Powder Technology 107 2000.

268272
www.elsevier.comrlocaterpowtec

Synthesis of high surface area tin oxide powders via water-in-oil

microemulsions
Ki Chang Song
a

a,)

, Jong Huy Kim

Department of Chemical Engineering, Konyang Uniersity, Nonsan, Chungnam 320-711, South Korea
b
Korea Institute of Energy Research, Yosong-gu, Taejon 305-343, South Korea
Received 18 February 1999; received in revised form 5 August 1999; accepted 18 October 1999

Abstract
Nanoparticles of tin oxide have been synthesized from water-in-oil microemulsions consisting of water, AOT surfactant., and
n-heptane oil.. The aqueous cores of microemulsions were used as constrained microreactors for the precipitation of precursor hydroxide
particles. When these precursor powders were calcined at 6008C for 2 h, they exhibited a higher BrunauerEmmetTeller BET. surface
area of 86 m2rg. On the other hand, the powders obtained from the conventional precipitation method showed a lower surface area of 19
m2rg. q 2000 Elsevier Science S.A. All rights reserved.
Keywords: Tin oxide particles; Water-in-oil microemulsions; High surface area; Pore structure

1. Introduction
Tin oxide powders have been attracting much attention
since they are highly conducting, transparent, and sensitive
to reducing gases. They have many important applications
including liquid crystal displays, photo-detectors, solar
cells, and gas sensors w13x. When tin oxide powders are
used as transparent conducting electrodes for display devices, a low sheet resistance and high transmission of
visible light are required. But for their applications as gas
sensors, fine particles with high surface area are required
to improve sensitivity.
The tin oxide powders used for gas sensors are usually
prepared from precursor hydroxides precipitated by the
addition of NH 4 OH to SnCl 4 aqueous solution w46x.
However, from this precipitation method, it is not easy to
obtain high surface area powders because of the high
degree of agglomeration and the irregular particle morphology in the tin oxide powders w6x. Therefore, a new
preparation method for tin oxide powders is necessary,
especially for gas sensor applications. In this study, the
authors used a water-in-oil microemulsion as a reaction
medium in order to obtain a higher surface area of tin
oxide powders.
)
Corresponding author. Institute of Process Engineering, Swiss Federal Institute of Technology, ETH Zentrum, Zurich 8092, Switzerland

It is well-known that the water-in-oil microemulsion

system, consisting of a surfactant phase, an oil phase, and
an aqueous phase, is a thermodynamically stable isotropic
dispersion of the aqueous phase in the continuous oil phase
w7x. The size of the aqueous droplets is usually in the range
of 150 nm in diameter. When a soluble metal salt is
incorporated in the aqueous phase of the microemulsion, it
will reside in the aqueous microdroplets surrounded by oil.
Chemical reactions can take place when droplets containing the desirable reactants collide with each other. Each of
these aqueous droplets will thus become a nanosized reactor for forming nanosized solid particles. For the preparation of a ceramic powder via the microemulsion method,
precursor particles of hydroxide or oxalate are first formed
in a microemulsion system. Subsequent calcination of the
precursor powder at an appropriate temperature will lead
to the formation of desired ceramic powder. A refinement
in particle size, particle size distribution, and the degree of
particle agglomeration is therefore expected in the resulting powder, due to the fact that the reaction is taking place
in nanosized domains w8x.
Water-in-oil microemulsions have been used to synthesize colloidal metals w9,10x, superconducting materials
w11,12x, and magnetic materials w1315x. However, descriptions concerning the preparation of tin oxide particles
for gas sensors by the use of water-in-oil microemulsion
have been limited. Song and Kim w16x recently reported the

0032-5910r00r$ - see front matter q 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 3 2 - 5 9 1 0 9 9 . 0 0 2 5 5 - 7

K.C. Song, J.H. Kim r Powder Technology 107 (2000) 268272

formation of tin oxide nanoparticles in the range of 23

nm in diameter together with a narrow particle size distribution by the water-in-oil microemulsion method. However, the details concerning the surface area and pore
structure of the tin oxide powders prepared by the microemulsion method were not discussed.
In this study, we report the successful synthesis of high
surface area tin oxide powders using the water-in-oil microemulsion method. Detailed discussion concerning the
pore structure of the tin oxide powders prepared by the
microemulsion method is also presented.

2. Experimental

269

The thermal behaviors of the powders were investigated at

a heating rate of 108Crmin in flowing air condition by
DTArTG STD2960, TA instruments.. The IR spectra of
the powders were measured in a wave number range from
400 to 4000 cmy1 using Fourier transform infrared FT-IR.
spectroscopy FTS155, Bio-Rad.. Adsorption study of the
calcined tin oxide powders was performed on a
Brunauer Emmet Teller BET . surface analyzer
ASAP2010, Micromeritics. using N2 as the adsorbent at
liquid nitrogen temperature. The surface area was determined by the multipoint BET method using the adsorption
data in the relative pressure PrPo . range of 0.050.25.
Desorption isotherm was used to determine the pore size
distribution using the BJH method.

2.1. Preparation of tin oxide particles

3. Results and discussion
Tin chloride pentahydrate 99%. and ammonia water
28%. by Aldrich were used. The n-heptane high purity
HPLC grade. by Sigma-Aldrich and sodium dioctylsulfosuccinate AOT, 98%. by Aldrich were also used.
Tin oxide particles were prepared via two processing
methods: i. the conventional precipitation method and ii.
the microemulsion method. In the precipitation method,
aqueous NH 3 solution was added to 0.1 M aqueous SnCl 4
solution until the pH of the mixture reached 12, where
white tin hydroxide precipitates were produced. The precipitates were dried in a drying oven at 1008C for 24 h.
Then the dried precipitates were calcined for 2 h at constant temperatures in an electric furnace in air.
For the microemulsion method, n-heptane was used as
the oil phase and AOT was used as the surfactant. Two
kinds of microemulsions type I and II of microemulsions.
with different aqueous phases were obtained. The aqueous
phase in the type I microemulsion was a solution of 0.1 M
SnCl 4 , while the aqueous phase in the type II microemulsion was a solution 1.0 M. of the precipitating agent,
NH 3 . The molar ratio of water to AOT was 15 and the
volume ratio of aqueous phase to oil phase was 0.054 in
two microemulsions. These two microemulsions were
mixed for 3 h under constant stirring speed. The water-clear
appearance of the mixture upon vigorous stirring indicated
the formation of the microemulsions. Powder samples
were obtained by flocculating the colloids with acetone,
followed by separation in a superspeed centrifuge at 8500
rpm for 30 min. The precipitates were washed with the
n-heptane to eliminate excess AOT. The precipitates were
then dried at 1008C for 24 h. The dried precipitates were
calcined at constant temperatures for 2 h to completely
convert the hydroxide into tin oxide.

Fig. 1a and b show the SEM micrographs of the dried

powders prepared by the microemulsion and precipitation
method, respectively. As shown in Fig. 1a, the microstructure of the powder prepared by the microemulsion method
shows a compact arrangement of uniform particles. The
size of the particle diameter is in the range of 3070 nm.
On the other hand, the powder prepared by the precipitation method has a totally different microstructure as shown
in Fig. 1b. The main elements of the structure are non-uni-

2.2. Characterization
Scanning electron microscopy SEM. photographs by
Philips 535M SEM were taken to examine the microstructural evolution and particle size of the dried precipitates.

Fig. 1. The SEM micrographs of the dried precipitates prepared by

different methods: a. the microemulsion method and b. the precipitation
method.

270

K.C. Song, J.H. Kim r Powder Technology 107 (2000) 268272

form agglomerates with larger diameters up to 1000 nm.

From this result, it can be seen that the powder from the
microemulsion method has a much smaller particle size
than that from the precipitation method. These results seem
to be caused by the fact that the powder prepared by the
microemulsion method was formed in a controlled manner
in nanosized domains, which kept the composition partition to a nanosized level. In contrast, the powder prepared
by the precipitation method was formed in an uncontrolled
manner, leading to an extensive growth and aggregation of
precipitated particles w6,16x.
Fig. 2 shows the DTA and TG curves for the dried
powder prepared by the microemulsion method. The TG
curve exhibits two apparent decreases in specimen weight;
the first is over the temperature range from room temperature to 1508C and the second is over the temperature range
from 1508C to 4008C. Presumably, the first fall in specimen weight is due to the loss of the residual solvent and
water in the as-dried precursor powder. This is evidenced
by an endotherm in the DTA curve over the same temperature range. The second fall in specimen weight, over the
temperature range from 1508C to 4008C, is believed to be
due to the decomposition of surfactant AOT. The removal
of surfactant is also confirmed by a strong exothermic
peak at 3008C in the DTA curve. Little further weight loss
in the TG curve is observed at temperatures above 4008C,
indicating the completion of any reaction involving a
weight change. From these results, it is clear that most of
the surfactant AOT in the pores of as-dried precursor
powder are eliminated at about 4008C.
This fact is confirmed by the IR spectra of the tin oxide
powders prepared by the microemulsion method as shown
in Fig. 3. The IR spectrum of dried powder at 1008C
contains peaks from the surfactant AOT. The peak at
28003050 cmy1 is due to CH 3 C, [CH, CH 2 chains of
AOT. The peak at 17001750 cmy1 is from CH 2 CO
OR of AOT. The hydrophobic group R-SO 3 Naq of AOT
is responsible for the peak at 10001100 cmy1 w17x. From
the above result, it is obvious that AOT molecules are
adsorbed in the pores of the powders heat-treated at 100

Fig. 2. The DTA and TG curves for the dried precipitates prepared by the
microemulsion method.

Fig. 3. The FT-IR spectra of the tin oxide powders prepared by the
microemulsion method at various calcination temperatures: '. marks
the characteristic peaks of surfactant AOT.

and 2008C. However, when the powders are calcined at

above 4008C, the peaks from surfactant AOT disappear.
The IR spectra also confirm that most of the surfactant
AOT are eliminated at above 4008C.
Fig. 4 is a plot of the specific surface areas of the
powders obtained from both the microemulsion and precipitation method as a function of calcination temperature.
The powder prepared by the precipitation method shows a
sharp decrease in the surface area as a calcination temperature increases. However, the powder prepared by the microemulsion method shows a slight increase of the surface
area with calcination temperature from 2008C to 6008C
and then a continuous decrease of the surface area until

Fig. 4. Changes in the specific surface areas of the powders obtained by

different methods with calcination temperatures: a. the microemulsion
method and b. the precipitation method.

K.C. Song, J.H. Kim r Powder Technology 107 (2000) 268272

8008C. This result suggests that the surface areas are

significantly higher for the microemulsion-derived powders than the precipitation-derived powders for high calcination temperatures above 4008C.. However, at a low
calcination temperature of 2008C, the surface area of the
microemulsion-derived powder shows a lower value than
that of the precipitation-derived powder.
For the powder from the precipitation method, it is
known that the surface area decreases sharply from 2008C
to 8008C due to the elimination of ammonia, physically
adsorbed water, and chemically bonded water in the precursor powder w6,18,19x. However, the powder prepared by
the microemulsion method shows a different surface area
behavior with calcination temperature. The powder heattreated at 2008C after being prepared from the microemulsion method contains the surfactant AOT in the pores, and
thus shows a lower surface area than the powder from the
precipitation method. However, as the calcination temperature increases, the surfactant incorporated in the pores is
removed, as confirmed by DTArTG and FT-IR analyses
in Figs. 2 and 3, and thus, the powder shows a slight
increase of the surface area with calcination temperature
from 2008C to 6008C. Hence, the surface areas of the
microemulsion-derived powders become larger than those
of the precipitation-derived powders at high calcination
temperatures above 4008C. It seems that the reason why
the surface area of the powder from the microemulsion
method is higher than that from the precipitation method at
the same calcination temperature is due to smaller size of
the particles from the microemulsion method, as shown in
Fig. 1.
Fig. 5 shows the adsorption and desorption isotherms of
the powders calcined at 6008C after being prepared from
both the microemulsion and precipitation methods. Both
powders have type IV isotherms and exhibit hysteresis
loops of type H2 w20,21x. Thus, we can know that the
powders are mesoporous and the pores are networked w20x.
The adsorption amount of N2 on the powder from the
microemulsion is much larger than that on the powder

Fig. 5. Nitrogen adsorption `. desorption v . isotherms for the powders calcined at 6008C for 2 h after being prepared by different methods:
a. the microemulsion method and b. the precipitation method.

271

Fig. 6. Pore size distribution curves for the powders calcined at 6008C for
2 h after being prepared by different methods: a. the microemulsion
method and b. the precipitation method.

from the precipitation, indicating that the powder from the

microemulsion has a much larger surface area than the
powder from the precipitation. This result is well consistent with the surface area data in Fig. 4; when calcined at
6008C for 2 h, the microemulsion-derived powder shows a
higher specific surface area of 86 m2rg, while the conventional precipitation-derived powder exhibits a lower surface area of 19 m2rg.
In order to compare the difference of the pore structures
of the powders prepared from the microemulsion and
precipitation methods, respectively, the pore size distribution curves of two powders calcined at 6008C are given in
Fig. 6. The pore size distributions were calculated by using
the desorption isotherms of Fig. 5. It is seen that both
powders exhibit mesoporous pore structures. The pore
structure of the precipitation-derived powder is
monomodal. There is a maximum for the pore diameters in
the pore size distribution curve at 10.8 nm. On the other
hand, the pore size distribution of the microemulsion-derived powder shows two maximum values at 3.7 and 4.7
nm. It is also seen that most of the pores in the powder
from the microemulsion method are about 36 nm in
diameter, while the powder from the precipitation method
has larger pores ranging up to about 12 nm.
It is well-known that the pore size is a good reflection
of the actual particle size for micrormesoporous systems
w22x. Thus, it seems that the reason why the pore diameter
of the powders from the microemulsion method showed
smaller than that from the precipitation method in Fig. 6 is
due to smaller size of particles from the microemulsion
method. And it is interesting that the microemulsion-derived powder calcined at 6008C shows two maximum
values in the pore size distribution, while the precipitation-derived powder exhibits a maximum value. We ascribe the existence of two maximum values in the pore
size distribution of the microemulsion-derived powder calcined at 6008C to the formation of new pores by removing
of AOT molecules contained in the pores as the calcination
temperature increases.

272

K.C. Song, J.H. Kim r Powder Technology 107 (2000) 268272

4. Conclusions
This work shows that it is possible to prepare tin oxide
powder with high surface area by the water-in-oil microemulsion method. When the precursor powders from
the microemulsion method were calcined at 6008C for 2 h,
they exhibited a higher surface area of 86 m2rg. On the
other hand, the powders obtained from the precipitation
method showed a lower surface area of 19 m2rg. Marked
differences have also been seen in the pore size distributions of the powders prepared by the microemulsion and
precipitation methods, respectively. The pore structure of
the precipitation-derived powder was monomodal. There
was a maximum for the pore diameters at 10.8 nm. On the
other hand, the pore size distribution of the microemulsion-derived powder showed two maximum values at 3.7
and 4.7 nm. It was also shown that the powder from the
microemulsion method had smaller pores than that from
the precipitation method.
References
w1x K.L. Chopra, S. Major, D.K. Pandya, Thin Solid Films 102 1983.
1.
w2x A.L. Dawar, J.C. Joshi, J. Mater. Sci. 19 1984. 1.
w3x J.C. Manifacier, L. Szepess, J.F. Bresse, M. Perotin, Mater. Res.
Bull. 14 1979. 163.

w4x N. Nitta, S. Otani, M. Haradome, J. Electron. Mater. 9 1980. 727.

w5x N. Yamazoe, Y. Kurokawa, T. Seiyama, Chem. Lett. 1982. 1899.
w6x K.C. Song, J.H. Kim, J.S. Sung, J. Korean Inst. Chem. Eng. 35
1997. 457.
w7x P.G. deGennes, C. Taupin, J. Phys. Chem. 86 1982. 2294.
w8x G.K. Lim, J. Wang, S.C. Ng, L.M. Gan, Mater. Lett. 28 1996. 431.
w9x M. Arturo, L. Quintela, J. Rivas, J. Colloid Interface Sci. 158 1993.
446.
w10x P. Barnickel, A. Wokaun, W. Sager, H.F. Eicke, J. Colloid Interface
Sci. 148 1992. 80.
w11x D. Burgard, C. Kroph, R. Nass, H. Schmidt, Mater. Res. Soc. Symp.
Proc. 346 1994. 101.
w12x S. Hingorani, D.O. Shah, M.S. Multani, J. Mater. Res. 10 1995.
461.
w13x V. Pillai, D.O. Shah, J. Magn. Magn. Mater. 103 1996. 243.
w14x N. Moumen, P. Veillet, M.P. Pileni, J. Magn. Magn. Mater. 149
1995. 67.
w15x J. Wang, P.F. Chong, S.C. Ng, L.M. Gan, Mater. Lett. 30 1997.
217.
w16x K.C. Song, J.H. Kim, J. Colloid Interface Sci. 212 1999. 193.
w17x Y.S. Byoun, C.S. Oh, S.C. Choi, Ceramic Processing Science and
Technology, The American Ceramic Society, Westerville, USA,
1995, p. 21.
w18x S. Kaneko, J. Kanamor, F. Imoto, J. Chem. Soc. Jpn. 6 1976. 906.
w19x T. Furusaki, J. Takahashi, H. Takaha, K. Kodaira, J. Ceram. Soc.
Jpn. 101 1993. 451.
w20x S. Lowell, J.E. Shields, Powder Surface Area and Porosity, Chapman and Hall, London, 1984.
w21x S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity,
Academic Press, London, 1982.
w22x K. Hellgardt, D. Chadwick, Ind. Eng. Chem. Res. 37 1998. 405.