response of the sensor, and reduce the measurement times. Therefore the dynamic response of the sensor for measuring and classifying aromas can be used.

Ensemble classifier strategy based on transient feature fusion in electronic nose

Mohammad Ali Bagheri , Islamic Republic of Tarbiat Modares University, Iran, ma.bagheri@gmail.com

Gholam Ali Montazer, Islamic Republic of Tarbiat Modares University, Iran, montazer@modares.ac.ir

Some studies have investigated the performance of different feature types and feature (sensor) selection (FS) methods in sensor arrays . However, few studies have been dedicated to the evaluation of the potential improvement in classification results that could be achieved through feature fusion realized by means of ensemble-based methods. While the FS seeks to find an optimal subset of features, the goal of classifier ensembles is to combine the results of accurate and diverse classifiers to achieve optimal accuracy. In this paper, we test the performance of several ensembles of classifiers, in which each base learner has been trained on different types of extracted features, and then the classifiers are combined by the majority voting method. Experimental results show the potential benefits introduced by the usage of simple ensemble classification systems for the integration of different types of transient features.

Biomimetic transducting support for enhanced explosive detection thresholds

Denis Spitzer, ISL, France, denis.spitzer@isl.eu

This research is devoted to the detection of explosives at ultra-trace levels. For this purpose, a biomimetic concept was used for the first time to design and develop very sensitive and selective transducting supports. A patent is currently under deposition on this innovative approach. For this reason, the architecture of the transducting support will not be given herein but will be further detailed at the symposium. The concept has its origin in the fact that some animal species have the ability to sense only a few molecules. The detection performances of several bio-inspired sensors will be discussed. In a first time, a system for generating vapours of explosives at extremely low concentrations (ppq levels) will be presented (figure 1). This experimental setup was used to generate traces of the most representative high explosives (TNT, RDX, and PETN). In a second step, the affinities of different nanostructured oxides towards these explosives have been investigated on 2 mg of sensing material. It has been shown that the adsorption of explosives on metallic oxides has characteristic properties depending on the oxide/explosive affinity, which is a fundamental point for the development of selective sensors. For example, nanostructured WO3 is more sensitive to RDX than to TNT vapours. This result is illustrated by the longer saturation time observed in the RDX break-through curve (figure 2). In a third step, advanced transducting systems were built from the sensitive nanomaterials. The first results obtained with these bio-inspired sensors will be compared to those reported with existing systems in terms of sniffing abilities.

A flexible gas sensor for the integration into smart textiles

Thomas Kinkeldei, ETH Zurich, Switzerland, kinkeldei@ife.ee.ethz.ch Kunigunde Cherenack, ETH Zurich, Switzerland, cherenack@ife.ee.ethz.ch

Gerhrad Troester, ETH Zurich, Switzerland, troester@ife.ee.ethz.ch The integration of gas sensors or arrays of gas sensors into textiles offers a smart solution to hide electronics in a daily environment. Currently, electronics are mostly integrated into textiles using ICs which provide a large range of electronic functions but alter the textile properties [1] or by weaving or sewing threads with limited functionality (i.e. metallic and optical fibers) into the textile. Examples of smart textile applications include wound monitoring [3] and monitoring of physiological properties such as the body micro climate [2]. We recently developed a textile containing a micro fabricated humidity and temperature sensor on a yarn like strip. Integrated into a curtain or into a textile filter in an air conditioning system one could unobtrusively monitor and control the room climate. One next step is to expand the monitoring function and detect gases or odors in the exhaust air of i.e. industrial production processes. For this purpose we have developed a conductive polymer (Pani) gas sensor on flexible Kapton strips and tested its response to CO. Further we studied the polymer properties during mechanical strain one key issue occurring during the textile integration process. We tested the influence of different applied strain directions to the orientation of the electrodes of the conductive polymer and performed bending tests. Polyaniline (PaniT, Panipol Oy) was spincoated onto interdigitated gold electrodes on Kapton foil substrates and patterned using a lift-off technique. Single sensor units (1mm x 1mm) were designed to fit onto a strip with a width of 1.1 mm and a length of 70 mm. The sensor strips were cut using a dicing saw. For gas measurements we used CO diluted in a concentration from 0-400 ppm in N2 as carrier gas with a flow rate of 500 sccm. The measurements in Figure 2 show a distinct response of a sensor to the concentration of CO at constant temperature. Response of the sensor is about 3% for a concentration of 400 ppm. The sensor signal shows a drift in the baseline value of the resistance over time. For testing the mechanical properties of polyaniline we used a test set-up consisting of two parallel Plexiglas plates mounted onto a linear stage. To test the influence of the strain direction onto the electrode orientation we designed 3 different scenarios: current parallel to the strain direction, current perpendicular to the strain direction and current 45° angle to the strain direction. The strain was applied in the direction of the length of the strip. Every 3 minutes the strain was increased by 1%. In Figure 2 strained strips with the mentioned three orientations are pulled to a final elongation of 10%. Pulling tests show first a decreasing resistance till a strain of about 3%. Further pulling increases the resistance and leads to a rupture of the strip and a loss of conductance at about 45% of strain. The orientation of the electrode shows no significant influence on the applied strain/resistance behavior.

. An on-chip multi-class support vector machine applied to portable electronic nose data classification

Yao-Sheng Liang, Department of Electrical Engineering National Tsing Hua University, Taiwan,

g9761512@oz.nthu.edu.tw Kea-Tiong Tang, Department of Electrical Engineering National Tsing Hua University, Taiwan,

kttang@ee.nthu.edu.tw In this paper, a multiple-class support vector machine chip applied to a portable electronic nose system is presented. The multiple-class method of the classifier was implemented with a “one-versus-one” method, and the kernel function for SVM is Gaussian kernel, which is highly common and usually demonstrates high quality performance. The power consumption of the chip is 118?W; therefore, this low power design is highly suitable for portable applications. The feasibility of this work was verified by classifying fruit gas data, which were collected in a gas experiment. From the post-simulation results, all of the parameters could be successfully trained, and all testing data were dispensed to correct categories.

Towards a low-power miniaturized micromechanical electronic nose

Sywert H. Brongersma, Holst Centre / IMEC High Tech Campus 31, Eindhoven, 5605 KN, Netherlands T: +31404020425, F: +31404020699, brongers@imec-nl.nl

D. Karabacak, J. Pettine, V. Petrescu, M. Patrascu, M Vandecasteele, M. Crego Calama, Holst Centre / IMEC

A novel scalable electronic nose technology that operates at room temperature is demonstrated for low power sensing of volatiles. The approach is based on a miniaturized dense array of high aspect (length/thickness) ratio suspended resonators that can be individually functionalized with polymer coatings. It allows for high yield fabrication in a standard cleanroom environment, followed by application specific inkjet printing of appropriate coatings.

Low-power operation is achieved by using integrated piezoelectric transducers (instead of optics) and a dedicated circuit design for actuation and tracking of the resonant frequency. Also, reversible absorption/desorption kinetics at room temperature render heating of the resonators redundant, leading to an overall power consumption of only a few microwatts.

Detection of sub-ppm level concentrations is enabled by exploiting not only the mass change due to absorption of volatiles (as is common in MEMS devices), but also the much larger stress component due to swelling of the polymers on doubly clamped resonators.

Presently, a first demonstrator with discrete component read-out is used for screening of polymers and further system optimization. The final device, with resonators and dedicated circuitry, will have a size of only a few square millimeters, making it suitable for integration in small autonomous devices for a wide range of applications.

Biosensor based on olfactory receptors immobilization for the detection of odorant compounds

Marta Sanmartí, IBEC, Spain, msanmarti@ibec.pcb.ub.es Patrizia Iavicoli, IBEC, Spain, piavicoli@ibec.pcb.ub.es Edith Pajot-Augy,INRA, France, edith.pajot@jouy.inra.fr Aurora Dols-Perez, IBEC, Spain, adols@ibec.pcb.ub.es Annalisa Calo, IBEC, Spain, acalo@ibecbarcelona.eu Gabriel Gomila, IBEC, UB, Spain, ggomila@pcb.ub.es

Josep Samitier, IBEC, UB, Spain, jsamitier@ibecbarcelona.eu

Summary: The olfactory system can discriminate quickly thousands of odorants at very low concentrations. With the aim of building a sensor that can detect odorants with high specificity and sensitivity, a new generation of artificial noses using olfactory receptors is being developed in our group. The possibility of using olfactory receptors for sensing in electronic noses was demonstrated in a previous work . Following those obtained results, we designed a new artificial olfactory system with the goal of developing a robust and reproducible odor-sensing device. This bioelectronic nose will be capable of detecting specific odorants which can be used in the area of food quality control. Methods and Results: The olfactory receptors specific for the odorants that have to be detected are expressed in yeast cells. After a process of cell disruption and sonication, nanometric liposomes (nanosomes) are prepared. The morphology of the nanosomes is characterized by cryo-fracture Transmission Electron Microscopy and cryo-Transmission Electronic Microscopy. The nanosomes, which carry the olfactory receptors in their active form on their lipidic bilayer, are immobilized on an array of nanoelectrodes. The nanosomes are grafted onto functionalized electrodes in a well controlled, efficient and homogeneous way. The immobilization step is crucial to preserve the functionality of the nanosomes and for the sensitivity, selectivity, reliability, and reproducibility of the results. Atomic Force Microscopy, Surface Plasmon Resonance and Electrochemical Impedance Spectroscopy (EIS) are used to follow the anchoring of the olfactory receptors to the electrodes. Similarly to what has been demonstrated by Y. Hou and colleagues , the odorant detection is electronically performed using EIS measurements. When the receptors bind a given odorant molecule, the electrode can detect electrical changes induced in the olfactory receptors and give a response during the EIS measurements. In a future work the nanoelectrodes will be assembled in an electronic and highly-miniaturized apparatus to acquire and process the signals arising from the nanobiosensor. The characterization of nanosomes and their immobilization onto the nanoelectrodes will be described herein.

Cystic fibrosis sweat patch

Andrea Pagliuca, State University of New York at Stony Brook, United States, nepenthe@optonline.net Gagan Jodhani, State University of New York at Stony Brook, United States, gaganjodhani@gmail.com

Perena Gouma, State University of New York at Stony Brook, United States, pgouma@notes.cc.sunysb.edu

Summary

This work focuses on the novel design for a chloride ion sensor in the form of a skin patch for the diagnosis of cystic fibrosis (CF). LEB-PANI (leucoemeraldine base polyaniline)-based films are employed as the ion-to-electron transducer. High concentrations of chloride ions (> 60mM) in human sweat are accepted as a diagnosis of cystic fibrosis. Current sweat tests involve bulky equipment, require a large amount of sweat and can take over 15 minutes to respond. Based upon our potassium ion band-aid sensor principle, where a skin patch utilizing polyaniline as the electroactive layer, in conjunction with an ion selective membrane is used, a promising a non-invasive, low cost and effective method for the diagnosis of CF is envisioned. Only a small amount of sweat will be required and the patch can be worn over time to collect a sample, minimizing patient discomfort. LEB-PANI-based films that are functionalized in such a manner as to be sensitive to chloride ions have been used. The films showed a change of 8.5MOhm in their electrical resistance of when introduced to 60mM KCl solution. Alumina substrates with gold interdigitated electrodes were employed and the KCl solution was pipetted onto the polymer films to simulate a sweat sample. These results suggest that it might soon be possible to achieve selectivity to Cl- in sweat in concentrations of clinical significance using this biosensor.

Molecularly imprinted polymer based sensor for the detection of theophylline

Guilherme de Souza Braga, EPUSP, Brazil, gbraga@lme.usp.br Leonardo Giordano Paterno, University of Sao Paulo, Brazil, lpaterno@lme.usp.br

Fernando Josepetti Fonseca, EPUSP, Brazil, fernando.fonseca@poli.usp.br Manel del Valle, Universitat Autònoma de Barcelona, Spain, manel.delvalle@uab.es

A molecularly imprinted polymer (MIP) impedance-based sensor was employed to detect theophylline in distilled water. Theophylline is a methylxanthine found in coffee, tea and chocolate . It is also used in lung disease treatments, such as asthma. Impedance measurements in diluted solution of theophylline (1mM) and distilled water were carried out (1Hz to 1MHz) using both MIP and NIP (reference non imprinted polymer) sensors. MIP showed higher sensitivity to theophylline than the NIP. This feature would show their suitability for developing an electronic tongue system for determination of methylxanthines.

Metal ion binding motifs in vertebrate olfactory receptors

Ken Suslick, University of Illinois at Ubana-Champaign, United States,ksuslick@uiuc.edu Jonathan Kemling, University of Illinois, United States,kemling2@uiuc.edu

Zaida Luthey-Schulten, University of Illinois at Ubana-Champaign, United States,schulten@scs.uiuc.edu The sense of smell is arguably the least understood of our senses, and the activation mechanism for olfactory receptors is still unclear. Here we provide evidence that vertebrate olfactory receptors are metallo?proteins, utilizing metal ion ligation in an extracellular loop as a primary interaction with odorants. The metal ion binding motif HXXC[D,E] is present in 62-77% of all mammalian olfactory receptors and QXXC[D,E] is present in 76-91% of birds, while neither is found in other G protein-coupled receptors. This suggests an important role for odorant binding to metal ions in olfactory signaling.

VLSI implementation of a bio-inspired olfactory spiking neural network

Hung-Yi Hsieh, National Tsing Hua University, Taiwan, hyhsieh@larc.ee.nthu.edu.tw

Kea-Tiong Tang, National Tsing Hua University, Taiwan, kttang@ee.nthu.edu.tw

This paper proposes a VLSI circuit implementing a low power, high-resolution spiking neural network (SNN) with STDP synapses, inspired by mammalian olfactory systems. By representing mitral cell action potential by a step function, the power consumption and the chip area can be reduced. By cooperating sub-threshold oscillation and inhibition, the network outputs can be distinct. This circuit was fabricated using the TSMC 0.18um 1P6M CMOS process. Post-layout simulation shows that the proposed SNN could correctly classify at least three odors. The average power consumption for the proposed SNN is less than 2.6uW at a 1V power supply.

Estimation of theaflavins (TF) and thearubigins (TR) ratio in black tea liquor using electronic

vision system

Amitava Akuli, Center For Development Of Advanced Computing, Kolkata , C-DAC (K) ,India,amitava.akuli@cdackolkata.in

Abhra Pal, Center For Development Of Advanced Computing, Kolkata , C-DAC (K) ,India,abhra.pal@cdackolkata.in

Nabarun Bhattacharryya, Center For Development Of Advanced Computing, Kolkata , C-DAC (K) ,India,nabarun.bhattacharya@cdackolkata.in

Arunangshu Ghosh, Jadavpur University, Kolkata, India, arunangshu.ghosh@gmail.com,

Rajib Bandhopadhyya, Jadavpur University, Kolkata, India rb@iee.jusl.ac.in Pradip Tamuly, Tea Research Association (TRA), Tocklai, Jhorhat, Assam, India,

tamulyp@yahoo.co.in

Quality of black tea is generally assessed using organoleptic tests by professional tea tasters. They assess the quality of black tea based on its appearance (in dry condition and during liquor formation), aroma and taste. Variation in the above parameters actually contributed by a number of chemical compounds like, Theaflavins (TF), Thearubigins (TR), Caffeine, Linalool, Geraniol, Phenylacetaldehyde etc. Among the above, TF and TR are the most important chemical compounds, which actually contribute to the formation of colour and brightness in tea liquor. Estimation of TF and TR in black tea is generally done using a spectrophotometer instrument. But, the analysis technique undergoes a rigorous and time consuming effort for sample preparation, also the operation of costly spectrophotometer requires expert manpower. To overcome above problems an Electronic Vision System based on digital image processing technique has been developed. The system is faster, low cost, repeatable and can estimate the amount of TF and TR ratio for black tea liquor with accuracy. The data analysis is done using Principal Component Analysis (PCA), Multiple Linear Regression (MLR) and Multiple Discriminate Analysis (MDA). A correlation has been established between colour of tea liquor images and TF, TR ratio. This paper describes the newly developed E-Vision system, experimental methods, data analysis algorithms and finally, the performance of the E-Vision System as compared to the results of traditional spectrophotometer.

Classification of optical-sensor response cues with a bi-dimensional wavelet-transform approach

Jose S. Murguia, Universidad Autónoma de San Luis Potosí, San Luis Potosí, México,

ondeleto@uaslp.mx Alexander Vergara, Bio Circuits Institute, University of California, San Diego, USA,

vergara@ucsd.edu Marco Trincavelli, AASS Research Centre, Örebro University, Sweden,

marco.trincavelli@oru.se Cecilia Vargas-Olmos, Universidad Autónoma de San Luis Potosí, San Luis Potosí, México,

cv_olmos@cactus.iico.uaslp.mx Ramon Huerta, Bio Circuits Institute, University of California, San Diego, USA,

ramon.huerta@gmail.com In the last years there has been a growing demand for and rapid increase in the quality/variety of chemo-sensing technologies, in which optical-based sensors, in general, and porous silicon film-based chemical sensor, in particular, is gaining more and more importance . Both, the high surface sensing area available and the nano-scale size of the porous have contributed to such a preference. The detection of chemical agents is usually performed by monitoring shifts in the porous silicon reflectance (wavelength) spectrum generated during exposure to parts-per-million levels of airborne chemicals (Fig 1 right panel). However, due to the aspects of the raw signal of the sensor response, there is some valuable information that such a general feature may be overlooking. In this direction, we propose an alternative way to extract more useful attributes for optical-based sensors, based on the 2D-discrete wavelet transform (2D-DWT), that will allow us to have a more in-deep sight of the sensor response. Using a linear support vector machine (SVM) as classifier, we evaluate our approach for a six-analyte discrimination problem.