The sensing properties of all as-fabricated sensors were investigated with different ethanol concentrations and operating temperatures.. At the region of low operating temperature below
Trang 1Mixed SnO 2 /TiO 2 included with carbon nanotubes for gas-sensing application Nguyen Van Duya, Nguyen Van Hieub,c, , Pham Thanh Huyb,c, Nguyen Duc Chienc,d,
M Thamilselvana, Junsin Yia
a
School of Information and Communication Engineering, Sungkyunkwan University, 300 Chunchun-dong, Jangan-gu, Suwon 440746, South Korea
b
International Training Institute for Materials Science (ITIMS), Hanoi University of Technology (HUT), No 1 Dai Co Viet Road, Hanoi, Vietnam
c
Hanoi Advanced School of Science and Technology (HAST), Hanoi University of Technology (HUT), Vietnam
d
Institute of Engineering Physics (IEP), Hanoi University of Technology (HUT), Vietnam
a r t i c l e i n f o
Article history:
Received 21 April 2008
Received in revised form
9 July 2008
Accepted 9 July 2008
Available online 25 September 2008
PACS:
61.48.De
07.07.df
81.07.De
Keywords:
Mixed SnO2–TiO2
Carbon nanotubes
Gas sensor
a b s t r a c t
TiO2and SnO2are the well-known sensing materials with a good thermal stability of the former and a high sensitivity of the latter Carbon nanotubes (CNTs) have also gas sensing ability at room temperature CNTs-included SnO2/TiO2material was a new exploration to combine the advantages of three kinds of materials for gas-sensing property In this work, a uniform SnO2/TiO2 solution was prepared by the sol–gel process with the ratio 3:7 in mole The CNTs with contents in the range of 0.001–0.5 wt% were dispersed in a mixed SnO2/TiO2matrix by using an immersion-probe ultrasonic The SnO2–TiO2and the CNTs-included SnO2–TiO2thin films were fabricated by the sol–gel spin-coating method over Pt-interdigitated electrode for gas-sensor device fabrication and they were heat treated at
500 1C for 30 min
FE-SEM and XRD characterizations indicated that the inclusion of CNTs did not affect the particle size as well as the morphology of the thin film The sensing properties of all as-fabricated sensors were investigated with different ethanol concentrations and operating temperatures An interesting sensing characteristic of mixed SnO2/TiO2sensors was that there was a two-peak shape in the sensitivity versus operating temperature curve At the region of low operating temperature (below 280 1C), the hybrid sensors show improvement of sensing property This result gives a prospect of the stable gas sensors with working temperatures below 250 1C
&2008 Elsevier B.V All rights reserved
1 Introduction
Semiconductor metal oxide gas sensors have been investigated
extensively since the past decades owing to their advantages of
high sensitivity to pollutant gases, fast response and recovery, low
cost, easy implementation, and small size[1,2] Gas sensors based
on SnO2materials have been commercially available[3,4]
Thin-film gas sensors have improved the gas-sensing properties from
bulk or thick- film ones They not only give a high sensitivity but
also have very fast response and recovery times However, there
still exist great disadvantages of SnO2and TiO2materials SnO2is
thermally unstable and its electrical properties can be
degener-ated upon prolonged thermal treatment in reducing the gas
atmosphere[1] On the other hand, in spite of the high thermal
and chemical stability, the gas sensors based on TiO2 materials
require high operating temperatures (normally up to 400 1C) This
would result in high power consumption and difficulty of packaging Mixed oxide has been studied to combine the advantages of the sensing property of each oxide component
[4–6] The formation of mixed oxide is classified into three types
as follows:
(1) Chemical compound
(2) Solid solution
(3) Mix of (1) and (2) types
SnO2–TiO2 falls into the second category The use of mixed oxides in gas detection has been tried successfully in some systems such as SnO2–WO3[5], TiO2–WO3[5–7], TiO2–SnO2[5,8] Among these mixed oxides, the SnO2–TiO2 system has been investigated more extensively for gas-sensing applications
[5,8–12] Carbon nanotubes (CNTs) have been the most actively studied material in recent years due to their unique electrical, mechanical and chemical properties, and much attention has been paid to their application in various fields of nanotechnology [13,14] Moreover, they have nanoscale size and large surface area that can
Contents lists available atScienceDirect
journal homepage:www.elsevier.com/locate/physe
Physica E
1386-9477/$ - see front matter & 2008 Elsevier B.V All rights reserved.
Corresponding author at: International Training Institute for Materials Science
(ITIMS), Hanoi University of Technology (HUT), No 1 Dai Co Viet Road, Hanoi,
Vietnam Tel.: +84 4 8680787; fax: +84 4 8692963.
E-mail addresses: hieu@itims.edu.vn , hieunv-itims@mail.hut.edu.vn
(N Van Hieu)
Trang 2provide excellent gas absorption properties These extreme
absorption properties make CNTs advantageous for use in many
areas of applications For example, the gas absorption of CNTs at
room temperature will change its electric properties with fast
response time, which can enable CNTs to be a good candidate of
gas-sensing applications[15–17] For the gas-sensing materials,
there are various approaches using CNTs as the solution such as
CNTs for dispersion, CNTs for composite, CNTs for filling, CNTs for
coating, etc[18–26] It has been recently reported in the literature
that single-wall CNTs (SWCNTs) doping on SnO2can significantly
improve SnO2gas-sensor performance, and especially the sensor
can function at room temperature with sufficient sensitivity[27]
Some other endeavors on including CNTs into tungsten tri-oxide
(WO3) [25], polymethylmethacrylate [28], polypyrrole[29], etc
have been published In our previous work, we have demonstrated
the improvement of performance of the TiO2-based sensor by
including CNTs[30]and high performance of the room-temperature
NH3 gas sensor by using SnO2/CNTs composites [31] In this
work, we explore possibilities to improve the performance and to
reduce the operating temperature of the SnO2–TiO2ethanol sensors
by adding CNTs
2 Experimental
2.1 Materials synthesis and characterizations
SnO2–TiO2 sol was prepared by the sol–gel method The
precursors used to fabricate the solutions were tetra propylortho
titanate Ti(OC3H7)4(99%), tin ethylhexanoate Sn(OOCC7H15)2and
isopropanol C3H7OH (99.5%) To synthesize the hybrid SWCNT/
SnO2–TiO2material, the SWCNTs with a diameter lower than 2 nm
and multi-wall CNTs (MWCNTs) with diameters ranging from 20
to 40 nm purchased from Shenzhen Nanotech Port Ltd Co
(Shenzhen China) were introduced in the SnO2–TiO2sol solution
by an ultrasonic shaker at a power of 100 W for 10 min The CNTs
content was varied in the range of 0.001–0.5 wt% The film was
deposited by spin coating on silica substrate at 4000 rpm for 20 s
and a film thickness of around 320 nm was obtained The sensors
realized with different SWCNTs contents were signed as S0–S7
Meanwhile, the sensors with various MWCNTs contents were
signed as M0–M7 As-deposited films were dried for 30 min at
60 1C and then they were annealed at 500 1C for 30 min The
morphology and the crystalline phase of the films were
char-acterized using a field emission scanning electron microscope
(FE-SEM; 4800 Hitachi, Japan) The microstructure of the sintered film
was characterized by X-ray diffraction (XRD), using a Bruker-AXS
D5005
2.2 Gas sensor fabrication and testing
The fabrication of the gas sensor was carried out in the
following manner: (i) the inter-digitated electrode was fabricated
using a conventional photolithographic method with a finger
width of 100mm and a gap size of 70mm The fingers of the
inter-digitated electrode were made by sputtering 10 nm Ti and 200 nm
Pt on a layer of silicon dioxide (SiO2) with a thickness of about
100 nm thermally grown on top of a silicon wafer; (ii) the sensing
layers were deposited on top of the electrode with subsequent
heat treatment at 500 1C for 30 min
The sensor under test was placed on top of a hot plate and held
by two tungsten needles Then they were loaded in a glass
chamber with a volume of 4 L as shown in Fig 1 More details
about the measurement set-up can be found elsewhere[31] The
desired ethanol concentrations were obtained by mixing ethanol
gas with air using a mass flow control system with computer control (AALBORG model GFC17S-VALD2-A0200) and subse-quently injected into the chamber The chamber was purged with air and the experiment was repeated The electrical resistance response during testing was monitored by a precision semicon-ductor parameter analyzer HP4156A, which can be used to detect
a very low electrical current (around 1012A) This allows us to measure the high resistance of the mixed oxide films The resistance responses of the sensor in air ambient and upon exposure of ethanol pulses were monitored The sensor response (S) was defined as the ratio of the sensor resistance in air (Ra) and
in ethanol gas (Rg)
3 Results and discussion
3.1 Microstructure characterizations
The formation of the SnO2–TiO2solid solution can be seen by
an XRD pattern inFig 2 With the mole ratio of SnO2:TiO2at 3:7, it shows that the diffraction peaks of the oxide solution follow Vegard’s law A similar result has been seen for SnO2–TiO2
deposited by sol–gel[11]and sputtering methods[5] The solution
is formed by mixing SnO2and TiO2lattices in the rutile phase in which both the materials are in the tetragonal structure From XRD peaks, we get the inter-planar spacing values of SnO2–TiO2 -mixed oxide as shown inTable 1 The peak shift is explained by the substitution of Sn4+for Ti4+in the TiO2crystal structure Because
of the larger radii of Sn4+, the lattice spacing increases when the substitution occurs In the sol–gel process, the chemical reaction controlled at low speed gives the possibility of a homogenous mixed solution Sn–O and Ti–O bonding disperse uniformly during stirring and hydrolysis reaction From the peak broadening, the crystallite size estimated by the Scherrer equation was found to be about 5.5 nm XRD was carried out with the highest SWCNTs content of 0.5 wt% (sample S7); it is understandable that the SWCNTs peaks were not detected in the XRD pattern
FE-SEM images show the surface morphology of the thin films after heat treatment They exhibit that the particle size is around
10 nm These results may be caused by the impeding of the polycrystalline aggregate process of each other SnO2 and TiO2 This grain size is approximately two times the Debye length for the depletion layer on the surface It implies that the surface-sensing mechanism is more effective in these films Another result
is that all the films’ surfaces are highly porous and uniform in granular shape The high porosity of the thin film makes it more easy to adsorb and desorb gas molecules All these characteristics promise good gas sensing properties of the material The FE-SEM
H P 4 1 5 6 A Delta Electronic
ES30-5 Power Supply
Exhaust
Target Gas
MFC
Mass Flow Controller
Fig 1 Apparatus for gas-sensor testing.
Trang 3images of S0 and S4 samples as shown inFig 3a and b indicate
that the film morphology was not clearly different between the
undoped and the SWCNTs-doped samples CNTs trace cannot be
seen in the FE-SEM image of 0.1 wt% SWCNTs/SnO2–TiO2(S6) after
annealing at 500 1C for 30 min We suggest that at low content of
CNTs, they are embedded in the oxide matrix In addition,
SWCNTs–TiO2 and SWCNTs–SnO2 bondings can be formed
naturally through some physicochemical interactions such as
Van der Waals force, H bonding and other bondings The
interaction between –OH groups in the course of the hydrolysis
reaction of Sn(OC7H15)2, Ti(OC3H7)4and –COOH, –OH groups on
SWCNTs formed by the purification process can be a case for
explanation This indicated that the crystallites would grow up
and enclose SWCNTs during the heat treatment Therefore, it is
very difficult to find CNTs on the film surface In general, the trace
of CNTs on the film surface could be seen in the composite
material in which the CNTs’ content would normally be higher
than 5 wt%
3.2 Ethanol sensing properties
We have measured the responses of all sensors to ethanol at
different concentrations ranging from 125 to 1000 ppm and at
operating temperature in a range from 210 to 400 1C to investigate
the gas-sensing properties The sensor responses at various
operating temperatures are shown in Fig 4 It was found that
the response and recovery times of the sensors are less than 10 s
We have observed that the metal oxide thin-film sensor show a
relatively low response-recovery time, and the hybrid CNTs/metal oxide thin-film sensor show even lower values This observation was also previously reported[20,24,26,32]
Table 1
The interplanar spacing values of SnO2–TiO2 mixed oxide calculated by Vegard’s
law are close to the measured values
d(11 0) (A˚) d(1 0 1) (A˚) d(2 0 0) (A˚) d(2 11) (A˚)
Calculation with
Vegard law
Fig 3 FESEM images depict the uniform and highly porous surface of blank (a) and hybrid (b) 0.1% SWCNTs/SnO2–TiO2 samples.
1M
10M 100M
1G
Air Air
500 ppm
1000 ppm
375 ppm
250 ppm
125 ppm
305°C 335°C 365°C 400°C
t (s)
Fig 4 Ethanol response characteristics of sensor S4 at different temperatures show fast response and recovery times less than 10 s.
2-Theta - Scale
Fig 2 X-ray diffraction pattern of SnO2–TiO2 (at ratio 3:7 in mole) shows the
diffraction peaks of solid solution following Vegard’s law Dot lines indicate SnO2
rutile peaks and dash lines indicate TiO2 rutile peaks.
Trang 4The stepwise decrease in electrical resistance obtained with
increasing ethanol concentration from air to 1000 ppm ethanol
gas in air, and after several cycles of the gas injection, the
resistance turns back to the original value when the sensor is
exposed to air These characteristics indicated that the hybrid
sensor has relatively stable response However, the high resistance
of around 109Oat an operating temperature below 300 1C is a
drawback of the hybrid material
Working temperature is one of the most important parameters
for gas sensors The conventional gas sensors based on SnO2and
TiO2 materials operate at the temperature region from 300 to
400 1C The response versus operating temperature (S–T) curves of
our sensors at 1000 ppm ethanol depict the two-peak shape
characteristic The first maximum in response appears at an
operating temperature of around 260 1C and the second peak is
around 380 1C This can be seen clearly with S1 and S4 sensors
based on the SWCNTs/SnO2–TiO2material, as shown inFig 5 For
the SWCNTs content of 0.001 wt% (S1) and 0.025 wt% (S4), we get
the response to 1000 ppm ethanol of 11.1 and 9.6 at operating
temperature of 260 1C, 32 and 41 at an operating temperature of
380 1C, respectively Meanwhile, at higher content of SWCNTs,
there is a strong degradation in the response This observation
cannot be clearly explained yet A plausible explanation for the
observed effect is that the addition of SWCNTs considerably increases the surface adsorption area of the mixed oxide and added more p/n junction of SWCNTs/SnO2–TiO2 as discussed below However, when the CNT content is sufficiently high, the SWCNTs begin to connect together and results in a shorter resistance path that shunts the gas-sensing current of the mixed oxide layer Thus, the gas sensitivity is reduced for a very high SWCNT content
The dependence of the response on ethanol concentration at operating temperatures of 260 and 380 1C is given inFig 6 It can
be seen that all the sensors present more or less linear characteristic in the investigated range from 125 to 1000 ppm ethanol, which makes their use more convenient Once again, S1 and S4 dedicate the best in slope than the others The slope values
of fit lines are given in Table 2 We have also surveyed the
200
0
5
10
15
20
25
30
35
40
45
/R ethanol
S0
S1
S3
S4
S7
0
5
10
15
20
25
30
35
T ( o C)
/R Ethanol
M0
M2
M3
M4
M7
420 400 380 360
T (°C)
340 320 300 280 260 240 220
200 220 240 260 280 300 320 340 360 380 400 420
Fig 5 The dependence of response on operating temperature depicts the two
maximum characteristics on both SWCNTs/SnO2–TiO2 (a) and MWCNTs/SnO2–TiO2
(b) systems The first peak is around 260 1C and the second one is around
5 10 15 20 25 30 35 40
45
S0 S1 S3 S4 S7
/R ethanol
0 2 4 6 8 10
(R Air /R Ethanol
S0 S1 S3 S4 S7
1000 800
600 400
200
Fig 6 Response versus on ethanol concentration characteristics in the range from
125 to 1000 ppm at operating temperatures of 240 (a) and 380 1C (b).
Table 2 Fitting slope of S–C curves at operating temperatures of 240 and 380 1C
At 240 1C (/100 ppm) 0.28 0.69 0.34 0.20 0.84 0.38 0.27 0.24
At 380 1C (/100 ppm) 1.91 3.78 2.12 2.78 2.74 2.07 2.07 1.09
Trang 5influence MWCNTs inclusion on the sensing properties of
the mixed oxide material The sensing properties of this hybrid
material are quite similar to that of SWCNTs/SnO2–TiO2 (see
Fig 5b) From the response values at operating temperatures of
240 and 380 1C given in Table 3, we can see that the best
improvement in ethanol sensing is obtained for 0.01 and
0.025 wt% MWCNTs However, the effect of MWCNTs on the
ethanol sensing property of mixed oxide is not as high as SWCNTs
To summarize all the results, we plotted the maximum
sensitivities versus the CNTs doping content, as seen inFig 7 It
is easy to see how better when the CNTs-doped mixed SnO2–TiO2
sensors are working at the low temperature The best
improve-ment for operating temperatures of 380 and 260 1C is achieved at
SWCNTs contents of 0.001% and 0.025%, respectively These
observations are the same as in the case of MWCNTs inclusion
These results of the sensing properties at a working temperature
below 250 1C even give ethanol detectability that is 20–25 times
smaller compared to the CNTs/SnO2composite sensor prepared by
electron beam evaporation[26]
3.3 Gas sensing mechanism
At first, one needs to discuss the two-peak shape of
response-operating temperature curves For the sensors based on tin oxide
and titanium oxide, such results have never been seen before We
assumed that the presence of both SnO2 and TiO2 makes the
mixed oxide material with combined properties At the operating
temperature below 500 1C, the surface sensing mechanism plays a
dominant role Ethanol vapor adsorbs on the surface grain
boundaries and reacts with the adsorbed oxygen ions on the surface It should be noted that the adsorbed oxygen ions trap electrons, inducing a surface depletion layer between the grains This means the surface density of the negatively charged oxygen decreases by the ethanol vapor absorption, so the barrier height in the grain boundary is reduced The reduced barrier height decreases sensor resistance We propose that these processes take place more easily for SnO2 than for TiO2 due to the lower working temperature of SnO2[9] The presence of both SnO2and TiO2 has two effective working temperature regions At an operating temperature of around 250 1C, the sensing properties
of the mixed oxide are due to SnO2, while TiO2is more sensitive at
a temperature around 380 1C
As described in the previous section, the CNTs inclusion has caused no obvious differences in surface morphology as well as particle size Consequently, the porosity and particle size cannot result in a remarkable improvement of the hybrid CNTs/SnO2 –-TiO2gas-sensor performance The improvement of the SnO2–TiO2
gas-sensor performance by including SWCNTs has not been well understood so far and not much work has been published on the subject The model proposed by Wei et al [27] seems to be reasonable for the explanation This model was applied for SWCNTs-doped SnO2 and somehow we can apply for our case The model has been hypothesized that CNTs-doped SnO2–TiO2
materials can build up p/n hetero junctions, which was formed by (n-oxide)/(p-CNT)/(n-oxide) Fig 8 schematically depicts the changes in the electronic energy bands for two depletion layers, one is on the surface of mixed-oxide particles and the other is at the interface between CNT and mixed oxide When the mixed oxide is exposed to ethanol gas, the gas molecules will react with oxygen ions previously adsorbed on the surface of mixed oxide This can simply be described as[33]
2C2H5OH þ O
2 ¼2CH3CHOþþ2H2O þ e The electrons released from the surface reaction transfer back into the conductance bands, which increase the conductivity of
1E-3 0
5
10
15
20
25
30
35
40
45
50
CNTs content (%)
SWCNTs, T=240-260°C SWCNTs, T=360-880°C MWCNTs, T=240-260°C MWCNTs, T=360-880°C
/R ethanol
0.5 0.05
0.01
Fig 7 Maximum response of two sensor systems at low and high operating
temperature regions, ethanol concentration of 1000 ppm.
CNT
CNT
n
d 1
d 2
d 3
d 4
n
E e
E f
E v
p
Depletion layer
Distance
Grain boundary
TiO2/SnO2
CH3/CHO
CH3/CHO
CH3/CHO
CH3/CHO
TiO2/SnO2 TiO2/SnO2
O2
O2
O2
O
Fig 8 Schematic of potential barriers to electronic conduction at grain boundaries and at p–n heterojunctions for CNTs/mixed oxide; d1 and d3 are depletion layer widths when exposed to ethanol; d2 and d4 are depletion layer widths in air.
Table 3
Two maximum values in response of MWCNTs/SnO2–TiO2 to 1000 ppm ethanol:
Sm1, Sm2
Trang 6the sensing material It is noted that the adsorption of the ethanol
gas may change the two depletions layers as described above
Before the ethanol gas is adsorbed, the widths of the depletion
layers at the interface between mixed oxide grains and mixed
oxide/CNT are given as d2 and d4, respectively After the
adsorption, the widths of these depletion layers are d1 and d3,
respectively The change in both the depletion layers at the oxide
grain boundaries and the n/p junction contributed to the
improved sensitivity of the sensing materials In other words,
n-type mixed oxide and p-type CNT form a heterostructure Like
the working principle of an n–p–n amplifier, the CNT works as a
base, blocked electrons transfer from n (emitter) to n (collector),
and thus lowering the barrier a little bit allows a large amount of
electrons to pass from the emitter to the collector [24] This
amplification effect may explain the fact that the hybrid materials
(SnO2/SWCNTs) can detect NO2at room temperature[27] So the
improvement of the gas sensor performance and the shift of
operation temperature toward the lower temperature region in
our work can be attributed to the amplification effect of the p–n
junctions in addition to the effect of the grain boundaries
Meanwhile, the fact that the contribution of MWCNTs
(20odo40 nm) is not as much as SWCNTs (do2 nm) can be
explained based on the quantum effect as follows The space
charge layer thickness (Debye length) is around 3 nm for the metal
oxides (for example SnO2) So the largest distance between
adjacent boundaries accessing gas molecules should be less than
6 nm [34] However, mixed oxide (SnO2/TiO2) grains are much
larger than 6 nm so that not all metal oxides can participate in the
reaction when gas absorbs on it Therefore, the mixed-oxide/
SWCNT material structure formed by inclusion of the SWCNTs
with diameter lower than 2 nm will produce quantum effects
between SWCNTs and mixed oxide nanoparticles The SWCNTs
with a diameter of o2 nm reduce the distance between two
adjacent gas-assessing and reaction surface to be less than the
space charge layer thickness
4 Conclusion
SnO2–TiO2 mixed oxide has been studied at the ratio 3:7 in
mole for ethanol-sensing properties At appropriate annealing
conditions, it has shown the formation of the solid solution from
two components by the XRD pattern All the film surfaces were
uniform and highly porous In addition, the grain size around
10 nm gave a high specific surface The new explorer in the
two-peak shape of the response versus operating temperature
characteristics has proved the combined behavior of the
mixed-oxide material SnO2 and TiO2are complementary to each other
for gas-sensing properties The inclusion of CNTs at specific
contents into the mixed oxide system improved the response of
the sensor in the low operating temperature region Further
studies on this type of material would make it a promising
candidate for gas sensing application that can work at around
250 1C with a high stability
Acknowledgements
This work was financially supported by HAST Project no 01 The authors also acknowledge Grant no 405006 (2006) from the Basic Research Program of the Ministry of Science and Technology (MOST) and for the financial support from Third Italian-Vietnamese Executive Programme of Co-operation in S&T for 2006–2008 under the project ‘‘Synthesis and Processing of Nanomaterials for Sensing, Optoelectronics and Photonic Applications’’
References
[1] J Puigcorbe, A Cirera, J Cerda, J Folch, A Cornet, J.R Morante, Sens Actuators
B 84 (2002) 60.
[2] G Eranna, B.C Joshi, D.P Runthala, R.P Gupta, Crit Rev Solid State Mater Sci.
29 (2004) 111.
[3] Figaro Eng Inc / http://www.figarosensor.com S.
[4] Microsens Inc / http://www.microsens.ch S.
[5] K Zakrzewska, Thin Solid Films 391 (2001) 229.
[6] P Nelli, L.E Depero, M Ferroni, S Groppelli, V Guidi, F Ronconi, L Sangaletti,
G Sberveglieri, Sens Actuators B 31 (1996) 89.
[7] G.N Chaudhari, A.M Bende, A.B Bodade, S.S Patil, V.S Sapkal, Sens Actuators
B 115 (2006) 297.
[8] K Zakrzewska, M Radecka, Thin Solid Films 515 (2007) 8332.
[9] H.-C Lee, W.-S Hwang, Appl Surf Sci 253 (2006) 1889.
[10] R.-J Wu, C.-Y Chen, M.-H Chen, Y.-L Sun, Sens Actuators B 123 (2007) 1077 [11] M Radecka, K Zakrzewska, M Rekas, Sens Actuators B 47 (1998) 194 [12] C.M Carney, S.Y Sheikh, A Akbar, Sens Actuators B 108 (2005) 29 [13] T.W Ebbesen, Annu Rev Mater Sci 24 (1994) 235.
[14] R.H Baughman, A.A Zakhidov, W.A de Heer, Science 297 (2002) 787 [15] A Modi, N Koratkar, E Lass, B Wei, Nature 424 (2003) 171.
[16] T Someya, J Small, P Kim, C Nuckolls, J.T Yardley, Nano Lett 3 (2003) 877 [17] J Li, Y Lu, Q Ye, M Cinke, J Han, M Meyyappan, Nano Lett 3 (2003) 929 [18] Y.-L Liu, H.-F Yang, Y Yang, Z.-M Liu, G.-L Shen, R.-Q Yu, Thin Solid Films
497 (2006) 355.
[19] L Zhao, L Gao, Carbon 42 (2004) 3251.
[20] Y.X Liang, Y.J Chen, T.H Wang, Appl Phys Lett 85 (2005) 666.
[21] A Yang, X Tao, R Wang, S Lee, C Surya, Appl Phys Lett 91 (2007) 133110 [22] E.H Espinosa, R Ionescu, B Chambon, G Bedis, E Sotter, C Bittencourt,
A Felten, J.J Pireaux, E Lolbet, Sens Actuators B 127 (2007) 137.
[23] R Ionescu, E.H Espinosa, R Leghrib, A Felten, J.J Pireaux, R Erni, G.V Tendeloo, C Bittencourt, N Canellas, E Llobet, Sens Actuators B 131 (2008) 174.
[24] J Gong, J Sun, Q Chen, Sens Actuators B 130 (2008) 829.
[25] C Bittencourt, A Felten, E.H Espinosa, R Ionescu, E Llobet, X Correig, J.-J Pireaux, Sens Actuators B 115 (2006) 33.
[26] A Wisitsoraat, A Tuantranont, C Thanachayanont, V Patthanasettakul,
P Singjai, J Electroceram 17 (2006) 45.
[27] B.-Y Wei, M.-C Hsu, P.-G Su, H.-M Lin, R.-J Wu, H.-J Lai, Sens Actuators B
101 (2004) 81.
[28] B Philip, J.K Abraham, A Chandrasekhar, Smart Mater Struct 12 (2003) 935 [29] K.H An, S.Y Jeong, H.R Hwang, J.H Lee, Adv Mater 16 (2004) 1005 [30] N Van Hieu, N Van Duy, N.D Chien, Physica E: Low-Dimensional Syst Nanostruct 40 (2008) 2950.
[31] N Van Hieu, L.T.B Thuy, N.D Chien, Sens Actuators B 129 (2008) 888 [32] H.C Wang, Y Li, M.J Yang, Sens Actuators B 119 (2006) 380.
[33] H Idriss, E.G Seebauer, J Mol Catal A: Chem 152 (2000) 201.
[34] Noboru Yamazoe, Sens Actuators B 5 (1991) 7.