Journal of Thermal Analysis and Calorimetry, Vol.

94 (2008) 2, 499–505

THERMAL STABILITY OF HEXAGONAL TUNGSTEN TRIOXIDE IN AIR

I. M. Szilágyi1*, Judit Pfeifer2, C. Balázsi2, A. L. Tóth2, Katalin Varga-Josepovits3,

J. Madarász4 and G. Pokol4
1
Materials Structure and Modeling Research Group of the Hungarian Academy of Sciences, Budapest University of Technology
and Economics, Department of Inorganic and Analytical Chemistry, 1111 Budapest, Szt. Gellért tér 4., Hungary
2
Ceramics and Nanocomposites Laboratory, Research Institute for Technical Physics and Materials Science, 1121 Budapest
Konkoly-Thege út 29–33, Hungary
3
Department of Atomic Physics, Budapest University of Technology and Economics, 1111 Budapest, Budafoki út 8, Hungary
4
Department of Inorganic and Analytical Chemistry, Budapest University of Technology and Economics, 1111 Budapest
Szt. Gellért tér 4, Hungary

We studied the thermal stability of different hexagonal tungsten trioxide, h-WO3 samples, which were prepared either by annealing
hexagonal ammonium tungsten bronze, (NH4)0.33–xWO3–y, or by soft chemical synthesis from Na2WO4. The structure and composi-
tion of the samples were studied by powder XRD, SEM-EDX, XPS and 1H-MAS NMR. The thermal properties were investigated
by simultaneous TG/DTA, on-line evolved gas analysis (TG/DTA-MS), SEM and in situ powder XRD. The preparative routes in-
fluenced the thermal properties of h-WO3 samples, i.e. the course of water release, the exothermic collapse of the hexagonal frame-
work and the phase transformations were all affected.

Keywords: evolved gas analysis by MS, hexagonal tungsten oxide, in situ studies, simultaneous TG/DTA, X-ray diffraction

Introduction that h-WO3 transforms into monoclinic (m-) WO3 in

an exothermic reaction between 250–500°C. Other
For applications in photoelectrochromical cells and studies on m-WO3 have revealed that m-WO3 trans-
chromogenic (electro-, photo- and thermochromic) forms reversibly into tetragonal (t-) WO3 between
devices [1–11], gas sensors [12–17] and catalysts 700–800°C [1, 38].
[18–21], tungsten oxides have attracted much atten- The different preparative routes result in differ-
tion in the past decades. Hexagonal tungsten trioxide, ent morphology and different composition with vary-
h-WO3 is one of the most researched tungsten oxides ing amounts of residual elements, and these might af-
due to its open-tunnel structure and intercalation fect the thermal properties of h-WO3. Therefore it
chemistry [22–24]. It has been prepared in several seemed advisable to compare the thermal behavior of
ways: by dehydration of tungsten oxide hydrates [24, differently prepared h-WO3 samples. This was also
25], by wet chemical synthesis using various precur- encouraged by the lack of evolved gas analytical
sors [26, 27], by thermal oxidation of tungsten metal (EGA) studies on h-WO3.
[28], by thermal or wet chemical oxidation of hexago- Thus, here we present a comparison of the ther-
nal ammonium tungsten bronze [22, 29–31], by ion mal stability of different h-WO3 samples, which were
exchanging template ions from hexagonal tungsten prepared either by thermal oxidation of hexagonal
bronze [32] or by spray pyrolysis [10]. ammonium tungsten bronze (HATB),
The thermal stability of h-WO3 is important in (NH4)0.33–xWO3–y (Sample 1), or by soft chemical syn-
the previously mentioned application fields, espe- thesis, i.e. acidification of Na2WO4 solution followed
cially there, where h-WO3 is used at elevated temper- by washing and hydrothermal treatment (Samples 2,
atures, e.g. by gas sensors. The thermal behaviour of 2a, 2b). Recently the authors managed to prepare
the title compound has already been studied by highly crystalline, monophase HATB,
thermogravimetry (TG) [24], differential scanning (NH4)0.07(NH3)0.04(H2O)0.09WO2.95 by heating ammo-
calorimetry (DSC) and differential thermal analy- nium paratungstate tetrahydrate, (NH4)10[H2W12O42]·
sis (DTA) [7, 29, 33], X-ray powder diffrac- 4H2O in H2 for 6 h at 400°C [39]. This gave us chance
tion (XRD) [34–36], transmission and scanning elec- to prepare good quality h-WO3 by oxidizing the
tron microscopy (TEM and SEM) [36, 37], Raman [7] highly crystalline HATB starting material in air. On
and FTIR spectroscopy [34]. The results have showed the other hand, the authors also had experience with

* Author for correspondence: imre.szilagyi@mail.bme.hu

1388–6150/$20.00 Akadémiai Kiadó, Budapest, Hungary

© 2008 Akadémiai Kiadó, Budapest Springer, Dordrecht, The Netherlands
 SZILÁGYI et al.

the soft chemical synthesis of h-WO3 from Na2WO4 not detect these species [39]. 1H-MAS-NMR experi-
[17, 24, 40, 41]. ments were carried out on a VARIAN NMR SYS-
The structure and composition of the samples was TEM spectrometer (600 MHz for 1H) using a 3.2 mm
studied by XRD, scanning electron microscopy with en- HXY VARIAN/Chemagnetics probe. 1H chemical
ergy dispersive X-ray analysis (SEM-EDX), X-ray pho- shifts were referenced to adamantane (d1H=0 ppm).
toelectron spectroscopy (XPS) and solid-state 1H-MAS All spectra were recorded under the same experimen-
(Magic Angle Spinning) NMR spectroscopy. The ther- tal conditions. 16 transients were acquired at 12 kHz
mal behavior of hexagonal tungsten oxides was investi- spinning rate and a recycle delay of 20 s was used.
gated by simultaneous TG/DTA, on-line coupled Background suppression DEPTH [42] was employed
evolved gas analysis with mass spectrometer to remove signals from the probe. The composition of
(TG/DTA-MS), SEM and in situ high temperature precursor HATB was determined earlier by chemical
X-ray powder diffraction (HT-XRD). analysis, XPS and 1H-MAS-NMR spectroscopy [39],
The two types of samples had differences in and in HATB the NH +4 /NH3 amount was 0.55 and
morphology and composition, i.e. (1) consisted of mm 0.32 mass%, respectively. Now we measured the
scale blocks of nanoparticles and contained NH +4 ion 1
H-MAS-NMR spectra of both HATB and 1, and by
residues and (2) had nm scale particles and contained comparing the normalized signals of NH +4 and NH3
Na+ ion residues. The differences in morphology and peaks [39], we could estimate semiquantitatively the
composition resulted in different thermal properties. NH +4 /NH3 amount of 1, which was ca. 0.11 and
0.04 mass%, respectively.
Powder X-ray diffraction (XRD) pattern of 1
Experimental was measured by a PANalytical X’pert Pro MPD
X-ray diffractometer equipped with an X’Celerator
Sample of h-WO3 (1) was prepared by heating highly detector using CuKa radiation. XRD patterns of 2, 2a
crystalline, monophase hexagonal ammonium tungsten and 2b were recorded by a Bruker AXS D8 Discover
bronze (HATB), (NH4)0.07(NH3)0.04(H2O)0.09WO2.95 [39] X-ray diffractometer using CuKa radiation.
in air (15 L h–1) at 10°C min–1 to 470°C and then keep- A LEO-1550 FEG SEM instrument was used
ing it there isothermally for 3 min in an open aluminium during the SEM characterisation of 1, 2, 2a, 2b and
crucible in a DuPont 910 DSC instrument. their decomposition products.
Samples of h-WO3 (2, 2a, 2b) were prepared by During the evolved gas analytical (TG/
soft chemical synthesis, i.e. by acidification of DTA-MS) measurements of 1, 2, 2a and 2b, open
Na2WO4 solution followed by washing and hydro- platinum crucible, a heating rate of 10°C min–1, sam-
thermal treatment of the precipitate [24, 40, 41]. By ple sizes of 30–100 mg and flowing air
this route first WO3·0.33H2O is prepared from (130 mL min–1) were used. The TG/DTA-MS appara-
Na2WO4 in aqueous phase, then the oxide hydrate tus consisted of an STD 2960 Simultaneous
transforms into h-WO3. 2 was made up of monophase DTA/TGA (TA Instruments Inc.) thermal analyzer
h-WO3. 2a contained some WO3·0.33H2O traces from and a Thermostar GSD 200 (Balzers Instruments)
the precursor. 2b contained high amount of residual quadrupole mass spectrometer. The details of the
sodium, ca. 1.2 mass% in the bulk and ca. 2.3 mass% TG/DTA-MS setup are discussed elsewhere [43, 44].
on the surface. 2a and 2b were included in the studies, In situ high temperature X-ray diffraction
because their composition showed deviations from (HT-XRD) patterns of 1 and 2 were collected by a
standard h-WO3 samples. Therefore by these samples PANalytical X’pert Pro MPD X-ray diffractometer
we could check the effect of different compositions equipped with an X‘Celerator detector and an Anton
on the thermal properties of h-WO3. Paar HTK-2000 high temperature XRD chamber us-
The sodium content of 2b was measured by both ing CuKa radiation. The scanning time was ca. 3 min
X-ray photoelectron spectroscopy (XPS) and energy for each pattern. HT-XRD experiments were per-
dispersive X-ray analysis (within a scanning electron formed in static air using 10°C min–1 heating rate be-
microscope) (SEM-EDX). XPS data were collected tween XRD measurements.
by a VG Microtech instrument consisting of a XR3E2
X-ray source, a twin anode (MgKa and AlKa) and a
CLAM 2 hemispherical analyser using MgKa radia- Results and discussion
tion. SEM-EDX measurements were performed in a
Röntec QuanTAX energy dispersive microanalyser Characterization of h-WO3 samples
system. The difference in the composition of the h-WO3 sam-
The ammonium ion and ammonia molecule con- ples was obvious. Besides tungsten and oxygen,
tent of 1 could only be measured by 1H-MAS-NMR 1 contained little amounts of ammonium ions and am-
spectroscopy; by XPS or FTIR spectroscopy we could

500 J. Therm. Anal. Cal., 94, 2008

 THERMAL STABILITY OF HEXAGONAL TUNGSTEN TRIOXIDE IN AIR

monia molecules (0.11 and 0.04 mass%, respec- up of 50–100 nm aggregated particles (Figs 1a and b).
tively), which had remained from the parent hexago- 2, 2a and 2b were greyish-white coloured and they
nal ammonium tungsten bronze phase. 2 and 2a consisted of 20–30 nm thick platelets and rods, which
contained certain amount (0.02–0.1 mass%) of so- aggregated into 1–2 mm particles (Fig. 1c). The differ-
dium, which was necessary for the formation of ent color of 1 and 2, 2a, 2b suggests different elec-
h-WO3 during the soft chemical synthesis tronic structure, which might affect applications. The
[17, 24, 40, 41]. In 2b the amount of sodium was rela- morphology of 2, 2a and 2b showed no significant
tively high. By SEM-EDX it was determined to be differences, thus only the SEM image of sample (2a)
ca. 1.2 mass%, and by XPS it was found to be is shown in Fig. 1c.
ca. 2.3 mass%. The two methods complement each There were only small differences in the XRD
other: in SEM-EDX the information depth of the anal- patterns of 1 and 2 (Fig. 2), and they were both identi-
ysis is 400–500 nm, while in XPS it is only about fied as pure h-WO3 (ICDD 85-2460 and 33-1387). As
5–10 nm, thus SEM-EDX gives roughly the bulk con- already discussed, WO3·0.33H2O (ICDD 35-0270)
centration and XPS gives the surface concentration of was also detected in 2a (Fig. 2). 2b had identical XRD
the elements. pattern as 2, and thus it is not shown in Fig. 2.
The different preparative routes strongly af-
fected the morphology. 1 consisted of 1–80 mm yel- Thermal behaviour of 1
low particles, which were cracked. Under higher
magnification we found that actually they were built The TG/DTA and evolved gas analytical (TG/
DTA-MS) curves of 1 are presented in Fig. 3. The
evolved gas analytical study showed only the evolu-
tion of water. Though 1 contained traces of ammo-
nium ions and ammonia molecules, their amount was
so small that their release could not be detected well
with the TG/DTA-MS system. The HT-XRD patterns
of 1 are shown in Fig. 4. The SEM image of the de-
composition product of 1 at 900°C is presented in
Fig. 5a.
According to the thermoanalytical curves, up to
200°C adsorbed and chemisorbed water were released
in two overlapping endothermic reactions. Between
200–400°C water release continued, but here no sig-
nificant heat effect was observed and the amount of
evolved water was much smaller than below 200°C.
Fig. 1 SEM images of a – h-WO3 (1) at lower (10k) magnifi- Based on the XRD patterns, the starting h-WO3 struc-
cation, b – h-WO3 (1) at higher (150k) magnification; ture remained nearly the same up to 450°C.
the black dashed circle in the middle of a – shows the
area from where b – was taken, c – h-WO 3 (2a)

Fig. 3 Simultaneous TG/DTA and evolved gas analytical

Fig. 2 XRD patterns of a – h-WO3 (1), b – h-WO3 (2) and curves of h-WO3 (1) measured in air (130 mL min–1,
c – h-WO3 (2a) 10°C min–1, open Pt crucible, 102.4 mg)

J. Therm. Anal. Cal., 94, 2008 501

 SZILÁGYI et al.

50–100 nm particles observed in the starting material.

The crystallization of the more stable m-WO3 from
h-WO3 and the further annealing of m-WO3 are
probably responsible for the growth of the particle size.

Thermal behaviour of 2
The TG/DTA and evolved gas analytical
(TG/DTA-MS) curves of 2 are presented in Fig. 6.
The evolved gas analytical study showed only the
evolution of water. The HT-XRD patterns of 2 are
shown in Fig. 7. The SEM image of the decomposi-
tion product of 2 at 900°C is presented in Fig. 5b.
When heating 2 at first adsorbed and
chemisorbed water were released in two overlapping
endothermic reactions. But here this stage extended
up to 250–300°C, and the mass loss corresponding to
adsorbed and chemisorbed water was also greater
(1.61% at 250°C) than in the case of 1 (0.49% at
190°C). The greater mass loss should be due to the
Fig. 4 In situ HT-XRD patterns of h-WO3 (1) in static air re-
corded from r.t. to 900°C

Fig. 5 SEM images of decomposition products at 900°C in air

of a – h-WO3 (1) and b – h-WO3 (2)

Between 450–500°C another water release step

took place. At 500°C the reflections of h-WO3 be-
came more diffuse in the XRD patterns. Then be-
tween 500–550°C the hexagonal structure collapsed
in an exothermic reaction, and transformed irrevers-
ibly into monoclinic (m-) WO3 (ICDD 43-1035).
Since the h-WO3 structure is metastable, it is easy to
understand that its transformation into the more stable
m-WO3 is an exothermic reaction – as was shown by
our measurements and previous results [29].
The m-WO3 phase was stable up to 700°C.
Around 750°C it transformed into tetragonal (t-) WO3
(ICDD 85-0807) in a reversible endothermic reaction.
At 750°C the HT-XRD analysis showed a mixture of
m- and t-WO3, while by 800°C only t-WO3 was pres-
ent. The well known reversibility [1, 38] of the m- and
t-WO3 transformation is reflected also by that when
the sample was cooling (not shown on Fig. 3), there
was an exothermic DTA peak where t-WO3 trans-
formed back into m-WO3. At room temperature (r.t.) Fig. 6 Simultaneous TG/DTA and evolved gas analytical
only m-WO3 was observed. curves of h-WO3 (2), (2a) and (2b) measured in air
The SEM image of the decomposition product of (130 mL min–1, 10°C min–1, open Pt crucible, 29.7,
1 at 900°C showed 100–500 nm particles instead of the 40.8 and 29.8 mg, respectively)

502 J. Therm. Anal. Cal., 94, 2008

 THERMAL STABILITY OF HEXAGONAL TUNGSTEN TRIOXIDE IN AIR

were formed in our experiments, with the available

ICDD cards. The Na2W4O13 (ICDD 27-1425,
21-1667) phase was well detected, while the presence
of tungsten bronzes was only guessed.
The Na2W4O13 phase was observed in small
amounts already at 700°C. The presence of Na2W4O13
is also supported by the fact that this phase also ap-
peared during the thermal decomposition of a tung-
sten sodium hydroxide, Na0.4WO3·0.6H2O [45]. The
amount of Na2W4O13 increased at 750°C, then de-
creased at 800°C, and by 850°C it disappeared com-
pletely. But in the sample cooled back to r.t. it was
observable again.
Between 700–900°C four endothermic heat ef-
fects were observed on the DTA curve of 2, and we
suggest an explanation for them on the basis of the
HT-XRD results. The Na2W4O13 phase formed up to
750°C, because its amount was the largest at this tem-
perature, so this endothermic peak at 735°C should be
connected with the appearance of Na2W4O13. The
main part of the m- to t-WO3 transformation took
Fig. 7 In situ HT-XRD patterns of h-WO3 (2) in static air re- place between 750–800°C, therefore we assume that
corded from r.t. to 900°C
the endothermic peak at 790°C might belong to this
reaction. Though, it has to be mentioned that this tem-
morphology of 2, because the smaller particles of 2
perature is 40°C higher than in the case of 1.
have larger specific surface and hence greater amount
By 850°C Na2W4O13 disappeared and only
of adsorbed and chemisorbed molecules. Between
t-WO3 was detected. Since after cooling back to r.t.
250–450°C the water release continued, similarly as
the sample contained Na2W4O13 again besides
in case of 1.
m-WO3, it means that this phase did not decompose,
Between 450–550°C the hexagonal structure
but melted between 800–850°C, and the endothermic
collapsed, and 2 transformed irreversibly in an exo-
heat at 815°C effect should correspond to the melting
thermic reaction into m-WO3, at 50°C lower tempera-
heat of the Na2W4O13 phase. The SEM image of the
ture than 1. This transformation was accompanied
decomposition product of 2 at 900°C also supports
also by the release of small amount of water, similarly
this idea (Fig. 5b). In this picture, the coexistence of
as for 1. The sample at 500°C contained mostly
m-WO3 nanoparticles (small particles) and Na2W4O13
m-WO3 and some h-WO3. Up to this point, the ther-
microparticles (large crystals) can be seen. The solidi-
mal decomposition of 1 and 2 involved the same pro-
fying of Na2W4O13 from its melt explains well, how
cesses, only their dynamics was different.
0.5–4 mm particles could be formed from the starting
When 2 was heated further, according to XRD
20–30 nm h-WO3 particles.
patterns, the m-WO3 structure was ordering, i.e. the
It needs further investigation to explain the source
intensity of reflections of m-WO3 increased. Only at
of the tiny heat effect at 885°C. The XRD pattern at
700°C, a well defined m-WO3 structure was present.
900°C was practically the same as at 850°C. The only
Then m-WO3 started its reversible transformation into
difference was that the CuK a1 and K a 2 reflections be-
tetragonal (t-) WO3 between 700–750°C. At 750°C
came split, particularly at higher 2q values, which
the sample still contained mainly m-WO3 with some
showed an increase in the crystalline order.
t-WO3. At 800°C the main component was t-WO3
with traces of m-WO3, and at 850°C only t-WO3 was
detected. Thermal behaviour of 2a and 2b
2 contained sodium besides tungsten and oxy-
Due to the WO3·0.33H2O content of 2a and the high
gen, and after the collapse of the hexagonal frame-
sodium content of 2b, a different thermal behavior
work, this sodium content was responsible for the for-
was found in both cases in comparison to 2. The
mation of sodium tungsten oxides. The analysis of the
TG/DTA and evolved gas analytical (TG/DTA-MS)
sodium containing phases had great uncertainty. On
curves of 2a and 2b are presented in Fig. 5. The
the one hand their amount was small, on the other
evolved gas analytical study showed only the evolu-
hand in some cases they were not well defined phases.
tion of water.
Therefore it was difficult to match the phases, which

J. Therm. Anal. Cal., 94, 2008 503

 SZILÁGYI et al.

The additional WO3·0.33H2O content of 2a was ble up to 400–450°C. Furthermore 2b was stable up to
responsible for the endothermic water release step be- 500°C due to its high sodium content, which stabi-
tween 250-370°C. Therefore thermal analysis is a lized the hexagonal tungsten oxide structure better.
good tool to check the state of the Upon further annealing, between 750-800°C
WO3·1/3H2O®h-WO3 conversion. monoclinic WO3 transformed reversibly into tetra-
By both 2 and 2b the water release ended parallel gonal WO3. Compared to 1, a separate feature of 2
to the collapse of the hexagonal structure, similarly as was that from 700°C Na2W4O13 was also detected,
in case of 1 and 2. As a significant difference this col- which was a result of the sodium content of 2. Above
lapse occurred nearly at 100°C higher temperature in 700°C 2a and 2b showed small differences in the oc-
the case of 2b, compared to 2 and 2a. This should be curring heat effects, compared to 2.
the result of the high sodium content of 2b, which
might have stabilized the crystal structure better. The
exothermic DTA peak belonging to the collapse of Acknowledgments
the hexagonal structure showed an additional over-
lapping reaction by 2b. The Authors are indebted to Péter Király (Institute of Struc-
Between 750–850°C there was only one endo- tural Chemistry, Chemical Research Center of the Hungarian
Academy of Sciences, Budapest, Hungary) for recording the
thermic DTA peak instead of two by 2a and 2b. The 1
H-MAS-NMR spectra. The financial support of the
greater sample mass of 2a, i.e. 40 mg of 2a instead of GVOP-3.2.1.-2004-04-0210/3.0 project is acknowledged.
30 mg as of 2, and the higher Na content of 2b,
i.e. larger melting heat of Na2WO4, might be responsi-
ble why the two endothermic peaks were not sepa- References
rated, but overlapped each other. In addition in the
case of 2b we did not observe any endothermic DTA 1 E. Lassner and W.-D. Schubert, Tungsten. Properties,
peak between 700–750 and 850–900°C. Chemistry, Technology of the Element, Alloys, and
Chemical Compounds, Kluwer Academic/Plenum
Publishers, New York 1999.
Conclusions 2 C. G. Grandqvist, Handbook of Inorganic Electrochromic
Materials, Elsevier, Amsterdam 1995.
The thermal behavior of differently prepared h-WO3 3 M. Grätzel, Nature, 414 (2001) 338.
4 S. Wang, X. Feng, J. Yao and L. Jiang, Angew. Chem. Int.
samples was studied by on-line evolved gas analysis
Ed., 45 (2006) 1264.
(TG/DTA-MS), in situ high temperature X-ray dif- 5 S. H. Baeck, T. Jaramill, G. D. Stucky and
fraction (HT-XRD) and scanning electron micros- E. W. McFarland, Nano Lett., 2 (2002) 831.
copy (SEM). Our measurements showed that the pre- 6 I. Turyan, U. O. Krasovec, B. Orel, T. Saraidorov,
parative routes influenced the thermal behavior of R. Reinsfeld and D. Mandler, Adv. Mater., 12 (2000) 330.
different h-WO3 samples. 7 C. Santato, M. Odziemkowski, M. Ulmann and
The h-WO3 samples were prepared either by J. Augustynski, J. Am. Chem. Soc., 123 (2001) 10639.
thermal oxidation of hexagonal ammonium tungsten 8 Y. He, Z. Wu, L. Fu, C. Li, Y. Miao, L. Cao, H. Fan and
bronze, (NH4)0.33–xWO3–y (Sample 1), or by soft B. Zou, Chem. Mater., 15 (2003) 4039.
9 B. Xue, J. Peng, Z. Xin, Y. Kong, L. Li and B. Li,
chemical synthesis from Na2WO4 (Samples 2, 2a, 2b).
J. Mater. Chem., 15 (2005) 4793.
The two preparative routes resulted in different mor- 10 J. M. Ortega, A. I. Martinez, D. R. Acosta and
phology and colour. In addition 2a contained also C. R. Magana, Sol. Energy Mater. Sol. Cells.,
WO3·0.33H2O besides h-WO3 and 2b had relatively 90 (2006) 2471.
high sodium content, i.e. 1.2 mass% 11 S. M. A. Durrani, E. E. Khawaja, M. A. Salim,
When heating the samples, as common feature M. F. Al-Kuhaili and A. M. Al-Shukri, Sol. Energy Mater.
up to 200–250°C, adsorbed and chemisorbed water, Sol. Cells., 71 (2002) 313.
then up to 450–500°C structural water were released. 12 J. Polleux, A. Gurlo, N. Barsan, U. Weimar, M. Antonietti
The small amount of WO3·0.33H2O in 2a caused an and M. Niederberger, Angew. Chem. Int. Ed.,
45 (2006) 261.
additional endothermic water release step between
13 X. L. Li, T. J. Lou, X. M. Sun and Y. D. Li, Inorg. Chem.,
250–370°C. Therefore thermal analysis is a good tool 43 (2004) 5442.
to check the state of the WO3·0.33H2O®h-WO3 14 G. Shaw, I.P. Parkin, K. F. E. Pratt and D. E. Williams,
conversion. J. Mater. Chem., 15 (2005) 149.
The starting hexagonal structures collapsed in an 15 C. J. Jin, T. Yamazaki, Y. Shirai, T. Yoshiwaza,
exothermic reaction between 450–600°C and trans- T. Kikuta, N. Nakatani and H. Takeda, Thin Solid Films,
formed into monoclinic WO3. Thus for sensor and 474 (2005) 255.
electrochromic applications our results showed that 1 16 A. Hoel, L. F. Reyes, S. Saukko, P. Heszler, V. Lantto and
was stable up to 450–500°C, while 2 and 2a were sta- C. G. Granqvist, Sensor. Actuat. B, 105 (2005) 283.

504 J. Therm. Anal. Cal., 94, 2008

 THERMAL STABILITY OF HEXAGONAL TUNGSTEN TRIOXIDE IN AIR

17 C. Balázsi, A. K. Prasad, J. Pfeifer, A. L. Tóth and 33 L. Seguin and M. Figlarz, Solid State Ionics, (1993) 437.
P. I. Gouma, SEMINANO2005 – First International 34 W. Han, M. Hibino and T. Kudo, B. Chem. Soc. Jpn.,
Workshop on Seminconductor Nanocrystals, Budapest, 71 (1998) 933.
Hungary 2005, Vol. 1, pp. 79–83. http://www.mfa.kfki.hu/ 35 N. Kumagai, N. Kumagai, Y. Umetzu, K. Tanno and
conferences/seminano2005/pdf/p079_Balazsi_man.pdf J. P. Pereira-Ramos, Solid State Ionics,
18 M. M. Natile, F. Tomaello and A. Glisenti, Chem. Mater., 86–88 (1996) 1443.
18 (2006) 3270. 36 M. Figlarz, B. Dumont, R. Gerand and B. Beaudoin,
19 T. W. Bitner and J. I. Zink, Inorg. Chem., 41 (2002) 967. J. Microsc. Spectrocs. Electron., 7 (1982) 371.
20 A. M. Garrido Pedrosa, M. J. B. Souza, D. M. A Melo and 37 A. Al Mohammad and M. Gillet, Thin Solid Films,
A. S.Araujo, J. Therm. Anal. Cal., 87 (2007) 349. 408 (2002) 302.
21 T. Onfroy, G. Clet and M. Houlla, J. Phys. Chem. B, 38 O. Yamaguchi, D. Tomihisa, H. Kawabata and
109 (2005) 3345. K. Shimizu, J. Am. Ceram. Soc., 70 (1987) C-94.
22 W. Han, M. Hibino and T. Kudo, Solid State Ionics, 39 I. M. Szilágyi, F. Hange, J. Madarász and G. Pokol, Eur. J.
128 (2000) 25. Inorg. Chem., 17 (2006) 3413.
23 O. Y. Khyzhun, Y. M. Solonin and V. F. Dobrovolsky, 40 C. Balázsi and J. Pfeifer, Solid State Ionics,
J. Alloys. Compd., 320 (2001) 1. 151 (2002) 353.
24 J. Pfeifer, C. Balázsi, B. A. Kiss, B. Pécz and A. L. Tóth, 41 J. Pfeifer, C. Guifang, P. Tekula-Buxbaum, B. A. Kiss,
J. Mater. Sci. Lett., 18 (1999) 1103. M. Farkas-Jahnke and K. Vadasdi, J. Solid State Chem.,
25 B. Gerand, G. Nowogrocki, J. Guenot and M. Figlarz, 119 (1995) 90.
J. Solid State Chem., 29 (1979) 429. 42 D. G. Cory and W. M. Ritchey, J. Magn. Reson.,
26 Y. Oaki and H. Imai, Adv. Mater., 18 (2006) 1807. 80 (1988) 128.
27 Z. Gu, Y. Ma, W. Yang, G. Zhang and J. Yao, Chem. 43 J. Madarász, I. M. Szilágyi, F. Hange and G. Pokol,
Commun., (2005) 3597. J. Anal. Appl. Pyrolysis, 72 (2004) 197.
28 Y. Wu, Z. Xi, G. Zhang, J. Yu and D. Guo, J. Cryst. 44 I. M. Szilágyi, J. Madarász, F. Hange and G. Pokol,
Growth, 292 (2006) 143. J. Therm. Anal. Cal., 88 (2007) 139.
29 K. H. Cheng, A. J. Jacobson and M. S. Wittingham, 45 N. M. Laptash, S. A. Polyshchuk, E. I. Melnichenko and
Solid State Ionics, 5 (1981) 355. Y. Y. Kravtzova, J. Thermal. Anal., 39 (1993) 1477.
30 B. Schlasche and R. Schöllhorn, Rev. Chim. Miner.,
19 (1982) 534.
31 A. M. De La Cruz, L. M. Torres-Martínez, Received: June 12, 2007
G. García-Alvaro, E. Morán and M. A. Alario-Franco, Accepted: May 12, 2008
Solid State Ionics, 84 (1996) 181. OnlineFirst: September 20, 2008
32 M. S. Wittingham, J. Li, J. D. Guo and P. Zavalij, Mater.
Sci. Forum, 152–153 (1994) 99.
DOI: 10.1007/s10973-007-8601-y

J. Therm. Anal. Cal., 94, 2008 505