Synthesis, characterization and gas sensitivity of MoO3 nanoparticles

Abstract

Nanoparticles of molybdenum oxide were synthesized using the citrate sol-gel method and characterized using scanning electron
microscopy and X-ray diffraction techniques. The sensitivity of the material to the presence of various gases was analysed
and the particles showed higher sensitivity towards NO2 gas.

Full text

Page 1

Bull. Mater. Sci., Vol. 30, No. 2, April 2007, pp. 183–185. © Indian Academy of Sciences.

183
Synthesis, characterization and gas sensitivity of MoO3 nanoparticles
ARNAB GANGULY and RAJI GEORGE*
M.S. Ramaiah Institute of Technology, MSRIT Post, Bangalore 560 054, India
MS received 5 February 2007; revised 13 March 2007
Abstract. Nanoparticles of molybdenum oxide were synthesized using the citrate sol–gel method and charac-
terized using scanning electron microscopy and X-ray diffraction techniques. The sensitivity of the material to
the presence of various gases was analysed and the particles showed higher sensitivity towards NO2 gas.

Keywords. Sol–gel citrate; metal oxide; gas sensing; nanoparticles; SEM.
1. Introduction
High sensitivity gas sensing assumes great importance in
view of the plethora of health hazards related to poisonous
gases that are associated with industrial and automobile
exhausts. Thus, the requirement of gas sensors to sense such
harmful gases is indubitable and with this requirement
comes the demand for highly sensitive detection systems
that detect trace levels of highly harmful gases.
One approach is the use of advanced materials techno-
logies. The sensitivity of the sensing device is directly
proportional to the surface area to the volume ratio of the
exposed sensing surface. Hence, nano sized particles used
as sensors would have the maximum ratio and provide the
best possible response. For example, Rella et al (1999)
demonstrated that by maintaining the grain size of SnO2
particles in the 10–100 nm range increased their sensitivity
towards CO. Ferroni et al (1999) demonstrated good re-
sponse in the use of solid solutions of TiO2 and WO3
when the grain size was maintained at 60 nm. Chung et al
(1999) showed that by increasing the firing temperature
the response of WO3 sensors to NOx reduced. Rossinyol
et al (2005) demonstrated the advantageous use of nano-
structured cerium oxide and tungsten oxide in gas sensing.
A number of nano structured materials are known to have
been used as metal oxide gas sensors amongst which the
following are a few: ZnO, Fe2O3, WO3, SnO2, ZrO2,
SrTiO3 and the like (Hoffheins et al 1996). The sensing
properties of resistive type of sensors were found to in-
crease by reducing the size of the oxide particles (Gouma
2003). Nanocrystalline oxide-based sensors were found to
be highly sensitive towards gaseous species that may
have evolved from various chemical processes. It was thus
thought pertinent to study the system of nanostructured
molybdenum oxide in a continuous effort of improving
gas sensitivity to NO2.
This paper involves the synthesis of nano structured
molybdenum oxide particles using the citrate sol–gel method.
The characterization of the samples prepared has been done
with the help of an SEM micrograph and X-ray diffraction
techniques. The relative sensitivity of the system obtained
has been tested for the passage of different gases.
2. Experimental
Molybdenum oxide nanoparticles were synthesized using
the citrate sol–gel method. Ammonium molybdate powder
(1⋅16 g) was dissolved in de-ionized water to which was
added citric acid crystals (0⋅38 g). The mixture was then
stirred carefully using a magnetic stirrer while ammonium
hydroxide was added to obtain a pH of 7. The mixture
was then heated in a furnace to a temperature of 250°C for
1 h. Initially a zero gel and finally a powder was obtained.
The powder was then heated to a temperature of 500°C
for 90 min to obtain a pale yellow powder.
To carry out the characterization, the prepared sample
was sonicated and then gold coated. MoO3 synthesized by
the sol–gel citration method was characterized using an SEM
(JEOL JSM 5600LV instrument) and X-ray analysis done
on a Philips X-ray diffractometer.
For carrying out the gas sensitivity studies, synthesized
MoO3 powder was converted to a paste using ethanol and
mixed thoroughly. The prepared paste was then applied onto
the micro-fabricated alumina substrate. The paste on the
substrate was then allowed to dry. The gas sensitivity test
was conducted by passing gas through a multi gas channel
controller MKS 1251C with appropriate electrical circuitry
to support it.
The 1251C multi gas controller from MKS system was
used for the sensitivity/cross sensitivity tests for the sen-
sor material. The sensor was exposed to N2 + O2 (80% +
20%) for 6 h at 300°C for curing. Different gases of desired
concentrations were opened at different times to check
the sensor material’s sensitivity towards each of them.

*Author for correspondence (rgeorgemsrit@rediffmail.com)

Page 2

Arnab Ganguly and Raji George

184
The base gas was 99⋅999% ultra pure N2 and the test
temperature was 400°C. Before gas testing, the settings in
the gas controller were programmed for all the gases for
desired time period so that there is no human error during
the test. The sequence of gas flow was N2, NO, N2, CO, N2,
O2 (each gas for 5 min). The electronics to amplify and
signal processing was a simple impedance match circuit
and was set for gain ‘4’ for all the gas adsorption signal
outputs.
The impedance (resistance) of the sensor changed drasti-
cally as the NO and NO2 gases were ‘ON’, whereas the
change was minimal for CO and O2 gases.
3. Results and discussion
3.1a Scanning electron microscopy: The SEM charac-
terization was conducted using a JEOL JSM 5600 LV
instrument. The type of electrons used was of the secondary
type. Fairly spherical nano-crystallites were observed in
the micrographs. It was also observed that the synthesis
carried out at temperatures above 750°C led to agglome-
rated particles being formed while the synthesis carried
out at lower temperatures (till 500°C) but for a longer
duration yielded better results. The samples prepared
have been found to be in the 10–100 nanometer range. The
particle size and shape analysis were carried out by Quan-
timet software from Leica, UK (Cambridge instruments).
Figure 1 shows SEM micrograph of the oxide sample
prepared using the above illustrated process.

3.1b X-ray diffraction: X-ray analysis of the oxide parti-
cles were performed using a Philips X-ray diffractometer.
Figure 2 represents the XRD patterns for MoO3 powders.
The wavelength used for the XRD analysis was CuKα,
–1⋅50546 Å. After the synthesis, it was found that the
peaks corresponding to (0 2 0), (1 1 0), (0 4 0), (0 2 1) are
of orthorhombic crystal structure of MoO3. It is noted



Figure 1. SEM micrograph of the oxide sample prepared.
that all the XRD peaks are identified as MoO3 peaks from
the JCPDS card 35-0609.
Figure 2 shows the XRD pattern for the sample pre-
pared. The peaks correspond to the orthorhombic crystal
structure of MoO3.

3.1c Gas sensitivity: Micro-fabricated sensor substrates
consisting of 1 × 1 cm alumina with inter-digitized elec-
trodes deposited on one side of them were used for gas
detection and the set up consists of 2 electrodes and 2
heaters connected across the substrate-coating. The passage
of gases was opened so that it passed through the sample.
The time vs voltage plot shown in figure 3 is used in deter-
mining the gas sensitivity.
A characteristic graph of sensitivity would consist of a
build up and a recovery phase as the channel is shut off.
The heater voltage is set at 7⋅5 V and the input voltage at
5 V. Gas sensing studies were conducted by passing the gases
in the following order: N2, NO2, N2, NO, N2, CO, N2, O2,
N2. The plot clearly indicates that the sample is not sensi-
tive to N2 gas (region 1 in figure 3), but as soon as NO2
gas is passed, there is a deflection (region 2) in the volt-
age–time plot till the saturation is reached (point 3). This
indicates that the sample is sensitive to the presence of
NO2 gas. Following this, the gas supply is shut off and
the recovery phase is initiated (region 4). The recovery
phase is followed by the sensing of NO gas (region 5).
However, the sensitivity to NO is much lesser, as indi-
cated by a smaller drop in the plot till the saturation is
reached (point 6). Similarly the next recovery phase is
followed by the sensing of CO gas till the saturation point
(point 7). Thus the sensitivity is again much lower, when
compared to the sensitivity to NO2 gas (point 3). It is,
therefore, evident that the sensitivity of the sample is more
towards NO2 than towards CO or N2 gases. It is very impor-
tant to observe the cross-sensitivity of the sensor to other
gases like CO, CO2, O2. There was a flow of these cross
sensitive gases for 5 min with 500 ppm concentration and
the graph clearly indicates that the response was flat. Though
the recovery of the sensor for the gases was relatively



Figure 2. XRD pattern for the sample prepared.

Page 3

Synthesis, characterization and gas sensitivity of MoO3 nanoparticles

185

Figure 3. Sensitivity plot for various gases.

slow, the problem was overcome by using heater control
techniques which switches the sensor temperatures to
higher values for 15 s. It may also be noted that the sensiti-
vity of gases is in the order of the thermal and electrical
conductivity of the gases: NO2 > NO > CO > N2.
The underlying principle behind this behaviour is prima-
rily the surface reaction that takes place between the gas
molecules and the material sensing it. As the size of the
material particles used for sensing tends towards nanometer
scale, the particles can be considered to be as a continu-
ous surface in three dimensions (Somarjai 1998) and this
in turn increases the surface reactivity due to the higher
surface area. The increase in reactivity can also be attri-
buted to the fact that there is an increase in the number of
defect sites which are responsible for the reactions taking
place on the surface. Oxygen ions are adsorbed onto the
surface, removing electrons from the bulk and creating a
potential barrier, this process is responsible for the varied
conductivity in the initial stages. When the gases to be
sensed are passed over the surface, they combine with this
oxygen, thus reducing the potential barrier created and
hence change the conductivity. This change in the con-
ductivity is a function of the characteristic properties of the
gases and is indicative of the presence of gas being sensed.
The gas detection by chemical sensors based on semi-
conducting materials is due to the variation in conductivity.
This is induced by the adsorption of gases on the semi-
conductor surface as mentioned earlier. When oxygen is
adsorbed on the semiconductor surface, negatively charged
oxygen ions are formed (Morrison 1994; Henrich and
Cox 1993). This can be represented as
O2 + 2e– → 2O–.
The oxygen ionosorption causes electron transfer from
the surface of the grain towards the adsorbed species
thus leading to the formation of an electron depleted sur-
face. This varies the electrical conductivity of the semicon-
ductor.
When a gas such as CO is adsorbed in the presence of
ionosorbed oxygen, a similar variation in conductivity is
seen and the reaction can be represented by
2CO + O–
2 → 2CO2 + e–,
CO + O–
2 → CO2 + e–.
Figure 3 shows sensitivity data for the sol–gel based
MoO3 powder for different gases. The plot indicates that the
sample has higher sensitivity towards NO2 when compared
to CO or other gases.
4. Conclusions
The present work provides a method for processing nano-
structured MoO3 with enhanced gas sensing capabilities.
Nano-sized MoO3 was synthesized using the citrate sol–gel
method and the same was characterized using electron
microscopy and X-ray diffraction techniques. The sample
was then subjected to gas sensitivity tests using a multi
gas channel controller. The sample showed exemplary
sensitivity towards NO2 gas while exhibiting lower sensiti-
vity towards gases like NO, CO etc.
Acknowledgement
The authors would like to thank Dr A R Raju, Sanjeeb
Tripathy and Dr Somnath Ganguly for their continuous
help and guidance during the study. The authors are grateful
to Prof. D D Sarma and Prof. T N Guru Row, Indian In-
stitute of Science, Bangalore, for providing laboratory
facilities. In addition, the authors gratefully acknowledge
the support provided by the M.S. Ramaiah Institute of
Technology, Bangalore.
References
Chung Y K et al 1999 Sensors and Actuators B60 49
Ferroni M et al 1999 Sensors and Actuators B58 289
Gouma P I 2003 Rev. Adv. Mater. Sci. 5 123
Henrich V E and Cox P A 1993 Appl. Surf. Sci. 72 277
Hoffheins B, Taylor R F and Schultz J S (eds) 1996 Solid state
resistive gas sensors, Handbook of chemical and biological
sensors (Philadelphia: Institute of Physics)
Morrison R S 1994 in Semiconductors sensors (ed.) S M Sze
(New York: John Wiley) p. 383
Rella R et al 1999 Sensors and Actuators B58 283
Rossinyol E et al 2005 Sensors and Actuators B109 57
Somarjai G A 1998 MRS Bull. 23