gas - sensing properties of tin oxide doped with metal oxides and carbon

Please cite this article in press as: N. Van Hieu, et al., Gas-sensing properties of tin oxide doped with metal oxides and carbon nanotubes:
A competitive sensor for ethanol and liquid petroleum gas, Sens. Actuators B: Chem. (2009), doi:10.1016/j.snb.2009.03.043
ARTICLE IN PRESS
G Model
SNB-11453; No.of Pages7
Sensors and Actuators B xxx (2009) xxx–xxx
Contents lists available at ScienceDirect
Sensors and Actuators B: Chemical
journal homepage: www.elsevier.com/locate/snb
Gas-sensing properties of tin oxide doped with metal oxides and carbon
nanotubes: A competitive sensor for ethanol and liquid petroleum gas
Nguyen Van Hieu
a,∗
, Nguyen Anh Phuc Duc
b
, Tran Trung
c
,
Mai Anh Tuan
a
, Nguyen Duc Chien
b
a
International Training Institute for Materials Science (ITIMS), Hanoi University of Technology (HUT), No. 1 Dai Co Viet Road, Hanoi, Viet Nam
b
Institute of Engineering Physics, Hanoi University of Technology, Hanoi, Viet Nam
c
Faculty of Environment and Chemistry, Hung Yen University of Technology and Education, Hung Yen, Viet Nam
article info
Article history:
Available online xxx

Keywords:
Gas sensor
Tin oxide
Carbon nanotubes
abstract
SnO
2
doped with metal oxides such as PtO
2
, PdO, La
2
O
3
CuO, and Fe
2
O
3
and multi-walled carbon nan-
otubes (MWCNTs) thin films were prepared by the sol–gel method. Thin film gas sensors were fabricated
by spin-coating the sol onto interdigitated microelectrodes. The microstructure and morphology of the
materials were characterized by XRD, FE-SEM, and TEM. The results reveal that their SnO
2
particle size
is lower than 10nm, and the MWCNTs doping is well embedded in the SnO
2
matrix. The response of all
the sensors was studied for different concentrations of ethanol and liquid petroleum gases (LPG) and at
different operating temperatures. Comparative results reveal that the (1 wt%) PtO
2
-doped SnO

2
sensor
exhibits higher sensitivity to ethanol gas and LPG than the sensors doped with the other dopants. Espe-
cially, the (1 wt%) PtO
2
-doped SnO
2
sensor shows higher selectivity to ethanol gas over LPG, while the
(0.1wt%, 20 <d < 40 nm)-doped SnO
2
shows higher selectivity to LPG over ethanol gas in the same testing
conditions.
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
The hybrid materials made of semiconductor metal oxides
(SMO) such as TiO
2
, SnO
2
and WO
3
and carbon nanotubes (CNTs)
have been given much attention in recent year for their various
applications such as photocatalysis, anode material for lithium-
ion batteries and gas sensor [1–14]. The nanoarchitectures forming
hybrid materials between SMO and CNTs have been conducted in
different ways such as SMO/CNTs composite [1–5], SMO-coated
CNTs [6–8], SMO-filled CNTs [9] and CNTs-doped SMO [10–14]. The
special geometries and properties of the hybrid materials facilitate
their great potential applications as high-performance gas sensors.

Previous works have demonstrated that the hybrid materials can
be used to detect various gases such as NH
3
,NO
2
,H
2
, CO, LPG,
and ethanol [4–6,10–14]. These works also reported that the hybrid
materials gas sensors have a better performance compared to the
sensors used SMO as well as CNTs as sensing materials. Interest-
ingly, the composite SnO
2
/CNTs and the CNTs-doped SnO
2
sensors
respond to NH
3
and NO
2
at room temperature, respectively [4,10].
This would reduce considerably the power consumption of the
∗
Corresponding author. Tel.: +84 4 38680787; fax: +84 4 38692963.
E-mail addresses: ,
(N. Van Hieu).
sensing-device. The CNTs are hollow nanotube and p-type semi-
conductor, therefore the enhancement of the sensing performance
of the sensors based on CNTs/SnO
2

hybrid materials in comparison
with the sensorsbased on the separated materials was attributed to
additional nanochannel for gas diffusion and p/n junctions formed
by CNTs and SnO
2
. These mechanisms were previously represented
in [4,5,10].
Ethanol gas sensors are extensively used for the control of
drunken driving and monitoring of fermentation and other pro-
cesses inchemical industries, whileLPG sensorsare frequently used
in the detection of the gas leakages to prevent accidental explo-
sion. The development of ethanol gas and LPG sensors based on
SnO
2
thin film technology offers great advantages such as high
sensitivity, fast response, and low cost. Therefore, much effort has
been devoted to improve its sensitivity and selectivity by intro-
ducing various dopants such as PtO
2
, CdO, La
2
O
3
, PdO, SiO
2
, and
RuO
2
[16–31] or by mixing with other metal oxides such as Nb
2

O
3
,
Fe
2
O
3
, andZrO
2
[23,32,33]. It was found thatamong additives, SnO
2
sensors doped with La
2
O
3
and CdO showed good performance to
ethanol gas [17–19], while the Pd-, Pt- and RuO
2
-doped SnO
2
sen-
sors showed good performance to LPG [20,24]. In this paper, for
the first time, we study and compare the performance of various
metal oxides-doped SnO
2
and MWCNTs-doped SnO
2
sensors for
the detection of ethanol gas and LPG. In the later, the MWCNTs with
different diameters were used for the doping.

0925-4005/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.snb.2009.03.043
Please cite this article in press as: N. Van Hieu, et al., Gas-sensing properties of tin oxide doped with metal oxides and carbon nanotubes:
A competitive sensor for ethanol and liquid petroleum gas, Sens. Actuators B: Chem. (2009), doi:10.1016/j.snb.2009.03.043
ARTICLE IN PRESS
G Model
SNB-11453; No.of Pages7
2 N. Van Hieu et al. / Sensors and Actuators B xxx (2009) xxx–xxx
Fig. 1. XRD pattern of typical SnO
2
thin film heat-treated at 500 (a), 600 (b), and
700
◦
C (c).
2. Experimental
Stannic acid gel was synthesized by hydrolyzing 0.2 M solution
of tin chloride (SnCl
4
) with ammonia, using the following reaction
at room temperature:
SnCl
4
+ 4NH
4
OH → SnO
2
·nH
2
O + 4NH
4

Cl + (2 − n)H
2
O (1)
The resulting precipitate was washed thoroughly by repeating the
procedures of suspending the gel into deionized water and collect-
ing it back by filtration to remove Cl
−
. The SnO
2
gel was suspended
in an aqueous ammonia solution (pH 10.5) and followed by stir-
ring for 2 h, with a calculated amount of SnO
2
gelinorderto
achieve5 equivalent wt%SnO
2
sol–gel solution.The suspensionwas
transferred to a Teflon-lined stainless steel autoclave and treated
hydrothermally at 200
◦
C for 10 h. The 1 equivalent wt% CuO-,
Fe
2
O
3
-, La
2
O
3
-, and PtO

2
-99 wt% SnO
2
sols were prepared by mix-
ing the required amount of dissolutions of Cu(NO
3
)
2
, Fe(NO
3
)
3
,
La(NO
3
)
3
and PtCl
4
(0.1M) to the pure SnO
2
sol.
Functionalized MWCNTs with different diameters (d <10nm,
d =20–40 nm, and d =60–100 nm) were used for the fabrication of
the MWCNTs-doped SnO
2
sensors with a calculated amount of the
MWCNTs in order to achieve 0.1 equivalent wt% MWCNTs–99.9 wt%
SnO
2

sol. The MWCNTs were functionalized by using a typical pro-
cedure described as follows: 200 mg MWCNTs were suspended in
35 mL concentrated nitric acid (15 M) and refluxed for 12 h in a
silicone oil bath maintained at 140
◦
C to modify the MWCNTs sur-
face, they were then rinsed with distilled H
2
O until the pH of the
solution was neutral, and finally they were dried at 80
◦
C in vac-
uum oven. The immersion-probe ultrasonic generator with a high
power up to 500 W (Model VC-505, Sonics, US) was used for disper-
sion of MWCNTs in SnO
2
sol. The morphology and the crystalline
phase ofthe filmswerecharacterized byusing afield emissionscan-
ning electron microscope (FE-SEM; 4800 Hitachi, Japan) and X-ray
diffraction (XRD, Philips XPert Pro), respectively. The dispersion of
the MWCNTs in theSnO
2
sol was characterized by TEM usinga JEM-
100cx instrument with an accelerating voltage of 80 kV. The details
on the gas sensor fabrication and characterization were describe d
in our previous works [4,14]. All sensors were tested with various
concentrations of ethanol gas (100–1000 ppm) and LPG (0.1–1%
or 1000–10,000 ppm), and with different operating temperatures
Fig. 2. TEM image of SnO
2

nanoparticles (a); FE-SEM image of a SnO
2
thin film heat-treated at 500
◦
C (b); TEM image of an MWCNTs-doped SnO
2
sol (c); FE-SEM image of
an MWCNTs-doped SnO
2
(d).
Please cite this article in press as: N. Van Hieu, et al., Gas-sensing properties of tin oxide doped with metal oxides and carbon nanotubes:
A competitive sensor for ethanol and liquid petroleum gas, Sens. Actuators B: Chem. (2009), doi:10.1016/j.snb.2009.03.043
ARTICLE IN PRESS
G Model
SNB-11453; No.of Pages7
N. Van Hieu et al. / Sensors and Actuators B xxx (2009) xxx–xxx 3
Fig. 3. Response of PtO
2
,Fe
2
O
3
,La
2
O
3
, and CuO-doped SnO
2
sensors to 2500 ppm (0.25%) LPG in air (a) and 250 ppm ethanol gas in air (b); the response to ethanol gas
(250 ppm) and LPG (2500 ppm) of (1 wt%) a PtO

2
-doped SnO
2
sensor (c); step wise decrease in electrical resistance obtained with a increase in ethanol concentration from
air to 1000 ppm ethanol gas in air for (1wt%) PtO
2
-doped SnO
2
sensor operating at 240
◦
C; (e) the sensor response versus ethanol gas concentration.
(190, 220, 240, 260, 290, 320, 340, and 360
◦
C). It should be noted
that technological application requires the ethanol and LPG sensor
to be able to detect at least 200 ppm alcohol (∼0.6g/L in the human
blood) and 0.24% (2400 ppm) LPG (lowest explosive level). Hence,
our gas-concentration range to be tested is within levels.
3. Results and discussion
Fig. 1a–c shows the XRD patterns of the SnO
2
samples after the
heat-treatment at temperatures of 500, 600, and 700
◦
C, respec-
tively. The XRD characterization was also carried out with the
MWCNTs-doped SnO
2
samples (not shown), but we observed no
differences. This is attributed to the use of very low doping con-

tent of MWCNT and the well-embedded MWCNTs in SnO
2
matrix,
which have already been reported in the literature [1,3,4,7].Itcan
be seen that the heat-treated samples are well crystallized with all
diffraction peaks which can be well indexed to the tetragonal rutile
structure of SnO
2
. The broad and well-defined reflections were
observed at 2Â =26.51, 33.67 and 51.78 corresponding to (11 0),
(1 01) and (21 1) planes, respectively, in the XRD spectrum of the
annealed SnO
2
thin films, which are in good agreement with the
previously reported [20,21], confirmingthe formation ofa polycrys-
talline SnO
2
thin film. The estimated value of the lattice constants
were found to be a = b = 4.734Å and c = 3.185 Å (JCPDS 21-1250). The
value of the crystallite size of the heat-treated SnO
2
thin film was
estimated by fitting the width of (1 1 0) reflection using Scherrer
formula d=K/ˇcos Â, where K is 0.94,  is the X-ray wavelength, ˇ
the peak full width half maxima (FWHM) and, Â is the diffraction
peak position.
The roughly estimated values of crystallite size of the sam-
ples heat-treated at 500, 600, and 700
◦
C are found to be about

5.8, 6.2, and 7.1 nm, respectively. This indicated that the crystallite
sizes do not significantly vary for heat-treating temperature rang-
ing from 500 to 700
◦
C. Actually, we have already investigated the
gas-sensing properties of blank SnO
2
films heat-treated at these
temperatures, and we have obtained similar responses to ethanol
gas and LPG (not shown here). Therefore, these characterizations
are in order to choose right heat-treated temperature for metal
oxides- and MWCNTs-doped SnO
2
thin films. In addition, it is well
known that MWCNTs materials can be burned out at heat-treated
temperatures higher than 550
◦
C.
Fig. 4. Response as a function of operating temperature of the PtO
2
-doped SnO
2
sensors with varying PtO
2
doping content.
Please cite this article in press as: N. Van Hieu, et al., Gas-sensing properties of tin oxide doped with metal oxides and carbon nanotubes:
A competitive sensor for ethanol and liquid petroleum gas, Sens. Actuators B: Chem. (2009), doi:10.1016/j.snb.2009.03.043
ARTICLE IN PRESS
G Model
SNB-11453; No.of Pages7

4 N. Van Hieu et al. / Sensors and Actuators B xxx (2009) xxx–xxx
Fig. 5. Response of (0.1wt%) MWCNTs (with d < 10 nm; 20 nm < d < 40 nm; 60 nm < d < 100 nm) -doped SnO
2
sensors to 2500 ppm (0.25%) LPG in air (a) and 250 ppm ethanol
gas in air (b); response to ethanol gas (250 ppm) and LPG (2500 ppm) of (0.1 wt%, 10 nm < d < 20 nm) MWCNTs-doped SnO
2
sensor (c); step wise decrease in electrical resistance
obtained with an increase in LPG concentration from air to 10000ppm (1%) LPG in air for (0.1wt%, 10 nm < d< 20 nm) MWCNTs-doped SnO
2
sensor operating at 240
◦
C; (e)
the sensor response versus LPG concentration.
The particle size and morphology of the SnO
2
thin film charac-
terized further by TEM and FE-SEM are shown in Fig. 2. The TEM
image (Fig. 2a) shows that the particle size is quite homogenous in
the range of 4–8 nm. The FE-SEM image (Fig. 2b) shows the mor-
phology of the SnO
2
thin film treated at 500
◦
C. It is shown that
the particle size is smaller than 10 nm. The MWCNTs-doped SnO
2
sample was characterized by TEM and FE-SEM as shown in Fig. 2c
and d, respectively. Fig. 2c shows the TEM image of the MWCNTs-
dispersed SnO
2

sol in which the black materials absorbed onto
the wall of the functionalized MWCNTs were believed to be SnO
2
nanoparticles from the SnO
2
sol. This is because it was recognized
that the MWCNTs-SnO
2
bonding can be formed naturally through
some physicochemical interactions such as Van der Waals force, H
bonding and other bonds. For example, the
OH group on SnO
2
may
possibly react with the
OH and COOH groups on the functional-
ized MWCNTs in removing the H
2
O contained in the wet material,
and thus the bonding C
O Sn or O C O Sn might form through
the dehydration reaction that happens among the groups on the
two materials. However, this was not strongly explained and more
intensive studies are needed to confirm this. The absorbed SnO
2
on the MWCNTs would grow up and enclose the MWCNTs during
the heat-treatment. This observation was consistent with previous
reports [1–3]. Fig. 2d shows the FE-SEM image of the MWCNTs-
doped SnO
2

film after heat-treatment at 500
◦
C. It can be seen that
the MWCNTs are well encapsulated with a SnO
2
matrix and is still
present after the heat-treatment at 500
◦
C.
The sensing characteristics of (1 wt%) metal oxides (PtO
2
,Fe
2
O
3
,
CuO, La
2
O
3
)-doped sensors to ethanol gas and LPG have indicated
in Fig. 3. The sensor responses as a function of operating tempera-
ture to LPG and ethanol gas are respectively shown in Fig. 3a and b.
It seems that the optimized operating temperatures of the sensor
to ethanol gas and LPG are around 350 and 250
◦
C, respectively.
It can be recognized that all the metal oxides-doped SnO
2
sen-

sors show an improvement in their response to ethanol gas, while
only CuO, PtO
2
, and La
2
O
3
-doped sensors show an improvement
in the sensor response to LPG compared with undoped SnO
2
sen-
sors. However, this also depends on the operating temperature to
be selected. These observations are consistent and have been rea-
sonably explained in the literature [30,31]. Our experimental data
show that (1 wt%) the PtO
2
-doped sensor has a better sensitivity to
ethanol gas and LPG as compared to that of the sensors doped with
the other dopants. For instance, at the operating temperature of
240
◦
C, the response to 250 ppm ethanol gas and 2500 ppm (0.25%)
LPG is around 101.9 and 2.1, respectively. These values are compa-
rable with the data reported in the literature [17–20,31].Wehave
plotted the responses to ethanol gas (250 ppm) and LPG(2500 ppm)
of the (1 wt%) PtO
2
-doped SnO
2
sensor as depicted in Fig. 3c. It can

be seen that the sensor has relatively good selectivity to ethanol gas
over LPG. Fig. 3d shows the electrical resistance variations obtained
with several steps of different ethanol concentration from air to
1000 ppm ethanol in air for the (1 wt%) PtO
2
-doped SnO
2
sensor at
an operating temperature of 240
◦
C. As can be seen, upon switching
on ethanol gas, the film reaches the saturated resistance R
g
in 50 s
and at the end of the injection cycle, when dry air is introduced, its
electrical resistance returns to the original value (R
a
). This fact is a
proof of the reversibilityof theprocess. Thestepwise decrease of the
electrical resistance of the film is very consistent with an increas-
ing amount of ethanol oxidation. Greater ethanol oxidation caused
the introduction of more electrons into the SnO
2
surface and the
film became less resistive. Fig. 3e depicts the correlation between
the ethanol gas concentration and the response of the (1 wt%) PtO
2
-
doped SnO
2

sensor.It seemsthat the correlation lines were not good
linear for such broad ethanol concentration (100–1000 ppm). How-
ever, for practical applications of this sensor, the linear fit can be
Please cite this article in press as: N. Van Hieu, et al., Gas-sensing properties of tin oxide doped with metal oxides and carbon nanotubes:
A competitive sensor for ethanol and liquid petroleum gas, Sens. Actuators B: Chem. (2009), doi:10.1016/j.snb.2009.03.043
ARTICLE IN PRESS
G Model
SNB-11453; No.of Pages7
N. Van Hieu et al. / Sensors and Actuators B xxx (2009) xxx–xxx 5
well made between the sensor response and ethanol concentration
for narrower ethanol concentration ranges (e.g. 100–500 ppm and
500–1000 ppm). The linear dependent of the ethanol concentration
on the sensor response of our PtO
2
-doped SnO
2
sensor for the nar-
row ethanol concentration range is consistent with that previous
reported [17,18].
Indeed, we have studied further the sensing properties of the
PtO
2
-doped SnO
2
sensors, in which the PtO
2
doping content was
varied from 0.2 to 2 wt% in comparison with SnO
2
weight. The sens-

ing characteristics of these sensors were measured with 250 ppm
ethanol at different operating temperatures, and the results are
shown in Fig. 4. It can be seen that the (1 wt%) PtO
2
-doped SnO
2
sensor shows a higher response. From this, it can be concluded that
the PtO
2
doping content of 1 wt% is the optimal value.
The 90% response time for gas exposure (t
90%(air-to-gas)
) and that
for recovery (t
90%(gas-to-air)
) were calculated from the resistance-
time data shown in Fig. 3d. The t
90%(air-to-gas)
values are around 23 s,
while the t
90%(gas-to-air)
value is around 46 s. These results are quite
comparable with those of SnO
2
-based sensors reported previously
[17–20].
Recently, hybrid CNTs/SnO
2
sensors have been extensively
investigated. Therefore, for comparison with metal oxides-doped

SnO
2
sensors, we have prepared and characterized MWCNTs-doped
SnO
2
sensors in the same route as metal oxides-doped sensors in
this work. Fig. 5 shows the sensing characteristics of SnO
2
sensors
doped with different kinds of MWCNTs. It can be recognized that
the responses to ethanol gas and LPG of all MWCNTs-doped SnO
2
sensors are improved at a low region of operating temperature.
This observation is consistent with previously reported data for
the hybrid sensors of CNTs/SMO [5,7,11,13,14]. To explain this, we
speculate that it may result from the fact that the doping of MWC-
NTs in the SnO
2
matrix can introduce nanochannels and additional
hetero-junction between SnO
2
(n-type) and CNTs (p-type). Both
these effects do not cause the response improvement of the hybrid
MWCNTs-doped sensor athigh operatingtemperatures because the
nanochannels formed by the MWCNTs may not play any role for
gas diffusion into the SnO
2
matrix at a high operating temperature.
Otherwise, we believe that SnO
2

(n-type)/MWCNTs (p-type) can
not functionalize well at a temperature higher than 350
◦
C due to
the transition from semiconductor behavior to metallic one of the
CNTs. More detail on this mechanism can be found further in recent
worksby us andothers [4,5,10,14]. Additionally,it also observed that
the effect of MWCNTs on the response of the MWCNTs-doped SnO
2
sensors is not significant in the detection of LPG and ethanol gas. It
seems that (d = 10–20 nm) MWCNTs-doped SnO
2
sensors have bet-
ter performance to LPG and ethanolgas atan operatingtemperature
range of 280–350
◦
C.
The specific surface area (SSA) of MWCNTs with diameter of <10,
20-40 nm and 60–100 nm were 242.2, 112.2, and 45.2m
2
/g, respec-
tively. In principle,the material with ahigher SSA would have better
gas response. However, we have observed that the doping content
is so small that it could not affect the SSA of the MWCNTs-doped
SnO
2
materials. Thus, the SSA factor cannot be a piece of evidence
on the difference in the sensor response. The observed effect can be
explained by the fact that the MWCNTs embedded in SnO
2

behave
as nanochannels for the gas diffusion in the SnO
2
bulk material.
However, a larger diameter of MWCNTS (e.g. d = 60–100 nm) can
result in the decrease of sensor response because such larger diam-
eter of MWCNTs could not be well embedded in the SnO
2
matrix,
and they begin to connect together, resulting in a shorter resistance
path of the MWCNTs-doped SnO
2
sensors.
From Fig. 5c, we can see that MWCNTs-doped SnO
2
sensors are
more selective to LPG than to ethanol gas at an operating tem-
perature range of 280–350
◦
C. This effect is completely different
with the metal oxides-doped SnO
2
sensors (see Fig. 3c) that will
be discussed further in the next paragraph. Fig. 5d depicts the
Fig. 6. Response of an undoped SnO
2
sensor to 250 ppm ethanol and 2500 ppm
(0.25%) LPG in air operating at a temperature range from 190 to 360
◦
C.

electrical resistance variations obtained with several steps of dif-
ferent LPG concentration from air to 1% LPG in air for the (0.1wt%,
20 < d < 40 nm) MWCNTs-doped SnO
2
sensor operating at 320
◦
C.
Similar to the PtO
2
-doped SnO
2
sensors in the detection of ethanol,
the MWCNTs-doped SnO
2
sensors shows a good reversibility in the
detection of LPG and the stepwise decrease of electrical resistiv-
ity of the MWCNTs-doped SnO
2
film is very consistent with the
increasing amount of LPG oxidation. More LPG oxidation caused
the introduction of more electrons into the SnO
2
surface and the
film became less resistive. Fig. 5e depicts the variation of response
with LPG concentration in air for the MWCNTs-doped SnO
2
sensors
at an operating temperature of 320
◦
C. It can be observed that the

response does not increase linearly for the concentration range of
0.1–0.6% (1000–6000 ppm). It seems that the response tends to sat-
urate for an LPG concentration higher than 0.5% (5000 ppm). This
can be attributed to the fact that there would be an insufficient
number of oxygen anions available on the surface of the MWCNTs-
doped SnO
2
materials for reaction with LPG.
The 90% response time for gas exposure (t
90%(air-to-gas)
) and that
for recovery (t
90%(gas-to-air)
) were calculated from the resistance-
time data shown in Fig. 5d. The t
90%(air-to-gas)
value is around 21 s,
while the t
90%(gas-to-air)
value is around 36 s. It can be seen that the
response times of the Pt- and MWCNTs-doped SnO
2
sensors are
similar, while the recovery time of the MWCNTs-doped sensor is
relatively shorter than that of the PtO
2
-doped SnO
2
sensors. This
could be attributed to the formation of the nanochannels in SnO

2
materials by doping CNTs that can enhance the diffusion in and out
of the gas molecules.
To study the effect of MWCNTs doping on the sensing proper-
ties to ethanol gas and LPG, we plotted the response of undoped
SnO
2
sensor to 250 ppm ethanol gas and 2500 ppm (0.25%) LPG as
shown in Fig. 6. It is indicated that the response of undoped SnO
2
sensors to 250 ppm ethanol gas is higher than that to 2500 ppm
(0.25%) LPG over an operating temperature range of 190–360
◦
C.
Therefore, this points out that the higher response of the MWCNTs-
doped SnO
2
sensors to LPG than to ethanol can be attributed to the
MWCNTs doping. This is an interesting finding that cannot yet be
clearly explained as of now. The pure SnO
2
sensor is more sensitive
to ethanol than LPG even though the ethanol gas concentration is
about 10 times lower than the LPG concentration. This has alsobeen
explored in previous works [15]. The sensing mechanism of the
ethanol and LPG has long been known and widely adopted in pre-
vious reports [24–30]. However, to explain why the ethanol is more
sensitive than LPG, even though the former has a lower concentra-
tion than the later, is still unclear. It has long been known that there
Please cite this article in press as: N. Van Hieu, et al., Gas-sensing properties of tin oxide doped with metal oxides and carbon nanotubes:

A competitive sensor for ethanol and liquid petroleum gas, Sens. Actuators B: Chem. (2009), doi:10.1016/j.snb.2009.03.043
ARTICLE IN PRESS
G Model
SNB-11453; No.of Pages7
6 N. Van Hieu et al. / Sensors and Actuators B xxx (2009) xxx–xxx
is a dehydrogenation step in the reaction between ethanol/LPG and
SnO
2
surface at elevated temperatures, and they can be described
as [25,30]:
Ethanol dehydrogenation
2CH
3
CH
2
OH + O
2
−
→ 2CH
3
CHO + H
2
O + e (2)
LPG dehydrogenation
2CH
3
CH
2
CH
3

+ O
2
−
→ 2CH
3
CH
2
CHO + H
2
O + e (3)
2CH
4
CH
2
CH
2
CH
3
+ O
2
−
→ 2CH
4
CH
2
CH
2
CHO + H
2
O + e (4)

2CH
3
CHCH
2
+ O
2
−
→ 2CH
3
CHCO + H
2
O + e (5)
It should be noted that the main constituent of LPG is propane
(∼85% by liquid volume), butane (∼2.5% by liquid volume) and
propene (∼5% by liquid volume). Among these constituents,
propane and butane are m ore stable than propene, which is an
unsymmetrical alkene containing a double bond. Thus, propene is
more prone to hydrogenation than to the dehydrogenation, which
can be used to compare with the dehydrogenation of ethanol gas.
For 2500 ppm LPG, the propene concentration can be estimated to
be about 125 ppm. This is the reason why the response of the SnO
2
sensors to 2500 ppm LPG is lower than that of 250 ppm ethanol gas.
It has been reported that the MWCNTs-doped SnO
2
sensors
show a better performance compared with undoped SnO
2
sen-
sor. In our case, for ethanol detection, the response of the (0.1 wt,

20 <d < 40 nm) MWCNTs-doped SnO
2
sensors (∼12.3 for 250ppm
at 260
◦
C, see Fig. 5b) is about three times higher than that of the
undoped SnO
2
sensor (∼3.4 for 250 ppm at 260
◦
C, see Fig. 5b).
The reason for this was previously explained in detail [4,5,10,14].
The question to raise here is that why the MWCNTs-doped SnO
2
is more sensitive to 2500 ppm LPG than to 250 ppm ethanol as
depicted in Fig. 5c. Further intensiveinvestigationshould be done to
understand this phenomenonmore comprehensively.The plausible
explanation for the observed effect can be based are as follows: (i)
the MWCNTs are hollow nanotubes that gas absorption could occur
in the inside andoutside of theMWCNTs[34], (ii) themethane (CH
4
)
molecules (e.g. propane and butane) can be physically adsorbed
on the outgassed nanotubes (i.e., nanotube after oxygen exposure)
[35], and (iii) the oxygen molecules are strongly adsorbed on the
defective sites of MWCNTs (adsorption energy is about 0.32 eV) [36]
that can serve as a reactive gas for the oxidation reactions of LPG.
These reasons can enhance Reactions (3) and (4). Additionally, the
consumption of the adsorbed oxygen can affect the electrical prop-
erties of the MWCNTs [36,37], and the electrical resistance of the

MWCNTs-doped SnO
2
film can be consequently changed.
4. Conclusion
We have systematically investigated and compared the perfor-
mance of metal oxide- and MWCNTs-doped SnO
2
thin film sensors
to LPG andethanol gas.We have found that the SnO
2
doped withthe
1 wt% PtO
2
sensor shows the highest response to ethanol gas and
LPG compared with that of SnO
2
doped with the other dopants.
Among carbon nanotubes-doped SnO
2
sensors, the sensor doped
with 0.1 wt% MWCNTs with a diameter ranging from 20 to 40 nm
exhibits the highest response to ethanol gas and LPG. An interesting
finding is that the PtO
2
-doped SnO
2
sensor shows good selectivity
to ethanol gas over LPG, while, the MWCNTs-doped SnO
2
sensor

shows good selectivity to LPG over ethanol gas, at the same testing
conditions. The gas-sensing mechanism of the hybrid sensor has
been discussed. However, further study is needed to understand
better the selectivity of the hybrid sensor to ethanol gas and LPG.
Acknowledgments
The work was supported by the National Foundation for Sci-
ence and Technology Development (NAFOSTED) of Vietnam (for
Basic Research Project: 2009-2012), the National Key Research Pro-
gram for Materials Technology (Project No. KC 02-05/06-10), the
research projects of Vietnam Ministry of Education and Training
(Code B2008-01-217 and B2008-21-09) and Key basic research pro-
gram for application orientation (2009-2012).
References
[1] M.H. Chen, Z.C. Huang, G.T. Wu, G.M. Zhu, J.K. You,Z.G. Lin, Synthesis and charac-
terization of SnO–carbon nanotube composite as anode material for lithium-ion
batteries, Mater. Res. Bull. 38 (2003) 831–836.
[2] W. Wang, P. Serp, P. Kalck, J.L. Faria, Photocatalytic degradation of phenol
on MWCNT and titania composite catalysts prepared by a modified sol–gel
method, Appl. Catal. B: Environ. 56 (2004) 301–308.
[3] Xie, K. Vijay, Varadan, Synthesis and characterization of high surface area tin
oxide/functionalized carbon nanotubes composite as anode materials, Mater.
Chem. Phys. 91 (2005) 274–280.
[4] N.V. Hieu, L.T.B. Thuy, N.D. Chien, Highly sensitive thin film NH
3
gas sensor oper-
ating at room temperature based on SnO
2
/MWCNTs composite, Sens. Actuators
B 129 (2008) 888–895.
[5] J. Gong, J. Sun, Q. Chen, Micromachined sol–gel carbon nanotube/SnO

2
nanocomposite hydrogen sensor, Sens. Actuators B 30 (2008) 829–835.
[6] W Q. Han, A. Zettl, Coating single-walled carbon nanotubes with tin oxide,
Nano Lett. 3 (2003) 681–683.
[7] Q. Kuang, S F. Li, Z X. Xie, S C. Lin, X.–H. Zhang, S Y. Xie, R B. Huang, L S.
Zheng, Controllable fabrication of SnO
2
-coated multiwalled carbon nanotubes
by chemical vapor deposition, Carbon 44 (2006) 1166–1172.
[8] Y L.Liu, H F.Yang,Y.Yang,Z M. Liu, G L. Shen, R Q. Yu,Gas sensing properties
of tin dioxide coated carbon nanotubes, Thin Solid Films 497 (2006) 355–360.
[9] L. Zhao, L. Gao, Filling of multi-walled carbon nanotubes with tin(IV) oxide,
Letters to the Editor, Carbon 42 (2004) 3251–3272.
[10] B Y.Wei,M C. Hsu, P G.Su,H M. Lin, R J. Wu, H J. Lai, A novel SnO
2
gas sensor
doped with carbon nanotubes operating at room temperature, Sens. Actuators
B 101 (2004) 81–89.
[11] A. Wisitsoraat, A. Tuantranont, C. Thanachayanont, V. Patthanasettakul, P. Sing-
jai, Electron beam evaporated carbon nanotubes dispersed SnO
2
thin films gas
sensor, J. Electroceram. 17 (200 6) 45–47.
[12] C. Bittencourt, A . Felten, E.H. Espinosa, R. Ionescu, E. Llobet, X. Correig, J J.
Pireaux, WO
3
films modified with functionalized multi-wall carbon nanotubes:
morphology, compositional and gas response, Sens. Actuators B 115 (2006)
33–41.
[13] E.H. Espinosa, R. Ionescu, B. Chambon, G. Bedis, E. Sotter, C. Bittencourt, A. Fel-

ten, J J. Pireaux, X. Correig, E. Llobet, Hybrid metal oxide and multiwall carbon
nanotube films for low temperature gas sensing, Sens. Actuators B 127 (2007)
137–142.
[14] N.V. Hieu, N.V. Duy, N.D. Chien, Inclusion of SWCNTs in Nb/Pt co-doped TiO
2
thin film for ethanol vapor detection, Phys. E 40 (2008) 2950–2958.
[15] F. Pourfayaz, A. Khodadadi, Y. Moryazavi, S.S. Mohajerzadeh, CeO
2
doped SnO
2
sensor selective to ethanol in presence of CO, LPG, and CH
4
, Sens. Actuator B
108 (2005) 172–176.
[16] W. Gopel, K.D. Schierbaum, SnO
2
sensors: current status and future prospects,
Sens. Actuators B 26–27 (1995) 1–12.
[17] P. Ivanov, E. Llobet, X. Vilanova, J. Brezmes, J. Hubalek, X. Correig, Development
of high sensitivity ethanol gas sensors based on Pt-doped SnO
2
surfaces, Sens.
Actuators B 99 (2004) 201–206.
[18] G. Neri, A. Bonavita, G. Micali, N. Donato, F.A. Deorsola, P. Mossino, I. Amato,
B. De Benedetti, Ethanol sensors based on Pt-doped tin oxide nanopowders
synthesised by gel-combustion, Sens. Actuators B 117 (2006) 196–204.
[19] C. Bittencourt, E. Llobet, P. Ivanov, X. Correig, X. Vilanova, J. Brezmes, J. Hubalek,
K. Malysz, J.J. Pireaux, J. Calderer, Influence of the doping method on the sen-
sitivity of Pt-doped screen-printed SnO
2

sensors, Sens. Actuators B 97 (2004)
67–73.
[20] M.H. Madhusudhana Reddy, A.N. Chandorkar, E-beam deposited SnO
2
, Pt–SnO
2
and Pd–SnO
2
thin films for LPG detection, Thin Solid Films 349 (1999) 260–265.
[21] Z. Tianshu, P. Hing, Y. Li, Z. Jiancheng, Selective detection of ethanol vapor and
hydrogen using Cd-doped SnO
2
-based sensors, Sens. Actuators B 60 (1999)
208–215.
[22] I. Stambolova, K. Konstantinov, S. Vassilev, P. Peshev, Ts. Tsacheva, Lanthanum
doped SnO
2
and ZnO thin films sensitive to ethanol and humidity, Mater. Chem.
Phys. 63 (2000) 104–108.
[23] I.T. Weber, R. Andrade, E.R. Leite, E. Longo, A study of the SnO
2
-Nb
2
O
5
system
for an ethanol vapor sensor: a correlation between microstructure and sensor
performance, Sens. Actuators B 72 (2001) 180–183.
[24] M.V. Vaishampayan, R.G. Deshmukh, I.S. Mulla, Influence of Pd doping on mor-
phology and LPG response of SnO

2
, Sens. Actuators B 131 (2008) 665–672.
[25] S. Majumder, S. Hussain, S.N. Das, R.B. Bhar, A.K. Pal, Silicon doped SnO
2
films
for liquid petroleum gas sensor, Vacuum 82 (2008) 760–770.
Please cite this article in press as: N. Van Hieu, et al., Gas-sensing properties of tin oxide doped with metal oxides and carbon nanotubes:
A competitive sensor for ethanol and liquid petroleum gas, Sens. Actuators B: Chem. (2009), doi:10.1016/j.snb.2009.03.043
ARTICLE IN PRESS
G Model
SNB-11453; No.of Pages7
N. Van Hieu et al. / Sensors and Actuators B xxx (2009) xxx–xxx 7
[26] M.S. Wagh, G.H. Jain, D.R. Patil, S.A. Patil, L.A. Patil, Surface customizationof SnO
2
thick films using RuO
2
as a surfactant for the LPG response, Sens. Actuators B
122 (2007) 357–364.
[27] S. Majumder, S. Dalui, R. Bhar, A.K. Pal, Synthesis of Pd/SnO
2
films by wet chem-
ical route for LPG sensor, Eur. Phys. J. Appl. Phys. 42 (2008) 193–202.
[28] C. Wang, C. Xiangfeng, W. Mingmei, Highly sensitive gas sensors based on hol-
low SnO
2
spheres prepared by carbon sphere template method, Sens. Actuators
B 120 (2007) 508–513.
[29] J J. Ho, Y.K. Fang, K.H. Wu, W.T. Hsieh, C.H. Chen, G.S. Chen, M.S. Ju, J J. Lin,
S.B. Hwang, High sensitivity ethanol gas sensor integrated with a solid-state
heater and thermal isolation improvement structure for legal drink-drive limit

detecting, Sens. Actuators B 50 (1998) 227–233.
[30] T. Jinkawa, G. Sakai, J. Tamaki, N. Miura, N. Yamazoe, Relationship between
ethanol gas sensitivity and surface catalytic property of tin oxide sensors mod-
ified with acidic or basic oxides, J. Mol. Catal. A 155 (2000) 193–200.
[31] D. Haridas, K. Sreenivas, V. Gupta, Improved response characteristics of SnO
2
thin film loaded with nanoscale catalysts for LPG detection, Sens. Actuators B
133 (2008) 270–275.
[32] O.K. Tan, W. Zhu, Q. Yan, L.B. Kong, Size effect and gas sensing characteristics of
nanocrystalline SnO
2
-(1-x)/Fe
2
O
3
ethanol sensors, Sens. Actuators B 65 (2000)
361–365.
[33] O.K. Tan, W. Cao, W. Zhu, J.W. Chai, J.S. Pan, Ethanol sensors base d on nano-
sized a-Fe
2
O
3
with SnO
2
, ZrO
2
,TiO
2
solid solutions, Sens. Actuators B 93 (2003)
396–401.

[34] A. Fuijwara, K. Ishii, H. Suematsu, H. Kataura, Y. Maniwa, S. Suzuki, Y. Achiba,
Gas adsorption in the inside and outside of single-walled carbon nanotubes,
Chem. Phys. Lett. 336 (2001) 205–211.
[35] L. Valentini, I. Armentano, L. Lozzi, S. Santucci, J.M. Kenny, Interaction of
methane with carbon nanotube thin films: role of defects and oxygen adsorp-
tion, Mater. Sci. Eng. C 24 (2004) 527–533.
[36] M. Grujicic, G. Cao, R. Singh, The effect of topological defects and oxygen adsorp-
tion on the electronic transport properties of single-walled carbon-nanotubes,
Appl. Surf. Sci. 211 (2003) 166–183.
[37] P.G. Collins, K. Bradley, M. Ishigami, A. Zettl, Extreme oxygen sensitivity of
electronic properties of carbon nanotubes, Science 187 (2000) 1801–1804.
Biographies
Nguyen Van Hieu received his MSc degree from the International Training Institute
for Materials Science (ITIMS), Hanoi University of Technology (HUT) in 1997 and PhD
degree from the Department of Electrical Engineering, University of Twente, Nether-
lands in 2004. Since 2004, he has been a research lecturer at the ITIMS. In 2007, he
worked as a post-doctoral fellow at Korea University. His current research inter-
ests include nanomaterials nanofabrications, characterizations and applications to
electronic devices, gas sensors and biosensors.
Nguyen Anh Phuc Duc received the BS degree in Engineering Physics from Institute
of Engineering in Physics, Hanoi University of Technology, Vietnam in 2005, his MSc
degree in Materials Science from the International Training Institute for Materials
Science (ITIMS), Hanoi University of Technology (HUT) in 2007, and he is currently
working toward his PhD degree at Leuven University, Belgium. His current research
interests include oxide semiconductors nanoparticle for gas-sensing applications.
Tran Trung received his MSc degree in 1994 and his PhD degree in 1998 from the
Department of Electrochemistry, Hanoi University of Technology. In 2000 and 2001,
he worked as a post-doctoral fellow in Pusan National University, Korea. At present
he is working as an Associate Professor at the Faculty of Environment and Chemistry,
Hung-Yen University of Technology and Education. His research activities are related

with the design, fabrication and characterization of organic–inorganic hybrids and
nanomaterials for application to electronic devices and battery systems.
Mai Anh Tuan received his MSc degree from the International Training Institute for
Material Science (ITIMS), Hanoi University of Technology (HUT) in 1999. In 2004,
he completed his PhD program at Universite Claude Bernard Lyon 1, France. Since
2000, he has been working as a lecturer at ITIMS, HUT. He is now biosensor group-
leader at ITIMS. His current research interests include biosensors for bio-medical
and environmental application, functional materials and IC packaging technology
(materials consideration).
Nguyen Duc Chien received the engineering degree in Electronic Engineering at
Leningrad Electrotechnical University, Russian, in 1976, and the MSc and PhD in
Microelectronics at Grenoble Polytechnique University, France, in 1985 and 1988,
respectively. He has been working as Professor at the Institute of Engineering Physics
(IEP), Hanoi University of Technology (HUT). From 1989 to 1990 he worked as a vis-
iting professor at the Grenoble University, France. From 1992 to 2006 he was a vice
director of the International Training Institute for Materials Science (ITIMS), HUT,
where he established the Laboratory of Microelectronics and Sensors. Since 2003 he
has been the Director of the IEP, HUT. His research interests include: characteriza-
tions and modeling of MOS devices, nanomaterials for chemical sensor, biosensor,
optoelectronic materials and devices, and MEMS devices. He has been the leader of
many national research projects related to microelectronic devices and functional
nanomaterials. Dr Nguyen Duc Chien is also a member of Physics Society of Vietnam
and the Vietnamese Materials Research Society.