70
J.
EIectrochem. Soc.:
ELECTROCHEMICAL SCIENCE AND TECHNOLOGY
January 1983
Chem. Sot.,
93, 6691 (1971).
10. W. Giggenbach,
Inorg. Chem.,
10, 1308 (1971).
11. J. P. Bernard, A. DeHaan, and H. Van der Poorten,
C. R. Acad. Sci. Paris,
276, 587 (1973).
12. J. Hi Kennedy and F. Adamo,
This Journal,
119,
1518 (1972).
13. B. Cleaver, A. J. Davies, and D. J. Schiffrin,
Elec-
trochim. Acta,
18, 747 (1973).
1~. J. R. Birk and R. K. Steunenberg,
Adv. Chem. Set.,
740, 786 (1975).
15. D. O. Raleigh, J. T. White, and C. A. Ogden,
This
Journal,
126, 1087 (1979).
16. K. W. Kam and K. E. Johnson,
J. Electroanal.
Chem. Interracial EIectrochem.,
115, 53 (1980).
17. G. Santarini,
C. R. Acad. Sci. Paris,
268, 457 (1979).
18. C. H. Liu, A. J. Zielen, and D. M. Gruen,
This Jour-
nal,
120, 67 (1973).
19. M. J. Weaver and D. Inman,
Electrochim. Acta,
20,
929 (1975).
20. R. S. Nicholson and I. Shain,
Anal. Chem.,
36, 706
(1964).
21. D. S. Polcyn and I. Shain,
ibid.,
38, 370 (1966).
22. C. P. Andrieux, L. Nadjo, and J. M. Saveant,
J. Electroanal. Chem. Interracial Electrochem.,
26,
147 (1970).
23. G. Mamantov, D. L. Manning, and J. M. Dale,
ibid.,
9, 253 (1965).
24. R. C. Vogel etal., Argonne National Laboratory
ANL-7775, 120 (1971).
25. M. S. Shuman,
Anal. Chem.,
41, 142 (1969).
26. M. L. Saboungi, J. J. Marr, and M. Blander,
This
Journal,
125, 1567 (1978).
27. M. L. Saboungi, J. J. Marr, and M. Blander,
Metall.
Trans.,
10B, 477 (1979).
28. T. Berzins and P. Delahay, J.
Am. Chem.
Soc., 75,
555 (1953).
Oxygen Transfer on Substituted ZrO2, Bi203, and CeO2 Electrolytes
with Platinum Electrodes
I. Electrode Resistance by D-C Polarization
M. J. Verkerk, 1 M. W. J. Hammink, and A. J. Burggraaf
Twente University of Technology, Department of Chemical Engineering, Laboratory of Inorganic Chemistry and Materials
Science, 7500 AE Enschede, The Netherlands
ABSTRACT
The electrode behavior of Ptosputtered and PT-gauze electrodes on ZrO2-Y203, Bi20~-Er20.~, and CeO~-Gd203 solid electro-
lytes was investigated by means of d-c measurements in the temperature region of 770-1050 K and in the oxygen partial
pressure region of 10 -~ - 1 atm O2 using N2-O2 mixtures. On these different materials the same electrode morphology was
realized and was preserved during the subsequent experiments. The electrode process is strongly influenced by the nature
of the electrolyte. The electrode resistance for Pt electrodes on Bi203-Er203 was found to be many times lower than on
ZrO2-Y20~ and CeO2-Gd20~. On zirconia- and ceria-based materials-diffusion of atomic oxygen on the Pt electrode is the
rate-determining step in the electrode process, whereas for bismuth sesquioxide-based materials diffusion on the oxide
surfaces is rate determining.
There is a considerable interest in solid electrolytes
for use in oxygen sensors, oxygen pumps, and high
temperature fuel cells. In these applications the elec-
trode polarization and the electrolyte resistance both
play an important role. These effects influence the re-
sponse of an oxygen sensor and give rise to energy
losses and therefore to a decreased efficiency, of oxygen
pumps and fuel cells. Many studies are performed on
electrode processes. However, the dominant elementary
steps of the overall process are still unknown in most
cases.
The kinetics of the electrode process is strongly in-
fluenced by the electrode structure (1-6), electrode ma-
terial (4, 5, 7-9), and the electrolyte (4, 5, 7). Informa-
tion about the role of the electrolyte in the electrode
process is very scarce and is limited to the role of the
dopant. Schouler (4) found that the oxygen partial
pressure dependence of the electrode resistance of a
sputtered platinum electrode depends on the Y203
content of the ThO2 electrolyte. Fabry and Kleitz (7)
found that the activation energy of the electrode re-
sistance of a point electrode on calcia-stabilized zir-
conia is higher than that on yttria-stabilized zirconia.
Wang anal Nowick (5) reported that the exchange
current of a Pt-paste electrode is not influenced by the
nature and the. concentration of the dopant of the CeO2
electrolyte.
Even less attention is paid to the influence of the
nature of the electrolyte on the overall electrode pro-
i Present address: PhUips Research Laboratories, Eindhoven,
The Netherlands.
Key words: electrode resistance, solid electrolytes, Pt elec-
trodes, electrode morphology.
cess. From model considerations this influence has to
be expected if particular elementary steps are domi-
nant. In most studies reported in the literature the
measurements are performed under different circum-
stances ~ and using different and badly characterized
electrode structures and consequently the results can
hardly be compared.
Therefore, we studied the influence of the electrolyte
conductivity and electrolyte composition on the elec-
trode kinetics using different oxygen ion conductors.
An important condition is that on these different mate-
rials the same electrode morphology has to be realized.
Sputtered and gauze electrodes, subjected to several
treatments, can meet this condition. The comparison of
a sputtered and a gauze electrode on the same electro-
lyte may be a tool for investigating the rate-deter-
mining step.
The following systems were investigated:
(ZrO2) 0.ss (YO1.5)o.17 (ZY17)
(CeO2) 0.9o (GdOl.~) 0.10 (CG10)
(Bi.20~) 0.s0 (Er~O~) 0.20 (BE20')
(Bi203) o.70 (ErgO3) 0.30 (BE30)
(Bi203)o.60 (Er~Os) 0.40 (BE40)
The bulk conductivity of the stabilized bismuth sesqui-
oxides is 10-100 times higher than that of the stabilized
zirconias and the substituted cerias (10, 11). Besides,
bismuth oxide compounds are used as oxidati,m cata-
lysts and a relative rapid transfer of oxygen on the
solid-gas interface is suggested. For these reasons, we
Downloaded 22 Jun 2009 to 130.89.112.86. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp
Vol.
130, No. I
OXYGEN TRANSFER
71
investigated the idea whether the electrode resistance
on the bismuth-based materials might be lower than
that on the zirconia- and ceria-based materials.
Special attention is paid to the possibility of using
the above-mentioned materials as a solid electrolyte in
an oxygen pump. Therefore, the measurements were
performed at relatively low temperatures and high
oxygen partial pressures.
In this paper a phenomenolpgical description of
the
influence of the nature of the electrolyte on the elec-
trode process is given (d-c study). In part II of this
study (12) the observed influence of the electrolyte is
analyzed with respect to its critical steps by means of
a frequency dispersion study.
Experimental
Sample preparation and characterization are given in
Table I. The composition of the samples was checked
with x-ray fluorescence. The diameter of the samples
was 5-10 mm and the thickness 0.7-1.5 mm.
Measurements were performed in the temperature
range 770-1050 K and with N2-O2 mixtures in the oxy-
gen partial pressure range of 10-5-1 atm O2. The oxy-
gen partial pressures in the range of 10-2-1 atm were
obtained by mixing appropriate O2/N2 mixtures and in
the range of 10-~-10-2 arm by an oxygen pump (15,
16). The oxygen partial pressure is measured using an
oxygen sensor based on stabilized zirconia. The gases
were supplied at a rate of 600 cm~/min.
The electrical circuit used in the d-c experiments is
shown in Fig. 1. A Wenking potentiostat was used. The
current was measured by the voltage drop across the
reference resistor. The voltages were measured with
an HP 3465A multimeter. The electrode resistance Rel
was measured using a voltage of 1-10 mV across the
sample and calculated according to
(
Rel -- ~ ) -- Rb [1]
I---> 0
The total bulk resistance Rb was obtained from a-c
impedance measurements (13).
The complex impedance measurements were per-
formed in the frequency range 106-10 -~ Hz using a
Solartron 1174 Frequency Response Analyzer with a
sample voltage of 10 inV. The circuit is described in
Ref. (13).
Electrode Preparation and Characterization
Platinum electrodes with a thickness of 0.75 ~m were
sputtered onto the polished samples. The sintering of
the electrode during thermal treatments is influenced
by the nature of the electrolyte. To obtain the same
electrode morphology on different electrolytes a sin-
tering temperature in the region of 1200-1400 K was
necessary (2-4 hr). To be sure that no visible changes
in the electrode morphology take place during the
measurements, the system was equilibrated for 48 hr
at
the highest measuring temperature, and an a-c
current treatment (1 kHz, 1V, 5 min) was applied. A
typical scanning electron microscope (SEM, JEOL
JSMU 3) picture is given in Fig. 2a. The morphology
of the electrode is characterized by the length of the
three-phase line and the surface area of the interfaces:
electrolyte/electrode, electrolyte/gas, and electrode/
m
..y
Fig. 1. Electrical circuit for the d-c measurements
gas. These data are given in Table II. The maximum
measuring temperature is restricted by the requirement
that the morphology of the electrode does not change
in time. This condition is satisfied if the maximum
measuring temperature is kept 150-200 K below the
sintering temperature of the metal electrode. At every
temperature the electrode was equilibrated for 16 hr
before a measurement was performed.
Platinum gauze electrodes (200 meshes/cm 2) were
plastically deformed before use to obtain a reproducible
contact surface. An overall picture of the gauze elec-
trode is shown in Fig. 2b and its characteristic data
are given in Table II. Figure 2c shows a representative
part of the contact surface of a gauze electrode used
on ZY17 or CG10 and Fig. 2d shows the same for BE20,
BE30, and BE40. From these figures it is clear that be-
cause of the roughness of the contact surface oxygen
molecules can penetrate this interface. Therefore, the
actual three-phase line will be much larger than cal-
culated on the basis of the circumference of the oval
flattened parts of the gauze electrode. The gauze elec-
trode was held with a constant pressure on the sample.
Before the measurements were performed the above-
described thermal and current treatments were applied.
After performing the d-c measurements (1-10 mV)
and subsequently the a-c measurements [reported in
part II of this study (12)], the electrode resistance was
again measured by d.c. (1-10 mV). The total measuring
time was about 600 hr. The electrode resistance of the
Pt-sputtered electrode on BE20 was increased by a
factor of two. Some Er~Q particles could be detected
on the electrode surface by means of SEM and x-ray
diffraction. The ReL for the other combinations was not
significantly changed during the measuring period.
SEM pictures showed no changes in the morphology of
the electrodes. Afterward the same samples were used
for d-c experiments in the region of 0-2V.
Reproducibility of the electrode resistance of sput-
tered and gauze electrodes was checked on three
nominally equal electrodes applied on ZY17. The elec-
tro.de resistance falls in the region of log Rel • log 1.2
and log Rel • log 2 for sputtered and gauze electrodes,
resuectively.
Preliminary experiments on ZY17 with grain sizes of
0.5, 2.5. and 20 pm showed that the electrode resistance
is not influenced by the grain size of the electrolyte.
Theory
In this section the relevant theory for the interpre-
tation of the experimental data is given.
Table II. Characteristic data of the morphology of the sputtered
(sp) and the gauze (g) electrode
Table I. Preparation and characterization of the samples
Sinter-
ing
temper- Den-
Sys- ature
sity
tern Preparation method (K) ( % )
ZY17 Alkoxide synthesis (13) 1673 99
CG10 Citrate synthesis (14) 1773 96
BE20 Solid-state reaction (I0) 119R 95
BE30 Solid-state reaction (10) 127~ 96
BE40 Solid-state reaction (10) 1323 95
Grain
size
2.5
5
30
30
30
Three-
Electrolyte/ Electro-
Electrode/
phase electrode lyte/gas gas
line contact contact
contact
Elec- length surface surface
surface
trode (m/m ~) (m2/m 2) (m~/m 2) (m~/rn~)
sp 1.2 • 10 ~ 0.64 0.36 0.8
g 1.4 • 10 ~* 0.64 0.96 1.6
g
Ratio ~ 860 16 0.38 0.50
sp
*
Determined from the circumference of the oval flattened
parts of the electrode.
Downloaded 22 Jun 2009 to 130.89.112.86. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp
72
J. E~ectrochem. Soc.:
ELECTROCHEMICAL SCIENCE AND TECHNOLOGY
January 1983
Fig. 2. Morphology of the platinum electrodes used in this study. Sputtered electrode (a, upper left), gauze electrode (b, upper right),
contact surface of a gauze electrode after use on a ZY17 or CG10 sample (c, lower left), and contact surfaces of a gauze electrode after
use on a BE20, BE30, or BE40 sample (d, lower right).
The elementary adsorption step in the electrode
process is given by
O~(g) + 2Vaas ~ 2 Pads [2]
where Vaas is a vacant adsorption site and Pads is an
adsorbed oxygen atom. The mass action relation for
this reaction can be written as (Langmuir adsorption)
0ads/(1- 0aas) =
KI(T)
X [Po2] ~" [3]
The equilibrium constant is given by
K1 (T) = Kto exp
(,SHads/RT)
[4]
where
hHaas
is the heat of adsorption.
The adsorption step is followed by diffusion to the
reaction site
Pads -{- Vrs ~ Ors "Jr Vads [5]
where Vrs and Ors are a vacant and occupied reaction
site, respectively.
At the reaction site the charge transfer process takes
place (given in KrSger-Vink notation)
Ors + 2e' + Vo" ~--- O,o x + Vrs [6]
If the charge process is rate limiting the relation be-
tween the current I and the overpotential ~ is given by
the Butler-Volmer equation (17)
I
--
Io[exp (~anV*)
--
exp (--acnV*)] [7]
V* =
~IF/RT
[8]
where ~a and ,e are the anodic and cathodic transfer co-
efficients, respectively, V* is a dimensionless potential,
Io the exchange current, and n number of charges
transferred.
However, if the anodic or cathodic process is limited
by mass transport the following equation holds (17)
I = [exp (~anV*)
--
exp
(--acnV*)]/
[/'o -I -1- Ila -1 exp (aanV*) + Ilc -1 exp
(--acnV*)
] [9]
where Ila and Ilc are the anodic and cathodic limiting
current, respectively, and are defined positive. At low
overpotentials, InV*] <~ 1, ohmic behavior is obtained
I -= nV*/(Io -1 + Ila -1 +
Ilc -1) [10]
Using [8] we find
Downloaded 22 Jun 2009 to 130.89.112.86. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp
Vol. 130, No. 1 OXYGEN TRANSFER
73
11
Rel -- -- --
Rot
"~ Rla + Rio [11]
I
RT 1 RT 1 RT 1
, Rla -- ~" "~'1, ' and Rlc = ....
with Rct -- nF Io nF la nF Ilc
From Eq. [9]-[11] it is clear that for Ila, Ilc >> Io the
electrode process is determined by the charge transfer
and for Io >>/la, Ilc by mass transport.
The oxygen atoms for the electrode process can be
supplied by dissociated adsorption of oxygen molecules
at the electrode or the electrolyte and the charge
transfer
will take place near the triple line (this will
be further discussed below). Now we assume a model
where the limiting current originates from diffusion of
atomic oxygen from the adsorption site at x = 5 to
the reaction site at x = 0. At x = 8 the concentration
oxygen atoms is constant [see part II of this study
(12) ]. We can write then for Ic and Ia
Co(ads) -- Co(rs)
Ic -- --DonF [12]
b'
Co (rs) -- Co (ads)
Ia : DonF [13]
8"
where Do and Co are the diffusion coefficient and the
concentration of the oxygen atoms. ~c and I~ are given
per surface unit of reaction area. A cathodic and anodic
limiting current is observed for Co(rs) ~ O and Co(rs)
-* Co'(rs), respectively, and are given by
Co (ads)
Ilc
: DonF [14]
5'
Co' (rs) -- Co (ads)
Ila : D onF [15]
8"
where Co'(rs) is the surface concentration of oxygen
atoms for ~ : 1.
Results and Discussion
Pt electrodes on ZY17.--For all samples the I-V
relation at small voltages shows an ohmic behavior, as
shown in Fig. 3 for ZY17. Due to a small d-c voltage
of 0-1 mV, probably originating from thermoelectric
effects, the curves do not pass through the origin. The
determination of the electrode resistance according to
Eq. [1] is not influenced by this small voltage.
The electrode resistance Rel measured as a function
of temperature for Po2 is equal to 1, 0.21. 1.6 • 10 -2,
and 9 • 10 -4 arm 02, respectively. Values of the ac-
tivation energy Ea and the pre-exponential term log Ro
are given in Table III. The results for a sputtered and
a gauze electrode at Po2 = 1.6 • 10 -2 arm are shown
8
6
p%
latm)
0.046
/ //o.,o
2 4 6 8 10
-
vtmv)
Fig. 3. Current-voltage characteristics for ZY17 at T ---- 983 K
using Pt-sputtered electrodes.
in Fig. 4. At about 960 K a bend in the Arrhenius plot
of Rel is observed and Ea changes from about 75 kJ
mo1-1 (high temperature part) to about 250 kJ mo1-1
(low temperature part). This bend is correlated with a
change from the regime where Re1 ~ Po~ -'/~ to the
regime where Rel ~ Po2 +~/~ (see below). Its position
on the temperature scale depends on the Po2 pressure
and its position on the Po2 scale on the temperature
(Fig. 5).
Figure 5a gives Rel as a function of Po2 at 983 K. For
comparison the Rel-Po2 relation calculated from Table
III for 908 K is also given in this figure. A minimum in
the ReI-Po2 curve is observed at a value of Po2 rain.. For
both sputtered and gauze electrodes at Po2 <: Po2 mira
one finds Rel ~ Po2- 2/2 and from Fig. 4 and Table III it
follows that in this region Ea has a value of 75-100 kJ
mol-L Whereas Po2 > Po2 rain" one finds Rel ~ Po2 +V=
and in this region Ea has a value of about 250 kJ mol-L
The value of Po2 rain. shifts to higher values with in-
creasing temperature.
These results are in good agreement with literature
data concerning platinum electrodes on stabilized zir-
conia. However, the reported experimental litera-
ture data are incomplete and only parts of the picture
described above were measured. A minimum in the
Table III. Activation energies and log Ro of the electrode resistance of sputtered (sp) and gauze (g) electrodes on several
electrolytes. The deviation is given in the 65% reliability interval.
Sy~em
1
arm O3 0.21 atm O2 1.6
• 10 -~ arm Oe 9 x 1O -~ arm Oe
--log Ea - log Ea -- log Ea -- log Ea
Ro (kJ Ro (kJ Range Ro (kJ Range Ro (kJ
(tim ~) tool-l) (tim 2) mol-~) (K) (~m ~) mol-~) (K) (~m ~) mol-~)
ZY17 (sp) 16.9-----0.9 270----+15 16.5 +-- 0.4 260• <970 15.3 ~ 0.6 230• <925
>970 6.6 ~-- 0.6 72 • lO >925
ZY17 (g) 13.8 • 0.6 245 • 10 15.8 • 0.7 280 -- 10 <960 9.9 167
>969 4 9 -- 0.4 75 • I0
CGIO (sp) 14,8--• 0.5 230• 14.9 ~ 0.5 321• <970 16.0• 245• <930
>970 7.7 • 03 91
'•
6 >930
CG10 (g) 12.3 • 0.3 212 -- 6 12.3 -- 0.3 298 • 6 <940 12.1 203
>940 5.1 • 0.3 77 • 5
BE20 (sp) 11.8• 145• 9.6• 115• 10.0 ~- 0.3 129•
BE20 (g) 9.7• 133• 9.6 • 0.4 137"• 8.1~ 0.4 116•
BE30 (sp) 11.1 • 0.1 138 _ 2 10.0 • 9.3 125 __- 5 9.2 • 0.2 121 • 3
BE30 (g) 8.3---4---0.3 112__.5 7.9• lC8• 6.3• 87•
BE40 (sp) 10.3+__0.4 129• 8.8__-0.5 110• 9.6-----0.3 125•
BE40 (g) 8.3• 119• 7.7• 110• 6.9+0.2 102+--3
BE20 (gold, 11.2 -~ 0.9 150 • 15
sp)
9.8 ~ 0.5 145 • 10
5.6 -- 0.5 70 + 10
5.5 • 0.1 98 • 5
10.3 --~ O.1 148 -- 1
5.4 • O.3 61 • 4
5.5 • 0.2 96- 4
2.2 • 1O-~ atm O-~
7.1 ---+ 0.3 1Ol -- 6
5.6 ---+ 0.6 83 • 10
6.8 • 0.6 99 • l0
3.4 • 0.4 45 +-- 7
5.7 -- 1.1 75 • 20
3.3• ~5 --'-Z- 6
Downloaded 22 Jun 2009 to 130.89.112.86. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp
74
J.
Electrochem. Soc.: ELECTROCHEMICAL SCIENCE AND TECHNOLOGY
January 1983
=
T(K)
lOOO 900
i i
~d
g-
E
9
0
~o 4
~o ~
~bo ~.~o ~;~o
.
IO00/T(K -1)
Fig. 4. Electrode resistance as a function of temperature for
Po~ = 1.6 X 10 -~ atm O~. O, Pt-sputtered electrode on ZY17;
O, Pt-gauze electrode on ZY17; [-], Pt-sputtered electrode on
CG10, II, Pt-gauze electrode on CGI0.
Rel-Po~
curve was reported by Schouler (4) for Pt-
sputtered and Pt-paste electrodes on (ZrO~)0.ss(YO1.5)0.1~
and by Hartung (6) for Pt-paste electrodes on
(ZrO2)0.s2(YO1.5)0.10(MgO)o.08. According to Bauerle
(2) at 1073 K, Rel ,~ Po2 -~ at low oxygen partial
pressures and becomes constant at high oxygen par-
tial pressures (0.21-1.00 atm O~) for Pt-sputtered
electrodes on yttria-stabilized zirconia (YSZ). In the
opinion of the authors the expe}imental data permit
the following interpretation: Po2 rain. has a value of
0.21-1.00 atm O2 and for Po~ <
Po2 min'
Rel ~ Po2 -0"5.
For Pt electrodes on YSZ an activation energy of about
100 kJ tool -1 was reported in Ref. (4 and 7) in the re-
gion where Rel ,~ Po2 -'/2 and a value of about 250 kJ
mol-~ in Ref. (2 and 4) in the region where Po2 >
Po2 rain.
The observed phenomena were ascribed by Schouler
(4) to oxidation of the platinum surface. This hy-
pothesis may be a good explanation for the high ac-
tivation energy found at high Po2. The observed rela-
tionship Rel ~ Po2 + '/~ at Po2 ~ Po2 min" cannot be ex-
plained with the formation of an oxidized Pt surface
because it is well known that this is catalytically less
active.
In our opinion the observed phenomena can be ex-
plained in the following way. The .oxygen atoms for
the electrode process are supplied by dissociative ad-
sorption of oxygen molecules at the Pt electrode. The
overall electrode reaction is rate determined by diffu-
sion of the oxygen atoms from the adsorption site to
the reaction side (see section on Theory). At Po~ <
Po2 min',
~ads is low and mass transport limitation occurs
at the cathode. At Po~ >
Po2 min', 0ads ~-~
1 and mass
transport limitation occurs at the anode. This hypoth-
esis is discussed in the following.
As shown by Wang and Nowick (5) the electrode re-
sistance varies as the --V4 or + 1/4 power of Poe if the
electrode process is determined by charge transfer (see
following section). The experimentally found powers
of
--V2 and 4-V2 point to mass transfer limitation and
can be derived from Eq. [14] and [15]. For ~aas << 1
holds (1 -- Oads) ~ 1 and assuming that ~' ~ 5" we
find I~ >> I~. From Eq. [3] it follows that ~ads =
K~(T) X
[Po2] '/" and using Eq. [10], [11], and [14] it
is found that
(b)
CG 10
(2)
(3)~"
-4 ' -2 ; -4 -~2
0
log po2 (atm)
Fig. 5. (a) Electrode resistance as a function of the oxygen partial
pressure at 983 K, 1; 908 K, 2; end 833 K, 3. Pt electrodes on ZY]7.
The drawn lines represent a Po2 -1/2 and Po2 ~1/2 dependence
respectively. (b) Pt electrodes on CG10. The drawn lines represent
a Po2 -1/2 and Po2 +1/4 dependence respectively. Open points:
sputtered electrodes. Closed points: gauze electrodes.
8
Rel --Rlc
-- -- Po2- '/~ [16]
DoV~/~
For
0ads
~ 1 we find that 11c >> lla and (1 -- 0ads) =
Kt (T) X [Po2]-'/2 and this results in
Rel -~- R1a "~ -- PO2 + '4 [17]
Do
According to Eq. [16] at Po2
< Po2 'nin" (i.e.,
Cads <<
1) mass transport limitation of oxygen atoms occurs at
the cathode and according to Eq. [1'7] at Po2 > Po2 rain"
(i.e.,
eads ~ 1) mass transport hmitation of oxygen
atoms occurs at the anode. The surface coverage of
oxygen atoms decreases with increasing temperature
and this is in a good agreement with the observed
change of Po2 rain. as a function of T. From Eq. [16] and
[17] it follows that in the region where Rel varies with
Po2-'/2 and Po2 + ,/2 the activation energy of Re1 is given
by Ea(Rel) = AHd -- 1/~AHads and Ea(Rel) -- AHd +
1/zAHads,
respectively, where
AH d
is the activation en-
thalpy for diffusion. In literature no precise data are
known of
AHd
and
AHads as
a function of the coverage.
Supposing that AHd and AHads are not strongly depend-
ent on the coverage, we calculate from the experimen-
tal values of Ea(Rel) for AHd and AHads values of 170
and 160 kJ mol -~, respectively. These calculated data
are in reasonable agreement with literature data con-
cerning alia and ~H~ds of oxygen on platinum. Lewis
and Gomer (19) found a surface diffusion enthalpy of
145 kJ mo1-1 for oxygen atoms on platinum at T >
500 K. The values of the enthalpy of adsorption found
by different authors scatter. Brennan
et aI.
(21) re-
ported that aH,as decreases with increasing surface
coverage from 285 to 150 k5 tool -1 (300 K). Netzer
and Gruber (18) reported for high surface coverages
a value for AHads of 135 k3 mo1-1 (670-770 K). Finally,
AHads can be calculated from the temperature depend-
ence of Po2 rain.. According to the Langmuir equation a
minimum in Rel is predicted for e -= 1/2 and with Eq.
[31 this result~ in
Ks(T) X Po2 rain' :
1. From Schouler's
(4) data it follows that
AHads =
280 --4-__ 15 kJ tool -1
and from our data AHads --= 160 _+ 30 kJ mol-L
Downloaded 22 Jun 2009 to 130.89.112.86. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp