Sf

CFa

b

a

ARRAA

KNCSG

1

aasbuywtccfcrpnnswntb

0d

Sensors and Actuators B 141 (2009) 97–103

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical

journa l homepage: www.e lsev ier .com/ locate /snb

elective detection of SO2 at room temperature based on organoplatinumunctionalized single-walled carbon nanotube field effect transistors

ristina C. Cida, Giselle Jimenez-Cadenaa, Jordi Riua,∗, Alicia Marotoa,. Xavier Riusa, Guido D. Batemab, Gerard van Kotenb

Department of Analytical and Organic Chemistry, Rovira i Virgili University, C/Marcel lí Domingo s/n, 43007 Tarragona, SpainChemical Biology & Organic Chemistry, Debije Institute for Nanomaterials Science, Faculty of Science, Utrecht University, 3584 CH Utrecht, The Netherlands

r t i c l e i n f o

rticle history:eceived 30 December 2008eceived in revised form 23 April 2009

a b s t r a c t

We report a field effect transistor (FET) based on a network of single-walled carbon nanotubes (SWCNTs)that for the first time can selectively detect a single gaseous molecule in air by chemically functional-izing the SWCNTs with a selective molecular receptor. As a target model we used SO2. The molecular

ccepted 12 May 2009vailable online 28 May 2009

eywords:anomaterialarbon nanotubes

synthetic receptor is a square-planar NCN-pincer platinum (II) complex (NCN is the N,C,N′-terdentate-coordinating monoanionic [C6H3(CH2NMe2)3-2,6]− ligand) to which SO2 selectively binds. Because ofthe strong electronwithdrawing character of SO2, it withdraws negative charge from the synthetic recep-tor thus affecting the electronic properties of the functionalized SWCNTs. The minimum concentrationdetected is 0.05% SO2 in air at room temperature. Interferences like CH4 and CO2 need to be present in a

n SO2

O2

as

higher concentration tha

. Introduction

In recent decades, interest in gas detection has consider-bly increased because of the environmental problems caused bytmospheric contamination and because of the high toxicity ofome industrial facilities. Interest in SO2 monitoring has increasedecause of its highly reactive nature and relative abundance. SO2 issually determined using chromatography, electrochemical anal-sis and spectroscopy, most of these techniques being combinedith solution-absorbing pre-treatment chemical analysis. These

echniques are expensive, time-consuming, and require sophisti-ated measuring equipment [1]. SO2 has also been determined withhemical sensors using different detection systems, but these sufferrom different problems such as high working temperatures (espe-ially in sensors based on metallic oxides), lack of selectivity, highesponse times or low sensitivity [2–4]. To overcome many of theseroblems, nanostructured materials have been incorporated intoew sensors. But despite the immense potential shown by theseew materials some performance characteristics of these new sen-ors, such as selectivity, are not always well defined, especially

hen detecting gaseous substances [5,6]. Single-walled carbonanotubes (SWCNTs) are nanostructures with outstanding elec-rical and mechanical properties. Several authors [7,8] noted thatare SWCNTs are sensitive to many gases, showing that bare SWC-

∗ Corresponding author. Tel.: +34 977 558 491; fax: +34 977 558 446.E-mail address: jordi.riu@urv.cat (J. Riu).

925-4005/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.snb.2009.05.021

to give a significant response.© 2009 Elsevier B.V. All rights reserved.

NTs cannot selectively detect a single gaseous molecule. Tans etal. [9] reported the first field effect transistor based on a singlesemiconducting SWCNT, and Kong et al. [10] found that the elec-trical conductivity of these carbon nanotube field effect transistors(CNTFETs) was sensitive to various gases such as ammonia or nitro-gen dioxide and that these CNTFETs can thus operate as sensitivechemical sensors [11,12]. These sensors are based on CNTFETs inwhich the conductor channel can be either a single semiconductingSWCNT or a network of SWCNTs [13]. However, selectivity of CNT-FET devices towards a single gaseous molecule is still an unsolvedproblem. A functionalization process with a specific receptor toobtain a selective CNTFET gas sensor has not yet been developed.Therefore, interferences and selectivity still need to be checkedfrom a practical point of view for the selective detection of smallcompounds using CNTFET devices, especially for small gaseous ana-lytes. Selective detection of small analytes using CNTFET devicesis still a challenging issue and so far the main efforts have beenaddressed to the detection of such analytes in solution. Zhao etal. [14] developed a CNTFET functionalized with a small syntheticreceptor, pyrenecyclodextrin, that was able to detect a range ofsmall chemicals in solution, although no experimental influencefrom any other compounds was reported in this paper. Very recentlySánchez-Acevedo et al. [15] developed a CNTFET functionalized

with a nuclear receptor for the selective detection of picomolarconcentrations of bisphenol A in water. In this paper we reportfor the first time that a CNTFET can be functionalized with anspecific receptor towards the successfull selective detection of atarget gaseous analyte. As a proof-of-concept, we used a field effect

9 Actua

tfrmFad

attfiiimNs

F(S

8 C.C. Cid et al. / Sensors and

ransistor based on networks of SWCNTs as the transducing layerunctionalized with an aryl platinum (II) complex [16] as a specificeceptor to selectively and quantitatively detect a single gaseousolecule, SO2, in gaseous samples at room temperature. This CNT-

ET obtains a linear relationship between instrumental responsend gas concentrations, which opens the way to quantitativelyetecting SO2 in gaseous samples.

The disadvantage of the CNTFETs reported so far in the liter-ture and based on non-functionalized SWCNTs is that, althoughhey can be very sensitive gas sensors [8,17–19], they are not selec-ive because they are sensitive to many gases [5,6]. The chemicalunctionalization of SWCNTs is a way to increase the selectiv-ty of CNTFETs. Two main strategies have been applied so far to

mprove the selectivity of these CNTFETs as gas sensors: (a) coat-ng SWCNTs with a polymer and (b) decorating SWCNTs with

etal nanoparticles. As an example of the first strategy, SWC-Ts have been coated with polyethylene imine (PEI), a polymer

howing alkaline properties, to detect acidic gases such as CO2

ig. 1. (a) Functionalization process of SWCNTs with PEI and the platinum complex. (b) Rc) Atomic force microscope (AFM) image of a typical SWCNT network obtained by CVD (lWCNTs is between 5.5 and 6 nm (right).

tors B 141 (2009) 97–103

[20] or NOx [21]. These acidic gases are able to react with theamino groups of PEI and to decrease the conductance of the func-tionalized SWCNTs. Selectivity between different acid gases is notachieved, although gases which are not able to donate or shareelectrons are effectively discriminated. In the second strategy, Staret al. [22] showed that a sensor array consisting of multiple CNT-FET devices with different catalytic metallic nanoparticles adsorbedonto the SWCNTs presented selectivity towards several gases. Thesensing mechanisms of CNTFETs applied to gas detection thatare described in the literature can be explained with two mainapproaches. [23] One is based on the modification of the Schot-tky barrier due to the adsorption of the target compound ontothe metal-nanotube junction [17,24], while the other one is based

on the charge transfer due to the adsorption of the target com-pound onto the nanotube’ sidewalls [13,25]. Both mechanisms canbe distinguished by passivating the metal-nanotube junction. Eventhough the gas molecules can penetrate through the pinholes of thepassivation layer, the response time increases when the sensing

eaction of the remaining free amino groups of PEI with the blocking molecule NAS.eft). The SWCNT have been already functionalized. The height of the functionalized

Actua

m[

tfitntaTdCSplttagpgiwiiwtefi

2

ww[bgwdolmtwsatto(

Fc

C.C. Cid et al. / Sensors and

echanism is caused by the modification of the Schottky barrier17].

The CNTFET described in this paper as a proof-of concept ofhe selective detection of a single gaseous analyte in air for therst time (SO2 in our case), is assembled using a three step func-ionalization process. In the first step, the SWCNTs are coatedon-covalently with a layer of PEI that is irreversibly adsorbed overhe whole surface of the SWCNTs. The polymer coating offers suit-ble functional groups for a further covalent binding of the receptor.he second step incorporates the synthetic receptor to selectivelyetect SO2. In our case the receptor is the platinum complex [PtI(4-OONC4O2H2-2,6-(CH2NMe2)2-C6H2] that selectively reacts withO2 by forming a coordinated bond between the platinum and sul-hur atoms at room temperature [16,26]. The platinum complex is

inked to the PEI-coated SWCNT network by a reaction betweenhe amino groups of the polymer and the succinimide ester ofhe platinum complex (Fig. 1a). The third step of the function-lization process consists of blocking the remaining free aminoroups of PEI that may not have reacted with the platinum com-lex and which could give rise to interferences by reacting withaseous substances other than SO2. For this reason we use the block-ng molecule N-acryloxysuccinimide (NAS). This molecule reacts

ith the remaining free amino groups of PEI (Fig. 1b) and preventsnterferences as a result of possible acid-base reactions with annterferent gas (for example with acidic gases such as CO2). In thisay, the unique recorded interaction between the CNTFET and the

est samples is between the SO2 and the specific receptor complexnabling the selective detection of a single gaseous analyte for therst time.

. Experimental

The network of SWCNTs that forms the channel of the CNTFETas grown by chemical vapour deposition (CVD) over Si substratesith a layer of SiO2 (500 nm thick) on top as described previously

27]. Source and drain electrodes (5 �m × 10 �m) were patternedy using optical lithography (Cr/Au 2/30 nm thick, respectively). Theap distance between the electrodes was 2.5 �m. The Si substrateas used as a gate electrode. For the functionalization process, theevices with the SWCNTs and the electrodes were first submergedvernight in a 20% (w/v) aqueous solution of PEI (high molecu-ar weight, water-free, from Sigma Aldrich, Tres Cantos, Spain). A

onolayer of polymer is irreversibly adsorbed on the sidewalls ofhe nanotubes [28,29]. The devices were then rinsed thoroughlyith water to remove the non-specific adsorbed polymer over the

ample surface, and dried with nitrogen. The platinum complex

cting as the specific receptor for SO2 was synthesized accordingo a literature procedure [30]. To anchor the platinum complexo the PEI-coated SWCNTs, the devices were soaked in a solutionf 3 mg/mL of the platinum complex [PtI(4-COONC4O2H2-2,6-CH2NMe2)2-C6H2] in dichloroethane and allowed to react (amide

ig. 2. (a) Gate dependence of the source–drain current of bare SWCNTs and after adsorbiompletely functionalized CNTFET device in air (0% of SO2) and after exposure to an SO2 c

tors B 141 (2009) 97–103 99

bond formation with the NH2 group of PEI) for 3 h at room temper-ature. The devices were rinsed thoroughly with dichloroethane andwater and dried with nitrogen. The last step of the functionaliza-tion process consisted of blocking the remaining amino groups ofPEI that did not react with the platinum complex. The devices wereimmersed in a solution of 3 mg/mL of NAS (from Sigma Aldrich, TresCantos, Spain) in water for 2 h at room temperature. Afterwards, thedevices were rinsed with water and dried with nitrogen. For the gasmeasurements, the completely functionalized CNTFET device wasthen placed in a gas cell and exposed to SO2 in an air atmosphere.The gas assays were performed at room temperature at a relativehumidity of 30–40%. SO2 (99.98%), CO2 (99.8%) and CH4 (99.0%)were purchased from Carburos Metálicos (Barcelona, Spain).

3. Results and discussion

Our CNTFET contains a network of SWCNTs grown using thechemical vapour deposition (CVD) method. The diameter of theSWCNTs is between 1.7 and 2 nm. The density of the network (seeFig. 1c) that connects the source and drain electrodes of the CNTFETis similar in all the devices and has several parallel pathways. Theheight of the completely functionalized SWCNTs (i.e. after linkingthe platinum complex to the PEI-coated SWCNTs) is between 5.5and 6 nm (Fig. 1c), what indicates that the height of the PEI coatingand the platinum complex is about 3.5–4 nm.

Fig. 2 shows the electrical characteristics of a device foreach functionalization step at a constant source to drain voltage,Vsd = 0.25 V at room temperature. The gate voltage (Vg) was sweptbetween −5 and 5 V using a semiconductor parameter analyzer(Agilent E5270A). Fig. 2 shows the average values of three instru-mental replicates. We can see that the as-grown SWCNT networkshows a p-type semiconductor behaviour [31]. Because the networkgrown by CVD is made of both metallic and semiconducting nan-otubes [32], the conductance in the off state will be governed bythe conductance of the metallic ones (Fig. 2). Low on/off ratios areobserved (Fig. 2a) which are characteristic of networks with a gapdistance of 2.5 �m between source and drain electrodes [33]. Thenon-covalent coating of PEI changes the semiconductor behaviourof the SWCNTs’ network from p-type to n-type (Fig. 2a) because ofthe electron donating ability of the amine groups of PEI [28]. Fig. 2bshows the current versus gate voltage characteristics for a function-alized CNTFET before and after exposure to 1.8% (v/v) of SO2 in an airatmosphere. When SO2 is present in the atmosphere, the platinumcomplex reacts reversibly to form a coordination bond between themetallic centre and the sulphur atom (Fig. 3). This reaction changesthe geometry of the metallic centre from square planar to square

pyramidal and creates a steric effect between the methyl groups ofthe nitrogen atoms and the SO2 molecule coordinated in the api-cal position of the complex [34,35]. The reversible bond formationbetween the sulphur atom in SO2 and Pt is a consequence of thenucleophilicity of the platinum (II) center and the Lewis acidity of

ng PEI onto the SWCNTs. (b) Source–drain current (I) versus gate voltage (Vg) of theoncentration (1.8%, v/v). Source to drain voltage was set at 0.25 V.

100 C.C. Cid et al. / Sensors and Actuators B 141 (2009) 97–103

F d thec struct

tbet

ftprndTLtasaa[aFt

rcafnqcpttNdt

ced

tinguish from the noise in the baseline) was 0.05%. We always usedthe same CVD parameters, however, the density of the networkof SWCNTs and the sensitivities may change slightly for differentdevices. Nevertheless, we were able to detect concentrations of SO2

ig. 3. Representation of the complex formed between the platinum complex anoordination bond between the metallic centre and the sulphur atom. The receptor

he sulphur atom in SO2. It is important to remark the role playedy the iodine atom in the complex since the metal-halide bond isssential for tuning the nucleophilic character of Pt and to ensurehe full reversibility of the SO2 bonding [34].

The decrease in the electrical current after the exposure of theunctionalised CNTFETs to SO2 can be explained by the strong elec-ronwithdrawing character of SO2. When SO2 interacts with thelatinum complex, the SO2 withdraws a negative charge from theeceptor and thus affects the n-type PEI-coated SWCNTs. Since the-type SWCNTs have electrons as major carriers, the electron with-rawing effect decreases the conductance of the CNTFETs [20,21].he charge transfer through single bonds has been described byiu and Diao [36] and Huang et al. [37] who explain the chargeransfer process between Au substrates and SWCNTs bridged byliphatic compounds. However, a second mechanism could be con-idered and the change in the carrier mobility of the CNTFET couldlso result from the modification of the Schottky barrier due to thedsorption of the target compound on the metal-nanotube junction16,20]. Fig. 2b shows that the device’s response to SO2 was highert positive gate voltages, leaving the negative section less affected.or on-time measurements the gate voltage was set at +5 V, withhe Vsd fixed at 0.25 V.

Control experiments were carried out to show both that theeceptor was correctly anchored to the substrate and that thehanges in the instrumental response were only due to the inter-ction between the platinum complex and SO2. In this way, weunctionalized SWCNTs with PEI and subsequently with NAS, butot with the platinum complex. These CNTFET devices were subse-uently exposed to up to 10% v/v of SO2 and CO2, with no significanthange to the electrical current (data not shown). Therefore, thelatinum complex is needed to generate an electrical response uponhe addition of SO2. In this way we proved, at the same time, thathe functionalization with PEI and NAS effectively shields the SWC-Ts from the interaction of acid gases such as CO2, and that theecrease in the electrical current is only observed when the syn-

hetic receptor and the SO2 interact.

To determine the response time of the CNTFET, the electricalurrent of the CNTFET was followed during the exposure to sev-ral concentrations of SO2. Fig. 4 shows how the electrical currentecreases for increasing concentrations of SO2 in an air atmosphere

SO2 (left). The platinum complex acts as a receptor reacting with SO2 to form aure changes from a square planar to a square pyramidal state (right).

in the gas cell at room temperature (at Vsd = 0.25 V and Vg = 5 V).Response times range from 16 min at lower concentrations of SO2to 8 min at higher concentrations. Before each new addition of SO2we always checked that the electrical current was stabilized givingrise to the horizontal sections seen in Fig. 4.

The inset in Fig. 4 shows a linear dependence for the electricalcurrent along the concentration range. The sensitivity (calculatedas the slope of the fitted straight line) was −0.028 �A/(%, v/v) witha standard deviation of 0.004 �A/(%, v/v). The range of concentra-tions of SO2 detected with this CNTFET was from 0.25% to 1.8%(v/v), which was similar, for instance, to the concentrations of CO2detected by Star et al. in a PEI-coated CNTFET device [20]. The min-imum concentration of SO2 that we were able to detect with ourCNTFET devices (i.e. the minimum concentration that we could dis-

Fig. 4. Time dependence of the source–drain current (I), at Vg = +5 V and Vsd = 0.25 V,for increasing concentrations of SO2. Vertical straight lines define the compositionover the time in the gas cell. The inset shows the source–drain current versus theconcentration of SO2. The error bars were obtained with two measurements withthe same device in two consecutive days.

Actuators B 141 (2009) 97–103 101

oTcttwct[sAaSbtgsppiapw

a[stsvSTtcwplfcdFcr

CabF(tcotwctaccbsttebp

Fig. 5. Change of the source–drain current (I) at Vg = +5 V and Vsd = 0.25 V in thepresence of interfering gases. Increasing concentrations of CH4 and CO2 were added.

C.C. Cid et al. / Sensors and

f the same order of magnitude using different CNTFET devices.he sensitivity of the sensor is also affected by the steric modifi-ations in the platinum complex [16]. A suitable modification ofhe metal center by tuning the ligand properties (e.g. changinghe electrophilicity of platinum either by introducing electron-ithdrawing or releasing para-substituents on the aryl ring or by

hanging the basicity of the nitrogen donor atoms) could increasehe charge transference between the platinum complex and SO226]. So with an optimized design of the synthetic pincer ligand pre-umably an improved sensitivity of the sensor could be achieved.nother way to increase the sensitivity of the sensor would be usingnother strategy to directly attach the molecular receptor to theWCNT instead of using an intermediate coating like PEI. This cane achieved for instance covalently linking the molecular recep-or to a pyrenyl group, and then directly adsorbing the pyrenylroup over the SWCNTs. The pyrenyl group is known to interacttrongly with the SWCNTs via �-stacking [38]. In this case furtherrotection would be needed to fill the gaps left between adsorbedyrenil groups over the SWCNTs to prevent the non-specific bind-

ng of other gaseous molecules. If the sensing mechanism is due tomodification of the charge transfer when SO2 is coupled to thelatinum complex, we can presumably think that the sensitivityould be higher if the molecular receptor is closer to the SWCNTs.

We can see in Fig. 4 that the CNTFET was successfully regener-ted in only 25 min by re-exposing it to air at room temperature34,35]. The following day the CNTFET was again exposed to theame increasing concentrations of SO2. The electrical current forhat day (data not shown) was recorded in a similar way to thathown in Fig. 4. The inset of Fig. 4 shows the average of the twoalues of the electrical current obtained for each concentration ofO2 taken on two consecutive days with the same sensing device.he error bars (inset in Fig. 4) show the difference between thesewo measurements on these two different days. Similar electricalurrent values are then obtained after regenerating the CNTFET,hat shows that the CNTFET can be used for quantification pur-oses during several days. The sensor has shown this stability at

east for seven days. This regeneration might be achieved evenaster by using other mechanisms to release the SO2 from theomplex formed with the receptor [26]. The recovery time of ourevices was similar to that observed for other functionalized CNT-ETs [21,39]. The recovery time of the baseline after regenerationould be explained by the process involving the reversibility of theeaction between the receptor and SO2.

The selectivity was checked using other gases such as CH4 andO2 that might be found in the same samples together with SO2nd that can also contribute to environmental problems [40]. As haseen previously indicated [20], sensors using polymer coated CNT-ETs need to eliminate the cross reactivity between similar gasesi.e. acidic gases like CO2 or SO2) to be able to discriminate amonghem. For these reasons these gases have been used as interferentompounds. Fig. 5 shows the selectivity of the device in the presencef CH4 and CO2 against the response for SO2. Increasing concentra-ions of the interfering gas (i.e. CH4 or CO2) in an air atmosphereere added to the gas cell in the absence of SO2 and the electrical

urrent was recorded at Vg = 5 V and Vsd = 0.25 V. Fig. 5 shows thathe electrical current did not significantly change for CH4 even up toconcentration of 15% (v/v). No appreciable changes in the electricalurrent were observed for successive additions of CO2 until the con-entration reached a value of 10%. In any case, these changes cannote attributed to the interaction of CO2 with the amino groups of PEI,ince one of the previously described control experiments showed

hat a device properly functionalized with PEI and NAS (withouthe platinum complex) did not respond to CO2. The change in thelectrical current after the addition of over 10% (v/v) of CO2 coulde related to a low interaction between CO2 and the platinum com-lex. This means that the sensor is selective but not specific for

Vertical dotted lines describe the gas cell composition over time. A 15% v/v CH4 didnot affect the base signal and a 10% (v/v) of CO2 makes a change in the current of10 nA. These two experiments for interfering gases are compared to the 1% (v/v) ofSO2 that reduces the CNTFET current to 60 nA.

SO2, and that there is a certain degree of cross-reactivity with CO2.Therefore although we prove for the first time that CNTFETs can beused for the selective detection of a single gaseous analyte, the spe-cific detection of the target gaseous analyte should be performedfunctionalizing the SWCNTs with a receptor specific for the targetanalyte.

The platinum complex is known to react with I2 [26], but I2 isusually found in extremely low concentrations in gaseous samples.Water does not affect the CNTFET as it could also be used in aque-ous solutions. The blocking molecule (NAS) effectively shields thePEI-coated SWCNT network from the presence of other gases suchas CO2 or CH4. The presence of the platinum complex makes ourdevices selective to SO2 because a higher concentration of othergases is needed to cause a significant and similar response to thatproduced by SO2. Therefore, the proposed chemical functionaliza-tion of SWCNTs operates in a twofold manner: (1) the syntheticreceptor makes the CNTFET selective for SO2 and (2) the function-alization process together with the NAS blocking step shields theCNTFET device from interfering gases.

4. Conclusion

In this paper we show for the first time that CNTFETs can besuccessfully functionalized for the selective detection of a singlegaseous substance. As a proof-of-concept, we describe a CNTFETdevice for selectively and quantitatively detecting a single gaseousmolecule, SO2, at room temperature by means of a synthetic recep-tor. The CNTFET is based on a reversible interaction between aplatinum complex and SO2. The SWCNTs have been non-covalentlyfunctionalized with a polymer (PEI) with functional groups to beaccessible for a desired receptor without damaging the SWCNT’selectrical properties. The complete functionalized network of theSWCNTs is properly blocked because acidic gases show no signifi-cant interaction with amine groups of PEI. The platinum complex isselective to SO2 because a higher concentration of interfering gasesis needed to obtain a significant response, although the degree ofselectivity is given by the cross-reactivity of the platinum complex

and interfering gases. The sensing mechanism could involve eithera modification of the Schottky barrier due to the absorption of thetarget compound on the metal-nanotube junction or a modifica-tion of the charge transfer when SO2 is coupled to the platinumcomplex. This strategy could be used for detecting other gaseous

1 Actua

mtcb

A

a6fiSAaBTf

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

02 C.C. Cid et al. / Sensors and

olecules by functionalizing the CNTFETs with a suitable recep-or. Moreover, the CNTFET devices functionalized with the platinumomplex could be used as a device to detect SO2, not only selectively,ut also quantitatively.

cknowledgements

The authors would like to thank the Spanish Ministry of Sciencend Education (projects NAN2004-09306-C05-05 and CTQ2007-7570/BQU) and the European Union (project STRP 01071) fornancial support. J. Riu would like to thank the Spanish Ministry ofcience and Technology for providing his Ramón y Cajal contract.. Maroto would also like to thank the Spanish Ministry of Sciencend Technology for providing her Juan de la Cierva contract. G. D.atema would like to thank the Council for Chemical Sciences ofhe Netherlands Organization for Scientific Research (NWO/CW)or financial support.

eferences

[1] D. Harvey, Modern Analytical Chemistry, McGraw-Hill, New York, 2000.[2] H. Li, Q. Wang, J. Xu, W. Zhang, L. Jin, A novel nano-Au-assembled amperomet-

ric SO2 gas sensor: preparation, characterization and sensing behavior, Sens.Actuators B 87 (2002) 18–24.

[3] E. Comini, Metal oxide nano-crystals for gas sensing, Anal. Chim. Acta 568(2006) 28–40.

[4] X. Liang, T. Zhong, B. Quan, B. Wang, H. Guan, Solid-state potentiometric SO2

sensor combining NASICON with V2O5-doped TiO2 electrode, Sens. ActuatorsB 134 (2008) 25–30.

[5] G. Jimenez-Cadena, J. Riu, F.X. Rius, Gas sensors based on nanostructured mate-rials, Analyst 132 (2007) 1083–1099.

[6] D.R. Kauffman, A. Star, Carbon nanotube gas and vapor sensors, Angew. Chem.Int. Ed. 47 (2008) 6550–6570.

[7] A. Goldoni, R. Larciprete, L. Petaccia, S. Lizzit, Single-wall carbon nanotubesinteraction with gases: sample contaminants and environmental monitoring,J. Am. Chem. Soc. 125 (2003) 11329–11333.

[8] A. Goldoni, L. Petaccia, L. Gregoratti, B. Kaulich, A. Barinov, S. Lizzit, A. Laurita,L. Sangaletti, R. Larciprete, Spectroscopic characterization of contaminants andinteraction with gases in single-walled carbon nanotubes, Carbon 42 (2004)2099–2112.

[9] S.J. Tans, A.R. Verschueren, C. Dekker, Room-temperature transistor based on asingle carbon nanotube, Nature 393 (1998) 49–52.

10] J. Kong, N.R. Franklin, C. Zhou, M.G. Chapline, S. Peng, K. Cho, H. Dai, Nanotubemolecular wires as chemical sensors, Science 287 (2000) 622–625.

11] P.G. Collins, K. Bradley, M. Ishigami, A. Zettl, Extreme oxigen sensitivity of elec-tronic properties of carbon nanotubes, Science 287 (2000) 1801–1804.

12] P. Qi, O. Vermesh, M. Grecu, A. Javey, Q. Wang, H. Dai, S. Peng, K.J. Cho, Towardlarge arrays of multiplex functionalized carbon nanotube sensors for highlysensitive and selective molecular detection, Nano Lett. 3 (2003) 347–351.

13] G. Grüner, Carbon nanotube transistors for biosensing applications, Anal.Bioanal. Chem. 384 (2006) 322–335.

14] Y.-L. Zhao, L. Hu, J.F. Stoddart, G. Grüner, Pyrenecyclodextrin-decorated single-walled carbon nanotube field-effect transistors as chemical sensors, Adv. Mater.20 (2008) 1910–1915.

15] Z.C. Sánchez-Acevedo, J. Riu, F.X. Rius, Fast picomolar selective detection ofbisphenol A in water using a carbon nanotube field effect transistor function-alized with estrogen receptor-�, Biosens. Bioelectron. 24 (2009) 2842–2846.

16] M. Albrecht, G. van, Koten, Platinum group organometallics based on “pincer”complexes: sensors, switches, and catalysts, Angew. Chem. Int. Ed. 40 (2001)3750–3781.

17] J. Zhang, A. Boyd, A. Tselev, M. Paranjape, P. Barbara, Mechanism of NO2 detec-tion in carbon nanotube field effect transistor chemical sensors, Appl. Phys. Lett.88 (2006) 123112.

18] J. Li, Y. Lu, Q. Ye, M. Cinke, J. Han, M. Meyyappan, Carbon nanotube sensors forgas and organic vapor detection, Nano Lett. 3 (2003) 929–933.

19] J. Andzelm, N. Govind, A. Maiti, Nanotube-based gas sensors—role of structuraldefects, Chem. Phys. Lett. 421 (2006) 58–62.

20] A. Star, T.R. Han, V. Joshi, J.C.P. Gabriel, G. Gruner, Nanoelectric carbon dioxidesensors, Adv. Mater. 16 (2004) 2049–2052.

21] O. Kuzmych, B.L. Allen, A. Star, Carbon nanotube sensors for exhaled breathcomponents, Nanotechnology 18 (2007) 375502.

22] A. Star, V. Joshi, S. Skarupo, D. Thomas, J.C.P. Gabriel, Gas sensor array based onmetal-decorated carbon nanotubes, J. Phys. Chem. B 110 (2006) 21014–21020.

23] D.R. Kauffman, A. Star, Electronically monitoring biological interactionswith carbon nanotube field-effect transistors, Chem. Soc. Rev. 37 (2008)1197–1206.

24] Z. Chen, J. Appenzeller, J. Knoch, Y.-m. Lin, P. Avouris, The role of metal-nanotubecontact in the performance of carbon nanotube field-effect transistors, NanoLett. 5 (2005) 1497–1502.

tors B 141 (2009) 97–103

25] Q. Fu, J. Liu, Integrated single-walled carbon nanotube/microfluidic devices forthe study of the sensing mechanism of nanotube sensors, J. Phys. Chem. B 109(2005) 13406–13408.

26] M. Albrecht, R.A. Gossage, M. Lutz, A.L. Spek, G. van Koten, Diagnosticorganometallic and metallodendritic materials for SO2 gas detection: reversiblebinding of sulfur dioxide to arylplatinum(II) complexes, Chem. Eur. J. 6 (2000)1431–1445.

27] C.C. Cid, J. Riu, A. Maroto, F.X. Rius, Carbon nanotube field effect transistorsfor the fast and selective detection of human immunoglobulin G, Analyst 133(2008) 1005–1008.

28] M. Shim, A. Javey, N.W. Shi Kam, H. Dai, Polymer functionalization for air-stablen-type carbon nanotube field-effect transistors, J. Am. Chem. Soc. 123 (2001)11512–11513.

29] A. Star, J.P. Gabriel, K. Bradley, G. Grüner, Electronic detection of specific proteinbinding using nanotube FET devices, Nano Lett. 3 (2003) 459–463.

30] B.M.J.M. Suijkerbuijka, M.Q. Slagta, R.J.M.K. Gebbinka, M. Lutzb, A.L. Spekb, G.van Koten, Platinum-pincer introduction using active ester chemistry, Tetrahe-dron Lett. 43 (2002) 6565–6568.

31] R. Martel, T. Schmidt, H.R. Shea, T. Hertel, P. Avouris, Single- and multi-wallcarbon nanotube field-effect transistors, Appl. Phys. Lett. 73 (1998) 2447–2449.

32] W. Kim, H.C. Choi, H. Shim, Y. Li, D. Wang, H. Dai, Synthesis of ultralong andhigh percentage of semiconducting single-walled carbon nanotubes, Nano Lett.2 (2002) 703–708.

33] E.S. Snow, J.P. Novak, P.M. Campbell, D. Park, Random networks of car-bon nanotubes as an electronic material, Appl. Phys. Lett. 82 (2003) 2145–2147.

34] M. Albrecht, R.A. Gossage, U. Frey, A.W. Ehlers, E.J. Baerends, A.E. Merbach, G.van Koten, Mechanistic aspects of the reversible binding of SO2 on arylplat-inum complexes: experimental and ab initio studies, Inorg. Chem. 40 (2001)850–855.

35] M. Albrecht, M. Schlupp, J. Bargon, G. van Koten, Detection of ppm quantitiesof gaseous SO2 by organoplatinum dendritic sites immobilised on a quartzmicrobalance, Chem. Commun. 18 (2001) 1874–1875.

36] Z. Liu, P. Diao, Electrochemistry at chemically assembled single-wall carbonnanotube arrays, J. Phys. Chem. B 109 (2005) 20906–20913.

37] H.-S.I.X.-J. Huang, O. Yarimaga, J.-H. Kim, D.-Y. Jang, D.-H. Lee, H.-S. Kim, Y.-K. Choi, Electrochemical behavior of needle-like and forest-like single-walledcarbon nanotube electrodes, J. Electroanal. Chem. 594 (2006) 27–34.

38] R.J. Chen, Y. Zhang, D. Wang, H. Dai, Noncovalent sidewall functionalization ofsingle-walled carbon nanotubes for protein immobilization, J. Am. Chem. Soc.123 (2001) 3838–3839.

39] B.N.B.-Y.Y. Megan, A.D. Marc, V.M. Nosang, Ting Zhang, Electrochemically func-tionalized single-walled carbon nanotube gas sensor, Electroanalysis 18 (2006)1153–1158.

40] Y.J. Lu, J. Li, J. Han, H.T. Ng, C. Binder, C. Partridge, M. Meyyappan, Room tempera-ture methane detection using palladium loaded single-walled carbon nanotubesensors, Chem. Phys. Lett. 391 (2004) 344–348.

Biographies

Cristina C. Cid obtained the BSc (2003), MSc (2007) and PhD (2009) in Chemistry,at Rovira i Virgily University, Tarragona, Spain. Her main interest is the developmentof sensors based on carbon nanotubes exploiting its particular properties combinedwith molecular recognition processes.

Giselle Jimenez-Cadena received her degree in chemistry from UniversidadNacional de Colombia in 2005 and her Master in Nanoscience and Nanotechnologyfrom Universitat Rovira i Virgili in Spain during 2007. Nowadays she is developingher PhD in Materials for Engineering in Università di Brescia in Italy. Her main activ-ities deal with synthesis of nanomaterials of metal oxides and their applications forsolar cells. She has worked in theoretical chemistry, synthesis of carbon nanotubesand their applications for gas sensing.

Jordi Riu obtained a BSc (1993) and PhD (1999) in Chemistry, both from Rovira i Vir-gili University, Tarragona, Catalonia, Spain. His main interests are the developmentof chemical and biochemical sensors based on carbon nanotubes and molecularrecognition processes.

Alicia Maroto obtained a BSc (1997) and PhD (2002) in Chemistry, both from Rovirai Virgily University, Tarragona, Spain. She has been a researcher at the departmentof Analytical and Organic Chemistry in the Rovira i Virgili University from 1998 until2008. She is currently an associate professor at ESCOM, Compiègne, France. Her maininterest is the development of sensors based on carbon nanotubes and molecularrecognition processes.

F. Xavier Rius is professor of Analytical Chemistry at the Rovira i Virgili University,Tarragona, Spain. He is the head of the Chemometrics, Qualimetrics and Nanosensors

group. His main interest is the development of electrochemical sensors, contain-ing chemo- and bio-recognition layers that take advantage of the new propertiesprovided by nanostructured materials.

Guido D. Batema obtained a MSc (2002) in Molecular Chemistry at the Universityof Amsterdam, The Netherlands, with a major in Inorganic Chemistry and Catalysis.

Actua

INc

GD

C.C. Cid et al. / Sensors and

n 2007 he received his PhD degree in Chemistry at the Utrecht University, Theetherlands. His main interest is the application of conjugated ECE-Pincer metalomplexes in new optical materials, bio-conjugates, biomarkers and sensors.

erard van Koten is Emeritus University Professor of Utrecht University. He wasean of the Faculty of Science at Utrecht University. In 2007 he became Distinguished

tors B 141 (2009) 97–103 103

Research Professor (part-time) at Cardiff University. His research interests comprisethe study of fundamental processes in organometallic chemistry and the applicationof organometallic complexes as homogeneous catalysts. Recent interests are theimmobilization of catalysts on dendrimers and on inorganic supports as well as theuse of these materials for tandem catalysis. This includes also the development ofnovel enzyme-organometallic catalyst hybrid materials.