An electronic nose based on solid state sensor arrays for ...

1 Sensors and Actuators B 101 (2004) 39 46An electronic nose based on solid state sensor arrays forlow-cost indoor air quality monitoring applicationsS. Zampolli , I. Elmi, F. Ahmed1, M. Passini, Cardinali, S. Nicoletti, L. DoriCNR-IMM Sezione di Bologna, Via P. Gobetti 101, 40129 Bologna, ItalyReceived 9 October 2003; received in revised form 9 October 2003; accepted 11 February 2004 Available online 28 March 2004 AbstractThe occurrence of illnesses related with poor ventilation has driven an increasing attention towards indoor air quality monitoring. Inbuildings equipped with climate control systems, the diseases related to the air quality can be significantly reduced if smart interveningprocedures, aiming to control the concentration of pollutants in the indoor air, can be implemented in the heating, ventilation air conditioningunit. When reliable information about both the indoor and outdoor air quality is made available, the climate control system can providethe most appropriate amount of ventilation, ensuring safe and comfortable living this paper, a dedicated, miniaturized, low-cost electronic nose based on state -of-the-art metal oxide sensors and signal processingtechniques was developed.

2 The proposed device is targeted to the quantification of carbon monoxide and nitrogen dioxide in mixtures withrelative humidity and volatile organic compounds by using an optimized gas sensor array and highly effective pattern recognition electronic nose was tested in an environment reproducing real operating conditions. Exploiting the unique response patterns of thedifferent sensors in the array and the capability of a simple fuzzy-logic system it was possible to identify and discriminate concentrationsas low as 20 ppb for NO2and 5 ppm for CO in the test gas environment, allowing to reach the necessary sensitivity towards the targetpollutants together with the selectivity towards the typical interfering gas species. 2004 Elsevier All rights :Indoor air quality; Gas sensors; electronic nose; Fuzzy logic1. IntroductionThe increasing interest in indoor air quality (IAQ) moni-toring is mainly due to the growing incidence of a new classof diseases, identified as building-related illnesses (BRI) andsick building syndromes (SBS), arising from long-term oc-cupancy of confined living spaces, like office buildings orhomes/apartments.

3 The occurrence of these diseases is re-lated with the presence of physical, biological and/or chem-ical contaminants inside the building. The effects on thehuman health due to indoor pollution include pathologiesfor which the etiologic agent has been precisely identifiedand pathologies not directly linked with a specific physical,chemical or biological specie, since the symptoms arise fromthe concomitant effect of several pollutants which are allpresent in the ambient, from low air quality[1]. Chem- Corresponding author. Tel.:+39-051-639-9109;fax:+ (S. Zampolli).1On leave from: Department of Physics, Jahangirnagar University,Savar, Dhaka 1342, contaminants, and in particular gaseous compounds,were found to be among the main responsible ones for boththe BRI and the SBS diseases, since in buildings with inad-equate ventilation they can accumulate with , we can find several hundreds of indoor chem-ical contaminants, including by-products of the combustion(NO2,SO2, CO, etc.)

4 , cigarette smoke, particulate matter,mineral fibers and a number of volatile organic spite of the very low concentrations, some of these com-pounds are extremely toxic, like NO2or CO; some other, likebenzene and formaldehyde, were proved to be , the monitoring of the air quality is of paramountimportance to keep safe and healthy buildings equipped with heating ventilation air condi-tioning (HVAC) systems, the pollutants are diluted by ven-tilation, which is generally operated on the basis of fixedduty cycles. This approach does not necessarily ensure animprovement of the indoor air quality, since in many cities fresh outdoor air can contain many pollutant species atconcentrations higher than the threshold values. Further-more, an uncontrolled preventive increase of the ventilationresults in an increment of the overall energy consumption,0925-4005/$ see front matter 2004 Elsevier All rights Zampolli et al.

5 / Sensors and Actuators B 101 (2004) 39 46 Table 1 Typical indoor air quality contaminantsAldehydesOzoneNitrogen dioxideCarbon dioxideParticulate matterRadonCarbon monoxideSulfur dioxideTotal VOCF ormaldehydeWater vaporLeadespecially when the temperature gap between indoor andoutdoor air is non-negligible[2].Several attempts to implement demand-controlled venti-lation have been reported in literature. Many of them rely onthe quantification of CO2, used as a tracer of human occu-pancy in confined living spaces or as surrogate of inhabitantgenerated pollution. However, this approach is inadequateto monitor IAQ, since several toxic compounds released bybuilding materials and furnishings or generated by humanactivities are also present in the indoor environment. Sinceeach compound has a different impact on human health,it is important to monitor their concentrations 1reports a list of the most common indoor contami-nants found in confined living spaces, which can be used astracers for the IAQ[1,3,4].

6 For this reason, the availability of reliable, low-cost sen-sors suitable to monitor both the indoor and outdoor airquality would allow the implementation of smart HVAC in-tervening procedures, considering not only the typical infor-mation about temperature and relative humidity but also theconcentration of a number of compounds used as air qual-ity tracers. This way, by driving the ventilation systems ondemand to keep the IAQ under control, it would be possibleto maintain acceptable healthy and comfortable conditionswhile minimizing the overall power consumption. Being theinformation about the quality, the humidity and the temper-ature available for the inside and the outside air, the climatecontrol system would intake fresh air, recycle the indoor airor operate active scrubbers, depending on the most favorableenergetic low-cost tools for IAQ analyses are nowadays com-mercially available, but they are not suitable to provide re-Table 2 Some gas species considered as IAQ tracers and their target concentration rangesCompoundConcentration rangeIndoor threshold values (8 h exposure)CO5 30 ppm9 ppmNO220 200 ppb53 ppbaVOC (benzene, toluene,m-xylene)60 600 vapor (RH)5 95%No thresholdaUS-EPA suggested value for 24 h exposure outdoor threshold.

7 No indications given for indoor standards have been set for VOCs in non-industrial indoor information, since the various pollutants are not pre-cisely identified and quantified. These devices usually giveindications about the overall IAQ, without any estimation ofthe concentration of each pollutant. In fact, only the detailedinformation about the concentration of the single gas wouldallow an optimized control of the HVAC system, includingthe activation of scrubbers to catalytically convert some pol-lutants, like, volatile organic compounds, in CO2andwater this work, an electronic nose (e-nose) based on a solidstate gas sensor array for the identification and the quantifica-tion of two typical indoor air quality tracers is presented. Thesimplicity of the proposed approach, which uses stand-alonegas sensors, simple driving electronics and fuzzy-logic pat-tern recognition algorithms, aims to realize a low-cost toolsuitable to monitor some of the compounds of interest forIAQ.

8 The availability of this type of devices is fundamentalfor a capillary analysis of the pollution level inside build-ings equipped with forced ventilation and air climate con-trol units. To validate the approach, only CO and NO2were taken into consideration as target compounds, whilesome VOCs and the humidity were considered as ExperimentalThe aim of this work is to develop a reliable tool for IAQmonitoring using a very simple system architecture basedon a stand-alone gas sensor array, driving electronics andsuitable pattern recognition algorithms, without any fluidiccomponents (valves, filters, pumps, etc.). Furthermore, thesystem has been designed to integrate additional sub-unitsable to monitor other parameters relevant for IAQ. To val-idate this approach, the number of compounds consideredwithin this work was limited to two gases chosen amongthose reliably detectable by metal oxide solid state gas 2reports these gases together with the target con-centration ranges considered for effective IAQ monitoringand the relevant indoor threshold values as defined in[3].

9 Inspite of their relevance for indoor comfort, some other com-pounds, like carbon dioxide (CO2), were not considered herebecause their detection is problematic with metal oxide gassensors. However, other sensing devices capable to detectthese compounds are nowadays commercially available andthey could be easily integrated into a modular Zampolli et al. / Sensors and Actuators B 101 (2004) 39 4641 Fig. 1. Schematic overview of the considered e-nose system developed for this work consists of an arrayof thin-film metal-oxide gas sensors, driven by ad hoc elec-tronic circuits suitable to power and measure each deviceindependently. The gas sensors were operated under differ-ent conditions and they provided distinct response commercial RH sensor monitored the relative humidity,which interferes in both CO and NO2detection. The discrim-ination between the different target gases and their quantifi-cation were achieved using state -of-the-art data processingand pattern recognition algorithms, which rely on the dis-tinct response patterns provided by the sensors.

10 In fact, thelow selectivity of the solid state gas sensors has been com-pensated by the combination of the response patterns, al-lowing the quantification of the single gas species. In thiswork, a fuzzy-logic system was used for pattern recogni-tion. A similar approach, based on a multi- sensor array andusing an artificial neural net (ANN) for pattern recognition,has been proposed in[5].A schematic drawing of the unit is shown inFig. 1. Thissimple architecture is targeted for the detection and quantifi-cation of CO and NO2, but it does not allow a precise quan-tification of volatile organic compounds (VOC), because ofthe lack of selectivity in metal oxide semiconductor gas sen-sors. We are currently addressing this problem following anapproach based on a gas-chromatographic architecture real-ized by MEMS technologies and developed for the detectionand quantification of VOC[6], which will be described ina future paper.