nano - sized pdo loaded sno2 nanoparticles by reverse micelle method

That is, the sensor response declined gradually with increasing the particle size of PdO although the maximum of the sensor response was obtained in PdO = 0.1 mol%.. In this study, we fo

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

for highly sensitive CO gas sensor

Masayoshi Yuasaa,∗, Takanori Masakib, Tetsuya Kidaa, Kengo Shimanoea, Noboru Yamazoea

aDepartment of Energy and Material Sciences, Faculty of Engineering Sciences, Kyushu University, 6-1 Kasuga-Koen, Kasuga, 816-8580 Fukuoka, Japan

bDepartment of Molecular and Material Sciences, Interdisciplinary Graduate School of Engineering Sciences, Kyushu University, Kasuga, Fukuoka, Japan

a r t i c l e i n f o

Article history:

Received 12 June 2008

Received in revised form 22 October 2008

Accepted 14 November 2008

Available online 30 November 2008

Keywords:

CO sensor

Nanoparticle

Reverse micelle method

a b s t r a c t

A reverse micelle method was investigated for preparing nano-sized PdO loaded on SnO2nanoparticles PdO–SnO2nano-composite was prepared by precipitating Pd(OH)2and Sn(OH)4inside a reverse micelle The microstructure and the gas sensing properties of obtained nanoparticles were investigated Although the particle size of SnO2was as same as ca 10 nm at each observed sample, the particle size of PdO got larger as increasing with loading amount of PdO because of agglomeration of PdO nanoparticles each other As a result of the gas sensing measurement, it was found that the particle size of PdO on SnO2nanoparticle influences the gas sensing property closely That is, the sensor response declined gradually with increasing the particle size of PdO although the maximum of the sensor response was obtained in PdO = 0.1 mol% In this method, small amount of PdO loading can be achieved as compared with PdO-loaded SnO2sensor prepared by the conventional impregnation method

© 2008 Elsevier B.V All rights reserved

1 Introduction

For semiconductor gas sensors, tin oxide (SnO2) has been one of

the attractive materials because of its high sensitivity and chemical

stability During the past decades, the physical and chemical

proper-ties of SnO2have been well studied with the aim of improving the

performance of SnO2-based gas sensors[1–6] More importantly,

recent extensive studies have found three basic factors concerning

the sensing properties of semiconductor gas sensors In particular,

the followings are proposed as the most influential factors: (1) grain

size of particles[7], (2) microstructure of the sensing body[8], and

(3) surface modification of particles (noble metal loading)[9–14]

For the grain size effect (first factor), Xu et al have reported that

the sensor response increased drastically as the grain size decreased

to less than 6 nm, which value is twice as large as the thickness of

depletion layers in SnO2[7] On the basis of these findings, we

pre-pared almost mono-dispersed SnO2nanoparticles (mean diameter:

4 nm) suspended in an aqueous solution by hydrothermal

treat-ment of tin hydroxide gel, and succeeded in achieving a significant

increase in the sensor response to H2gas[15] In addition, we

con-tinued to investigate the nature of the grain size effect and reported

that small crystals can be depleted of conduction electrons beyond

the scheme of convention depletion theories[16] On the other

hand, for the second factor, it has been proved experimentally and

theoretically that the sensor performances such as sensitivity and

∗ Corresponding author Tel.: +81 92 583 7539; fax: +81 92 583 7538.

E-mail address:yuasa@mm.kyushu-u.ac.jp (M Yuasa).

response speed depend largely on the rates of diffusion of a tar-get gas and its surface reaction (with oxygen adsorbed on SnO2) [17–22] This clearly demonstrates the importance of microstruc-ture control of sensing layers, and indicates that sensing layers with porous structures allow detection of larger sized gases by facilitat-ing their diffusion deep inside the sensfacilitat-ing layer Indeed, we have achieved a higher response to CO larger than H2by controlling the microstructure of the sensing body[8] For the surface modifica-tion effect (third factor), it is now well accepted that loading of small amounts of noble metals, such as Pd and Pt on the SnO2, promotes gas response as well as the rate of response In partic-ular, Pd has frequently been loaded on commercial SnO2-based gas sensors In this case, the sensitization originated in electronic inter-action between Pd (actually PdO) and SnO2, as follows The loading

of PdO on SnO2increases the electric resistance often by about one order of magnitude, because PdO acts as a strong acceptor of elec-trons and remove elecelec-trons from the oxide On the other hand, the resistance, when PdO is reduced to Pd on contact with the reducing gases, decreases by back electron transfer from Pd to SnO2 The dif-ference in the electric resistance of SnO2induced by a change in the oxidized and reduced states of Pd is often large, giving rise to a large increase in response to the reducing gases Matsushima et al suc-ceeded in loading fine PdO particles of 3–20 nm on SnO2particles (mean diameter: 40 nm) by several routes including impregnation, colloid adsorption, and chemical fixation methods, and observed more than ten times higher H2gas response by the loading[23,24] Hence, according to the above three factors concerning the sen-sor performance, it can be proposed that the loading of Pd onto porous sensing layers composed of SnO2nanoparticles would offer 0925-4005/$ – see front matter © 2008 Elsevier B.V All rights reserved.

a way to further improve the sensor response However, the loading

of fine Pd nanoparticles onto SnO2nanoparticles is not still

chal-lenged in contrast to the case where larger SnO2 particles were

used as the matrices[23,24] Thus, development of a new loading

method of PdO onto SnO2nanoparticles is required to selectively

bind PdO nanoparticles onto SnO2 nanoparticles without

coagu-lation In this study, we focused on a reverse micelle method as

a new route for loading PdO nanoparticles onto SnO2

nanoparti-cles It is well known that when small amounts of aqueous solution

and surfactant are mixed together with organic solvent in a

des-ignated ratio, nano-sized water droplets called as reverse micelles

stabilized by the surfactant were formed in the organic solvent

This method utilizes the reverse micelles (abbreviated as “RM”) as

nano-spaces for chemical reaction Preparation of nano-sized

par-ticles by RM methods has been reported for various materials, e.g

metals[25–27], metal oxides[28–32], sulfides[33], supported

cat-alysts[34–36]and nanocomposites[37] Recently, we adopted a

RM method for preparing carbon-supported LaMnO3

nanoparti-cles as an electrocatalyst for an oxygen reduction electrode[38,39]

The promising feature of the RM methods is that nanoparticles

can be prepared in nano-sized water droplets as a reaction place

Furthermore, other nanoparticles can be deposited subsequently

on the nanoparticles formed in nano-sized water droplets, would

result in homogeneous deposition of Pd nanoparticles onto SnO2

nanoparticles In this study, PdO-loaded SnO2nanoparticles were

prepared by precipitating Pd(OH)2on Sn(OH)4of nano-sized

pre-cursor with hydroxides inside RMs, followed by calcination The

obtained nanocomposites (PdO-loaded SnO2) were used to form

thick film type sensor devices, and the relationship between the gas

sensing properties and nanostructure of PdO-loaded SnO2particles

are discussed

2 Experimental

2.1 Material preparation

The preparation procedure of PdO-loaded SnO2 nanoparticles

by using RM method is schematically shown inFig 1 Three

dif-ferent RM solutions were prepared in this preparation process The

molar ratio of water to surfactant was in the range of 3–12 The

mix-ing ratio of surfactant to organic solution was 4:6 in weight ratio

for all RM solutions The amount of aqueous solutions used was

constant at 10 mL At first, a RM solution containing [Sn(OH)4]2 −

(RM-A) was prepared by mixing cyclohexane (C6H12), non-ionic

surfactant (NP-6, polyoxyethylene (6) nonylphenyl ether), and an

aqueous solution containing [Sn(OH)4]2 −(0.1 M) Mixing was

per-formed at 10◦C until the solution became colorless The solution

containing [Sn(OH)6]2−was prepared by dissolving Sn(CH3COO)4

in a tetramethylammonium hydroxide solution (10%) The pH of the

aqueous solution was around 13, and as such [Sn(OH)6]2 −was likely

formed at this high pH To precipitate precursor Sn(OH)2, the

solu-tion RM-A was mixed together with a RM solusolu-tion containing an

aqueous HNO3solution (6%, pH 2) (RM-B) The mixing results in a

decrease in the pH of the aqueous phase by collision between

dif-ferent reverse micelles, precipitating Sn(OH)4nanoparticles inside

the reverse micelles collided The final pH of the aqueous phase in

the mixed solution was around 9 Then, the mixed solution was

fur-ther mixed with the RM solution containing an aqueous Pd(NO3)2

solution (0.1–5.0 mM, pH 4) (RM-C) to precipitate Pd(OH)2particles

onto Sn(OH)4particles inside RMs The loading amount of Pd was

controlled between 0.05 and 5.0 mol% by changing the

concentra-tion of Pd ions in the soluconcentra-tion RM-C By adding ethanol to break RMs

containing Pd(OH)2–Sn(OH)4, the resulting precipitates were

col-lected by centrifugation and they were washed with ethanol After

drying at 120◦C, the obtained powder was calcined at 600◦C for 3 h

to form PdO–SnO nanocomposites

Fig 1 Schematic diagram of the preparation method of PdO-loaded SnO2 nanopar-ticles by a reverse micelle method.

2.2 Material characterization

The diameter of RMs in solutions containing Sn(OH)4or Pd(OH)2 particles was measured by a dynamic light scattering analyzer (DLS) (ELS6000/8000, Otsuka electronics Co., Ltd.) The morphology of the composites was observed by TEM (JEM-2000EX, JEOL Co., Ltd., Japan) Qualitative and quantitative analyses of PdO in the obtained samples were performed by a wavelength dispersion-type X-ray fluorescence spectrometer with LiF analyzing crystals and Pd K␣ X-ray source (ZSX-mini, Denki Co Ltd., Japan) The crystalline size of the samples was calculated by Scherrer’s formula from their XRD patterns measured by an X-ray diffractmeter with nickel-filtered

Cu K␣ (1.5418 Å) source (RINT2100, Rigaku Denki Co., Ltd., Japan)

2.3 Sensor fabrication and measurement

Sensor devices were fabricated by a screen-printing method The obtained PdO (0.05–5 mol%)–SnO2powders were mixed mechani-cally with diethanolamine as a binder to form pasts for printing The PdO–SnO2powders were pasted on alumina substrates attached with a pair of comb-type Au electrodes (at a space of 90␮m between the electrodes) through patterned-screens to fabricate sensor devices Then, the devices were heat-treated at 600◦C for

3 h in air to burn the organic binder The sensor device thus fabri-cated was settled in a quartz tube and heated by an electric furnace for sensing property measurements The sensor device was con-nected with a standard resistor in series, and the voltage across

the standard resistor was measured under an applied voltage of dc

4 V to evaluate the electrical resistance of the device The

electri-cal signal of the sensor devices was acquired with an electrometer

The electric resistances of the devices in air and in air containing

target gas (200 ppm CO) were measured at 300◦C which was the

most suitable operation temperature As a target gas, we choose CO

which don’t generate a by-product in order to evaluate only loading

effect Sensor response (S = Rair/Rgas) was defined as the ratio of the

electric resistance in air (Rair) to in target gas (Rgas)

3 Results and discussion

3.1 Characterization of PdO–SnO 2 nanocomposites

For reverse micelle formation, the molar ratio of water to

surfac-tant called as Rwvalue (Rw= [H2O]/[surfactant]) is a critical factor;

the size of water droplets substantially depends on this value

The stability of reverse micelles is also affected by the Rwvalue

Hence, to prepare precursor hydroxide particles with desired sizes,

the effects of the Rw on the diameter of reverse micelles were

first examined.Fig 2shows the dependences of the mean

diame-ters of reverse micelles containing precursor Pd(OH)2and Sn(OH)4

nanoparticles The reverse micelles containing Pd(OH)2were

pre-pared by mixing the solution RM-C with a reverse micelle solution

containing a tetramethylammonium hydroxide solution (10%) as

the precipitating agent The diameters of the two different reverse

micelles increased monotonically with increasing their Rwvalues

This tendency can be explained as follows: for reverse micelles to

form, the head of the hydrophilic group of surfactant molecules

(here, –(CH2CH2O)–) has to adsorb on the surface of nano-sized

water droplets in an organic solvent[40,41] When the amount of

surfactant molecules is decreased, the small water droplets seem to

cohere for reducing the interface free energy between water droplet

and organic solvent Accordingly, the diameter of the water droplets

inside reverse micelles tends to increase with increasing the Rw

value For the revere micelles containing Sn(OH)4, they were

sta-ble when the Rwvalue was 6–12 In this case, the diameter of the

reverse micelles was in the range of 4–13 nm On the other hand,

for Pd(OH)2, stable reverse micelle solutions were obtained when

the Rw value is around 3–9 This difference is likely due to the

difference in the pH of the aqueous phase in the two RMs

Con-sidering the size and the stability of reverse micelles, the Rwvalue

of 9 was selected as appropriate for the preparation of precursor

Sn(OH)4–Pd(OH)2 composites.Fig 3shows the particle size

dis-tribution of reverse micelles (Rw= 9) in the RM solution containing

Pd(OH)2(1.0 mol%)–Sn(OH)4 The size distribution was very narrow

Fig 2 Dependence of the diameter of reverse micelles containing Pd(OH)2 (open

circle) or Sn(OH) particles (closed circle) on the molar ratio of water to surfactant.

Fig 3 The particle size distribution of reverse micelles containing Pd(OH)2 –Sn(OH) 4

(1.0 mol%).

and no agglomeration was observed The above results suggest that Pd(OH)2–Sn(OH)4 nanocomposites of 7–12 nm in diameter was successfully obtained in nanosized water droplets inside reverse micelles

The PdO-loaded SnO2 nanoparticles obtained by calcination

of the above composite powder were characterized.Fig 4shows the XRD pattern of PdO (1.0 mol%)–SnO2 calcined at 600◦C In the pattern, only peaks ascribable to SnO2 (tetragonal structure,

a = b = 4.7382 Å, c = 3.1871 Å, JCPDS 41-1445) were seen,

suggest-ing the successful conversion of precursor Sn(OH)4 to SnO2 No peaks of PdO were observed because of its small loading amount (1.0 mol%) The crystalline size of SnO2 calculated with Scherer’s formula using the XRD peaks was 13.5 nm This is in nearly good agreement of the size of the precursor composite as shown inFig 2 The results suggest that no significant crystal growth occurred in the composite The qualitative and quantitative analyses of PdO in the composite were performed by X-ray fluorescence (XRF) analy-sis.Fig 5shows a representative XRF spectrum (Pd K␣) of the PdO (1 mol%)–SnO2nanocomposite, indicating the presence of Pd in the sample The ratio of Pd to Sn was also determined by the calibration curve obtained with reference samples For the 1.0 mol% Pd-loaded sample, it was confirmed that the determined loading amount was within 1.0% deviation from the nominal amount Thus, it is sug-gested that Sn and Pd ions were almost completely precipitated from the precursor solutions in the present method, although the small amount of Pd below 1 mol% loading could not be precisely quantified because of difficulty in separating noise from signal of XRF

Fig 4 XRD pattern of PdO (1.0 mol%)-loaded SnO2 nanoparticles prepared by the

Fig 5 Representative XRF spectrum (Pd K␣) of PdO (1.0 mol%)-loaded SnO 2

nanoparticles.

The morphology and the particle size of SnO2 and PdO were

observed by TEM.Fig 6shows TEM and high-resolution (HR)-TEM

images of PdO–SnO2nanocomposites with different PdO loading

amounts The obtained TEM images show that the particle size of

SnO2in all samples was as same as ca 10 nm, in good agreement

with the XRD results Thus, the observed particles are judged to be of

single crystalline without significant sintering even after high

tem-perature calcination This suggests the effectiveness of the present

method for preparing thermally stable SnO2nanoparticles On the

other hand, the particle size of PdO was different, depending on its

loading amount To differentiate between SnO2and PdO particles,

lattice images were taken by HR-TEM For 0.5 mol% PdO loading, the

particle size of PdO was observed to be less than 5 nm With

increas-ing the loadincreas-ing amount, PdO particles tended to be agglomerated

each other and grew up larger For smaller 0.1 mol% PdO loading,

no PdO particles with clear lattice images were observed However,

this is supposed to be owing to smaller particle size of PdO, probably

Fig 7 The dependence of the electric resistance in air at 300◦C on the loading amount of PdO for the devices prepared by the reverse micelle (closed circle) and impregnation methods (closed square).

less than 1 or 2 nm In addition, from the results of electric resis-tance in air, as shown later, it is understood that smaller particles

of PdO are loaded on nano-sized SnO2 This is in marked contrast to the reported case where PdO particles with a wider size distribu-tion (3–20 nm) by an impregnadistribu-tion method were loaded on larger SnO2particles (ca 50–100 nm)[42]

3.2 Gas sensing properties of PdO–SnO 2 nanocomposite films

Fig 7shows the dependence of the electric resistance of sensor films in air at 300◦C on the loading amount of PdO For compar-ison, the electric resistances of the sensor films prepared by the conventional impregnation method[42]were also shown in this figure

In the conventional impregnation method, stannic acid precipi-tated from an aqueous solution of SnCl4with ammonia solution was

Fig 8 The dependence of the sensor response to 200 ppm CO at 300◦C on the

loading amount of PdO for the devices prepared by the reverse micelle method.

calcined at 900◦C for 5 h to obtain the SnO2powder Then, PdCl2

solution was impregnated to the above SnO2powder, and then the

solution evaporated to dryness and reduced in a flow of H2gas for

3 h The particle size of SnO2and Pd in the conventional method

was 50–100 nm and 3–10 nm, respectively The electric resistance

of SnO2nanoparticles increased with loading PdO, and reached the

maximum at 0.1 mol% loading The increase at 0.1 mol% was more

than one order of magnitude The increase observed in the

elec-tric resistance is similar to the trend reported in the literature[42],

and can be interpreted in terms of the electric interaction between

PdO and SnO2, for which PdO attracts electrons from SnO2and

pro-duces electron depleted layers on the SnO2surface The obtained

results thus confirm that PdO nanoparticles were effectively loaded

on SnO2 nanoparticles as observed in TEM images On the other

hand, loading more than 0.1 mol% decreased the resistance The

further loading of PdO may lead to the agglomeration of PdO

par-ticles and impede the formation of effective contacts between PdO

and SnO2 Note that the observed dependence of the resistance on

amount of PdO loading is somewhat different from the reported

dependences for samples prepared through impregnation, colloid

adsorption, and chemical fixation methods[42] This means that

the electrical resistance depends on preparation methods, namely,

the dispersion state of PdO over the SnO2surface In addition, for

the reverse micelle method, the loading amount at the maximum

resistance was 15 times lower than those for the above methods

It is considered that finer dispersion of PdO was attained by the

present method, reducing the optimum PdO loading amount for

maximizing the depletion effects

Fig 8shows the dependence of the sensor response to 200 ppm

CO at 300◦C on amount of PdO loading The maximum sensor

response was obtained at 0.1 mol% PdO loading, reaching a high

value of S = 320 On the other hand, the sensor response was

decreased with further increasing the loading amount Such a trend

is in good accordance with the dependence of the electric resistance

as shown above This good consistency between the resistance

and the sensor response indicates that the electrical interaction

between PdO and SnO2 is dominant for the improvement of the

sensor response rather than the catalytic effect of PdO that assists

the combustion of CO with adsorbed oxygen

As revealed in this study, the developed method can improve

the sensor response even by a smaller loading amount of PdO, as

compared with the other reported methods The reduction of the

loading amount is the favorable feature of the present method

Note that the size of PdO was decreased by reducing the loading

amount, as deduced by the TEM observations Moreover, based on the obtained results, it can be suggested that the sensor response is associated with the number of contacts between PdO and SnO2 par-ticles It is speculated that the number of the contacts was increased

by the size reduction of PdO To examine the possibility of this effect, the number of PdO loaded on SnO2 was roughly estimated using the representative sizes of PdO observed in the HR-TEM images

The number of PdO particles per mass (N) for each sample can be

calculated using the following equation under the assumption that their sizes are constant for each sample:

N =the total volume of PdO per mass

For the estimation of N, the total volume of PdO per mass in the 0.5 mol%-loaded sample is abbreviated as V Likewise, those val-ues for 1.5 and 5.0 mol%-loaded samples are expressed as 3V and 10V, respectively The representative particle sizes of PdO in the 0.5,

1.5 and 5.0 mol% PdO-loaded samples was approximately 4,6 and

10 nm, respectively Thus, when the particle diameter for 0.5 mol%

is abbreviated as D, then those for 1.5 and 5.0 mol% can be expressed

as 1.5D and 2.5D, respectively By using these values for Eq.(1), the number of PdO particles for each sample can be calculated as follows:

(4/3)(D/2)3 = 6 × V

D3



(2)

(4/3)(1.5D/2)3 = 3.96 V

D3



(3)

(4/3)(2.5D/2)3 = 2.88 V

D3



(4)

The above simple calculation results indicate that the number of PdO particles tends to increase with decreasing the loading amount Such an increase in the number of PdO particles is readily expected

to cause an increase in the density of contacts between PdO and SnO2particles, provided that the size of SnO2is constant Conse-quently, the surface depletion effect, induced by the formation of PdO–SnO2junctions, is enhanced This significantly increases the electric resistance as well as the sensor response even by the quite low PdO loading

4 Conclusions

PdO-loaded SnO2nanoparticles were prepared by the reverse micelle method Stable and mono-disperse reverse micelles of ca

10 nm containing both Pd(OH)2 and Sn(OH)4 were obtained at

Rw= 9 The calcination of the collected hydroxide composites at

600◦C produced PdO-loaded SnO2nanoparticles; the particle size

of SnO2 was ca 10 nm irrespective of the PdO loading amount Nano-sized PdO particles of ca 4 nm were prepared at 0.5 mol% loading However, with increasing the PdO loading amount, PdO particles agglomerated each other and grew up larger It was found that both of the electric resistance and the sensor response of PdO-loaded SnO2were dependent on the loading amount The max-imum electric resistance and sensor response were obtained at 0.1 mol% PdO loading The optimum amount of PdO for maximiz-ing the sensor response was fairly smaller for the reverse micelle method, as compared with those for conventional methods This is suggested to be owing to that the number of PdO particles in con-tact with SnO2increased by reducing the size of the PdO particles with the developed preparation method The obtained results con-firmed that the electrical resistance as well as the sensor response

is significantly dependent on dispersion states of PdO particles on SnO

This work has been financially supported by a Grant-in-Aid for

Scientific Research (B) (No 18350075) from the Ministry of

Educa-tion, Science, Sports and Culture of Japan

References

[1] N Yamazoe, J Fuchigami, M Kishikawa, T Seiyama, Interaction of tin oxide

surface with O 2 , H 2 O and H 2 , Surf Sci 86 (1979) 335–344.

[2] J Tamaki, M Nagaishi, Y Teraoka, N Miura, N Yamazoe, Adsorption behavior

of CO and interfering gases on SnO 2 , Surf Sci 221 (1989) 183–196.

[3] V Lantto, P Romppainen, S Leppavuori, A study of the temperature dependence

of the barrier energy in porous tin dioxide, Sens Actuators 14 (1988) 149–163.

[4] S Roy Morrison, Selectivity in semiconductor gas sensors, Sens Actuators 12

(1987) 425–440.

[5] K.D Schierbaum, U Weimar, W Gopel, R Kowalkowski, Conductance, work

function and catalytic activity of SnO 2 -based gas sensors, Sens Actuators B:

Chem 3 (1991) 205–214.

[6] N Barsan, M Schweizer-Berberich, W Göpel, Fundamental and practical

aspects in the design of nanoscaled SnO 2 gas sensors: a status report, Fresenius

J Anal Chem 365 (1999) 287–304.

[7] C Xu, J Tamaki, N Miura, N Yamazoe, Grain size effects on gas sensitivity of

porous SnO 2 -based elements, Sens Actuators B: Chem 3 (1991) 147–155.

[8] K Yoshida, K Shimanoe, N Yamazoe, Preparation of SnO 2 cluster by

hydrother-mal treatment for highly sensitive thin film sensor, in: Proceedings of the

Technical Digest of the 6th East Asian Conference on Chemical Sensor, 2005,

pp 515–516.

[9] N Yamazoe, Y Kurokawa, T Seiyama, Effects of additives on semiconductor gas

sensors, Sens Actuators 4 (1983) 283–289.

[10] S Matsushima, T Maekawa, J Tamaki, N Miura, N Yamazoe, Role of additives

on alcohol sensing by semiconductor gas sensor, Chem Lett (1989) 845–848.

[11] T Maekawa, J Tamaki, N Miura, N Yamazoe, Sensing behavior of CuO-loaded

SnO 2 element for H 2 S detection, Chem Lett (1991) 575–578.

[12] J.G Duh, J.W Jou, Catalytic and gas sensing characteristics in Pd-doped SnO 2 , J.

Electrochem Soc 136 (1989) 2740–2747.

[13] J Kappler, N Bârsan, U Weimar, A Dièguez, J.L Alay, A Romano-Rodriguez, J.R.

Morante, W Göpel, Correlation between XPS, Raman and TEM measurements

and the gas sensitivity of Pt and Pd doped SnO 2 based gas sensors, Fresenius J.

Anal Chem 361 (1998) 110–114.

[14] K.H Cha, H.C Park, K.H Kim, Effect of palladium doping and film thickness on

the H 2 gas sensing characteristics of SnO 2 , Sens Actuators B: Chem 21 (1994)

91–96.

[15] N.S Baik, G Sakai, N Miura, N Yamazoe, Hydrothermally treated sol solution

of tin oxide for thin-film gas sensor, Sens Actuators B: Chem 63 (2000) 74–79.

[16] N Yamazoe, K Shimanoe, Roles of shape and size of component crystals in

semi-conductor gas sensors I Response to oxygen, J Electrochem Soc 155 (2008)

J85–J92.

[17] N.S Baik, G Sakai, K Shimanoe, N Miura, N Yamazoe, Hydrothermal

treat-ment of tin oxide sol solution for preparation of thin-film sensor with enhanced

thermal stability and gas sensitivity, Sens Actuators B: Chem 65 (2000) 97–

100.

[18] G Sakai, N.S Baik, N Miura, N Yamazoe, Gas sensing properties of tin oxide thin

films fabricated from hydrothermally treated nanoparticles dependence of CO

and H 2 response on thickness, Sens Actuators B: Chem 77 (2001) 116–121.

[19] V.V Kissinea, S.A Voroshilovb, V.V Sysoeva, A comparative study of SnO 2 and

SnO 2 :Cu thin flms for gas sensor applications, Thin Solid Films 348 (1999)

304–311.

[20] G Sakai, N Matsunaga, K Shimanoe, N Yamazoe, Theory of gas-diffusion

con-trolled sensitivity for thin film semiconductor gas sensor, Sens Actuators B:

Chem 80 (2001) 125–131.

[21] N Matsunaga, G Sakai, K Shimanoe, N Yamazoe, Formulation of gas

dif-fusion dynamics for thin film semiconductor gas sensor based on simple

reaction–diffusion equation, Sens Actuators B: Chem 96 (2003) 226–233.

[22] D.D Vuong, G Sakai, K Shimanoe, N Yamazoe, Preparation of grain

size-controlled tin oxide sols by hydrothermal treatment for thin film sensor

application, Sens Actuators B: Chem 103 (2004) 386–391.

[23] S Matsushima, Y Teraoka, N Miura, N Yamazoe, Electronic interaction between

metal additives and tin dioxide in tin dioxide-based gas sensors, Jpn J Appl.

Phys 27 (1988) 1798.

[24] S Matsushima, J Tamaki, N Miura, N Yamazoe, TEM observation of the

disper-sion state of Pd on SnO 2 , Chem Lett (1989) 1651.

[25] P Bartnickel, A Wokaun, W Sager, H.-F Eicke, Size tailoring of silver colloids

by reduction in W/O microemulsions, J Colloid Interf Sci 148 (1992) 80.

[26] I Lisiecki, M.P Pileni, Synthesis of copper metallic clusters using reverse

micelles as microreactors, J Am Chem Soc 115 (1993) 3887–3896.

[27] J Solla-Gullon, V Montiel, A Aldaz, J Clavilier, Electrochemical characterisa-tion of platinum nanoparticles prepared by microemulsion: how to clean them without loss of crystalline surface structure, J Electroanal Chem 491 (2000) 69–77.

[28] V Chhabra, V Pillai, B.K Mishra, A Morrone, D.O Shah, Synthesis, character-ization, and properties of microemulsion-mediated nanophase TiO 2 particles, Langmuir 11 (1995) 3307–3311.

[29] K Osseo-Asare, F.J Arriagada, Preparation of SiO 2 nanoparticles in a non-ionic reverse micellar system, Colloids Surf 50 (1990) 321–339.

[30] W Que, S Buddhudu, Y Zhou, Y.L Lam, J Zhou, Y.C Chan, C.H Kam, L.H Gan, G Roshan Deen, Preparation and characterization of erbium oxalate and erbium oxide nanoparticles by microemulsion technique, Mater Sci Eng C 16 (2001) 51–54.

[31] K.C Song, J.H Kim, Preparation of nanosize tin oxide particles from water-in-oil microemulsions, J Colloid Interf Sci 212 (1999) 193–196.

[32] D Chen, L Gao, Novel synthesis of well-dispersed crystalline SnO 2 nanopar-ticles by water-in-oil microemulsion-assisted hydrothermal process, J Colloid Interf Sci 279 (2004) 137–142.

[33] P.S Khiew, S Radiman, N.M Huang, Md Soot Ahmad, Studies on the growth and characterization of CdS and PbS nanoparticles using sugar–ester nonionic water-in-oil microemulsion, J Cryst Growth 254 (2003) 235–243.

[34] A Towata, Y Uwamino, M Sando, K Iseda, H Taoda, Synthesis of titania pho-tocatalysts dispersed with nickel nanosized particles, Nanostruct Mater 10 (1998) 1033–1042.

[35] T Masui, K Fujiwara, Y Peng, K Machida, G Adachi, Carbon monoxide oxidation characteristics over the Al 2 O 3 -supported CeO 2 –ZrO 2 catalysts prepared by the microemulsion method, Chem Lett (1997) 1285.

[36] T Miyao, N Toyoizumi, S Okuda, Y Imai, K Tajima, S Naito, Preparation of Pt/SiO 2 ultra-fine particles in reversed micelles and their catalytic activity, Chem Lett (1999) 1125.

[37] T Kida, G Guan, Y Minami, T Ma, A Yoshida, Photocatalytic hydrogen produc-tion from water over a LaMnO 3 /CdS nanocomposite prepared by the reverse micelle method, J Mater Chem 13 (2003) 1186–1191.

[38] M Yuasa, G Sakai, K Shimanoe, Y Teraoka, N Yamazoe, Reverse micelle-based preparation of carbon-supported La 1−x SrxMn 1−y FeyO3+dfor oxygen reduction electrode, J Electrochem Soc 151 (2004) A1690–A1695.

[39] M Yuasa, K Shimanoe, Y Teraoka, N Yamazoe, Preparation of carbon-supported nano-sized LaMnO 3 using reverse micelle method for energy-saving oxygen reduction cathode, Catal Today 126 (2007) 313–319.

[40] M Nakamine, M Abe, K Ogino, The effects of alkyl chain lengths of sodium alkyl sulfates and n-alkanes on microemulsion formation, J Colloid Interf Sci.

135 (1990) 449–454.

[41] S.K Ghosh, P.K Khatua, S.C Bhattacharya, Physicochemical characteristics of reverse micelles of polyoxyethylene nonyl phenol in different organic solvents,

J Colloid Interf Sci 279 (2004) 523–532.

[42] S Matsushima, T Maekawa, J Tamaki, N Miura, N Yamazoe, New methods for supporting palladium on a tin oxide gas sensor, Sens Actuators B: Chem 9 (1992) 71–78.

Biographies

Masayoshi Yuasa has been an assistant professor at Kyushu University Since 2005.

He received his ME degree in materials science in 2003 His current research interests include the development of chemical sensors and active electrocatalysts for oxygen reduction and oxygen evolution.

Takanori Masaki received his ME degree in materials science in 2007 from Kyushu

University.

Tetsuya Kida has been an associate professor at Kyushu University since 2006 He

received his ME degree in materials science in 1996 and his Dr Eng degree in 2001 from Kyushu University His current research interests include the development

of chemical sensors, nanoparticle synthesis, and self-assembles inorganic–organic hybrid materials.

Kengo Shimanoe has been a professor at Kyushu University since 2005 He received

the BE degree in applied chemistry in 1983 and the ME degree in 1985 from Kagoshima University and Kyushu University, respectively He joined Nippon Steel Corp in 1985, and received PhD in engineering in 1993 from Kyushu University His current research interests include the development of gas sensors and other functional devices.

Noboru Yamazoe had been a professor at Kyushu University since 1981 until he

retired in 2004 He received his BE degree in applied chemistry in 1963 and PhD in engineering in 1969 from Kyushu University His research interests were directed mostly to development and application of functional inorganic materials.