Full text of DOI:10.1557/JMR.2000.0405

Measurement of Gibbs energy of formation of LaGaO3 using
a composition-graded solid electrolyte
K. Thomas Jacoba)
Department of Metallurgy and Materials Research Center, Indian Institute of Science,
Bangalore 560 012, India
Niladri Dasgupta
Ceramic Technology Institute, Bharat Heavy Electricals Limited, Bangalore 560 012, India
Helfried Na¨fe and Fritz Aldinger
Max-Planck-Institut fu¨r Metallforschung, Heisenbergstrasse 5, D-70569 Stuttgart, Germany
(Received 24 July 2000; accepted 16 September 2000)
A composition-graded solid electrolyte (LaF3)y и (CaF2)1−y was used for the
measurement of the standard Gibbs energy of formation of LaGaO3 from its
component oxides. An equimolar mixture of CaO and CaF2 was employed as the
reference electrode. The composition of the working electrode depended on temperature.
A three-phase mixture of LaGaO3 + Ga2O3 + LaF3 was used in the temperature range
from 910 to 1010 K, while a mixture of LaGaO3 + Ga2O3 + LaO1−xF1+2x was
employed from 1010 to 1170 K. Both the reference and working electrodes were
placed under pure oxygen gas. Because of the high activity of LaF3 at the working
electrode, there was significant diffusion of LaF3 into CaF2. The composition-graded
electrolyte was designed to minimize the electrode–electrolyte interaction. The
concentration of LaF3 varied across the solid electrolyte; from y ס 0 near the
reference electrode to a maximum value y ס 0.32 at the working electrode. For the
correct interpretation of the electromotive force at T > 1010 K, it was necessary to use
thermodynamic properties of the lanthanum oxyfluoride solid solution. The standard
Gibbs energy of formation of LaGaO3 from its component oxides according to the
1
1
2
reaction, ⁄
2
La2O3 (A-rare earth) + ⁄ Ga2O3 (␤) → LaGaO3 (rhombohedral) can be
represented by the equation: ⌬Gof,(ox)/J mol−1 ס −46 230 + 7.75 T/K (±1500).
I. INTRODUCTION
trolyte, La2O3 + LaF3 as the reference electrode, and
LaF3 + LaGaO3 + Ga2O3 as the working electrode. Both
electrodes were placed under flowing oxygen gas. The
formation of lanthanum oxyfluoride from a mixture of
La2O3 and LaF3 is well established in the literature.5–7
Therefore, the reference electrode used by Azad et al.4 is
unstable in the temperature range of their measurements.
Because Azad et al.4 ignored the formation of the oxy-
fluoride phase, their results are unreliable. Kanke and
Navrotsky8 have measured the enthalpy of formation of
LaGaO3 from oxides as −50.86 (±2.92) kJ mol−1 using
high-temperature solution calorimetry.
Recently, the thermodynamic properties of the system
La2O3–LaF3 were measured by the authors in the tem-
perature range from 910 to 1170 K.9 A solid solution
with fluorite structure LaO1−xF1+2x, 0 ഛ x ഛ 0.33, is
present at high temperature. On cooling the stoichiomet-
ric LaOF transforms to a rhombohedral structure (R3m)
at temperatures below 778 (±3) K. The solid solution
containing excess fluorine has a tetragonal structure at
room temperature.5,7
Doped LaGaO3 is an excellent oxide ion conductor
with potential application in solid-oxide fuel cells.1,2
Doping of Sr on the La site and Mg on the Ga site is an
effective method for increasing oxygen vacancies and
enhancing ionic conductivity of LaGaO3. Thermody-
namic data for pure and doped LaGaO3 are useful for
evaluating the stability of the electrolyte under reducing
conditions and its compatibility with electrode materials.
Phase diagram for the system La2O3–Ga2O3 shows the
presence of two congruent melting interoxide com-
pounds; LaGaO3 melting at 1988 (±20) K, and La4Ga2O9
melting at 1977 (±20) K.3
Azad et al.4 attempted to measure the standard Gibbs
energy of formation of LaGaO3 from its constituent ses-
quioxides using a cell based on CaF2 as the solid elec-
a)Address all correspondence to this author.
e-mail: katob@metalrg.iisc.ernet.in
2836
J. Mater. Res., Vol. 15, No. 12, Dec 2000
© 2000 Materials Research Society
http://journals.cambridge.org
Downloaded: 18 Mar 2015
IP address: 140.182.176.13

K.T. Jacob et al.: Measurement of Gibbs energy of formation of LaGaO3 using a composition-graded solid electrolyte
II. EXPERIMENTAL METHODS
A. Materials
Both oxide and fluoride solid electrolytes can be used
to measure the standard Gibbs energy of formation of
LaGaO3 at high temperatures. The oxygen chemical poten-
tials established by phase mixtures Ga + La2O3 + La4Ga2O9
and Ga + La4Ga2O9 + LaGaO3 can be determined using
a cell incorporating yttria-doped thoria as the solid elec-
trolyte. At the low oxygen potentials established by these
mixtures, stabilized-zirconia solid electrolytes exhibit
significant electronic conductivity (te > 0.01). For the
design of a suitable solid-state cell based on a fluoride
solid electrolyte, it is important to know the composition
of the fluoride phase in equilibrium with LaGaO3 and
Ga2O3. The aim is to measure the activity of La2O3 in the
two-phase mixture containing LaGaO3 and Ga2O3.
Presented in this article is the measurement of the
Gibbs energy of formation of LaGaO3 from its com-
ponent oxides using a solid electrolyte based on CaF2.
Preliminary experiments indicated that the fluoride
phase that coexists with LaGaO3 and Ga2O3 was LaF3
for T < 1010 K and an oxyfluoride solid solution
LaO1−xF1+2x at T > 1010 K. Therefore, working elec-
trodes consisting of LaGaO3 + Ga2O3 + LaF3 and
LaGaO3 + Ga2O3 + LaO1−xF1+2x were used in a solid-
state cell incorporating single-crystal CaF2 as the solid
electrolyte and CaO + CaF2 as the reference electrode.
Both electrodes were placed under flowing oxygen gas.
However, there was significant interaction between the
working electrode and the solid electrolyte, with migration
of La3+ from the electrode into the electrolyte. A com-
Optical-grade single crystals of CaF2, in the form of
disks 1.5 cm in diameter and 0.2 cm thick, were obtained
from Harshaw Chemical Company. Ultrapure anhydrous
powders of CaF2, LaF3, CaO, La2O3, and Ga2O3 were
supplied by Johnson Matthey Ltd., London, UK. The
oxides were heated at 1273 K in dry inert gas for 10 ks
before use. For the preparation of LaGaO3, stoichiomet-
ric amounts of La2O3 and Ga2O3 were ball milled, pel-
letized, and reacted at 1673 K for 36 ks. The pellet was
reground, repelletized, and reacted again at the same tem-
perature for an identical period. The formation of
LaGaO3 was confirmed by powder x-ray diffraction
(XRD). At room temperature, LaGaO3 has an orthorhom-
bic unit cell (space group Pbnm) with a ס 0.5524, b ס
0.5492, and c ס 0.7775 nm. At 1023 K, XRD revealed a
unit cell with rhombohedral symmetry (space group
R3c), with lattice parameters a ס 0.5561 and c ס
1.3567 nm in the hexagonal setting.
Solid solutions, LaO1−xF1+2x, were prepared by solid-
state reaction. Fine powder of La2O3 was mixed with
LaF3 in the required ratio. The fully homogenized mix-
tures were compacted into pellets, buried in a powder
mixture of the same composition, and reacted in a pre-
purified inert-gas atmosphere first at 1275 K for 36 ks
and subsequently at 1475 K for 18 ks:
a LnF3 + b Ln2O3 → (a + 2b) LnO1−xF1+2x , (1)
position-graded solid electrolyte (LaF3)y и (CaF2)(1−y)
,
with axial variation of y, was designed to minimize the
effect of La3+ diffusion on cell performance. The com-
position of the graded electrolyte in equilibrium with
LaF3 was defined by y ס 0.32. The composition of the
graded electrolyte in equilibrium with LaO1−xF1+2x was
found by trial and error. The cells can be represented as:
with x ס (a − b)/(a + 2b). The high-purity Ar gas
(99.999%) used was first dehydrated by passing through
anhydrous Mg(ClO4)2 and P2O5, and then deoxidized by
passing through Cu turnings at 750 K and Ti granules at
1150 K. Pellets were pulverized after heat treatment. The
formation of LaO1−xF1−2x was confirmed by XRD. The
oxyfluoride solid solution has cubic structure at high
temperature and tetragonal structure at low temperature.
The high-temperature phase can be retained by rapid
quenching into chilled mercury. The lattice parameters of
the tetragonal phase are plotted as a function of compo-
sition in Fig. 1. The lattice parameters of the cubic phase
at room temperature vary linearly from a ס 0.5765 nm
for x ס 0 to a ס 0.5793 nm for x ס 0.3.
Lanthanum oxyfluoride solid solutions of various
compositions were equilibrated at different tempera-
tures with an equimolar mixture of LaGaO3 and
Ga2O3. The fluoride phase in equilibrium was identi-
fied by XRD. Below 1010 K, LaF3 was the equilibrium
phase. At higher temperatures, the oxyfluoride solid
solution was stable. The equilibrium composition of
the oxyfluoride phase was obtained from its lattice
parameter as shown in Fig. 1. The equilibrium compo-
sition of the lanthanum oxyfluoride phase is defined
(Cell I) Pt, O2, CaO
+ CaF2 ࿣ CaF2 ࿣ (LaF3)y и (CaF2)(1−y) ࿣ LaGaO3
y = 0
y = 0.32
+ Ga2O3 + LaF3, O2, Pt
(Cell II) Pt, O2, CaO
+ CaF ࿣ CaF ࿣ LaF ͒ и CaF ͒
࿣ LaGaO3
͑
͑
2
2
3 y
2
͑
1−y
͒
y = 0
y = 0.28
+ Ga2O3 + LaO1−xF1+2x, O2, Pt␶
with the right-hand electrodes positive. Single-crystal
CaF2 was coupled with a polycrystalline composition-
graded solid electrolyte to form a bielectrolyte combina-
tion. The single-crystal CaF2, placed adjacent to the
CaO + CaF2 reference electrolyte, prevented grain-
boundary diffusion of CaO into the solid electrolyte.
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 18 Mar 2015
2837
IP address: 140.182.176.13
http://journals.cambridge.org

K.T. Jacob et al.: Measurement of Gibbs energy of formation of LaGaO3 using a composition-graded solid electrolyte
pare such a structure. In the present study, pellets were
prepared in which the composition was varied in steps of
5 mol%; homogeneous solid solutions of CaF2 contain-
ing 5, 10, 15, 20, 25, 28, and 32 mol% LaF3. Homoge-
neous solid solutions were first prepared by solid-state
diffusion. Intimate mixtures of fine powders (d < 1 ␮m)
of CaF2 and LaF3 in the required proportions were pel-
letized and heat treated at 1475 K for approximately
400 ks under prepurified Ar gas.
The functionally graded solid electrolyte was prepared
by repeated consolidation of layers, each having a uni-
form composition. A weighed quantity of CaF2 powder
was pressed in a steel die at 150 MPa, solid solution
having the next composition (5 mol% LaF3) was then
placed over it and consolidated again at the same pres-
sure. The procedure was repeated with successive com-
positions of the solid solution until the graded structure
visualized in Fig. 2 was obtained of the type. A similar
graded structure with terminal composition y ס 0.28 was
used in cell II. After final compacting at 200 MPa, the
composite pellet was sintered under prepurified Ar gas
at 1375 K for approximately 300 ks. Processing of
the solid solutions containing lanthanum was carried out
in a glove box to prevent contamination, especially by
moisture.
FIG. 1. Lattice parameters of the solid solution LaO1−xF1+2x with
tetragonal structure as a function of composition (x) at room tempera-
ture. The composition of the oxyfluoride solid solution in equilibrium
with LaGaO3 + Ga2O3 at 1150 K is shown by the vertical dotted line.
by x ס 0.281 (or XLaF ס 0.685) at 1150 K. The equi-
3
librium value of x was found to very from 0.313 at
1010 K to 0.280 at 1170 K.
Similarly, the composition of the fluoride solid solu-
tion (LaF3)y и (CaF2)(1−y) in equilibrium with the oxy-
fluoride phase LaO0.72F1.56 was determined by direct
equilibration of the phases at 1150 K under prepurified
inert gas, and subsequent analysis by XRD. A large ex-
cess of the oxyfluoride phase was used so that its com-
position was relatively unaltered. The lattice parameters
of the phases present in the quenched sample were de-
termined by XRD at room temperature. The lattice
parameter of the solid solution (LaF3)y и (CaF2)(1−y) has
been determined earlier as a function of composition
(a/nm ס 0.5465 + 0.0554y).10 Using this information,
the equilibrium value y was found to be 0.25 at 1150 K.
The equilibrium value of y also changed with tempera-
ture, from 0.32 at 1010 K to 0.24 at 1170 K. An average
value of y ס 0.28 was used at the terminal composition
of the solid electrolyte in cell II. Small adjustments in
composition at the electrode occurred during equilibra-
tion at each temperature of measurement.
Although it is preferable to have a composition-graded
solid electrolyte (LaF3)y и (CaF2)(1−y) with linear varia-
tion in composition (y), in practice it is difficult to pre-
FIG. 2. Profile of the composition-graded solid electrolyte
(LaF3)y и (CaF2)1−y. The value of y varies from 0 to 0.25 in steps of
0.05.
2838
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 18 Mar 2015
http://journals.cambridge.org
IP address: 140.182.176.13

K.T. Jacob et al.: Measurement of Gibbs energy of formation of LaGaO3 using a composition-graded solid electrolyte
Oxygen gas of purity greater than 99.999% was dried
by passing it through towers containing silica gel and
anhydrous magnesium perchlorate, and finally over boats
containing anhydrous phosphorus pentoxide before use
in the solid-state cell.
foils between them. The cell assembly was mounted in-
side a vertical outer alumina tube, the ends of which were
closed with water-cooled brass caps. After assembling
the cell and fitting the brass caps with O-ring seals to the
outer alumina tube, the tube was evacuated and leak
tested. The outer alumina tube was then suspended in a
vertical resistance furnace such that the electrodes were
in the constant temperature zone (±1 K). The cell was
maintained under dry oxygen gas at a pressure of
0.1 MPa, flowing at 2.5 ml s−1. The flow rate was set by
a mass flow controller. Even trace amounts of water va-
por in the gas were found to react with CaF2, which
produced tiny specks of CaO on the surface of CaF2.
Unless water vapor was eliminated, the transparent single
CaF2 crystals soon became opaque. The composition-
graded electrolyte was also very susceptible to degrada-
tion by moisture.
B. Procedure
The reference electrodes of cells I and II were prepared
by heating a compacted equimolar mixture of CaO and
CaF2 under dry oxygen at 1280 K. The two working
electrodes were prepared in a similar way using
equimolar mixtures of Ga2O3 + LaGaO3 + LaF3 and
Ga2O3 + LaGaO3 + LaO0.72F1.56. Intimate mixing of the
fluoride and oxide phases at the electrodes was required
to generate the equilibrium fluorine potential at each
electrode under an atmosphere of oxygen gas.
A schematic diagram of the apparatus used in this
study is shown in Fig. 3. The electrode pellets were
spring loaded on either side of the bielectrolyte assembly,
with a thin platinum gauze placed between the electrolyte
and each electrode. Platinum wires, spot welded to the
platinum gauze, were used as electrical leads to a high-
impedance (1016 ⍀) voltmeter. The working electrode
was placed in contact with the composition-graded elec-
trolyte. The pellets were held together under pressure by
a system that consisted of a closed-end alumina tube, an
alumina rod, and springs that were attached to the brass
cap. Direct contact of the electrode pellets with the alu-
mina rod or tube was prevented by inserting platinum
The electromotive force (EMF) of cell I was measured
in the temperature range from 910 to 1010 K; cell II from
1010 to 1170 K. A Pt/Pt–13%Rh thermocouple that was
calibrated against the melting point of gold was used for
measuring the temperature of the cell. All temperatures
in this study refer to the ITS-90 scale.
The time required to attain steady EMF value varied
from approximately 60 ks at a temperature of 910 K to
about 18 ks at a temperature of 1170 K. The reversibility
of the cell EMF was checked by passing a small direct
current of approximately 10 ␮A in either direction
through the cell for approximately 0.6 ks using an exter-
nal potential source, and verifying that the EMF returned
to the steady value before each microcoulometric titra-
tion. The chemical potentials at each electrode were dis-
placed from equilibrium by an essentially infinitesimal
amount during each titration. Because the cell EMF re-
turned to the same value after successive displacements
in opposite directions, the attainment of equilibrium at
each electrode was confirmed. The EMF was indepen-
dent of the flow rate of oxygen through the cell in the
range 1.5 to 5 ml s−1. After each run, the electrode pellets
were examined by XRD and energy-dispersive x-ray
analysis. No significant change in composition of the
electrodes was detected.
III. RESULTS AND DISCUSSION
A. Interpretation of the EMF of the solid-state cell
The excess anions in the cubic solid solution
LayCa1−yF2+y occupy interstitial sites as they do in solu-
tions of YF3 in CaF2.11 The solid solution has higher
ionic conductivity than pure CaF2. Because fluoride ions
are the mobile species in CaF2 and LayCa1−yF2+y, the
EMF of a solid-state cell (E) based on these electrolytes
is given by
FIG. 3. Schematic diagram of the apparatus used for EMF
measurement.
␩FE ס ␮Ј − ␮Љ
,
(2)
F2
F2
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 18 Mar 2015
2839
http://journals.cambridge.org
IP address: 140.182.176.13

K.T. Jacob et al.: Measurement of Gibbs energy of formation of LaGaO3 using a composition-graded solid electrolyte
where ␮Ј and ␮Љ (␮Ј > ␮Љ ) are the chemical potentials
For cell II, the temperature dependence of EMF can be
expressed by the relation
F2
F2
F2
F2
of fluorine at the two electrodes, ␩ ס 2 is the number of
electrons associated with the electrode reactions, and
F ס 96 485.3 C mol−1 is the Faraday constant. Under the
experimental conditions used in this study, the transport
number of fluoride ion is greater than 0.99 in both pure
and doped CaF2.12,13
The Nernstian response of a cell incorporating the
composition-graded solid electrolyte LayCa1−yF2−y has
been demonstrated in an earlier study.10 It has been shown
both by experiment and theoretical analysis that the EMF
of a cell that incorporates a composition-graded electro-
lyte is determined by the chemical potentials of the
neutral form of the mobile species at the electrodes,
when there is only one mobile ion with transport
number close to unity.14–18 The concentration gradient
of relatively immobile ions does not result in a diffu-
sion potential.14–18
EII/mV ס 552.1 − 0.0986 T/K (±0.43)
.
(4)
The uncertainty limit corresponds to twice the standard
error estimate. When attempts were made to measure the
EMF of cell I at T > 1010 K, the EMF drifted down
gradually and approached the EMF of cell II after ap-
proximately 50 ks. This indicates gradual conversion of
LaF3 to LaO1−xF1+2x at T > 1010 K. In an earlier inves-
tigation of the system La2O3–LaF3, the authors measured
the EMF of the cell,
(Cell III) Pt, O2, CaO
+ CaF2 ࿣ CaF2 ࿣ (LaF3)y и (CaF2)(1−y)
࿣ ,
y = 0
y
LaO1מxF1+2x, O2, Pt
as a function of temperature and composition (x) of the
oxyfluoride phase.9 The reference electrode was the
same in cells I, II, and III. The EMF of cell III is related
to the ratio of activities, (aLa O /a2LaF ), in the oxyfluoride
B. Gibbs energy of formation of LaGaO3
The reversible EMFs of cells I and II are plotted as a
function of temperature in Fig. 4. The EMF of cell I is
almost constant at 452.5 (±0.1) mV. The linear least-
squares regression analysis gives the relation
2
3
solid solution. Figure 5 shows3the EMF of cell II super-
imposed on the plot of EIII as a function of composition
of the oxyfluoride (XLaF ) at 1150 K. The point of inter-
3
EI/mV ס 453.3 − 7.6 × 10−4 T/K (±0.1)
.
(3)
FIG. 5. Superimposition of the EMF of cell II on the EMF of cell III
FIG. 4. Temperature dependence of the reversible EMF of cells I
and II.
as a function of composition (XLa F ) of the lanthanum oxyfluoride
solid solution at 1150 K.
3
2840
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 18 Mar 2015
http://journals.cambridge.org
IP address: 140.182.176.13

K.T. Jacob et al.: Measurement of Gibbs energy of formation of LaGaO3 using a composition-graded solid electrolyte
section of the EMFs defines the composition of the oxy-
fluoride phase in equilibrium with the phase mixture
Ga2O3 + LaGaO3 at the working electrode of cell II. As
seen from Fig. 5, the composition is defined by
XLaF ס 0.685 (x ס 0.281) at 1150 K. This value is
ident3ical to that obtained from lattice parameter meas-
urement of the sample quenched from 1150 K.
2
1
LaO1−xF1+2x + ⁄
2
O2 + 2 e−
1 + 2x͒
͑
1
→
La2O3 + 2 F−
.
(13)
1 + 2x͒
͑
The activity of La2O3 at the electrode is defined by the
dissociation of LaGaO3 (two-phase equilibria involving
LaGaO3 and Ga2O3):
The electrochemical reaction at the working electrode
on the right-hand side of cell I can be written as
2
1
LaGaO3 →
Ga2O3
−
2
1
1
2
3
1 + 2x͒
1 + 2x͒
͑
͑
⁄
3
LaF3 + ⁄
3
Ga2O3 + ⁄
2
O2 + 2 e → ⁄
LaGaO3
(5)
+ 2 F−
.
1
+
La2O3
.
(14)
1 + 2x͒
͑
The corresponding reaction at the reference electrode
on the left-hand side of cell I can be represented as
The overall reaction at the working electrode, obtained
by combining Eqs. (13) and (14), is
−
1
CaO + 2 F → CaF2 + ⁄
2
O2 + 2 e−
.
(6)
2
1
1
LaO1מxF1+2x
+
Ga2O3 + ⁄
2
O2
(15)
When the oxygen partial pressure is the same over both
electrodes, the EMF of cell I is related to the standard
Gibbs energy change for the virtual cell reaction
1 + 2x͒
1 + 2x͒
͑
͑
2
1 + 2x͒
+ 2 eמ
→
LaGaO3 + 2 Fמ
.
͑
2
LaF3 + 1⁄
⌬G07 րJ mol−1 = −2FEIր1000
3
Ga2O3 + CaO → 2⁄
3
LaGaO3 + CaF2
,
(7)
Basically, the lanthanum oxyfluoride phase converts the
activity of La2O3 established by the phase mixture
LaGaO3 + Ga2O3 into an equivalent fluorine potential
under an atmosphere of pure oxygen. Reaction at the
reference electrode of cell II is identical to that of cell I,
which has been discussed above. Because the oxygen
partial pressure is the same over both electrodes, the
EMF of cell II is related to the standard Gibbs energy
change for the virtual cell reaction
⁄3
͑
͒
= −87 473 + 0.15 T K ±20͒ .
ր
͑
(8)
For the reaction
1
2
CaF2 + ⁄
3
La2O3 → CaO + ⁄
3
LaF3
,
(9)
standard Gibbs energy change in the temperature range
from 900 to 1200 K, computed from data in the com-
pilation of Knacke et al.,19 can be represented by the
equation
2
1
LaO1−xF1+2x
2
+
Ga2O3
1 + 2x͒
1 + 2x͒
͑
͑
+ CaO →
LaGaO3 + CaF2
,
(16)
⌬G09
ր
͒
J mol−1 = 56 645 + 5.02 T
ր
K ±1500͒
.
͑
1 + 2x
͒
͑
͑
(10)
⌬G016 րJ mol−1 = −2FEII ր1000
͑
͒
By combining Reactions (7) and (9), one obtains the
reaction that represents the formation of rhombohedral
LaGaO3 from ␤–Ga2O3 and La2O3 with A-rare-earth ox-
ide structure:
= −106 540 + 19.03 T K ±85͒ .
ր
͑
(17)
The Gibbs energy of mixing of LaO1−xF1+2x solid
solution relative to its component oxide (La2O3) and
fluoride (LaF3) has been determined recently as a func-
tion of temperature and composition (x).9 By combining
Gibbs energy change for Reactions (9) and (16) with
the Gibbs energy of mixing of the lanthanum oxyfluo-
ride solid solution, one obtains the standard Gibbs energy
of formation of LaGaO3 from Ga2O3 and La2O3, repre-
sented by Reaction (11):
1
1
2
⁄2
La2O3 A-rare earth͒ + ⁄
Ga2O3 ␤͒
͑
͑
→ LaGaO3 rhombohedral͒
,
(11)
(12)
͑
⌬G0
J mol−1 = −46 240 + 7.76 T
ր ր
K ±1500͒ ,
͑
͑11͒
in the temperature range from 910 to 1010 K. The rela-
tively large uncertainty limit arises from inaccuracies in
the thermodynamic data for oxides and fluorides of cal-
cium and lanthanum reported in the literature.19 When
more precise data for these compounds become avail-
able, it will be possible to derive more accurate data for
LaGaO3.
⌬G(o11)/J mol−1 ס −46220 + 7.74 T/K (±1500)
,
(18)
in the temperature range from 1010 to 1170 K. Thus, the
Gibbs energies of formation of LaGaO3 obtained from
the EMFs of cells I and II in different ranges of tempera-
ture are almost identical.
The electrochemical reaction at the working electrode
on the right-hand side of cell II can be written as
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 18 Mar 2015
2841
IP address: 140.182.176.13
http://journals.cambridge.org

K.T. Jacob et al.: Measurement of Gibbs energy of formation of LaGaO3 using a composition-graded solid electrolyte
The Gibbs energy change for reaction (11) can also be
evaluated by an alternate procedure involving EMFs of
cells II and III. By combining EMF of cell III with a
modified version of the Gibbs–Duhem equation,20 the
variation of chemical potentials of La2O3 and LaF3 with
composition in the binary system La2O3–LaF3 has been
delineated.9 From this information, a plot of the EMF of
cell III as a function of chemical potential can be con-
structed as shown in Fig. 6. Superimposing the EMF of
cell II on this graph directly yields the chemical potential
of La2O3 corresponding to the phase mixture
Ga2O3 + LaGaO3. The value of the Gibbs energy of
formation is equal to half the chemical potential of
o
1
La2O3: ⌬G(11) ס ⁄ ⌬␮La O . Values obtained by this
2
2 3
method are almost identical to that given by Eq. (18).
The standard Gibbs energy of formation of LaGaO3
obtained in this study is compared with the value re-
ported by Azad et al.4 in Fig. 7. The values obtained in
this study are more negative by approximately 19 kJ mol−1.
In their attempt to measure the Gibbs energy of formation
of LaGaO3, Azad et al.4 used the cell
(Cell IV) Pt, O2, La2O3 + LaF3 ࿣ CaF2 ࿣ LaF3
FIG. 7. Comparison of the Gibbs energy of formation of LaGaO3
+ LaGaO3 + Ga2O3, O2, Pt
.
obtained in this study with that reported by Azad et al.4
Their reference electrode consisted of a mixture of
LaF3 and La2O3 in approximately equal mass ratio. At
high temperature, this would have resulted in a single-
phase oxyfluoride solid solution characterized by
XLaF ס 0.625 (or x ס 0.182). According to the results
3
of this study, LaF3 is also not stable at their working
electrode at T > 1010 K; it would have equilibrated with
the other phases Ga2O3 and LaGaO3 to form an oxyfluo-
ride solid solution rich in LaF3. If the equilibrium phases
had formed during measurement, the EMF produced by
their cell can be estimated from the results of this study.
For example, at 1150 K their EMF would be the differ-
ence between EII, and EIII at XLaF ס 0.625. Their cell,
3
interpreted to function in the manner described above,
should have produced an EMF of 65.1 mV at 1150 K.
This is very close to the value of 65.4 mV at 1150 K
reported by Azad et al.4. The coincidence is remarkable,
especially because the composition of their reference
electrode is not precisely defined. However, the Gibbs
energy values reported by Azad et al.4 are clearly
incorrect.
The temperature-independent term in Eq. (12) or (18)
is related to the enthalpy of formation of LaGaO3 from
La2O3 and Ga2O3 at a mean temperature of 1040 K;
⌬Hof(ox)/kJ mol−1 ס −46.23 (±4.5). This is in reasonable
agreement with the value of −50.86 (±2.92) kJ mol−1 at
977 K reported by Kanke and Navrotsky.8 The corre-
sponding value from the EMF study of Azad et al.4 is
−24.5 (±4.5) kJ mol−1. The enthalpy of formation of
FIG. 6. EMF of cell III as a function of the relative chemical potential
of La2O3 in the lanthanum oxyfluoride solid solution LaO1−xF1+2x. The
value of ⌬␮
in the phase mixture LaGaO3 + Ga2O3 is defined
La2O3
when EII ס EIII
.
2842
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 18 Mar 2015
http://journals.cambridge.org
IP address: 140.182.176.13

K.T. Jacob et al.: Measurement of Gibbs energy of formation of LaGaO3 using a composition-graded solid electrolyte
ACKNOWLEDGMENT
LaGaO3 from the elements computed from the results of
this study using the Neumann–Kopp rule is ⌬Hfo(298.15)
/
One of the authors (K.T.J.) thanks the Max-Planck-
Institut fu¨r Metallforschung, Stuttgart, for a Visiting Pro-
fessorship, which made possible the collaborative
research.
kJ mol−1 ס −1488.25 (±5.5). The enthalpies of forma-
tion of Ga2O3 and La2O3 at 298.15 K from Knacke
et al.19 are used in this evaluation. The temperature-
dependent term in Eq. (12) or (18) is related to the en-
tropy of formation of LaGaO3 from oxides; ⌬Sof(ox)
/
kJ mol−1 K−1 ס −7.75 (±4.4). Assuming that the heat
capacity of LaGaO3 is the sum of the heat capacities of
La2O3 and Ga2O3, the standard entropy of LaGaO3 at
REFERENCES
298.15 K can be estimated as Sof(298.15)/J mol−1 K−1
ס
1. T. Ishihara, H. Matsuda, and Y. Takita, J. Am. Chem. Soc. 116,
3801 (1994).
2. M. Feng and J.B. Goodenough, Eur. J. Solid State Inorg. Chem.
31, 663 (1994).
3. M. Mizuno, T. Yamada, and T. Ohtake, Yogyo-Kyokai-Shi, 93,
[6] 295 (1985).
98.38 (±5). It would be useful to measure the heat
capacity of LaGaO3 accurately at both low and high tem-
peratures in order to obtain an independent assessment of
its entropy.
4. A.M. Azad, R. Sudha, and O.M. Sreedharan, Mater. Res. Bull. 26,
97 (1991).
IV. CONCLUSION
5. W.H. Zachariasen, Acta Crystallogr. 4, 231 (1951).
6. D.B. Shinn and H.A. Eick, Inorg. Chem. 8, 232 (1969).
7. K. Niihara and S. Yajima, Bull. Chem. Soc. Jpn. 45, 20 (1972).
8. Y. Kanke and A. Navrotsky, J. Solid State Chem. 141, 424 (1998).
9. K.T. Jacob, N. Dasgupta, H. Na¨fe, and F. Aldinger, J. Am. Ceram.
Soc. (in press).
10. K.T. Jacob, N. Dasgupta, and Y. Waseda, J. Am. Ceram. Soc. 81,
1926 (1998).
11. R.W. Ure Jr., J. Chem. Phys. 26, 1363 (1957).
12. J. Delect, R.J. Heus, and J.J. Egan, J. Electrochem. Soc. 125, 755
(1978).
13. K. Ono, T. Fujimura, and J. Moriyama, J. Jpn. Inst. Met. 42, 469
(1978).
14. S. Mukhopadhyay and K.T. Jacob, J. Electroanal. Chem. 395, 107
(1995).
15. S. Mukhopadhyay and K.T. Jacob, J. Electrochem. Soc. 140, 2629
(1993).
16. S. Mukhopadhyay and K.T. Jacob, Metall. Mater. Trans. A 25A,
173 (1994).
17. S. Mukhopadhyay and K.T. Jacob, in Proceedings of the Second
International Symposium on Ionic and Mixed Conducting Ceram-
ics, edited by T.A. Ramanarayanan, W.L. Worrell, and H.L. Tuller
(Electrochemical Society, Pennington, NJ 1994), pp. 106–126.
18. K.T. Jacob and S. Mukhopadhyay, J. Electrochem. Soc. 142, 161
(1995).
19. Thermochemical Properties of Inorganic Substances, 2nd ed., ed-
ited by O. Knacke, O. Kubaschewski, and K. Hesselmann
(Springer-Verlag, Berlin, 1991), Vols. I and II.
20. K.T. Jacob and J.H.E. Jeffes, High Temp.-High Press. 4, 177
(1972).
A composition-graded solid electrolyte was used for
the determination of the standard Gibbs energy of for-
mation of LaGaO3 in the temperature range from 910 to
1170 K. The value of y in the composition-graded
(LaF3)y и (CaF2)1−y polycrystalline solid electrolyte was
varied from 0 near the reference electrode, consisting of
a mixture of CaO + CaF2, up to 0.32 at the working
electrode. The terminal composition of the graded elec-
trolyte was adjusted such that the activity of LaF3 in
electrolyte was approximately equal to that at the work-
ing electrode. The cell essentially measures the activity
of La2O3 in the two-phase region LaGaO3 + Ga2O3 in the
binary system La2O3–Ga2O3. Under pure oxygen gas, the
equilibrium fluoride phase at the working electrode con-
verts the activity of La2O3 into an equivalent fluorine
chemical potential that can be measured using a cell es-
sentially based on doped CaF2.
The values of Gibbs energy of formation obtained in
this study are significantly different (approximately
19 kJ mol−1) from those reported earlier by Azad et al.4
The use of phase-incompatible mixtures, especially at the
reference electrode, by Azad et al.4 is the prime reason
for the difference. Their measurements, when correctly
interpreted, support the results obtained in this study.
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 18 Mar 2015
2843
IP address: 140.182.176.13
http://journals.cambridge.org