Breath acetone monitoring by portable Si:WO3 gas sensors

Page 1

B

MAa

b

c

h

�

�

�

�

�

a

ARRAA

KBCMFNC

1

e

0h

Analytica Chimica Acta 738 (2012) 69– 75

Contents lists available at SciVerse ScienceDirect

Analytica Chimica Acta

j ourna l ho me page: www.elsev ier .com/ locate /aca

reath acetone monitoring by portable Si:WO3 gas sensors

arco Righettonia, Antonio Tricoli a, Samuel Gassa, Alex Schmidb,c,nton Amannb,c, Sotiris E. Pratsinisa,∗

Particle Technology Laboratory, Department of Mechanical and Process Engineering ETH Zurich, CH-8092 Zurich, SwitzerlandUniv.-Clinic for Anesthesia, Innsbruck Medical University, A-6020 Innsbruck, AustriaBreath Research Institute of the Austrian Academy of Sciences, A-6850 Dornbirn, Austria

i g h l i g h t s

Portable sensors were developed andtested for monitoring acetone in thehuman breath.Acetone concentrations down to20 ppb were measured with shortresponse times (<30 s).The present sensors were highlyselective to acetone over ethanol andwater.Sensors were applied to humanbreath: good agreement with highlysensitive PTR-MS.Tests with people at rest and duringphysical activity showed the sensorrobustness.

g r a p h i c a l a b s t r a c t

r t i c l e i n f o

rticle history:eceived 16 April 2012eceived in revised form 28 May 2012ccepted 3 June 2012vailable online 12 June 2012

eywords:iosensor

a b s t r a c t

Breath analysis has the potential for early stage detection and monitoring of illnesses to drasticallyreduce the corresponding medical diagnostic costs and improve the quality of life of patients suffer-ing from chronic illnesses. In particular, the detection of acetone in the human breath is promising fornon-invasive diagnosis and painless monitoring of diabetes (no finger pricking). Here, a portable acetonesensor consisting of flame-deposited and in situ annealed, Si-doped epsilon-WO3 nanostructured filmswas developed. The chamber volume was miniaturized while reaction-limited and transport-limited gasflow rates were identified and sensing temperatures were optimized resulting in a low detection limit of

hemo-resistive gas sensoretal oxide

lame spray pyrolysisanoparticlesross-sensitivity to humidity

acetone (∼20 ppb) with short response (10–15 s) and recovery times (35–70 s). Furthermore, the sensorsignal (response) was robust against variations of the exhaled breath flow rate facilitating application ofthese sensors at realistic relative humidities (80–90%) as in the human breath. The acetone content inthe breath of test persons was monitored continuously and compared to that of state-of-the-art protontransfer reaction mass spectrometry (PTR-MS). Such portable devices can accurately track breath acetone

an al

concentration to become

. Introduction

Noninvasive detection of diseases by breath analysis is a fast,conomically viable and simple alternative to blood analysis and

∗ Corresponding author. Tel.: +41 446323180.E-mail address: [email protected] (S.E. Pratsinis).

003-2670/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.aca.2012.06.002

ternative to more elaborate breath analysis techniques.© 2012 Elsevier B.V. All rights reserved.

endoscopy [1]. The breath, however, contains several hundredvolatile organic compounds (VOC) with concentrations rangingfrom ppt to ppm [2]. The cellular and biochemical origin of many ofthese VOCs has not been determined and some of them might be of

exogenous origin [3]. So the field of exhaled breath analysis is stillin its infancy. Exceptions are isoprene and acetone, which appear inrelatively high concentrations of ∼100 and ∼500 ppb, respectively[4]. Acetone has the potential of supplementing information on the

Page 2

7 a Chim

st[cimaa

camse([dsechcccm

sh

FtoA

0 M. Righettoni et al. / Analytic

tatus of patients suffering from diabetes [5]. The acetone concen-ration in the breath varies from 300 to 900 ppb in healthy people4] to more than 1800 ppb for diabetics [6]. Nevertheless, acetoneoncentrations in the breath are not simply related to glucose levelsn the blood and additional research is necessary to make it a viable

arker compound for clinical routine. For example, exchange ofcetone between blood and breath is taking place not only in thelveoli but also in the upper airways [7].

Even though the broad application of breath analysis is stillhallenging, the development of small, hand-held devices for reli-ble and continuous, real-time measurement of important breatharkers such as acetone is desirable. Several methods have demon-

trated a remarkable potential for such measurements [8]. Forxample, gas chromatography with flame ionization detectionFID) [9], proton transfer reaction-mass spectrometry (PTR-MS)10] selected ion flow tube-mass spectrometry (SIFT-MS) [11], andifferential mobility spectroscopy (DMS) [12] have shown highelectivity, sufficient sensitivity and low limit of detection for sev-ral VOCs. In general, however, most of these methods are quiteostly and have rather limited portability except for DMS thatas also high potential for miniaturization [13]. In this regard,hemo-resistive gas sensors based on semiconductor nanoparti-les are attractive for breath analysis [14] offering low fabricationosts, high sensitivity, sufficiently low limit of detection and further

iniaturization potential [15].Recently, Cr- or Si-doped WO3 nanoparticles have shown high

ensitivity and selectivity to acetone [16], even up to 90% relativeumidity [17]. This has allowed detection of low concentrations

ig. 1. Schematic of the T-shaped chamber (L = 75, H = 50 and D = 18 mm) and thermocouungsten oxide (WO3) film onto interdigitated Pt electrodes laid on an Al2O3 substrate aln its back. (b) Schematic of the experimental set-up during breath analysis. The breath flll the data (dashed lines) are collected and analyzed by the computer.

ica Acta 738 (2012) 69– 75

of acetone (down to 20 ppb) with high signal to noise ratio andhigh selectivity to ethanol and water vapor [17]. These promisingresults, however, were obtained in large heated chambers and sim-ulated breath conditions. The application of such nanoparticles asacetone detectors in portable devices is not trivial [18] as reachingsufficiently high temperatures (300–450 ◦C) at reasonable powerconsumption requires a locally heated substrate that may resultin inhomogeneous temperatures and concentration profiles withinthe sensing film [19]. Thus, the development of a portable device forbreath acetone detection is still challenging [8] but quite attractive.

Here, a portable, chemo-resistive sensor has been developed andapplied to real breath acetone detection. The sensor was based ona back-heated substrate with a sensing film consisting of optimallyperforming 10 mol% Si-doped WO3 nanoparticles [17]. The sensi-tivity and selectivity of this device to acetone was investigated atvarious concentrations, temperatures, relative humidities and gasflow rates. Using this device, the breath acetone concentration oftest persons at rest or during physical activity was measured andcompared to that measured by highly accurate PTR-MS.

2. Materials and methods

A flame spray pyrolysis (FSP) reactor was used for synthesis

and direct deposition of 10 mol% Si-doped WO3 nanoparticle films[20] onto Al2O3 substrates featuring a set of interdigitated elec-trodes (Fig. 1a). A solution of ammonium (meta)tungstate hydrate(Aldrich, purity > 97%) and hexametyldisiloxane (HMDSO, Aldrich,

ple (TC) placed inside. A Macor piece is holding the sensor prior to deposition of aong with a resistance temperature detector (RTD) on the front side and a Pt heaterow (grey lines) is controlled by the mask and kept constant by the PTR-MS pump.

Page 3

a Chim

p(gp[tgrOpos1ocdPhsW

i(Gnsaecltpihasama(ts3c

fitoCrlota

aLm(p(wakoS

M. Righettoni et al. / Analytic

urity >99%) was created as dictated by the final SiO2 content10 mol%), and diluted in a 1:1 (volume ratio) mixture of diethylenelycol monobutyl ether (Fluka, purity > 98.5%) and ethanol (Fluka,urity > 99.5%) with a total metal (Si and W) concentration of 0.2 M20]. This solution was supplied at a rate of 5 mL min−1 throughhe FSP nozzle and dispersed to a fine spray with 5 L min−1 oxy-en (pressure drop 1.5 bar). That spray was ignited by a supportinging-shaped premixed methane/oxygen flame (CH4 = 1.25 L min−1,2 = 3.2 L min−1). An additional 5 L min−1 sheath oxygen was sup-lied from an annulus surrounding that flame to ensure excessxidant flow [20]. The nanoparticles collected downstream of theensor substrate consisted of pure �-WO3 with crystal size of about0 nm [20]. This nanoparticle film composition was selected for itsptimal thermal stability, selectivity and sensitivity even in humidonditions [17]. The alumina substrate (0.8 mm thick) had inter-igitated Pt lines (sputtered, 350 �m width and spacing) and at resistance temperature detector (RTD) on one side and a Pteater on the other side. The overall dimensions of the aluminaubstrate were 15 mm × 13 mm (Electronic Design Center, Case

estern Reserve University, Cleveland, OH, USA).The WO3 crystal size and phase composition were character-

zed by X-ray diffraction (XRD). Prior to sensing tests, the sensorsFig. 1a) were kept in an oven (Carbolite GmbH, Ubstadt-Weiher,ermany) at 500 ◦C for 5 h at ambient pressure to stabilize theanoparticle size and avoid further sintering during sensor mea-urements. The simulated breath was prepared by mixing syntheticir (Pan Gas, 99.999%) and humid air (at 20 ◦C) with acetone orthanol (10 ppm in synthetic air, Pan Gas 5.0) to obtain the desiredoncentration, as described in more detail elsewhere [17]. To attainower flow rates (<0.2 L min−1), a supplementary mass flow con-roller and a pressure release valve were added. The sensors werelaced inside a T-shaped tube chamber, 50 mm in height, 75 mm

n length and 18 mm in diameter (Fig. 1a) and placed onto a Macorolder connected to a voltmeter (Keithley, 2700 Multimeter/Datacquisition system) to measure film resistance and to a powerource (Hopesun, DC power supply) to heat the sensors. The oper-ting temperature (TO) was varied between 250 and 390 ◦C andeasured with the embedded Pt RTD on the substrate. Addition-

lly the chamber temperature was measured by a thermocouplen-type) placed about 15 mm away from the sensor. The chamberemperature increases at the beginning and then stabilizes to a con-tant value following always that by the Pt RTD. For example, at50 ◦C (RTD) at steady state and at 1 L min−1 inlet gas flow rate, thehamber temperature was about 60 ◦C.

The sensor response (S) is: S = Rair/Ranalyte − 1 where Rair is thelm resistance in air at a given relative humidity (RH) and Ranalyte ishat resistance at a given concentration of analyte (acetone, ethanolr water vapor). The cross-sensitivity to humidity (CS) is [21]:S = abs[(Sdry − SRH)/Sdry] × 100 where Sdry and SRH are the sensoresponses in dry air and at a given RH, respectively, as defined ear-ier [17]. The sensor response time is the time needed to reach 90%f the sensor response to 100 ppb acetone. The recovery time ishe time to recover 90% of the sensor response to that of 600 ppbcetone [17].

Real breath measurements (Fig. 1b) were performed with theid of a respiratory flow controlled mask (Cortex Biophysik GmbH,eipzig, Germany) that allowed sampling of specific breath seg-ents into the Si:WO3 sensors and a high-sensitivity PTR-MS

Ionicon Analytik GmbH, Innsbruck) [22]. Gas sampling was accom-lished by a heated Teflon tube using an insulated heating wireTNI Medical, Freiburg, Germany). More precisely, the temperatureas kept above 40 ◦C along the entire tube length, to minimize

ny water condensation. A constant flow rate of 70 mL min−1 wasept during all breath measurements while the RTD temperaturef the Si:WO3 sensor was set at 350 ◦C. The calibration lines for thei:WO3 sensors have been calculated by linear regression from the

ica Acta 738 (2012) 69– 75 71

sensor response of acetone at 90% RH with 70 mL min−1 inlet flowrate. The overall goodness of fit (R2) was above 0.97 indicating highaccuracy. Measurements during physical activity are carried outon a computer-controlled, semi-supine medical ergometer (eBikeL,GE Medical Systems, Milwaukee, USA) operating at constant lev-els of power independently of the pedal speed. A supporting bedstabilizes the torso of the volunteer thereby reducing movementartifacts appearing in the acquired physiological signals. Five malepersons (subjects) in good health between 25 and 35 years old weretested. The acetone and isoprene signals were the counts per secondmeasured at mass-to-charge ratio m/z = 59 and 69, respectively. Theresponse time calculated during human breath measurements isdefined as the time needed to reach 90% of the final average acetoneconcentration.

3. Results and discussion

3.1. Device characterization

Fig. 1a shows a schematic of the device consisting of aback-heated substrate with a sensing film of silica-doped WO3nanoparticles and a T-shaped chamber. The baseline resistance ofthe sensors in dry air was investigated as a function of operatingtemperature. Their baseline was within ca. 2 M� variation for alltemperatures, consistent with the expected variability of chemo-resistive gas sensors [23]. The average baseline decreased from 22.5to 1.3 M� by increasing the operating temperature from 150 to390 ◦C. The baseline of the present device at 350 ◦C was about 3 M�:more than 300 times lower than the that of SnO2-based micro-hotplate sensors operated at even higher temperatures (450 ◦C)[18] favoring, thus, their integration in monolithic sensing devices[24], where typical target baselines are below 1 G�.

With respect to acetone sensing, both the catalytic activity ofthe semiconductor WO3 surface and its electrical properties arestrongly influenced by temperature [25]. The sensor response to500 ppb acetone (circles) and ethanol (diamonds) was investigatedas function of film temperature (Fig. 2a). The maximum responseto acetone was at 350 ◦C, requiring about 9 W to heat up the sub-strate. The observed reduction in sensor response at temperaturesabove this optimum was attributed to the increased combustionof acetone in the upper layers of the sensing film [26], reducingthe acetone amount penetrating to lower layers. At 350 ◦C, theresponse of the back-heated sensor was comparable to that inheated chambers [17] at 400 ◦C. This is advantageous as lower oper-ating temperatures require less power improving the portabilityand lowering the energy needs of the device [27].

The average ethanol concentration in the breath [28] is approx-imately 196 ppb with a standard deviation of 224 ppb, making ita potentially disturbing analyte during breath acetone measure-ment. Here, the sensor response to 500 ppb ethanol (Fig. 2a) wasconsiderably lower than that to 500 ppb acetone at all temperaturesdemonstrating good acetone selectivity. Furthermore, the maximalresponse to ethanol was at slightly lower temperatures (∼325 ◦C)than for acetone. This difference in optimum sensing tempera-ture between acetone and ethanol increased the acetone selectivitytowards ethanol from 6.9 at 300 ◦C to 13.4 when operating thedetector at 350 ◦C.

Fig. 2b shows the sensor response to acetone (circles) andethanol (diamonds) as a function of their concentration at 350 ◦C.The acetone sensor response (Fig. 2b, circles) was always 9–13times higher than that of ethanol (diamonds), resulting in consid-

erably higher acetone selectivity here than that (4.7–6.7) withinthe heated chamber [17]. This difference is attributed to the differ-ent operating temperature and chamber geometry. Furthermore,the operating temperature in heated chambers was above the

Page 4

72 M. Righettoni et al. / Analytica Chimica Acta 738 (2012) 69– 75

Fig. 2. Sensor response (a) to 500 ppb acetone (circles) or ethanol (diamonds) asa function of sensor temperature in dry air, with an optimum at 350 ◦C; and (b)to different acetone (circles) and ethanol (diamonds) concentrations at 350 ◦C. Thee1

athart5wiss

crwaH1imsae(0i

Fig. 3. Sensor response to acetone concentration as a function of inlet total gasflow rate (from 0.05 to 1 L min−1). The sensor response is constant above 0.2 L min−1

depending on age, sex, and health [31]. As a result, small response

rror bars represent the variability of sensor response to acetone (circles) by three0 mol% Si-doped WO3 sensors.

utoignition temperature of ethanol (365 ◦C) that could have ledo its partial decomposition to more reactive species, yielding aigher response. Allowing a maximal measurement error of thecetone concentration (500–1500 ppb) of 5%, the sensor responseatio between acetone and the average breath ethanol concentra-ion (196 ppb) must be above 20. For example, at the lowest limit of00 ppb acetone concentration, the sensor response is 1.54 (Fig. 2b)hile that of ethanol at 196 ppb is about 0.06 (Fig. 2b) resulting

n a sensor response ratio of 25.6. The error bars on the acetoneensor response (Fig. 2b, circles) show the variability of differentensors.

The gas flow rate over the sensing film might be anotherhallenge as its variation may affect the sensor output [29] if aeaction-limited response is not reached [26]. Design of devicesith sufficiently high flow velocity can be achieved in line with

diffusion-reaction model [26] for chemo-resistive gas sensors.ere, variation in the gas flow rate from 0.05 (Fig. 3, crosses) to

L min−1 (circles) has been investigated. For a gas to be detected,t needs first to diffuse into the sensing film and react on the

etal oxide surface. For sufficiently high analyte flux into theensing film, a reaction-limited response is reached that is mainly

function of film temperature, morphology and material prop-rties. If the analyte flux, however, is not sufficient, a transport

diffusion)-limited sensor response is obtained. Flow rates between.2 and 1 L min−1 (Fig. 3, black lines), corresponding to flow veloc-

ties of 0.015 and 0.08 m s−1 respectively, supplied to the present

(black symbols) as it is reaction-limited. Below that the sensor response becomestransport (diffusion)-limited and decreases with decreasing inlet gas flow rate (greysymbols).

back-heated sensors, resulted in sufficiently high analyte mass fluxto reach a reaction-limited response: this results in almost identi-cal sensor responses (<4% difference). By decreasing, however, theflow rate substantially below 0.2 L min−1 (grey lines) yields a flowrate dependency similar to that in heated chambers [29] indicat-ing a transport (diffusion)-limited response. The sensor responseis reduced by about 15% when decreasing the flow rate from 0.2(black squares) to 0.1 L min−1 (grey diamonds) and decreases con-tinuously with lower flow. Flow rates higher than 0.2 L min−1

(reaction-limited) are more relevant to breath analysis as averageexpiratory flow rates of healthy persons are above 2.4 L min−1 [30].However, state-of-the-art measurement devices such as PTR-MSutilize lower (e.g. 70 mL min−1) flow rates. Here the latter were usedfor the Si:WO3 sensors to facilitate their comparison to establishedPTR-MS.

The present sensor response to water vapor from 10 to 90% RHwas minimal (<0.3) in agreement with results obtained in exter-nally heated chambers [17]. Despite this high selectivity againstH2O, it must be noted that the water vapor content of the inlet flowreduced considerably the temperature of the sensor, so its powersupply needs to be slightly adjusted to maintain the same sensortemperature (350 ◦C).

Fig. 4a shows the cross sensitivity (CS) to humidity as a functionof acetone concentration for RH ranging from 0 to 90%. The most sig-nificant reduction in sensor response was observed when increas-ing the RH from 0 to 20%, a rather unrealistic RH range for breathanalysis. Beyond this point, the CS to humidity decreased consid-erably. So, the CS was 54% between 0 and 90% RH but only 4.5%between 80 and 90% RH. This indicates that the acetone concentra-tion in the breath could be determined using this sensor with suffi-cient accuracy even without additional RH measurement. Further-more, the sensor response to the water content of the breath (90%RH) from the background ambient condition (40% RH) is less than0.03 (with S = R40%RH/R90%RH − 1, where R40%RH = 2.43 ± 0.03 M� andR90%RH = 2.43 ± 0.02 M�), which is within the sensor baseline vari-ability and rather small compared to acetone response: about 1.85for 600 ppb acetone for 1 L min−1 (Fig. 3).

A patient’s exhalation time is limited by his or her forced vitalcapacity (FVC) and expiratory flow, both of which vary considerably

and recovery times facilitate the application of the device in breathanalysis. Fig. 4b shows the response (open circles) and recovery(filled circles) times of these Si:WO3 sensors as a function of RH

Page 5

M. Righettoni et al. / Analytica Chimica Acta 738 (2012) 69– 75 73

Fig. 4. (a) Sensor response upon exposure to increasing acetone concentration atvarious RH. The sensor response to acetone decreased with increasing RH showinghigh CS (54%) between 0 and 90% RH. Between 80 and 90% RH, however, the CS wasonly 4.5% showing the robustness of such sensors at the typical RH of the humanb(

afrricfltt(tptta

tstmt5tp

person during the same day by the Si:WO3 sensor did not change

reath. (b) Response time to 100 ppb (open circles) and recovery time to 600 ppbfilled circles) acetone as a function of RH at 350 ◦C.

t 350 ◦C. The response time was not affected by variations in RHrom 0 to 90% and remained constant to ca. 14 s. In contrast, theecovery time (filled circles), increased from 36 to 62 s in that RHange. These response times were 5 to 25 times smaller than thosen heated chambers [17]. This is attributed to two effects: (1) heatedhambers are characterized by large volumes of almost stagnantuid and peripheral position of the sensor and thus the responseime is more a measurement of the transient concentration withinhe chamber rather than a characteristic feature of the device [19].2) Due to rapid inlet velocity of the impinging jet and resultingurbulence here, the analyte mass flux to the sensing film of theresent device was considerably higher. While it might be possibleo achieve faster measurements by improving the electronics [32],he present device already has sufficiently short response time tocetone.

Fig. 5 shows a clear reduction of sensor resistance from 2.51o 2.41 M� in response to even 20 ppb acetone at 90% RH corre-ponding to a sensor response of 0.042 which is comparable tohat (0.08) in bulkier, heated chambers at 400 ◦C [17]. Further-

ore, this sensor exhibited a high signal to noise ratio (>10) andhe ability to differentiate in sensor response (10%) even between

0 ppb and 60 ppb of acetone, further indicating the accuracy ofhe present sensor and that even lower limits of detection areossible.

Fig. 5. Sensor resistance at 90% RH and upon exposure to ultra low concentrationsof acetone (20, 50, 60 and 80 ppb) at 350 ◦C. The signal noise arises from smalltemperature fluctuations.

3.2. Exhaled breath analysis

Undoubtedly, real breath measurements are needed to validatethe present devices for breath acetone monitoring. Here, humanbreath was analyzed simultaneously by this novel sensor deviceand by state-of-the-art, high-sensitivity PTR-MS. As the alveolarlevels of blood-borne volatile species, such as acetone and isoprene,are best reflected by the end tidal fraction of each exhalation phase[33], here, the end tidal volume of the breath was sampled by aflow-controlled valve (Fig. 1b).

Fig. 6a shows the concentration of acetone in tidal part of therespiratory cycle of a healthy test person at rest measured by thepresent Si:WO3 sensor (thick solid line) and PTR-MS for acetone(thin solid line) and isoprene (dotted line). The acetone concentra-tion measured by Si:WO3 sensor was calculated from the calibratedsensor response at 90% RH and at 70 mL min−1 flow. At the start ofbreath sampling (∼3 min), the sensor resistance decreased rapidlyand recovered to the initial value after stopping the breath flow(∼8 min). The calibrated sensor response corresponded to about970 ppb acetone concentration on the average at 3–8 min. This wasin good agreement (>98%) with the acetone concentration readingof the PTR-MS, 980 ppb (thin solid line). The present Si:WO3 sensormeasured a similar acetone concentration evolution as the PTR-MSand even had a higher signal to noise ratio (60 and 9, respectively).

The visible drift for the acetone sensor response (Fig. 6a, thicksolid line) is probably due to a catalytic surface reaction of otherspecies present in the breath (e.g. CO) that locally increase the tem-perature causing also the initial overshoot of the Si:WO3 response.This leads to an initial drop in the resistance that later stabi-lizes toward that corresponding predominantly to acetone. Thishas been previously observed by Pt-doped SnO2 sensors duringCO detection [18]. A comparable but smaller drift is also visi-ble for the average PTR-MS signal. The response time to humanbreath (Fig. 6a), was of 27 and 28 s for the Si:WO3 sensor andPTR-MS, respectively. This is comparable but longer than that mea-sured for simulated breath condition (Fig. 4b, open circles). Asexpected, the isoprene concentration measured by PTR-MS (dottedline) remained rather constant during normal breathing withoutphysical activity (Fig. 6a).

The acetone breath concentration measured from the same

much during physical activity (Fig. 6b, thick solid line) comparedto that in Fig. 6a in agreement also with the acetone concentrationmeasured here by PTR-MS (thin solid line). A similar overshoot

Page 6

74 M. Righettoni et al. / Analytica Chimica Acta 738 (2012) 69– 75

Fig. 6. Expected acetone concentration by the Si:WO3 sensor (thick solid line) andacetone (thin solid line) and isoprene (dotted line) concentrations measured by PTR-MS during breathing of a test person (a) at rest and (b) during physical activity.Eas

iv9o1imalasiqoPfa8arb

bsas

ven though the isoprene concentration (dotted line) varied notably during physicalctivity, it does not influence the Si:WO3 sensor response indicating that its acetoneignal is quite robust and selective.

n the acetone concentration measured by the Si:WO3 sensor isisible as discussed above and its average concentration was about70 ppb, as in Fig. 6a (thick solid line). In contrast, the concentrationf isoprene, measured by PTR-MS (Fig. 6b) increased from about00 to above 240 ppb in agreement with the reported release of

soprene during physical activity [33]. The acetone concentrationeasured by the Si:WO3 sensor (Fig. 6b, thick solid line) followed

gain the acetone concentration evolution by PTR-MS (thin solidine) and was hardly affected by the presence of isoprene. Thelveolar ventilation increased from about 10 to 20 L min−1 at thetart of the physical activity and decreased back to 10 L min−1 atts end. These results show that the present Si:WO3 sensors areuite robust against changes in ventilation rate and in the presencef isoprene. Furthermore, these sensors were in agreement withTR-MS for acetone concentrations (ranging from 590 to 980 ppb)rom different healthy test persons (for 4–6 min). The sensor wasble to detect differences in breath acetone concentrations from80 to 980 ppb. Nevertheless, testing the breath in the morningfter breakfast with lower acetone (∼600 ppb) concentrationsesulted in an inferior (∼85%) but still comparable agreementetween this sensor and PTR-MS.

The response of the Si:WO3 sensors was also tested to short

reath pulses (Fig. 7), without controlling valve, as may be neces-ary in applied breath analysis. The sensors had response times ofbout 10 s and thus were able to fully capture the pulse profile. Theensor response to breath pulses (Fig. 7) of each test person were

Fig. 7. Si:WO3 sensor response to short pulses of three different healthy test personswith similar breath acetone concentrations.

fairly reproducible and sensitive to even small acetone concentra-tion changes, always recovering the initial baseline for each testperson. Test person 3, for instance, showed higher sensor responsethan that of the two other test persons indicating higher acetoneconcentration in the breath. However, due to different transienttime (e.g. sampling flow rate) it was not possible to compare itdirectly with PTR-MS measurements.

4. Conclusions

Portable acetone sensors made of 10 mol% Si-doped WO3nanoparticles were developed and tested for breath analysis. Theydemonstrated a strong potential for detection of acetone, a breathmarker for diabetes, both in ideal (dry air) and realistic conditions(90% RH). The sensors were highly selective over ethanol and sensi-tive to acetone, regardless of background RH. The sensor responsetimes were below 15 s at application typical conditions, makingthese devices attractive for breath analysis. Acetone concentra-tions as low as 20 ppb were measured with high signal to noiseratios (>10). Furthermore, the sensor signal (response) was robustagainst variation in gas (or breath) flow rates down to 0.2 L min−1

facilitating application of such sensors in real breath measurementconditions.

These sensors were applied to breath acetone monitoring offive different test persons and were in agreement (>98%) to high-sensitivity PTR-MS measurements of the same breath samples. Thebreath acetone concentrations measured by the Si:WO3 sensorshad very high signal to noise ratio (60) and the response time wasabout 27 s. The Si:WO3 sensors were selective to acetone both atrest and during physical activity, independently of the respiratorypace or presence of isoprene that is associated with physical activ-ity. Furthermore, these sensors also responded very rapidly to shortbreath pulses further supporting their potential for breath analysis.

Acknowledgments

This research was supported by the Swiss National ScienceFoundation, grant 200021 130582/1 and the European ResearchCouncil. The authors are grateful to J. King (BRI, Austria) for hisassistance during the breath analysis experiments.

References

[1] A. Amann, M. Corradi, P. Mazzone, A. Mutti, Expert. Rev. Mol. Diagn. 11 (2011)207–217.

[2] M. Phillips, K. Gleeson, J.M.B. Hughes, J. Greenberg, R.N. Cataneo, L. Baker, W.P.McVay, Lancet 353 (1999) 1930–1933.

Page 7

a Chim

[[[

[

[

[[[[

[

[

[[

[

[[[

[

[[[

M. Righettoni et al. / Analytic

[3] W. Filipiak, A. Sponring, A. Filipiak, C. Ager, J. Schubert, W. Miekisch, A. Amann,J. Troppmair, Cancer Epidemiol. Biomarkers Prev. 19 (2010) 182–195.

[4] A.M. Diskin, P. Spanel, D. Smith, Physiol. Meas. 24 (2003) 107–119.[5] W. Ping, T. Yi, H.B. Xie, F.R. Shen, Biosens. Bioelectron. 12 (1997) 1031–1036.[6] C.H. Deng, J. Zhang, X.F. Yu, W. Zhang, X.M. Zhang, J. Chromatogr. B 810 (2004)

269–275.[7] J. King, K. Unterkofler, G. Teschl, S. Teschl, H. Koc, H. Hinterhuber, A. Amann, J.

Math. Biol. 63 (2011) 959–999.[8] T.H. Risby, S.F. Solga, Appl. Phys. B 85 (2006) 421–426.[9] J.M. Sanchez, R.D. Sacks, Anal. Chem. 75 (2003) 2231–2236.10] K. Schwarz, W. Filipiak, A. Amann, J. Breath Res. 3 (2009) 027002.11] D. Smith, P. Spanel, Mass Spectrom. Rev. 24 (2005) 661–700.12] M. Shnayderman, B. Mansfield, P. Yip, H.A. Clark, M.D. Krebs, S.J. Cohen, J.E.

Zeskind, E.T. Ryan, H.L. Dorkin, M.V. Callahan, T.O. Stair, J.A. Gelfand, C.J. Gill, B.Hitt, C.E. Davis, Anal. Chem. 77 (2005) 5930–5937.

13] C.E. Davis, M.J. Bogan, S. Sankaran, M.A. Molina, B.R. Loyola, W. Zhao, W.H.Benner, M. Schivo, G.R. Farquar, N.J. Kenyon, M. Frank, IEEE Sens. J. 10 (2010)114–122.

14] M. Fleischer, E. Simon, E. Rumpel, H. Ulmer, M. Harbeck, M. Wandel, C. Fietzek,U. Weimar, H. Meixner, Sens. Actuator B: Chem. 83 (2002) 245–249.

15] A. Tricoli, M. Righettoni, A. Teleki, Angew. Chem. Int. Ed. 49 (2010) 7632–7659.16] L. Wang, A. Teleki, S.E. Pratsinis, P.I. Gouma, Chem. Mater. 20 (2008) 4794–4796.17] M. Righettoni, A. Tricoli, S.E. Pratsinis, Anal. Chem. 82 (2010) 3581–3587.18] A. Tricoli, M. Graf, F. Mayer, S. Kuhne, A. Hierlemann, S.E. Pratsinis, Adv. Mater.

20 (2008) 3005–3010.

[[[

ica Acta 738 (2012) 69– 75 75

19] A.M. Lezzi, G.P. Beretta, E. Comini, G. Faglia, G. Galli, G. Sberveglieri, Sens. Actu-ator B: Chem. 78 (2001) 144–150.

20] M. Righettoni, A. Tricoli, S.E. Pratsinis, Chem. Mater. 22 (2010) 3152–3157.

21] A. Tricoli, M. Righettoni, S.E. Pratsinis, Nanotechnology 20 (2009) 315502.22] A. Bajtarevic, C. Ager, M. Pienz, M. Klieber, K. Schwarz, M. Ligor, T. Ligor, W.

Filipiak, H. Denz, M. Fiegl, W. Hilbe, W. Weiss, P. Lukas, H. Jamnig, M. Hackl, A.Haidenberger, B. Buszewski, W. Miekisch, J. Schubert, A. Amann, BMC Cancer 9(2009) 348.

23] L. Madler, A. Roessler, S.E. Pratsinis, T. Sahm, A. Gurlo, N. Barsan, U. Weimar,Sens. Actuator B: Chem. 114 (2006) 283–295.

24] M. Graf, U. Frey, S. Taschini, A. Hierlemann, Anal. Chem. 78 (2006) 6801–6808.25] K. Aguir, C. Lemire, D.B.B. Lollman, Sens. Actuator B: Chem. 84 (2002) 1–5.26] T. Becker, S. Ahlers, C. Bosch-v.Braunmuhl, G. Muller, O. Kiesewetter, Sens.

Actuator B: Chem. 77 (2001) 55–61.27] M. Graf, D. Barrettino, S. Taschini, C. Hagleitner, A. Hierlemann, H. Baltes, Anal.

Chem. 76 (2004) 4437–4445.28] C. Turner, P. Spanel, D. Smith, Rapid Commun. Mass Spectrom. 20 (2006) 61–68.29] M. Righettoni, A. Tricoli, J. Breath Res. 5 (2011) 037109.30] B. Mahut, C. Delacourt, F. Zerah-Lancner, J. De Blic, A. Harf, C. Delclaux, Chest

125 (2004) 1012–1018.31] J.B. Schoenberg, G.J. Beck, A. Bouhuys, Respir. Physiol. 33 (1978) 367–393.32] M.I. Baraton, Sens. Actuator B: Chem. 31 (1996) 33–38.33] J. King, A. Kupferthaler, K. Unterkofler, H. Koc, S. Teschl, G. Teschl, W. Miekisch,

J. Schubert, H. Hinterhuber, A. Amann, J. Breath Res. 3 (2009) 027006.