microstructure characterization and no2 sensing properties of tungsten oxide

Likely, tungsten oxide in nanostructures like nanowires, nanosheets and nanorods were investigated [16–18], and they revealed good sensing properties while detecting toxic and haz-ardous

Sensors and Actuators B xxx (2010) xxx–xxx

Contents lists available atScienceDirect Sensors and Actuators B: Chemical

j o u r n a l h o m e p a g e :w w w e l s e v i e r c o m / l o c a t e / s n b

Microstructure characterization and NO 2 -sensing properties of tungsten oxide nanostructures

Yuxiang Qin∗, Ming Hu, Jie Zhang

School of Electronics and Information Engineering, Tianjin University, No 92, Weijjin Road, Nankai District, Tianjin 300072, PR China

a r t i c l e i n f o

Article history:

Received 16 November 2009

Received in revised form 28 June 2010

Accepted 29 June 2010

Available online xxx

Keywords:

Tungsten oxide

Nanowires

Nanosheets

Solvothermal synthesis

Gas sensors

a b s t r a c t

Nanowires and nanosheets of tungsten oxide were synthesized by solvothermal method with differ-ent tungsten hexachloride (WCl6) concentrations in 1-propanol solvent The morphology and crystal structure of the tungsten oxide nanostructures were investigated by means of field emission scanning electron microscope, X-ray diffraction and transmission electron microscope The specific surface area and pore size distribution were characterized by Brunauer–Emmett–Teller gas-sorption measurements One-dimensional W18O49nanowire bundles were synthesized at the WCl6concentration of 0.01 M With the concentration increasing to 0.02 M, the structure of the pure two-dimensional WO3nanosheets was formed The NO2gas sensing properties of W18O49nanowires and WO3nanosheets were investigated at

100◦C up to 250◦C over NO2concentration ranging from 1 to 20 ppm Both nanowires and nanosheets exhibit reversible response to NO2gas at different concentrations In comparison to WO3nanosheets,

W18O49nanowire bundles showed a much higher response value and faster response–recovery charac-teristics to NO2gas, especially a much quicker response characteristic with response time of 19 s at the concentration of 5 ppm

© 2010 Published by Elsevier B.V

1 Introduction

With the industrial development, air pollution is becoming more

and more serious Especially, nitrogen oxide NOx(NO2or NO) which

results from combustion and automotive emissions is a main source

of acid rain and photochemical smog [1] So detection of toxic

NOxgas is very important for the environmental protection and

human health Thus far, several kinds of solid-state NO2 sensors,

such as resistive[2], capacitive[3], and surface acoustic wave (SAW)

[4]types have been developed In particular, resistive-type

sen-sors based on metal oxide semiconductors are well suited for NO2

detection owing to their remarkable gas sensing performance and

simple structures[2,5] Among various metal oxide

semiconduc-tors, tungsten oxide (WO3−x), which is a wide band-gap n-type

semiconductor, has been found to be a promising material for

detection of NO2 gas[6,7] However, most tungsten oxide

sen-sors based on nanocrystalline powders or films have been studied

widely and shown too slow response–recovery time; and their

sensitivity still needs to make further improvement for practical

application

In these years, novel nanostructures such as nanowires,

nan-otubes, nanorods and nanobelts, have been evaluated as ideal

candidates for gas sensing applications due to their larger

spe-∗ Corresponding author Tel.: +86 22 27402372; fax: +86 22 27401233.

E-mail address: qinyuxiang@tju.edu.cn (Y Qin).

cific surface area and smaller dimensions than the Debye length [8,9] In fact, gas sensing materials such as SnO2[10,11], ZnO[12] and In2O3 [13]with well-established nanostructure have exhib-ited higher sensitivity and quicker response to detect gases at low concentrations than the corresponding thin film materials [14,15] Likely, tungsten oxide in nanostructures like nanowires, nanosheets and nanorods were investigated [16–18], and they revealed good sensing properties while detecting toxic and haz-ardous gases For example, very good results for H2S gas sensor based on tungsten oxide nanowires and nanosheets have been reported Chen co-workers[19]found that the single-crystalline potassium-doped tungsten oxide nanosheets could exhibit high sensitivity, fast response time and good stability to H2S, acetone and Cl2 The investigation of Rao co-workers[20]indicated that the WO2.72 nanowires had much higher response value to H2S than the nanoparticles or nanoplatelets of WO3 Very recently, Ger-litz et al.[21]reported that the sensor based on tungsten oxide nanotubes can detect dilute NO2as low as 200 ppb at 200◦C and exhibit response two to three orders-of-magnitude higher than the one based on WO3thin film These results clearly demonstrate the potential of tungsten oxide nanostructures in toxic gas detection

In this work, we synthesized one-dimensional bundled nanowires and two-dimensional nanosheets of tungsten oxide by the sim-ple solvothermal method and evaluated their sensing properties towards NO2 gas ranging from 1 to 20 ppm at operating temper-ature of 100–250◦C Our study indicates that both nanowires and nanosheets of tungsten oxide have high sensitivities to NO2 gas; 0925-4005/$ – see front matter © 2010 Published by Elsevier B.V.

doi: 10.1016/j.snb.2010.06.063

acteristic with response time of 19 s to 5 ppm NO2.

2 Experimental

Nanowires and nanosheets of tungsten oxide were synthesized

by solvothermal method with tungsten hexachloride (WCl6) as

precursor and 1-propanol as solvent First, a certain amount of

WCl6 was dissolved in a little ethanol to form a solution in a

beaker The ratio of WCl6mass to ethanol volume is 0.1 g/ml Then,

1-propanol was added to the solution which was subsequently

transferred to and sealed in a 100 ml Teflon-lined stainless steel

autoclave The volume of 1-propanol is 80 ml and the concentration

of WCl6 in 1-propanol varied from 0.01 to 0.02 M in our

experi-ments The solvothermal reaction was conducted at 200◦C for 9 h

in an electric oven After that, the autoclave was cooled naturally to

room temperature The final products were centrifuged and washed

sequentially by deionized water and ethanol several times, and the

obtained powder was dried at 70◦C for 6 h in air

The morphology, crystal structure, and phase composition of the

tungsten oxides were characterized using a field emission scanning

electron microscope (FESEM, FEI Nanosem 430), a X-ray

diffrac-tometer (XRD, RIGAKU D/MAX 2500V/PC, Cu K␣ radiation) and

a field emission transmission electron microscope (FETEM,

TEC-NAI G2F-20) In order to evaluate the specific surface area and

pore size distribution of the products, Brunauer–Emmett–Teller

(BET) gas-sorption measurements were carried out using

Quan-tachrome NOVA automated gas-sorption system after the samples

were vacuum-dried at 200◦C for 10 h The specific surface area was

estimated by nitrogen gas isotherms at a relative pressure (P/P0)

ranging from 0.005 to 0.1 Pore size distribution was obtained from

the analysis of the desorption branch of nitrogen gas isotherms

using the Barrett–Joyner–Halenda (BJH) model, and total pore

volume was determined by the amount of nitrogen adsorbed at

P/P0= 0.99

The gas sensors were fabricated by spin coating the slurry of

synthesized tungsten oxide nanostructures on the cleaned alumina

substrates which were attached with a pair of interdigitated Pt

electrodes with a thickness of 100 nm.Fig 1(a and b) shows the

schematic diagrams of the interdigitated electrodes and the sensor

respectively The electrodes were deposited using RF magnetron

sputtering method The coating slurry was prepared by

ultrasoni-cally dispersing tungsten oxide powders in mixed organic solvents

of terpineol and ethanol with 2:1 volume ratio for 2 h A physical

mask is used during spin coating to avoid the presence of slurry

at the end of the substrate The coated sensing films were dried in

air for 30 min subsequently annealed at 300◦C for 1 h at ambient

atmosphere in order to burn out the organic solvent used in

prepa-ration of coating paste and to enhance the adherence of the sensing

film to the sensor substrates Temperature was raised from

ambi-ent to 300◦C using a slow ramp of 2.5◦C/min in order to avoid the

occurrence of cracks in the films

The NO2 gas sensing measurements were carried out in a

home-built computer-controlled static gas sensing

characteriza-tion system consisting of a glass test chamber, a flat heating plate,

a professional digital multimeter and a data acquisition system

The schematic diagram of the gas sensing test system is shown in

Fig 1(c) The sensors were placed on the heating plate fixed in test

chamber The pure NO2gas was injected into the chamber directly

to get the desired concentration, and the sensor was recovered by

opening the top cover of the test chamber and setting up the electric

blower fixed at the bottom of the chamber An UNI-T UT70D

pro-fessional digital multimeter with the function of measuring range

automatic adjustment was used for continuously monitoring the

resistance change of the sensors during the whole measurement

process The electrically connection between the Pt electrodes and

Fig 1 Schematic diagrams of the interdigitated Pt electrodes (a), the sensor (b) and

the gas sensing test system (c).

the digital multimeter was realized by a pair of elastic Au-coated copper probes The acquired resistance data were stored in a PC for further analysis The sampling interval was set to 1 s The operating temperature of the sensing films was changed from 100 to 250◦C

by adjusting the temperature controller of heat plate The sensor response (S) was defined as S = (Rg− R0)/R0, where Rgand R0are the resistance of the sensor in the measuring gas and that in clean air, respectively The response time is defined as the time required for the resistance rising to 90% of the equilibrium value since the test gas is injected Conversely, the recovery time is the time for the resistance in equilibrium to go down to 10% of the original value in air since the test gas is released

3 Results and discussion

3.1 Structural characterization The morphologies of tungsten oxide nanostructures synthe-sized at different WCl6 concentrations after annealing at 300◦C for 1 h were shown inFig 2(a–d) It can be seen fromFig 2(a),

Y Qin et al / Sensors and Actuators B xxx (2010) xxx–xxx 3

Fig 2 (a, c, d) SEM images of tungsten oxide nanostructures synthesized at WCl6concentration of 0.01 M, 0.015 M and 0.02 M, respectively, after annealing at 300 ◦ C for 1 h (b) TEM image of the annealed tungsten oxide synthesized at 0.01 M The insets in (a) and (d) are the SEM images of the corresponding product before annealing.

the product synthesized at the WCl6concentration of 0.01 M after

thermal annealing exhibited one-dimensional nanowire bundles

features with diameters in 70–90 nm and lengths in 500–1000 nm

Further TEM examination can identify their bundled feature,

giv-ing evidence that many nanowires with diameters of about 10 nm

assembled along their main growth direction and formed a bundled

structure, as shown in theFig 2(b) Compared with the bundle

mor-phology before annealing shown in the inset inFig 2(a), it can be

seen that the thermal annealing results in the nanowire bundles

becoming thicker, shorter and straighter, indicating that a

possi-ble agglomeration had occurred Increasing WCl6concentration to

0.015 M, apparent evolution of the morphology can be observed

The SEM image shown inFig 2(c) exhibited the annealed

prod-uct was a mixture strprod-ucture of nanowires and nanosheets Up to

a high WCl6concentration of 0.02 M, a pure nanosheets structure

with thicknesses of 10–30 nm was formed (inset inFig 2(d)), and

showed unobvious change after annealing treatment (Fig 2(d))

From above results, it can be speculated that the WCl6

concentra-tion has a great effect on the specific morphologies of tungsten

oxide nanostructures synthesized by solvothermal method This is

in good agreement with the previous report[22] Lower solution

concentration contributed to the lower supersaturation of tungsten

source, promoting the growth of tungsten oxide nanowires[23]

At higher concentration, the highly saturated WCl6could prohibit

the growth of tungsten oxide nanowires along the main growth

direction

XRD analysis was carried out to identify the crystalline

struc-ture of the tungsten oxide before and after annealing at 300◦C for

1 h.Fig 3(a, c, e) and (b, d, f) respectively shows the XRD pat-terns of the as-synthesized and annealed samples As shown in Fig 3(a), the main diffraction peaks of the bundled nanowires syn-thesized at WCl6 concentration of 0.01 M can be well indexed as the monoclinic cell of W18O49with cell parameters of a = 18.32 Å,

b = 3.79 Å, c = 14.04 Å and ˇ = 115.04◦ (JCPDS No 65-1291) The

strongest peak intensity of (0 1 0) plane indicates that the

crys-Fig 3 XRD patterns of tungsten oxide nanostructures: (a, c, e) nanowires, mixture

and nanosheets synthesized at WCl 6 concentrations of 0.01 M, 0.015 M and 0.02 M, respectively, before annealing, (b, d, f) nanowires, mixture and nanosheets after annealing.

Measured resistances and sensor responses in the presence of 5 ppm of NO 2 for tungsten oxide nanowires and nanosheets as a function of temperature.

Baseline resistance (M)

Equilibrium resistance (M)

resistance (M)

Equilibrium resistance (M)

Sensor response

Fig 4 BET plots of N2 adsorption isotherms for tungsten oxide nanostructures after

annealing at 300◦C for 1 h.

tal grows preferentially along the b-axis, i.e the [0 1 0] direction

The XRD pattern of the nanosheets obtained at WCl6concentration

of 0.02 M corresponds to the monoclinic structure of WO3 with

lattice of a = 7.297 Å, b = 7.539 Å, c = 7.688 Å andˇ = 90.91◦ (JCPDS

No 43-1035), seen inFig 3(e) From this XRD pattern, the two

strongest diffraction peaks appear at 2 = 23.58◦ and 2 = 24.34◦

corresponding to (0 2 0) and (2 0 0) facets and the peak intensity of

the (0 0 2) reflection is much weaker, which implies the nanosheets

grow along the [0 1 0] and [1 0 0] crystallographic direction and

is enclosed by ±(0 0 1) facets The sample synthesized at WCl6

concentrations of 0.015 M is a mixture of monoclinic W18O49and

monoclinic WO3according to the XRD analysis inFig 3(c), which

is in accordance with the SEM characterization result shown in

Fig 2(c).Fig 3shows the comparison between the XRD patterns

of the samples before and after thermal annealing It is obvious

that the crystal structures of the W18O49 nanowires and WO3

nanosheets remained unchanged by the annealing treatment at

300◦C However, the much sharper peaks observed in the XRD

pat-terns of the annealed tungsten oxides indicated an increase degree

of crystallinity

3.2 Physical adsorption–desorption measurements

To examine the porous structure of the W18O49nanowire

bun-dles and WO3nanosheets, the specific surface area and pore size

distribution of the samples annealed at 300◦C for 1 h are

deter-mined by the physical adsorption–desorption measurements in N2

gas BET plots of N2gas adsorption isotherms for tungsten oxide

nanostructures are shown inFig 4 Here, W is the weight of N2

gas adsorbed at a relative pressure P/P0 P/P0is the pressure of N2

gas divided by its saturation vapor pressure It can be seen that

the data points are on a straight line for every sample,

suggest-Fig 5 Pore size distributions per unit mass of tungsten oxide nanowires and

nanosheets after annealing at 300◦C for 1 h.

ing that the specific surface area determined by BET analysis is reliable

The pore size distributions per unit mass of the annealed sam-ples determined by adsorption–desorption isotherms of N2 gas using the BJH model are represented inFig 5 Y-axis represents the pore surface area when the pore size is in a certain range The pore surface area and the pore size distribution above 8 nm appear sim-ilar between W18O49nanowires and WO3nanosheets But when the pore size is less than 8 nm, some discrepancy occurs W18O49 nanowires show larger pore surface areas than WO3 nanosheets Meanwhile, it can be seen that for W18O49nanowires, the pore size distribution has a peak at about 2.1 nm, while the pore size distri-bution peak shifts to a larger size of 2.3 nm for WO3nanosheets The measured specific surface area and pore volume of the annealed bundled W18O49 nanowires are 72.03 m2/g and 0.13 cc/g, whilst for the WO3 nanosheets they are 41.85 m2/g and 0.12 cc/g, respectively These values are slight lower than those obtained from the samples before annealing For the as-synthesized W18O49nanowires and WO3nanosheets, the specific surface area and pore volume are respectively 89.89 m2/g/0.14 cc/g and 46.67 m2/g/0.12 cc/g The relative higher specific surface area

of W18O49nanowires than that of WO3nanosheets is related to the bundles feature The individual nanowires comprising the bundles were observed in thin and long one-dimensional nanostructured materials from the SEM and TEM images shown inFig 2(a and b) It’s obvious that the high specific surface area of the bundled nanowires is in part ascribed to a combination of the ultra-thin feature of individual nanowires and the unique packing character-istic of the bundles themselves[24] Also, it is associated with the sizes and distributions of the pores[25], which can be proved from Fig 5 The formation of nanosheets structure will consequently lead

to the decrease of pore surface area, which in turn resulted in the decreased pore volume and specific surface areas

Y Qin et al / Sensors and Actuators B xxx (2010) xxx–xxx 5 3.3 NO2-sensing properties

The gas sensing properties of the sensors based on tungsten

oxide nanostructures towards 5 ppm NO2were tested at operating

temperatures ranging from 100 to 250◦C.Table 1shows the

mea-sured baseline resistances, equilibrium resistances and calculated

sensor responses for the W18O49nanowires and WO3nanosheets,

respectively, as a function of operating temperature It should be

noted that three continual tests were preformed to the same

sen-sor sample at every operating temperature in order to ensure the

reliability of the testing data The baseline resistances and the

equi-librium resistances in three tests were found to be similar, and

every data shown inTable 1 is the average value of three data

obtained from three tests respectively The standard deviations for

each value were also shown in the table FromTable 1, it can be seen

that the baseline resistances of tungsten oxide decrease along with

increasing temperature, which is consistent with typical

semicon-ductor materials It is well known that the response of the sensor

is much dependent on the operating temperatures Such relation

is illustrated inTable 1 The response tests’ results show

maxi-mum values of 151.2 and 107.3 at 150◦C for the W18O49nanowires

and WO3nanosheets, respectively When operating temperature

is above 200◦C, the nanowires and nanosheets lose their response

ability quickly Especially, for W18O49 nanowires, the response

value at 250◦C only was 1.2, which is less than 1% percent of that

at 150◦C The above results can be understood as following: As

been reported, tungsten oxide is a typical n-type semiconductor,

and its gas sensing mechanism belongs to the surface-controlled

type, and the change of conductivity is dependent on the species

and the amount of chemisorbed oxygen on the surface[26] At low

temperature, oxygen species on the film surface are not active, so

that a low interaction happens between adsorbed oxygen species

and detected NO2 gas Thus, the response of the tungsten oxide

film is low Conversely, some of the adsorbed oxygen species may

be desorbed from the film surface at high temperature, which also

leads to low response value As a result, there should be an optimal

operating temperature to balance the above two effects in order

to achieve the maximum gas response It is clear fromTable 1that

the measurement carried out at temperatures ranging from 150

to 200◦C can obtain relatively high NO2response, and the highest

response is achieved at 150◦C FromTable 1, W18O49nanowires

exhibited much higher NO2response than WO3nanosheets at

dif-ferent operating temperature ranging from 100 to 200◦C While it

is noteworthy that, at 250◦C, the NO2response of WO3nanosheets

sensor is almost 10 times higher than that of W18O49nanowires

sensor

The response and recovery time of the W18O49nanowires and

WO3nanosheets to 5 ppm NO2at various operating temperatures

are shown inFig 6 From this figure, the response time and recovery

time of the two samples are both decrease quickly with increasing

operating temperature When the operating temperature rise to

200◦C or above, both samples show the much fast response and

recovery characteristics, despite of their low response shown in

Table 1 It is also clear from this figure that the W18O49nanowires

show faster response–recovery than the WO3nanosheets at various

operating temperatures

Fig 7shows the dynamic responses of tungsten oxide

nanos-tructures to NO2 gas in varying concentration The operating

temperature is 200◦C Fig 7(a and b) shows the results of the

W18O49nanowires and WO3nanosheets synthesized at WCl6

con-centrations of 0.01 M and 0.02 M, respectively As shown in this

figure, the measured resistances increased upon exposure to NO2

gas This result is expected because the oxidizing analyte NO2

withdraws electrons from the n-type tungsten oxide surface and

induces the formation of electron-depleted space-charge layers

[27] Notably, the resistances could almost recover to its initial

Fig 6 Response time and recovery time of tungsten oxide nanowires and

nanosheets to 5 ppm NO 2 as a function of operating temperature.

value after NO2removal, indicating a good reversibility of these nanostructure materials From this figure, it also can be seen that the increase in the resistance of the W18O49nanowires upon expo-sure to NO2is much larger than that of the WO3nanosheets, which indicates that the W18O49nanowires have higher NO2response The response values of the W18O49nanowires upon exposure to 1,

5, 10 and 20 ppm NO2are 13.4, 123.6, 203.4 and 332.3, while those

of the WO3nanosheets are 13.3, 73.5, 144.2 and 279.8, respectively

As has been reported, the gas sensing mechanism of tung-sten oxide belongs to the surface-controlled type in which the surface states and oxygen adsorption play an important role [26,28] Atmospheric oxygen adsorbs electrons from the conduc-tion band of the sensing metal oxide and occurs on the surface

Fig 7 Dynamic response of (a) bundled W18O 49 nanowires, (b) WO 3 nanosheets to varying NO 2 concentration at an operating temperature of 200◦C.

Fig 8 Response time and recovery time curves of W18O 49 nanowires and WO 3

nanosheets to different NO 2 concentration at an operating temperature of 200 ◦ C.

in the form of O−, O2− and O2 − The reaction between NO2 and

the surface adsorbed species (O2−, O− and O2 −, etc.) induces

the formation of electron-depleted space-charge layers inside the

tungsten oxide surfaces and thus the increase in the resistance

[29] According to the BET measurements, the W18O49nanowires

showed much larger specific surface area (72.03 m2/g) than the

WO3nanosheets (41.85 m2/g) The larger surface area can provide

more adsorption–desorption sites and a larger amount of surface

adsorbed oxygen species interacting with detected gas molecules

Thus, W18O49nanowires with higher specific surface area show

much larger change in resistance upon exposure to NO2than the

WO3 nanosheets with lower specific surface area Besides,

non-stoichiometric crystal structure of W18O49 is another important

factor for high response[30] There exist much more oxygen

vacan-cies in the crystal structure of non-stoichiometric W18O49than fully

oxidized WO3[31], and the large amounts of oxygen vacancies can

serve as adsorption sites of gas molecular and effect on the

elec-tron density in oxide, which is beneficial to achieving much higher

gas response[32,33] Above analysis can explain why the W18O49

nanowires exhibit higher response than the WO3nanosheets

How-ever, as shown inTable 1, the NO2 response of WO3nanosheets

is higher than that of W18O49 nanowires for almost 10 times at

the operating temperature of 250◦C This result implies that some

other factor dominates the gas sensing performance of the one- and

two-dimension tungsten oxide nanostructure It is found fromFig 2

that thermal treatment resulted in a more evident change in the

microstructure of W18O49nanowires than that of WO3nanosheets,

indicating that the microstructure of W18O49nanowires is much

more sensitive to temperature than WO3nanosheets Therefore, it

is possible that the microstructure change of W18O49nanowires

(e.g further agglomeration) at high operating temperature affects

the gas diffusion and then induces the much low response

Accord-ing to this analysis, WO3 nanosheets might show much better

stability in the gas response due to much better thermal stability

in the microstructure in comparison with W18O49nanowires when

operating at high temperature

Fig 8 shows the response and recovery time curves of the

W18O49 nanowires and WO3 nanosheets to different

concentra-tion of NO2at 200◦C It can be seen that, for the tungsten oxide

nanowires and nanosheets, the response times decrease whereas

the recovery times increase with rising concentration of NO2

Worth of mention is the faster response–recovery of the W18O49

nanowires compared to that of the WO3 nanosheets when both

expose to the same NO2 concentration Especially, the W18O49

nanowires exhibit very fast response characteristic to NO2gas with

the response times of 42, 19, 16 and 14 s to 1, 5, 10 and 20 ppm NO2,

respectively

One- and two-dimensional tungsten oxide nanostructures were synthesized at 200◦C by solvothermal method with tungsten hex-achloride (WCl6) as precursor and 1-propanol as solvent The synthesis processes were preformed at different WCl6 concen-trations (0.01, 0.015 and 0.02 M, respectively) One-dimensional

W18O49 nanowire bundles were obtained at a WCl6 concen-tration of 0.01 M, while the structure of pure two-dimensional

WO3nanosheets was formed at concentration of 0.02 M Thermal annealing at 300◦C could not change the crystal structure of the nanowires and nanosheets BET measurements showed the specific surface areas and pore volumes were 72.03 m2/g and 0.13 cc/g for annealed W18O49nanowire bundles and 41.85 m2/g and 0.12 cc/g for annealed WO3nanosheets, respectively The gas sensing prop-erties measurements indicated that both W18O49nanowires and

WO3nanosheets exhibit reversible response to different concen-trations of NO2 Compared to WO3nanosheets, W18O49nanowires showed quicker response–recovery and higher response value to different concentration of NO2 gas due to the high specific sur-face area and the non-stoichiometric crystal structure, and their response time, recovery time and response value are 19 s, 112 s and 123.6 to 5 ppm NO2at 200◦C, respectively These results indicate the one-dimensional W18O49nanowire is a promising gas sensing material for high performance NO2gas sensor

Acknowledgments

This work was financially supported by the National Natural Science Foundation (No 60801018, 60771019), Tianjin Natural Science Foundation (No 09JCYBJC01100) and the New Teacher Foundation of Ministry of Education (No 200800561109) of China

References

[1] G Eranna, B.C Joshi, D.P Runthala, R.P Gupta, Oxide materials for development

of integrated gas sensors-a comprehensive review, Crit Rev Solid State Mater Sci 29 (2004) 111–188.

[2] A Ruiz, A Cornet, G Sakai, K Shimanoe, J.R Morante, N Yamazoe, Prepara-tion of Cr-doped TiO 2 thin film of P-typeconduction for gas sensor application, Chem Lett 31 (2002) 892–893.

[3] C Zamani, K Shimanoe, N Yamazoe, A new capacitive-type gas sensors com-bining an MIS with a solid electrolyte, Sens Actuators B 109 (2005) 216–220 [4] M.S Nieuwenhuizen, A.J Nederlof, Preliminary results with a silicon-based surface acoustic wave chemical sensor for NO 2 , Sens Actuators B 19 (1989) 385–392.

[5] G Sberveglieri, P Benussi, G Coccoli, S Groppelli, P Nelli, Reactively sputtered indium tin oxide polycrystalline thin films as NO and NO 2 gas sensors, Thin Solid Films 186 (1990) 349–360.

[6] M Akiyama, J Tamaki, N Miura, N Yamazoe, Tungsten oxide-based semicon-ductor sensor highly sensitive to NO and NO 2 , Chem Lett (1991) 1611–1614 [7] Y.K Chung, M.H Kim, W.S Um, H.S Lee, J.K Song, S.Ch Choi, K.M Yi, M.J Lee, K.W Chung, Gas sensing properties of WO 3 thick film for NO 2 gas dependent

on process condition, Sens Actuators B 60 (1999) 49–55.

[8] Z.W Pan, Z.R Dai, Z.L Wang, Nanobelts of semiconducting oxides, Science 291 (2001) 1947–1949.

[9] Y Cui, C.M Lieber, Functional nanoscale electronic devices assembled using silicon nanowire building blocks, Science 291 (2001) 851–853.

[10] A Kolmakov, D.O Klenov, Y Lilach, S Stemmer, M Moskovits, Enhanced gas sensing by individual SnO 2 nanowires and nanobelts functionalized with Pd catalyst particles, Nano Lett 5 (2005) 667–673.

[11] Y.B Shen, T Yamazaki, Z.F Liu, D Meng, T Kikuta, N Nakatani, M Saito, M Mori, Microstructure, H 2 gas sensing properties of undoped and Pd-doped SnO 2

nanowires, Sens Actuat B 135 (2009) 524–529.

[12] Y.Z Lv, L Guo, H.B Xu, X.F Chu, Gas-sensing properties of well-crystalline ZnO nanorods grown by simple route, Physica E 36 (2007) 102–105.

[13] C.Y Wang, M Ali, T Kups, C.-C Röhlig, V Climalla, T Stauden, O Ambacher, NO x

sensing properties of In 2 O 3 nanoparticles prepared by metal organic chemical vapour deposition, Sens Actuators B 130 (2008) 589–593.

[14] Z.F Liu, T Yamazaki, Y.B Shen, T Kikuta, N Nakatani, Y.X Li, O 2 and CO sensing

of Ga 2 O 3 multiple nanowire gas sensors, Sens Actuators B 129 (2008) 666–670 [15] L.C Tien, P.W Sadik, D.P Norton, L.F Voss, S.J Pearton, H.T Wang, B.S Kang, F Ren, J Jun, J Lin, Hydrogen sensing at room temperature with Pt-coated ZnO thin films and nanorods, Appl Phys Lett 87 (2005) 222106.

Y Qin et al / Sensors and Actuators B xxx (2010) xxx–xxx 7 [16] S.B Sun, Z.D Zou, G.H Min, Synthesis of bundled tungsten oxide nanowires

with controllable morphology, Mater Charact 60 (2009) 437–440.

[17] R Hu, H.S Wu, K.Q Hong, Growth of uniform tungsten oxide nanowires with

small diameter via a two-step heating process, J Cryst Growth 306 (2007)

395–399.

[18] R.F Mo, G.Q Jin, X.Y Guo, Morphology evolution of tungsten trioxide nanorods

prepared by an additive-free hydrothermal route, Mater Lett 61 (2007)

3787–3790.

[19] B Zhang, J.D Liu, S.K Guan, Y.Z Wang, Y.Z Zhang, R.F Chen, Synthesis of

single-crystalline potassium-doped tungsten oxide nanosheets as high-sensitive gas

sensors, J Alloys Compd 439 (2007) 55–58.

[20] C.S Rout, M Hegde, C.N.R Rao, H 2 S sensors based on tungsten oxide

nanos-tructures, Sens Actuators B 128 (2008) 488–493.

[21] R.A Gerlitz, K.D Benkstein, D.L Lahr, J.L Hertz, C.B Montgomery, J.E Bonevich,

S Semancik, M.J Tarlov, Fabrication and gas sensing performance of parallel

assemblies of metal oxide nanotubes supported by porous aluminum oxide

membranes, Sens Actuators B 136 (2009) 257–264.

[22] J Moon, M.L Carasso, H.G Krarup, J.A Kerchner, J.H Adair, Particle-shape

con-trol and formation mechanisms of hydrothermally derived lead titanate, J.

Mater Res 14 (1999) 866–875.

[23] H.G Choi, Y.H Jung, D.K Kim, Solvothermal synthesis of tungsten oxide

nanorod/nanowire/nanosheet, J Am Ceram Soc 88 (2005) 1684–1686.

[24] J Pfeifer, E Badaljan, P TekulaBuxbaum, T Kovacs, O Geszti, A.L Toth, Growth

and morphology of W 18 O 49 crystals produced by microwave decomposition of

ammonium paratungstate, J Cryst Growth 169 (1996) 727–733.

[25] Y.M Zhao, W.B Hu, Y.D Xia, E.F Smith, Y.Q Zhu, C.W Dunnill, Preparation and

characterization of tungsten oxynitride nanowires, J Mater Chem 17 (2007)

4436–4440.

[26] J Mizsei, How can sensitive and selective semiconductor gas sensors be made,

Sens Actuators B 23 (1993) 173–176.

[27] C.O Park, S.A Akbar, Ceramics for chemical sensing, J Mater Sci 38 (2003)

4611–4637.

[28] A Rothschild, Y Komem, The effect of grain size on the sensitivity of nanocrys-talline metal-oxide gas sensors, J Appl Phys 95 (2004) 6374–6380 [29] J Tamaki, Z Zhang, K Fujimori, M Akiyama, T Harada, N Miura, N Yama-zoe, Grain-size effects in tungsten oxide-based sensor for nitrogen oxides, J Electrohem Soc 141 (1994) 2207–2210.

[30] Y.M Zhao, Y.Q Zhu, Room temperature ammonia sensing properties of W 18 O 49

nanowires, Sens Actuators B 137 (2009) 27–31.

[31] C.C Liao, F.R Chen, J.J Kai, Annealing effect on electrochromic properties of tungsten oxide nanowires, Sol Energy Mater Sol Cells 91 (2007) 1258–1266 [32] W Göpel, K.D Schierbaum, SnO 2 sensors: current status and future prospects, Sens Actuators B 26–27 (1995) 1–12.

[33] I Jiménez, M.A Centeno, R Scotti, F Morazzoni, A Cornet, J.R Morante, NH 3

interaction with catalytically modified nano-WO 3 powders for gas sensing applications, J Electrochem Soc 150 (2003) 72–80.

Biographies

Yuxiang Qin received a Ph.D in microelectronics and solid-state electronics from

Tianjin University in 2007 She is currently an associate professor in Department of Electronics Science and Technology in Tianjin University Her research interest is in the areas of oxide semiconductor gas sensor, field emission materials and devices.

Ming Hu received a M.S in microelectronics and solid-state electronics from Tianjin

University in 1991 She is now a professor in Department of Electronics Science and Technology in Tianjin University Her research interests include MEMS, gas sensor, functional film devices.

Jie Zhang received her Bachelor degree in microelectronics and solid-state

elec-tronics from Tianjin University in 2008 She is now a graduate student at Tianjin University Her current research is focused on the tungsten oxide based gas sensor and material adsorption properties simulation.