vol%, and then decreased to 0.29 at 40 vol%. It is notable that
well dispersed nanorods of CsxWO3 with the Cs/W atomic ratio
of almost 0.33 which is the ideal value of tungsten bronze
structure were directly produced by the facile solvothermal
reaction in 80 vol% ethanol–20 vol% acetic acid mixed solution.
Based on above results, it might be suggested that the water
controlled-release process is useful to fabricate the homogenous
Cs0.33WO3 nanorods. In other words, the insertion amount of
Cs+ into the WO3 crystal is determined by the amount of
released water, i.e., when the amount of water is few, the
formation of CsxWO3 does not proceed effectively and CsxWO3
with low Cs/W atomic ratio is formed with low yield. On the
other hand, in the presence of excess amount of water, Cs/W
atomic ratio decreases due to the elution of Cs+ in the solution by
the ion exchange reaction shown by eqn (5).
CsxWO3 + yH2O ¼ CsxꢁyWO3 + yCs+ + yOHꢁ
(5)
In our experimental conditions, the amount of acetic acid was
much more than that of OHꢁ originated from eqn (5). Therefore,
the deprotonation of acetic acid was negligible.
The formation mechanism of CsxWO3 nanorods by WCRP
process can be explained as follows: for fabrication of homoge-
nous nanorods, appropriate reaction rate and quantity of water
generation are required. In the case of pure ethanol solvent, due
to the low reaction rate of the dehydration condensation reaction
of ethanol to generate water, the rate of CsxWO3 formation via
reactions (3) and (4) is slow, therefore, small amounts of nuclei
are formed to generate large particles of CsxWO3 (Fig.4(a)). By
adding acetic acid in ethanol, since the rate of water generation
increases, the amount of crystal nuclei increases to decrease the
particle size of CsxWO3. In addition, acetate ion adsorbed on the
surface of the CsxWO3 may play an important role to avoid
the particle agglomeration by the electrostatic repulsion force
and such capping process also facilitated the CsxWO3 to grow
along one-direction. However, when excess amount of acetic acid
is added in ethanol, the water-releasing rate becomes too high to
result in the formation of aggregated non-homogeneous CsxWO3
particles with low Cs/W atomic ratio (Fig. 4(d)). Therefore, the
optimum acetic acid content exists to form well dispersed
nanoparticles of Cs0.33WO3 by the WCRP process in ethanol–
acetic acid mixed solutions.
Fig. 6 Amounts of acetyl acetate, diethyl ether and water, Cs/W atomic
ratio in CsxWO3 and CsxWO3 yield as a function of acetic acid content in
ethanol at 240 ꢀC for 20 h with a Cs/W atomic ratio of 0.5.
The amounts of acetyl acetate and diethyl ether and water, Cs/
W atomic ratio in CsxWO3 determined by an energy-dispersive
X-ray spectrometer (EDS) analysis and CsxWO3 yield are shown
in Fig. 6 as a function of acetic acid content, where the amount of
water was calculated as the total of ethyl acetate and diethyl ether
according to the eqn (1) and (2). It can be seen that the amount of
ethyl acetate and water increased with an increase in acetic acid
content, while that of diethyl ether was decreased. Although eqn
(1) and (2) released water with the same stoichiometry, the
tendency of them was quite different. The etherification reaction
(2) was easier to be taken place than that of reaction (1). It is
reasonable to suggest that the addition of acetic acid promotes the
formation of ethyl acetate and depresses thatof diethylether, since
the esterification and dehydration condensation reactions shown
by eqn (1) and (2) are the equilibrium reactions. It is notable that
the CsxWO3 yield was greatly increased with an increase in the
amount of water formed by the addition of acetic acid in ethanol.
This indicated that water generated by the reactions (1) and (2) as
well as ethanol might play an important role to form CsxWO3
nanoparticles, i.e., water promotes the hydrolysis of WCl6 and
ethanol promotes the reduction of W6+ to W5+ to form CsxWO3 as
shown by eqn (3) and (4). In addition, acetic acid acts to depress
the agglomeration of particles by the electrostatic repulsion force
by forming negatively charged particles. It is reasonable to suggest
that the WO3 was a short-lived reaction intermediate and reacted
further to CsxWO3 as soon as it is formed.
The typical X-ray photoelectron spectrum of the core level
tungsten (W4f) of (a) as-synthesized CsxWO3 nanorods is shown
in Fig. 7 together with those of (b) WO2.72 and (c) WO3 for
comparison. The curves can be fitted as two spin-orbit doublets,
W4f7/2 and W4f5/2 for the interval of about 2.1 eV. The peaks at
34.6 and 36.7 eV, and 35.8 and 37.9 eV were attributed to W5+
and W6+, respectively, which agreed with the reported values.12,13
Tungsten trioxide (WO3) possessed a wide band gap of 2.62 eV
and was transparent in visible and NIR light range (Fig. 8(a)).
Fig. 8(b)–(e) shows the transmittance spectra of CsxWO3
synthesized in various ethanol–acetic acid mixed solutions. The
sample synthesized in pure ethanol showed lower visible light
transparency and NIR shielding performance than those
synthesized in ethanol–acetic acid mixed solutions mainly due to
its irregular large particles and low Cs/W atomic ratio (Fig. 8(b)).
It can be seen that the sample synthesized in ethanol–acetic acid
mixed solution with 40 vol% acetic acid showed high
WCl6 + 3H2O ¼ WO3 + 6 HCl
(3)
WO3 + xCsOH + x/4C2H5OH ¼ CsxWO3 + x/4CH3COOH +
3x/4H2O (4)
The atomic ratio of Cs/W in CsxWO3 increased at first from
0.26 to 0.32 with an increase in acetic acid content from 0 to 20
5102 | J. Mater. Chem., 2011, 21, 5099–5105
This journal is ª The Royal Society of Chemistry 2011