Thermal decomposition of mixed Ce and Gd oxalates and thermal properties of mixed Ce and Gd oxides

Article abstract of DOI:10.1007/s10973-005-7209-3

The aim of the present work is to study the thermal decomposition of the mixed oxalates (Ce_1-xGd_x)₂(C₂O ₄)₃?nH₂O. The mechanisms of decomposition of Ce and Gd oxalate are different, and mixed oxalates behave in an intermediate way. Their dehydration stages are more similar to those of Gd oxalate, as not all the molecules of water are equivalent like the cerium oxalate. The decomposition leads to (Ce_1-xGd_x)O_2-x/2. For x close to 0 or to 1 two solid solutions exist, while for the central composition, the presence of a biphasic region can not be excluded.

Full text of DOI:10.1007/s10973-005-7209-3

Journal of Thermal Analysis and Calorimetry, Vol. 84 (2006) 1, 207–211
THERMAL DECOMPOSITION OF MIXED Ce AND Gd OXALATES AND
THERMAL PROPERTIES OF MIXED Ce AND Gd OXIDES
A. Ubaldini1, C. Artini1, G. A. Costa1*, M. M. Carnasciali2 and R. Masini3
1INFM and DCCI, Via Dodecaneso 31, 16146 Genova, Italy
2INSTM and DCCI, Via Dodecaneso 31, 16146 Genova, Italy
3CNR–IMEM, Via Dodecaneso 33, 16146 Genova, Italy
The aim of the present work is to study the thermal decomposition of the mixed oxalates (Ce1–xGdx)2(C2O4)3×nH2O. The
mechanisms of decomposition of Ce and Gd oxalate are different, and mixed oxalates behave in an intermediate way. Their
dehydration stages are more similar to those of Gd oxalate, as not all the molecules of water are equivalent like the cerium oxalate.
The decomposition leads to (Ce1–xGdx)O2–x/2. For x close to 0 or to 1 two solid solutions exist, while for the central composition, the
presence of a biphasic region can not be excluded.
Keywords: rare earths mixed oxide, thermal decomposition
Introduction
The mixed Ce–Gd oxides are generally produced
by a coprecipitation method, a synthetic route which
offers an easy way to obtain homogenous samples
[8, 9]. An intermediate, normally an oxalate, is
prepared adding a precipitating agent to a solution of
the two cations. Then the mixed oxide is obtained
from the thermal decomposition of the oxalate [7].
The aim of the present work is to present a discus-
sion on the thermal decomposition of the mixed Ce–Gd
oxalates into the correspondent oxides. In particular,
compositions close to 50% were studied in order to re-
port some results on the stability and the structure of
the oxides in the critical part of the diagram.
Solid oxides fuel cells (SOFCs) have been studied in-
tensely because of their great technological interest;
they have a high energy conversion and a very low
emission of pollutants in the environment [1]. SOCFs
are composed by inteconnectors, anode, cathode and of
course by a solid electrolyte, generally an oxide with
high ionic conductivity. One of the most known and
used is the yttria stabilised zirconia (YSZ) for its excel-
lent properties. Unfortunately it requires very high op-
erating temperatures, about 1000°C [2, 3]. Extensive
researches were performed to seek alternative materi-
als requiring lower temperatures. Ceria doped by triva-
lent rare earth elements is considered as one of most
promising materials as electrolyte, because of its high
ionic conductivity at relatively low temperature (500
to 700°C) [2–5]. The substitution of cerium by a
heterovalent cation leads to the formation of vacancies
in the oxygen sublattice that keep the electroneutrality
of the system; the high conductivity is due to the high
mobility of the vacancies. Gd is considered one of the
most effective dopants [6, 7] because of the closeness
of the ionic radius of Gd3+ to that of Ce4+, leading to
large fields of existence of solid solutions [7–10].
However, there is not a perfect agreement about the
wideness of these solid solutions. Grover et al. [10] re-
ported that for 0.5£x£1 the C-type structure exists,
while for 0£x<0.5 the fluorite structure exists. On the
contrary, Tianshu et al. [7] and Ikuma et al. [8] re-
ported that the solubility of Gd in the CeO2 is smaller
and that the solid solution exists if x<0.3, then a
biphasic region appears.
Experimental
All the (Ce1–xGdx)2(C2O4)3×nH2O samples were pre-
pared by a coprecipitation method, starting from com-
mercial Ce and Gd2O3 (Aldrich, 99.99%). They were
weighed in the wished amounts in order to have a ratio
between Ce and Gd (x) equal to 0, 0.4, 0.5, 0.7 and 1.
Two solutions were prepared using a slight excess of
HCl (10% m/V) and mixed. The precipitation of a
mixed oxalate was achieved rapidly adding to the solu-
tion of the two cations a solution of oxalic acid previ-
ously prepared. An excess of oxalic acid was used to
ensure a complete precipitation and to avoid any possi-
ble concentration gradient. The precipitate was filtered
and washed with deionized water; it was then dried in
air at 80°C for 24 h. It was analysed by SEM–EDAX in
order to observe the particles dimensions, shape and
composition. The purity of the samples and the struc-
*
Author for correspondence: Costa@chimica.unige.it
1388–6150/$20.00
Akadémiai Kiadó, Budapest, Hungary
Springer, Dordrecht, The Netherlands
© 2006 Akadémiai Kiadó, Budapest

UBALDINI et al.
ture of the mixed oxides were checked by room tem-
XRD pattern of Gd oxalate kept at 100°C 12 h. On the
basis of the mass variation, it results that the latter re-
leases two molecules of water, becoming
Gd2(C2O4)3×8H2O. It is possible that during the drying
stage of the preparation the mixed oxalates lose several
molecules of water and that the hydration from the en-
vironment is a slow process. Therefore, the initial
amount of water in these samples is lower than 10.
perature X-ray diffraction using a Philips PW 1830
diffractometer. Thermal decomposition data were col-
lected by means of a Netzsch STA 409 thermal
analyser: the samples were heated to 1000°C at
5°C min–1 in flowing oxygen or air. The mass of
samples was about 100 mg.
Results and discussion
Thermal decomposition of oxalates
Oxalates characterisation
Figures 2a and b show the DTA-TG analysis of the
oxalates with x=0.4, 0.5 and 0.7 performed in oxygen;
heating rate equal to 5°C min–1.
XRD analysis shows that mixed oxalates of Ce and
Gd are crystalline (Fig. 1). Their XRD pattern is very
similar to those of other RE2(C2O4)3×10H2O and all
peaks can be indexed assuming that they crystallise in
the monoclinic system with space group P21/c [11].
The reported lattice constants for Gd oxalate
(a=11.04 , b=9.63 , c=10.09 , b=114.10°) [11]
are very close to the calculated ones (a=11.02 ,
b=9.64 , c=10.043 , b=113.70°).
Fig. 1 XRD patterns of mixed Ce–Gd oxalate
Cerium can exist in two different oxidation states,
namely Ce4+ and Ce3+. The XRD pattern of the cerium
oxalate is very similar to that one of Gd oxalate and the
observed mass variation between 25 and 1000°C
(52.02%) is very close to the theoretical mass variation
due to the transformation of Ce oxalate decahydrate
into CeO2 (52.48%). The result of the precipitation
with oxalic acid is Ce2(C2O4)3×10H2O. The cell param-
eters (a=11.26 , b=9.68 , c=10.41 , b=114.36°)
are greater than those of Gd oxalate, in agreement with
the fact that Ce3+ is larger than Gd3+. SEM–EDAX
analyses, performed on the Ce–Gd oxalates, have
shown that the average composition of the precipitates
is very close to the nominal one: the experimental Ce
percentage, determined in many points of 5×5 mm2 ran-
domly selected, for the oxalates with x=0.4, 0.5, 0.7
was 63.2, 50.9 and 31.0%, respectively.
Fig. 2 a – DTA and b – TG analysis performed on mixed Ce–Gd
oxalates in O2 (DTA and TG curves are shifted for clarity)
For x=0, two DTA peaks are present, to which
two steps in the TG curve are associated. The former
is endothermic, presents the onset at about 100 and
the maximum at 134°C. In correspondence of this
peak there is a mass diminution equal to the calcu-
lated one for the complete release of water. In other
words, the anhydrous Ce(III) oxalate forms in this
range of temperature [12]. The second peak is a very
high exothermic peak whose onset is at 275°C and the
maximum at 295°C. In this stage there is the complete
conversion of the initial oxalate into CeO2, as the total
experimental mass loss is 52.12%, close to the theo-
retical one (52.48%).
The
(Ce1–xGdx)2(C2O4)3×nH2O samples are different from
XRD
patterns
of
the
The case of the Gd oxalate is more complex, as in
this case the molecules of water are not equivalent as
those of decahydrated RE oxalate, but similar to the
208
J. Therm. Anal. Cal., 84, 2006

MIXED Ce AND Gd OXALATES AND MIXED Ce AND Gd OXIDES
the dehydration leading to the anhydrous Gd oxalate
loss of two molecules of water (5.71%). Then there is a
second endothermic peak with onset at 200 and the
maximum at 218°C. The mass variation (11.49%) is al-
most equal to theoretical one due the complete release
of water (11.43 %). These two steps are present also in
the other case, but their temperature increases as the
Gd content increases. The first peak has for x=0.5
and 0.7 the maximum at 137 and 144°C, respectively,
while for the second one the maximum shifts to 222
and 231°C. Moreover, the area of the second peak in-
creases. The behaviour of the anhydrous mixed oxa-
lates is complex, and different solid intermediate
phases can form. For x=0.4 and 0.5 two exothermic
signals can be noticed, while for x=0.7 the former is al-
most absent. For the first signal Tonset (310°C for x=0.4
and 330°C for x=0.5) and its Tmax (330°C for x=0.4 and
340°C for x=0.5) increase as a function of x, as a possi-
ble consequence of the fact that Gd is heavier than Ce.
Moreover, its area decreases for Gd rich samples. The
second peak is the most intense and analogously to the
former, it is observed at higher temperatures as the Gd
content increases. The sample mass decreases reaching
the final mass in a complex way. For x=0.4 the sample
mass is almost constant between 235 and 300°C, then
between 300 and 360°C, there is a continuous mass de-
crease equal to 15%. Then, in a very narrow interval,
between 360 (the onset of the most intense exothermic
DTA peak) and 373°C, there is a mass release equal
to 8%. At the end of the DTA signal the sample decom-
position is almost complete. For x=0.7, on the contrary
the sample mass is almost constant until the Tonset of the
DTA peak, about 370°C. Between 370 and 390°C the
sample loses 18% of its mass. Then the slope of TG
curve decreases and the sample reaches slowly its final
mass. The oxalate with x=0.5 behaves in an intermedi-
ate way. These results can indicate that the decomposi-
tion follows a different route changing the value of x.
In any case, it seems very probable that during the de-
composition several solid intermediates can form. Fur-
ther analyses are required in order to determine their
nature and to present a rigorous decomposition path.
The partial pressures of the gaseous substances
involved in the decomposition are an important param-
eter in the decomposition of RE oxalates, in agreement
with the Eq. (1). Figure 3 shows DTA analysis per-
formed on the oxalate with x=0.4 in oxygen and in air.
The total mass loss is almost the same in all the
cases. Moreover, the XRD analyses have shown that
the final product of the decomposition is the same. The
same number of DTA peaks and steps in TG curve can
be observed meaning that the same solid intermediate
phases form during the decomposition. It is possible to
notice that the different atmospheres have not any im-
portant effect on the dehydration steps, as it could be
expected. Actually, the only gaseous species produced
occurs in many passages, partially overlapped. From
room temperature to around 100°C there is a continu-
ous mass decreasing of about the 4.3%, i.e. close to the
calculated mass variation (4.73%) for the loss of two
molecules of water. After this, there are three endother-
mic DTA signals. The first has the onset at around 100
and the maximum at 115°C. The second DTA peak has
the onset at 155 and the maximum at 169°C and the
mass release between 25 and 200°C is 18.5%, i.e. simi-
lar to that one associated to the transformation
Gd2(C2O4)3×10H2O®Gd2(C2O4)3×2H2O (18.73%). Fi-
nally, the last peak presents Tonset 200°C and Tmax
at 222°C. During this step the total loss of water oc-
curs, as the experimental mass variation between 25
and 300°C is 23.5%, while the theoretical one is
the 23.7%. The Gd2(C2O4)3 is stable in a narrow range
of temperatures, between about 300 and 390°C. More-
over, the sample mass slight decreases even in this
temperature range, meaning that the decomposition is a
continuous process. From about 390 to 550°C there is a
great mass loss, very similar to the expected one for the
transformation of the sample into the oxocarbonate
Gd2O2(CO)3. In this interval there are two exothermic
DTA signals, the former at about 415°C, the latter, the
most intense, at 520°C, and it should be noticed that
the slope of the TG curve is not constant. According to
a previous work [13], it results that the decomposition
of Gd2(C2O4)3 passes initially through the formation of
Gd2O2(CO)3 and Gd2(CO3)3. Finally there is a small
endothermic process at about 575°C, where the forma-
tion of Gd2O3 occurs.
In the case of the mixed oxalates the decomposi-
tion is completed below 600°C. The observed total
mass losses between 25 and 700°C for the samples
with x=0.4, 0.5, 0.7 are, respectively, 45.32, 44.48
and 45.43%. In the presence of oxygen the Ce–Gd oxa-
lates convert into (Ce1–xGdx)O2–x/2 following this gen-
eral mechanism:
(Ce1–xGdx)2(C2O4)3×nH2O + (2 – x/2)O2®
(1)
2(Ce1–xGdx)O2–x/2 + 6CO2 + nH2O
The observed mass variations are close to the cal-
culated ones assuming n=4, being for x=0.4, 0.5, 0.7,
respectively, 44.20, 44.79 and 45.38%. Probably the
equilibrium value of n is 10 [10] like in the pure oxa-
lates of Ce and Gd, but it seems that these sample form
with a lower amount of water. The substitution of Ce
by Gd makes the molecules of water not equivalent. In-
stead of a single stage during which all the water is re-
leased as for the Ce oxalate, more steps are observed.
In the case of the sample with x=0.4, there is a small
endothermic peak at 128°C and the sample mass
changes between 25 and 170°C of about 5.25%, i.e. a
quantity close to the theoretical variation due to the
J. Therm. Anal. Cal., 84, 2006
209

UBALDINI et al.
Fig. 3 DTA analysis performed on Ce–Gd oxalate with x=0.4
in O2 and in air
Fig. 4 XRD patterns of the (Ce1–xGdx)O2–x/2 with x=0, 0.4, 0.5,
0.7 and 1; lines mark the peaks of the F-type structure
during these processes is water and thus the chemical
equilibria do not depend on the oxygen partial pres-
sure. On the contrary the peaks due to the decomposi-
tion of anhydrous oxalates depend on the oxygen par-
tial pressure. Increasing oxygen partial pressure leads
to a decrease of their temperatures. It results that the
former one shifts towards slightly higher temperatures
in air in comparison to the analysis performed in O2:
the maximum temperatures are at about 340 and
334°C, respectively. For the second process, the differ-
ence of temperature in air and in oxygen is smaller, be-
ing about 3°C.
and 0.015, respectively. This behaviour is due to the
fact that the small peak is due only to C-type
(Ce1–xGdx)O2–x/2 solid solution, while the main peak is
due both to the contribution of this phase and of
fluorite-type solid solution. However, in no cases the
main peaks are divided in two peaks, as it could be
expected if the system was formed by two separate
phases, namely a Ce saturated C-type solid solution
and a Gd saturated F-type one.
This could be due to the fact that this separation
can not be instrumentally solved, in our experimental
conditions, or to the fact that it does not exist. As a
consequence, we can not exclude the presence of a
biphasic region or the existence of a superstructure.
Several reports exist on the structures of these mixed
oxides [7–10]. The authors are in agreement on the
fact that for values of x close to 0 a solid solution
characterized by the fluorite structure exists, while
for x close to 1 a solid solution with C-type structure
forms, but some disagreements exist about the range
of existence of these solid solutions. It is important to
notice that the cell constants reach the greatest value
for x close to 0.5, starting from both sides of the dia-
gram, so deviating from the Vegard’s law. For x>0.5
this phenomenon can be explained considering that
the Ce ionic radius is slightly larger than that of Gd,
while for x<0.5 considering that the oxygen content
decreases to keep the electroneutrality of the system
and so the repulsion force among the cations in-
creases. Considering the presence of the F-phase as
well as of the C-phase, the calculated lattice parame-
ters are reported in Table 1. They are in agreement
with the values presented in literature independently
from the synthesis method [10, 15].
Structure of Ce–Gd oxides
CeO2 is a fluorite-like cubic oxide (F-type), belonging
to the spatial group Fm-3m and the metal atoms are
eight coordinated. Gd2O3 crystallises in a body center
cubic structure, called C-type, belonging to the spatial
group Ia3. For this structure Z=16 and so there
are 32 cations and 48 oxygen atoms in the unit cell. The
structure of the C-type RE oxides is formally derived
from that of REO2 oxides (F-type) removing one fourth
of the oxygen atoms from this network along
nonintersecting strings in the four <111> direc-
tions [9, 14]. Every cation is six coordinated [9]. As the
structures and the reduced cell volumes of Gd2O3 and
CeO2 are very similar, it can be expected that their XRD
patterns are also similar. Actually, the main peaks of the
diffractogram of Gd2O3 are at almost the same angles as
those of CeO2, but in the former many weak reflections
are present that do not appear in the latter, as it is possi-
ble to observe in Fig. 4.
The results of this work show that for all the
intermediate compositions many small peaks exist,
which can be referred to the C-structure. It results that
the ratio between the areas of these secondary peaks
and the main peak increases as function of x.
The main reflection is at about 28° and one of the
most intense peaks not present in the pure CeO2 is at
about 43°: the ratio between these peaks for
x=0.7, 0.5 and 0.4 decreases, being 0.033, 0.021
Table 1 Calculated lattice parameters of (Ce1–xGdx)O2–x/2
x
F-phase
C-phase
0.4
0.5
0.7
5.430(1) 
5.425(3) 
5.428(2) 
10.860(2) 
10.850(5) 
10.846(3) 
210
J. Therm. Anal. Cal., 84, 2006

MIXED Ce AND Gd OXALATES AND MIXED Ce AND Gd OXIDES
4 J. F. Q. Ray and E. N. S. Muccillo, J. Eur. Ceram. Soc.,
24 (2004) 1287.
Conclusions
5 S. Zha, C. Xia and G. Meng, J. Power Sources,
115 (2003) 44.
Several mixed Ce–Gd oxalates were prepared by a
coprecipitation method. The thermal decomposition of
these mixed oxalates was studied. All samples are hy-
drated, but the initial water content can be lower than
the equilibrium content. Ce oxalate decahydrated de-
composes into CeO2 following a two steps mechanism,
passing through the anhydrous oxalate. The Gd oxalate
decomposes according to a more complex mechanism,
with many dehydration steps, as not all the molecules
of water are equivalent, and through the formation of
several intermediates. The substitution of Ce by Gd
makes not equivalent the molecules of water. Oxygen
partial pressure is an important parameter for the ther-
mal decomposition. In all these cases the decomposi-
tion product is (Ce1–xGdx)O2–x/2. For x close to 0 or to 1
two solid solutions exist and for the central composi-
tions the presence of a biphasic region or the existence
of a superstructure can be inferred.
6 J. Ma, T. S. Zhang, L. B. Kong, P. Hing and S. H. Chan,
J. Power Sources, 132 (2004) 71.
7 Z. Tianshu, P. Hing, H. Huang and J. Kilner,
Solid State Ionics, 148 (2002) 567.
8 Y. Ikuma, K. Takao, M. Kamiya and E. Shimada,
Mater. Sci. Eng. B, 99 (2003) 48.
9 R. G. Haire and L. Eyring, Handbook on the Physics and
Chemistry of Rare Earths, Eds K. A. Gschneidner Jr. and
L. Eyring, North-Holland Publishing Company, 1994,
Vol. 18, p. 413.
10 V. Grover and A. K. Tyagi, Mater. Res. Bull., 39 (2004) 859.
11 A. Elizabeth, C. Joseph, I. Paul, M. A. Ittychen,
K. T. Mattew, A. Lonappan and J. Jacob, Mater. Sci. Eng. A,
391 (2005) 43.
12 Y. Altas and H. Tel, J. Nucl. Mater., 298 (2001) 316.
13 G. A. M. Hussein, M. H. Khedr and A. A. Farghali,
Colloids Surf. A, 203 (2002) 137.
14 E. Antic-Fidancev, J. Hölsa and M. Lastusaari,
J. Alloys Compd., 341 (2002) 82.
15 D. J. M. Bevan and E. Summerville, Handbook on the Phys-
ics and Chemistry of Rare Earths, Eds K. A. Gschneidner
Jr. and L. Eyring, North-Holland Publishing Company,
1979, Vol. 3, p. 401.
References
1 Y. Okawa and Y. Hirata, J. Eur. Ceram. Soc., 25 (2005) 473.
2 J. Ma, T. Zhang, L. Kong, P. Hing, Y. Leng and S. Chan,
J. Eur. Ceram. Soc., 24 (2004) 2641.
DOI: 10.1007/s10973-005-7209-3
3 V. V. Kharton, F. M. B. Marques and A. Atkinson,
Solid State Ionics, 174 (2004) 135.
J. Therm. Anal. Cal., 84, 2006
211