Review Biosensors in air monitoring† K. J. Mattias Sandstro�m*abc and Anthony P. F. Turnerc aUmea° University, Department of Public Health and Clinical Medicine, Occupational Medicine, S-905 81 Umea° , Sweden. E-mail: mattias.sandstrom@niwl.se; Fax:+46 90 786 50 27, Tel:+46 90 786 90 70 bNational Institute for Working Life, Department of Chemistry, P.O. Box 7654, S-907 13 Umea° , Sweden cCranfield Biotechnology Centre, Cranfield University, Cranfield, Bedford, UK MK43 0AL Received 9th April 1999, Accepted 18th June 1999 1 Introduction ments, and such instruments can be developed using biosensor 2 Methane technology. Biosensors can also oVer greater selectivity than 3 Carbon monoxide the commonly used direct-reading instruments.Many attempts 4 Formaldehyde have been made to produce biosensors with these kinds of 5 Ethanol characteristics for air monitoring and biosensors are now 6 Phenol important tools that can successfully be used for air 7 Pesticides and other hazardous chemicals monitoring. 8 Odours The first biosensor designed for air monitoring was described 9 Other sensors in 1974 in a paper by Goodson and Jacobs.1 It was a sensor 10 Conclusions for toxins that inhibit the cholinesterase enzyme.The sensor 11 Acknowledgements was used by pumping air, together with a reagent solution, 12 References through an electrochemical cell with an enzyme pad between two electrodes. The pad contained cholinesterase, which is Mr Sandstro�m gained aMSc inhibited by some chemicals. If the air sample contained a in chemistry at Umea° cholinesterase inhibitor, competition between the reagent and University in 1995.His PhD- the inhibitor occurred. Since the reagent produced an easily studies begun in 1997 with a oxidised product after reaction with cholinesterase and the project aiming to develop an inhibitors did not, the measured potential increased when an inexpensive, easy to use per- inhibitor was present.The sensor was produced for both air sonal exposure monitor using and, with some modifications, water sampling and could be biosensor technology. The used for 8 h without changing the enzyme pad. project is a collaboration between Umea° University Sweden, National Institute 2 Methane for Working Life, Sweden Another early biosensor for air monitoring used a reactor and Cranfield University, containing immobilised methane consuming micro-organisms UK.The National Institute for Working Life has a long to sample methane in air and was described by Okada and experience of air sampling co-workers in two similar articles.2,3 The sample gas was techniques, both active and pumped through the reactor and to an oxygen electrode. A reference reactor was also used to monitor the oxygen concen- passive sampling.At Cranfield University the research and development of biosensors has been successful for many years. tration in a reactor without micro-organisms. Microbial metab- Mr Sandstro�m’s research area is mainly the combination olism of methane requires oxygen resulting in a decreased of specific biological analysing methods incorporated into concentration of dissolved oxygen, which was monitored by sampling devices to simplify exposure measurements in the oxygen electrode.The sensor had a response time of 1 min occupational environments. and was said to give a constant response for 20 d. The minimum concentration of methane in air that could be detected was calculated to 13.1 mM and the linear range was up to 6.6 mM.No interfering agents were examined, however, 1 Introduction as the micro-organisms were said to use methane as their only The methods that are commonly used in air sampling today source of energy. sometimes lack some of the properties that would be considered favourable for a certain application; such properties might include high selectivity, high sensitivity, real-time moni- 3 Carbon monoxide toring, inexpensive analysis, one-step analysis, etc.Commonly An enzyme-based carbon monoxide sensor was described by used chromatographic methods usually require an air sampling Turner and co-workers.4,5 The biosensor was based on the procedure prior to the analysis. This time-consuming two-step oxidation of carbon monoxide to carbon dioxide by the analysis can be avoided by using on-site, direct-reading instruenzyme carbon monoxide oxidoreductase.Carbon monoxide oxidoreductase was placed on a conducting gel and covered with a membrane. The conducting gel consisted of graphite, †Presented at AIRMON ’99, Geilo, Norway, February 10–14, 1999. J. Environ. Monit., 1999, 1, 293–298 293mediator and liquid paraYn, and was in contact with a dehyde and requires visual comparison with a colour code that is supplied with the device.platinum electrode. The gas permeable membrane was used to keep the enzyme at the surface and to make it possible for In 1996, three biosensors for monitoring formaldehyde in air were described. Ha�mmerle and co-workers described a carbon monoxide to pass through to the enzyme. 1,1¾- Dimethylferrocene was used as a mediator and was oxidised biosensor based on an electrochemical cell divided into two parts by a dialysis membrane to prevent migration of the at the electrode surface at 150 mV versus Ag/AgCl.The amperometric response reached a steady-state current in less enzyme.9 FADH was put on the working electrode and prior to sampling the cell was filled with electrolyte containing than 15 s and the current in device decrease by 12% per hour.Most of the reported experiments, however, were performed cofactor and mediator. Since the electrolyte was not added until the time of sampling, the device could be stored for a in solutions. long time. The device was tested in a controlled atmosphere by measuring the equilibrium gas phase above an aqueous 4 Formaldehyde formaldehyde solution and the limit of detection was 0.3 ppm. A linear response was achieved up to 6 ppm using steady state Monitoring formaldehyde is of great importance since it is widely used in industry.It is also a known irritant and a measurements and the range was improved when initial rate data were used. The device could also be used for 7 h without possible carcinogen. In 1983, Guilbault described a biosensor for the determination of formaldehyde in air using formal- any loss of activity.Another biosensor for formaldehyde was described by dehyde dehydrogenase (FADH) coated on a piezoelectric crystal.6 The piezoelectric crystal technique has been widely Dennison and co-workers, who utilised enzymes and cofactors immobilised in a reversed micelle medium on screen-printed used in the construction of biosensors and is based on crystals that oscillate at a certain frequency when they are exposed to electrodes.10 The biosensor used FADH for the determination of formaldehyde but a similar construction was also used to an electric field.If the mass on the crystal changes, a shift in the frequency can be observed.Biological material can thereby determine alcohols with ADH. The re-oxidation of NADH to NAD+ was measured amperometrically at 0.8 V versus be attached to the surface to create a piezoelectric biosensor and the change in, or adsorption on, the biological material is Ag/AgCl. The reversed micelle medium was used to prevent water loss as the silicone oil acted as a barrier against monitored by measuring the frequency shift.The oxidation of formaldehyde to formic acid catalysed by FADH necessitated evaporation. The biosensor was found to be suitable for gasphase sensing when it was tested in controlled atmospheres. the presence of NAD+ and reduced glutathione. The stability for this device was, according to the author, 3 d or 100 Formaldehyde permeation tubes and ethanol diVusion vials, connected to a gas rig, were used to create the atmospheres.analyses, but this could be increased to 10 d if the enzyme and cofactors was chemically bound to the crystal surface. The gas concentrations were calibrated using Dra�ger tubes or by measuring the loss of sample gravimetrically. The linearity However, this was not recommended since the crystal in this case would not be reusable. The response to formaldehyde biosensors was estimated to be 1.3 ppb–1.2 ppm for formaldehyde and 50–250 ppm for ethanol and the biosensor air was linear from 10 ppb to 10 ppm.The test atmosphere was generated by a syringe injection, of a known volume of could be stored for 60 h at 4 °C without a decrease in response.An ion-sensitive field-eVect transistor (ISFET) was used by gas, into a controlled airflow and validated by formaldehyde sampling tubes and fluorimetric analysis. The biosensor was Vianello and co-workers in biosensor for formaldehyde measurement.11 The ISFET monitors H+ produced when specific to formaldehyde. No significant interference was seen from other aldehydes or alcohols.It is notable that no further formaldehyde is oxidised by FADH with NAD+ as a cofactor. The formaldehyde was removed from the atmosphere by development of this biosensor has occured. However, other researchers have used FADH in devices to monitor formal- pumping air through a glass coil together with an aqueous solution. The solution dissolved the formaldehyde and acted dehyde in air.FADH was one of the enzymes used in the diVusion badges as a carrier of the formaldehyde to the ISFET. A membrane containing FADH covered the ISFET and the solution con- developed by Rindt and Scholtissek.7 They used various enzymes lyophilised on to sintered glass rods put into vessels taining formaldehyde was transferred directly to the surface. The enrichment factor of this sampling technique was containing buVer–reagent solutions and covered with gas permeable membranes.The diVusion badges contained the 8000-fold but there were some problems with the immobilisation of the enzyme, which complicated the evaluation of buVer solutions to overcome the problem of drying. Since the device was constructed from two parts, the glass rod with the sensor. enzyme and the vessel with buVer–reagent solution, they could be stored separately.The enzyme could be stored dry, which 5 Ethanol increased the storage time. In addition to formaldehyde, the compounds determined using this type of construction were Detecting ethanol in air has an important application in determining breath alcohol. Barzana and co-workers devel- hydrogen peroxide, acetaldehyde and ethanol and the enzymes used were diaphorase, aldehyde dehydrogenase, alcohol oped a device which changes colour when it is exposed to ethanol.12 The detection was based on a visual observation of dehydrogenase (ADH) and horseradish peroxidase.The reaction between analytes and enzymes caused a dye to change a colour change. This gives a crude indication of the amount of alcohol in the breath.For quantification, the device was colour. This colour change was documented photographically and the colour was stable for hours after exposure. The gas tested using a densitometer for a more precise determination of ethanol vapour. The device was constructed by adding mixtures, used to test the badges, were generated with a perfusion vessel at controlled temperatures and verified with alcohol oxidase (AOD), peroxidase (POD) and 2,6-dichloroindophenol (DCIP) to microcrystalline cellulose.diVerent types of enzymatic reactions. The constant flow of buVer–reagent to the top of the device not only kept it moist The ethanol was oxidised by AOD and acetaldehyde and hydrogen peroxide were produced. The hydrogen peroxide but also concentrated the enzyme at the top of the glass rod, because the water slowly but constantly evaporated through reacted with POD and at the same time, the reduction of DCIP caused a colour change in the device.To make the the gas permeable membrane. The technique used in this device was later developed into a commercial product called device fast and simple to use it was optimised to give a sharp colour change after 1 min if the ethanol concentration was Bio-Check F (Dra�gerwerk AG, Lu�beck, Germany).8 The Bio- Check F is used for quantitative measurements of formal- over the legal limit for driving.Since AOD also has the ability 294 J. Environ. Monit., 1999, 1, 293–298to oxidise formaldehyde it can also be used to detect this NADH and placed in cuvettes.Fluorescence from NADH was measured with a spectrophotometer. Since the enzymatic reac- compound. However, for this application both methanol and ethanol would be serious sources of interference. tion was based on the fact that the reaction between alcohol and ADH with NAD+ as cofactor gave the corresponding A sensor for determination of alcohol and sulfur dioxide in air was described by Matuszewski and MeyerhoV.13 It was aldehyde and NADH in a reversible reaction, it would be possible to detect both alcohol and aldehyde.It could also be mainly constructed for the continuous electrochemical detection of hydrogen peroxide. By dissolving gaseous H2O2 in a possible to regenerate the sensor after exposure to one species by exposing the sensor to the other species.The device was buVer solution using coiled tubing with an internal buVer flow the H2O2 concentration could be measured when the buVer tested by gas-phase exposure to ethanol-containing gasoline and human breath containing ethanol. In both cases a detect- was pumped over an electrochemical cell. By adding an enzyme reactor containing H2O2-producing enzymes, such as AOD or able diVerence to ethanol-free control samples was achieved.sulfite oxidase (SOD), prior to the electrochemical cell, dissolved alcohol or sulfur dioxide could be detected. To increase 6 Phenol the sensitivity of the continuous flow measurements, a stopped- flow approach was investigated. The buVer flow through the Phenol is a chemical widely used in industry and exposure to phenol is known to cause irritations.Air monitoring of phenol coiled tubing was stopped for a certain time, which accumulated the compounds of interest during sampling. The ethanol is therefore very important. Saini and co-workers investigated the possibility of using biosensors to monitor phenol in air.17 and sulfur dioxide atmospheres were generated with a permeation tube and a commercial gas emitter, respectively, and An interdigitated microband electrode was chosen as the transducer and polyphenol oxidase (PPO) was immobilised diluted with air.The limit of detection was calculated for sulfur dioxide as 0.50 ppb with continuous flow and 0.15 ppb on the electrode in two diVerent materials, Nafion and tetrabutylammonium toluene-4-sulfonate (TBATS). Various electro- with 2 min stopped flow and for ethanol as 1.0 ppb with continuous flow and 0.5 ppb with 2 min stopped flow.After chemical techniques were used to investigate the device with respect to parameters such as thermodynamics and kinetics. storage for 2 weeks the SOD reactor lost more than 50% of its activity whereas the AOD reactor kept its activity for more With a view to health and safety monitoring, Dennison and co-workers further developed the biosensor for phenol.18 Their than 1 month.Mitsubayashi and co-workers constructed another type of device was constructed by immobilising PPO on a gold microelectrode using a glycerol-based gel. The phenol vapour reacts biosensor for the determination of ethanol in air.14 A reaction cell consisting of both gas- and liquid-phase compartments with the enzyme and the product (catechol ) takes part in a redox recycling reaction at the electrode surface.The authors separated by a diaphragm membrane was used in the sensor. AOD was immobilised in a cross-linked acrylamide gel and reported that good sensitivity was achieved partly by this recycling of the catechol–quinone redox couple. The limit of placed on a Clark-type oxygen electrode and covered with a polycarbonate membrane.The ethanol atmosphere in the test detection was estimated to be 29 ppb and the response was linear up to 13 ppm, both at 40% relative humidity. The chamber was generated by a gas generator connected to a computer controlled mass flow system and the calculated phenol atmosphere was generated with a phenol high-emission permeation tube mixed with humidified air.The phenol con- atmosphere was compared with a commercially available semiconductor gas sensor. According to the authors, the centration wasified with a method using an impinger to trap phenol, which was determined spectrophotometrically. biosensor measured ethanol down to 0.357 ppm and had a linear response from 1.57 to 41.5 ppm for steady state measure- Since glycerol is hygroscopic it had the ability to maintain the water content of the gel.The glycerol gel was also particularly ments and from 15.7 to 1242 ppm for maximum response slope measurements. The biosensor response decreased with suitable for phenol determination because of its ability to concentrate phenol in the biosensor.time. After 4 d, the output was 25% of the initial response. Interferences were measured for only a few compounds and In two papers, Kaisheva and co-workers described a biosensor for monitoring phenol in both the liquid and gas did not include any of the compounds known to react with AOD (methanol, propanol, formaldehyde, etc.). phase. The first paper19 mainly described the performance of the sensor in the liquid phase but preliminary experiments in In 1995, another biosensor for ethanol vapour was developed by Park and co-workers.15 This sensor, which was also con- the gas phase was also described.The second paper20 described experiments performed in the gas phase, also investigating structed mainly for measuring breath alcohol, used ADH and NAD+ immobilised on screen-printed electrodes with a mix- p-cresol and 4-chlorophenol vapours.The enzyme used in the sensor was tyrosinase, which catalyses both the reaction of ture of hydroxyethylcellulose, ethylene glycol and carbon powder. Ethanol reacts with the enzyme and at the same time phenol to catechol and the reaction of catechol to o-quinone. The electrochemical reduction of o-quinone back to catechol the NAD+ is reduced to NADH.NAD+ is then regenerated at the electrode surface from NADH. This amperometric then produced a measurable signal at the electrode. The sensors were evaluated in the gas phase over aqueous samples regeneration of NAD+ was carried out at a potential of 0.65 V versus Ag/AgCl. The simple and inexpensive technique of but the gas-phase concentrations were not calculated.However, linear calibration curves were achieved in the range screen-printing makes the sensors both disposable and easy to mass produce. The storage stability was dependent on the 5×10-7–1×10-4 M for phenol, 5×10-5–5×10-3M for pcresol and 5×10-4–5×10-2 M for 4-chlorophenol for the amount of enzyme in the biosensor. However, it could be stored for more than 35 d if the ADH/NAD+ ratio was >6.aqueous standards. The sensor could also be stored for 20 d without losing its activity. The sensor had a linear response up to 250 ppm and the vapour was generated by bubbling nitrogen through an ethanol solution. Preliminary tests were performed with people mainly 7 Pesticides and other hazardous chemicals to investigate if there were any interfering compounds in human breath.No interference was found. Monitoring of pesticides has long been of interest since they are highly toxic and widely used. Biosensors for monitoring Williams and Hupp developed a sensor based on sol–gel encapsulated ADH for the determination of alcohols and pesticides have mainly been developed for the liquid phase but some sensors have been described for gas-phase monitoring.aldehydes in air.16 The sol–gel consisted of silica and the enzyme was encapsulated in a transparent material. The ADH Ngeh-Ngwainbi and co-workers used antibodies against parathion (a known pesticide) attached to a piezoelectric crystal to was immobilised in the sol–gel with the cofactors NAD+ and J. Environ. Monit., 1999, 1, 293–298 295monitor parathion in air.21 The response to parathion was remained after 10 d.Tetrachloroethylene was used as an example of a compound in aerosol form. The aerosols were fast, usually 1–2 min, and the time to return to the baseline was 2–5 min. The linearity of the device was shown to be in produced with an atomiser, and a fan in the sampling chamber distributed the aerosols.The limit of detection for tetrachloro- the range 2–35 ppb. However, since there was a problem with the generation of parathion at higher concentrations, experi- ethylene was calculated to be 10 ppm and a linear calibration curve was achieved in the range 0–250 ppm. The device was ments were performed only up to 35.5 ppb. The test atmosphere was generated by bubbling carrier gas through a trap said to be suitable as an early warning system to protect workers from harmful chemicals, particularly where the chemi- containing liquid sample.The vapour-saturated carrier gas was then diluted with pure carrier gas and the concentration cals have not yet been identified. of the sampling atmosphere was verified using gas chromatography. Some interferences were seen from other pesticides, but 8 Odours with lower responses.It took between 3 and 20 times more of the diVerent interferents to give the same response as para- When odours are detected, the sensor is usually referred to as an electronic nose. A device using coated piezoelectric crystals thion. The lifetime of the crystals was approximately 1 week, after which the response decreased rapidly.was developed by Okahata and Shimizu to detect odours and perfumes in the gas phase.29 DiVerent coatings were tested Non-specific adsorption of compounds on antibody coated piezoelectric crystals was addressed and utilised by Rajakovic and it was found that a lipid bilayer had the best characteristics for the odours, represented by b-ionone. The response time and co-workers.22 They used piezoelectric crystals coated with diVerent proteins (valproic acid antiserum, parathion antibody, for the device was 5 min when exposed to a saturated atmosphere of b-ionone.IgG and bovine serum albumin) and exposed these to atmospheres containing diVerent hazardous compounds (valproic Piezoelectric crystals coated with four diVerent lipid films for the detection of odours represented by eight organic acid, o-nitrotoluene, toluene, parathion, malathion and disulfoton).The results showed that there was a higher sensitivity compounds (e.g., amyl acetate, b-ionone, methanol ) were described by Muramatsu and co-workers.30 The crystals were to the three pesticides (parathion, malathion and disulfoton) than valproic acid, o-nitrotoluene and toluene.The pesticides fixed in a vessel and the samples were injected into the vessel as liquids. The patterns for the diVerent odours were then also adsorbed better on an uncoated crystal. This was explained by the ability of organosulfur compounds to chemisorb normalised and compared. Mixtures of asolectin and cholesterol were used by strongly to metal surfaces. However, this does not explain the higher sensitivity of the sensors towards pesticides. The paper Muramatsu and co-workers in a device developed for odour recognition.31 The lipids were coated on piezoelectric crystals described the problems of non-specific adsorption to antibodies when used in gas-phase monitoring and demonstrated that it and the frequency shift was measured when the odours, represented by eight organic compounds, adsorbed on the was important to consider these kinds of interactions, but it also demonstrated that it was possible to construct non-specific coated crystal.The odours were vaporised by injecting liquid sample into the vessel in which the crystal was positioned and antibody biosensors for air monitoring. Another simple device is the C-probe film badge described the resonant frequency and the resonant resistance was measured both before and after injection.The patterns for the by Case and Crivello.23 This device required visual observation and might be suitable as a hazard indicator. The device was diVerent compounds were then compared with the aim of recognising the odours. described as a biological layer between a film base and a layer of dye.The chemical agents reacted with the biological layer Wu described a device for odour detection using olfactory receptors coated on a piezoelectric crystal.32 Olfactory receptor and were converted into active intermediates that triggered a colour change in the dye. The badge was said to respond to proteins (ORP) were used in an attempt to mimic the human sense of smell. Crude ORPs and ORPs fractionated into five 130 organic and inorganic compounds with a high correlation to carcinogenic hazards.It could be stored for 3 months and groups were coated on the crystals to establish the patterns from six organic compounds (e.g. caproic acid, isoamyl acetate, the sampling time was up to 8–15 h. In three papers, Albery and co-workers described an linalool ) used as odours.According to the authors the sensor did not lose sensitivity after storage for 5 months and it could inhibited enzyme electrode. The three papers deal with (a) a theoretical model for an electrochemical sensor measuring the be used continuously for 10 weeks without a decrease in sensitivity. inhibition of the enzyme activity,24 the kinetics of the cytochrome c and the cytochrome oxidase enzyme systems25 and (c) a description of an application where the sensor was 9 Other sensors used to analyse HCN and azide ion in the liquid phase and H2S in the gas phase.26 The sensor was based on the inhibition Okada and co-workers developed a biosensor for the determination of NO2 in air.33 Nitrite oxidising bacteria were immobi- of the enzyme cytochrome oxidase.H2S inhibited the enzymatic reaction producing a decrease in the current from the lised on an acetylcellulose membrane. The membrane was then attached to an oxygen electrode and covered with a gas- gold electrode. In the gas phase H2S could be measured down to 1 ppm and the linearity was said to be good up to 20 ppm. permeable Teflon membrane.The sample was prepared in a gas bag, pumped into the system and dissolved in a buVer Naessens and Tran-Minh described a whole-cell biosensor that can be used to monitor organic compounds in both which was pumped through the sample cell of the biosensor. The decrease in oxygen, caused by an increased activity of the vapours27 and aerosols.28 The sensor used a Clark oxygen electrode to monitor the oxygen produced during the photosyn- micro-organisms when NO2 was present, was measured.The minimum concentration that could be determined was calcu- thesis of immobilised micro-algae. When the algae were exposed to the organic compounds, as vapour or aerosol, lated as 0.51 ppm and the calibration curve was linear below 255 ppm. The sensor was said to be re-usable for 400 assays photosynthesis was inhibited and a decrease in oxygen was measured.The algae were flashed with an external light source or 24 d and to respond only to NO2. No experiments were performed with other inorganic gases (e.g. NO, SO2, NH3). for 1 min every 5 min to start the photosynthetic process. Methanol was used as an example of a gaseous compound A biosensor for nitrogen monoxide was described by Aylott and co-workers.34 It consisted of a sol–gel containing and the sampling was carried out in a thermostated cell with a gas–liquid equilibrium.The calculated detection limit for cytochrome c spin-coated on to a glass substrate. A gas flowthrough cell covered the sol–gel for the gaseous sample to methanol was 30 ppm and more than 50% of the algal activity 296 J.Environ. Monit., 1999, 1, 293–298come in contact with the enzyme. When the NO attached to ethanol and the signal could be measured as a steady-state current or as initial reaction rate. cytochrome c, a shift of the absorption wavelength occurred, which was measured spectrophotometrically. Since the bond between NO and cytochrome c was reversible, the sensor could 10 Conclusions be used for repeated exposures of NO.The standard deviation When looking at biosensors for air monitoring, the conclusion was calculated to be 1% of the response when five repeated can be drawn that this application has not attracted the same exposures to 10 ppm of NO were made. The limit of detection attention as other areas. This can possibly be attributed to the was calculated to be 1 ppm and the range of detecting NO fact that the major commercial area for biosensors has so far was 1–25 ppm.The authors found no evidence of interference been the field of medicine. However, biosensors are steadily from oxygen, nitrogen or carbon monoxide. However, NO2 being developed into useful tools for air monitoring but there was found to bind to cytochrome c and therefore to give rise are still some requirements that have to be fulfilled for them to interference.The authors therefore concluded that to be accepted as instruments for air monitoring. There are, cytochrome c only could be used for detecting NOx and not for instance, few articles that have a well described system for NO selectively. generating test atmospheres and there is generally a lack of A sensor for monitoring sulfur dioxide in air was described reliable reference methods to determine the gaseous concen- by Matuszewski and MeyerhoV.13 It was also used to monitor trations that are used for the tests.In addition, most of the alcohol and it has been mentioned earlier. A gas-phase sensors have not been suYciently validated, which is a require- biosensor for the direct determination of gaseous sulfur dioxide ment from the EU when it comes to developing devices for in the atmosphere was also developed by O’Sullivan.35 A air monitoring.38–40 mixture of agarose and carboxymethylcellulose was chosen The sensors themselves can oVer exquisite sensitivity and from a range of matrices as the medium for immobilisation of specificity, but the instability of isolated biological systems is sulfite oxidase.Agarose (1% w/v) and carboxymethylcellulose aggravated by the need to operate in air. Nevertheless, the (1% w/v) retained a relatively high proportion of water over literature shows a number of innovative approaches to engina 3 h period, thus preventing enzyme dehydration and allowing eering solutions to those problems and niche applications of eYcient dissolution of SO2.The sensor method was compared biosensors for air monitoring that can be expected to materialwith the standard method for SO2 determination and a corre- ise as a commercial reality in due course. The field of biosensor lation coeYcient of 0.999 was obtained, indicating eYcient research and development is rapidly expanding.The use of dissolution of SO2 in the matrix, accurate production of air– biosensors in air monitoring is mainly targeted on real-time sulfur mixtures by the gas rig and eYcient functioning of the devices for monitoring atmospheric pollution or research biosensor. The reproducibility of the biosensor was stated to applications, but there is also a need for fast and simple be extremely good; an RSD of 0.96% was obtained for n=10.measurement devices for personal exposure measurements in The linear range was 0–13.5 ppm and the LOD was 73.9 ppb. occupational environments; biosensor technology can be an There is a need to monitor not only chemical compounds important tool for this purpose. Another major advantage in air but also micro-organisms in air.In some industrial with biosensors is that the manufacturing process can be environments workers can be exposed to high levels of micro- inexpensive owing to the mass-production technology that is organisms. Biological warfare is another application that is in now widely available. Increased demand for more frequent need of fast sampling methods for micro-organisms in air.For and more varied analyses in the workplace can be expected to this purpose, Ligler and co-workers developed a light-weight catalyse biosensor developments in this area and we can expect biosensor that was monitored in a remotely piloted aero- to see commercially available devices in due course. plane.36 Aerosolised bacteria were sampled using a plastic cyclone air sampler, with a constant addition of buVer solution. 11 Acknowledgement A portion of the liquid sample was pumped over an optical fibre coated with polyclonal antibodies against the bacteria in The authors are indebted to Professor Jan-Olof Levin and question. The micro-organisms attached to the antibodies and Dr. Anna-Lena Sunesson for valuable discussions. a reagent solution containing fluorescent-labelled antibodies against the same bacteria was pumped over the optical fibre.References The fluorescent signal from the labelled antibodies was meas- 1 L. H. Goodson and W. B. Jacobs, in Enzyme Enginneering, ed. ured with a laser fluorimeter. All this equipment was mounted E. K. Pye and L. B. Wingard, Jr., Plenum Press, New York, 1974, in the aeroplane to collect, identify and transmit continuous vol. 2, pp. 393–400. information to an operator on the ground. The system was 2 T. Okada, I. Karube and S. Suzuki, Eur. J. Appl. Microbiol. tested by releasing harmless bacteria in various amounts in Biotechnol., 1981, 12, 102. 3 I. Karube, T. Okada and S. Suzuki, Anal. Chim. Acta, 1982, 135, the air while sampling with the aeroplane. The detection limit 61.of the fibre optic probe was calculated to be 3000 cfu ml-1 4 A. P. F. Turner, W. J. Aston, I. J. Higgins, J. M. Bell, J. Colby, G. (where cfu=colony forming units) when liquid samples were Davis and H. A. O. Hill, Anal. Chim. Acta, 1984, 163, 161. used and the dried probes could be rehydrated after several 5 A. P. F. Turner, W. J. Aston, G. Davis, I. J. Higgins, H.A. O. Hill months without a significant decrease in activity. With some and J. Colby, in Microbial Gas Metabolism, ed. R. K. Poole and modifications the system could probably be developed into a C. S. Dow, Academic Press, London, 1985, pp. 161–170. 6 G. G. Guilbault, Anal. Chem., 1983, 55, 1682. sensor for monitoring micro-organisms in air for occupational 7 K.-P. Rindt and S. Scholtissek, in Biosensors.Applications in and environmental purposes. Medicine, Environmental Protection and Process Control, ed. R. D. A patent has been published for manufacturing biosensors Schmid and F. Scheller, GBF Monographs, vol. 13, VCH, for the measurement of gas-phase chemicals by Park and Weinheim, 1989, pp. 405–415. Lee.37 The method is based on screen-printing of conducting 8 K.-P.Rindt, A. Flauss, R. Feldbru� gge, U. Priess and K. Kock, Enzyme DiVusion Badges as Biochemical Devices for the Dosimetric electrodes on to an insulating substrate consisting of alumina Measurement of Toxic Compounds in Air, Dra� gerwerk, Lu�beck, (Al2O3) or polymer material [e.g. poly(vinyl chloride)]. An 1992. enzyme was then immobilised on the device, covering the 9 M.Ha� mmerle, E.A. H. Hall, N. Cade and D. Hodgins, Biosens. electrodes. The device was tested by immobilising alcohol Bioelectron., 1996, 11, 239. dehydrogenase in the biosenor and measuring ethanol in air. 10 M. J. Dennison, J. M. Hall and A. P. F. Turner, Analyst, 1996, 121, 1769. A good response was achieved in the range 0–3000 ppm J. Environ. Monit., 1999, 1, 293–298 29711 F.Vianello, A. Stefani, M. L. Di Paolo, A. Rigo, A. Lui, B. 27 M. Naessens and C. Tran-Minh, Biosens. Bioelectron., 1998, 13, Margesin, M. Zen, M. Scarpa and G. Soncini, Sens. Actuators, B, 341. 1996, 37, 49. 28 M. Naessens and C. Tran-Minh, Anal. Chim. Acta, 1998, 364, 153. 12 E. Barzana, A. M. Klibanov and M. Karel, Anal. Biochem., 1989, 29 Y. Okahata and O. Shimizu, Langmuir, 1987, 3, 1171. 182, 109. 30 H. Muramatsu, E. Tamiya and I. Karube, Anal. Chim. Acta, 1989, 13 W. Matuszewski and M. E. MeyerhoV, Anal. Chim. Acta, 1991, 225, 399. 248, 379. 31 H. Muramatsu, E. Tamiya and I. Karube, Anal. Chim. Acta, 1991, 14 K. Mitsubayashi, K. Yokoyama, T. Takeuchi and I. Karube, Anal. 251, 135. Chem., 1994, 66, 3297. 32 T.-Z. Wu, Biosens. Bioelectron., 1999, 14, 9. 15 J.-K. Park, H.-J. Yee and S.-T. Kim, Biosens. Bioelectron., 1995, 33 T. Okada, I. Karube and S. Suzuki, Biotechnol. Bioeng., 1983, 10, 587. 1641. 16 A. K. Williams and J. T. Hupp, J. Am. Chem. Soc., 1998, 120, 34 J. W. Aylott, D. J. Richardson and D. A. Russell, Chem. Mater., 4366. 1997, 9, 2261. 17 S. Saini, W. Surareungchai, A. P. F. Turner and M. E. A. Downs, 35 C. K. O’Sullivan, PhD Thesis, Cranfield University, 1996. Biosens. Bioelectron., 1995, 10, 945. 36 F. S. Ligler, G. P. Anderson, P. T. Davidson, R. J. Foch, J. T. 18 M. J. Dennison, J. M. Hall and A. P. F. Turner, Anal. Chem., Ives, K. D. King, G. Page, D. A. Stenger and J. P. Whelan, 1995, 67, 3922. Environ. Sci. Technol., 1998, 32, 2461. 19 A. Kaisheva, I. Iliev, R. Kazareva, S. Christov, U. Wollenberger 37 J. K. Park and H. J. Lee, Eur. Pat., 94 630044.9, 1994. and F. Scheller, Sens. Actuators, B, 1996, 33, 39. 38 Ambient Air Quality—DiVusive Samplers for the Determination of 20 A. Kaisheva, I. Iliev, S. Christov and R. Kazareva, Sens. Concentrations of Gases and Vapours—Requirements and Test Actuators, B, 1997, 44, 571. Methods, PrEN XXX5draft, Comite� Europee�n de Normalisation, 21 J. Ngeh-Ngwainbi, P. H. Foley, S. S. Kuan and G. G. Guilbault, Brussels. J. Am. Chem. Soc., 1986, 108, 5444. 39 Workplace Atmospheres—General Requirements for the 22 L. Rajakovic, V. Ghaemmaghami and M. Thompson, Anal. Chim. Performance of Procedures for the Measurement of Chemical Acta, 1989, 217, 111. Agents, EN48251994, Comite� Europee�n de Normalisation, 23 G. D. Case and J. F. Crivello, Proc. Natl. Conf. Hazard. Wastes Brussels, 1994. Hazard. Mater., 1990, 7, 229. 40 Workplace Atmospheres—DiVusive Samplers for the Determination 24 W. J. Albery, A. E. G. Cass and Z. X. Shu, Biosens. Bioelectron., of Gases and Vapours—Requirements and Test Methods, 1990, 5, 367. EN83851995, Comite� Europee�n de Normalisation, Brussels, 1995. 25 W. J. Albery, A. E. G. Cass and Z. X. Shu, Biosens. Bioelectron., 1990, 5, 379. 26 W. J. Albery, A. E. G. Cass, B. P. Mangold and Z. X. Shu, Biosens. Bioelectron., 1990, 5, 397. Paper 9/02835H 298 J. Environ. Monit., 19