An oxide ion and proton co-ion conducting Sn0.9In 0.1P2 O7 electrolyte for intermediate- temperature fuel cells

Article abstract of DOI:10.1149/1.2988135

The ionic conductivity of Sn_0.9In_0.1P₂ O₇ ceramic was investigated under various atmospheres within the temperature range of 130-230°C. Similar to mixed-conductive perovskite oxides at high temperatures (such as SrCe_0.95Yb_0.05 O _3-α, La_0.9 Sr_0.1Ga_0.8 Mg _0.2 O_3-α at 600-1000°C), Sn_0.9In _0.1P₂ O₇ can conduct both protons and oxide ions at low temperatures (130-230°C). The conductivity of Sn _0.9In_0.1P₂ O₇ reaches 0.019 Scm at 200°C in wet nitrogen. Its transport numbers determined by steam concentration cells are around 0.76 for a proton and 0.12 for an oxide ion. The performance of direct methanol fuel cells at 170°C using mixed-ion conductive Sn_0.9In_0.1P₂ O₇ electrolyte is higher than that at 235°C using pure proton conductive CsH₂PO₄ electrolyte. This is attributed to direct oxidation of CO at the anode by the oxide ions generated at the cathode and moved through the Sn_0.9In_0.1P₂ O₇ electrolyte.

Full text of DOI:10.1149/1.2988135

An Oxide Ion and Proton Co-Ion Conducting Sn0.9In0.1P2O7
Electrolyte for Intermediate-Temperature Fuel Cells
Xilin Chen, Chunsheng Wang, E. Andrew Payzant, Changrong Xia and Deryn Chu
J. Electrochem. Soc. 2008, Volume 155, Issue 12, Pages B1264-B1269.
doi: 10.1149/1.2988135
Receive free email alerts when new articles cite this article - sign up in the box at the
top right corner of the article or click here
Email alerting
service
To subscribe to Journal of The Electrochemical Society go to:
http://jes.ecsdl.org/subscriptions
© 2008 ECS - The Electrochemical Society
Downloaded on 2013-02-22 to IP 132.203.235.189 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

B1264
Journal of The Electrochemical Society, 155 ͑12͒ B1264-B1269 ͑2008͒
0013-4651/2008/155͑12͒/B1264/6/$23.00 © The Electrochemical Society
An Oxide Ion and Proton Co-Ion Conducting Sn0.9In0.1P2O7
Electrolyte for Intermediate-Temperature Fuel Cells
,z
Xilin Chen,a Chunsheng Wang,a, E. Andrew Payzant,b Changrong Xia,c and
*
Deryn Chud
aDepartment of Chemical and Biomolecular Engineering, University of Maryland, College Park,
Maryland 20742, USA
bMaterials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge,
Tennessee 37831-6064, USA
cDepartment of Materials Science and Engineering, University of Science and Technology of China, Hefei,
Anhui, 230026, China
dU.S. Army Research Laboratory, Sensors and Electronic Devices Directorate, Adelphi,
Maryland 20783-1197, USA
The ionic conductivity of Sn0.9In0.1P2O7 ceramic was investigated under various atmospheres within the temperature range of
130–230°C. Similar to mixed-conductive perovskite oxides at high temperatures ͑such as SrCe0.95Yb0.05O3−␣,La0.9Sr0.1
Ga0.8Mg0.2O3−␣ at 600–1000°C͒, Sn0.9In0.1P2O7 can conduct both protons and oxide ions at low temperatures ͑130–230°C͒. The
conductivity of Sn0.9In0.1P2O7 reaches 0.019 S/cm at 200°C in wet nitrogen. Its transport numbers determined by steam concen-
tration cells are around 0.76 for a proton and 0.12 for an oxide ion. The performance of direct methanol fuel cells at 170°C using
mixed-ion conductive Sn0.9In0.1P2O7 electrolyte is higher than that at 235°C using pure proton conductive CsH2PO4 electrolyte.
This is attributed to direct oxidation of CO at the anode by the oxide ions generated at the cathode and moved through the
Sn0.9In0.1P2O7 electrolyte.
© 2008 The Electrochemical Society. ͓DOI: 10.1149/1.2988135͔ All rights reserved.
Manuscript submitted August 7, 2008; revised manuscript received August 28, 2008. Published October 7, 2008.
Recently, intermediate temperature solid state fuel cells, which
ion-conductive ceramic-supported Pt or Pt–Ru catalyst has a very
high CO oxidation ability, even at low temperatures provided an
oxygen ͑or oxide ion͒ source is present. It has been reported that
Sn0.9In0.1P2O7 ITFCs can tolerate up to 10% CO in H2; i.e., 10%
CO added to H2 at the anode did not decrease the performance of
Sn0.9In0.1P2O7 ITFC at 250°C.14,15 The excellent CO tolerance at
the anode for Sn0.9In0.1P2O7 ITFCs may be attributed to the oxide
ions transferring from the cathode through oxygen vacancies in
Sn0.9In0.1P2O7 because of the existence of oxygen vacancies in
Sn0.9In0.1P2O7.14,16
In this work, the proton and oxide ion co-ion conductivity of
Sn0.9In0.1P2O7 and the performance of Sn0.9In0.1P2O7 fuel cells with
methanol and/or hydrogen fuel were investigated. A fuel cell using a
co-ion conducting Sn0.9In0.1P2O7 membrane, operating at tempera-
tures of 170–200°C, combined the advantages of both the SOFC
and PEMFC, and complemented existing proton and oxide-ion fuel
cell systems. The goal of introducing oxide-ion transport in a proton
conducting ceramic is to oxidize CO and methanol in an ITFC,
are operated in the temperature range of 150–300°C, attracted wide-
spread interest because of their higher CO tolerance, faster electro-
chemical reaction kinetics, and simpler thermal and water manage-
ment as compared to proton exchange membrane fuel cells
͑PEMFCs͒. Because the working temperatures of intermediate tem-
perature fuel cells ͑ITFCs͒ are lower than that of solid oxide fuel
cells ͑SOFCs͒, ITFCs would allow the use of oxidation-resistant
metallic alloys or even plastics as interconnect materials. They
would also reduce the operating cost, increase the durability, extend
the service life, and permit more frequent temperature and perfor-
mance cycling.1,2
Many anhydrous proton conductors have been investigated as the
electrolytes for intermediate temperature H2/O2 fuel cells, such as
CsHSO4,3 CsH2PO4,4-7 NH4PO3-based composites,8-13 and indium-
doped SnP2O7.14-16 Although the CO tolerance of CsH2PO4 fuel
cells operating above 230°C ͑the transition temperature from a low-
to a high-conductive phase͒ is much better than PEMFCs operating
at 80°C, it is still lower than SOFCs because the hydrocarbon fuels
can be directly electro-oxidized by the oxide ions in SOFCs17,18 and
the operation temperatures ͑above 600°C͒ of SOFC are much higher
than that of ITFCs. A method to enhance the CO tolerance of ITFCs
is to use a proton and oxide ion co-ion conductive electrolyte mem-
brane to replace the currently used proton conductive membrane. In
such a cell, oxide ions are generated by the electrochemical reduc-
tion of dioxygen at the cathode and diffuse through the mixed oxide
ion/proton membrane to the anode, where they may efficiently and
directly oxidize any CO at the anode.
which is different from prior works on mixed proton/oxide ion con-
23
ducting perovskite oxides ͑such as SrCe0.95Yb0.05O3−␣
,
24
La0.9Sr0.1Ga0.8Mg0.2O3−␣ ͒, for which proton conductivity is ther-
mally activated at a lower temperature than that required for oxide
ion conduction25 and hence enables a lower SOFC operating tem-
perature.
Experimental
Most hydrocarbon oxidation reactions follow the Mars–van
Krevelen mechanism,19-22 in which surface lattice O2− is the actual
oxidation agent. During these reactions, the O2− reactant plays two
important roles: ͑i͒ being responsible for the removal of H from
C–H bonds to give water as in oxidative dehydrogenation, and ͑ii͒
being capable of insertion into the intermediate complex to give the
oxygenated product and water ͑mild oxidation͒. This mechanism is
more efficient than that in the direct methanol fuel cell, in which
methanol is directly electro-oxidized at low temperature by adsorbed
hydroxide on a Pt–Ru electrocatalyst. It is the reason why an oxide-
In3+-doped SnP2O7 was prepared using SnO2 ͑Sigma-Aldrich,
99.9%͒, In2O3 ͑Aldrich, 99.99%͒, and H3PO4 ͑Fisher Scientific,
85%͒ to form Sn0.9In0.1P2O7. The starting materials were mixed and
held with stirring at 300°C until the mixture formed a paste with a
high viscosity. The pastes were calcined at 650°C for 2.5 h and then
ground into powders with a mortar and pestle. The compound was
pressed at 250 MPa into pellets and then coated with Ag for con-
ductivity measurement and with a Pt catalyst for fuel cell testing.
The pellets were 13 mm in diameter and around 0.8 mm in thick-
ness. For comparison, CsH2PO4 was also prepared as described in
Ref. 4 using Cs2CO3 ͑Sigma-Aldrich 99%͒ and H3PO4 ͑Fisher Sci-
entific 85%͒ as precursors. The crystalline structure of the com-
pounds was analyzed by X-ray diffraction ͑XRD, PANalytical
X’Pert Pro MPD diffractometer with Cu radiation͒ and the micro-
*
Electrochemical Society Active Member.
z E-mail: cswang@umd.edu
Downloaded on 2013-02-22 to IP 132.203.235.189 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 155 ͑12͒ B1264-B1269 ͑2008͒
B1265
͑15 wt %͒. The catalyst ink was prepared by mixing Sn0.9In0.1P2O7
͑or CsH2PO4͒, distilled water, Pt/C catalyst, isopropanol ͑Fisher
Scientific, 99.9%͒, and PTFE. The ink was brushed onto the diffu-
sion layer and dried to form an electrode ͑anode or cathode͒. The
platinum loading was 4 mg/cm2. The electrolyte membrane
͑Sn0.9In0.1P2O7 or CsH2PO4͒ was sandwiched by the anode and
cathode to form a single cell. Humidification of gas was achieved by
saturating nitrogen/hydrogen/oxygen with water at a given tempera-
ture, resulting in a given water partial pressure, such as 76°C for
40% H2O. To add methanol into fuel gases, hydrogen/nitrogen was
saturated with a mixture of methanol and water at a given tempera-
ture.
(a)
(b)
Sn0.9In0.1P2O7
CsH2PO4
Conductivity measurements were carried out using a Solartron
͑SI1287 and SI1260͒ electrochemical impedance system in the fre-
quency range of 0.001–106 Hz. At each temperature, the sample was
held about 30 min prior to measurement.
Results and Discussion
20
30
40
50
60
70
Figure 1 shows typical XRD patterns for Sn0.9In0.1P2O7 and
CsH2PO4 at room temperature. The peaks observed are identical to
those reported in PDF# 00-29-1352 ͑SnP2O7͒ and PDF# 00-35-0746
͑CsH2PO4͒. SnP2O7 doped with 10% In3+ showed the same struc-
ture as the undoped SnP2O7, characterized by a cubic or pseudocu-
bic structure with SnO6 octahedral at the corners and P2O7 units at
the edges.26 However, Nagao et al. reported that 10% In3+ doping
increased the lattice constant from 7.945 to 7.950 Å16 and created
oxygen vacancy defects.14,16 The chemical composition of
Sn0.9In0.1P2O7 was confirmed by energy dispersive analysis by
X-rays. All peaks in Fig. 1b could be indexed to CsH2PO4, indicat-
ing the synthesized material is pure CsH2PO4.
2θ
Figure 1. XRD patterns for ͑a͒ Sn0.9In0.1P2O7 and ͑b͒ CsH2PO4.
morphology of the powders and pellets was observed by scanning
electron microscopy images using an FEI Quanta 200 with an elec-
tron diffraction spectrometer.
Both the anode and cathode contained two overlapped conduc-
tive porous layers: a diffusion layer and a catalyst layer. The diffu-
sion layer was formed from carbon cloth ͑E-TEK, B1A͒ which was
smoothed by coating a mixture of 85 wt % carbon powder ͑Vulcan
XC72͒ and 15 wt % poly͑tetrafluoroethylene͒ ͑PTFE, Aldrich, 60%
dispersion͒. The carbon powder loading was 2 mg/cm2. The catalyst
layer was formed from either Sn0.9In0.1P2O7 or CsH2PO4
͑35 wt %͒, 40 wt % Pt/C catalyst ͑50 wt %, E-TEK͒, and PTFE
Figure 2 shows the morphologies of powders and morphologies
of cross sections for Sn0.9In0.1P2O7 and CsH2PO4 pellets. The cross
sections were obtained by cutting the pellets with a sharp blade.
Although the particle size of Sn0.9In0.1P2O7 is smaller than that of
(a)
(b)
Figure 2. Micromorphologies of ͑a͒
Sn0.9In0.1P2O7 powder, ͑b͒ CsH2PO4 pow-
der, ͑c͒ cross section for Sn0.9In0.1P2O7
pellet, and ͑d͒ cross section for CsH2PO4
pellet.
(c)
(d)
Downloaded on 2013-02-22 to IP 132.203.235.189 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

B1266
Journal of The Electrochemical Society, 155 ͑12͒ B1264-B1269 ͑2008͒
t /oC
180
The interaction ͑Reaction 1͒ between water vapor and oxygen
vacancies ͑and/or electron holes͒ may be responsible for protonic
conduction in Sn0.9In0.1P2O7,14,16 which is similar to the mech-
anism of proton incorporation in perovskite oxides such as
SrCe0.95Yb0.05O3−␣. The increase in conductivity of Sn0.9In0.1P2O7
with humidity ͑Fig. 3͒ supports the hypothesis of the proton con-
ducting mechanism. During the operation of fuel cells, protons gen-
erated at the anode ͑Reaction 2͒ move through the Sn0.9In0.1P2O7
membrane to the cathode to react with oxygen ͑Reaction 3͒, whereas
oxide ions transit in the opposite direction from the cathode ͑Reac-
tion 4͒ to the anode to oxidize the CO or CH3OH ͑Reaction 5͒. CO
and methanol can also be chemically oxidized with water through a
water–gas-shift reaction ͑Reaction 6͒ and reforming ͑Reaction 7͒ at
intermediate temperatures.
(a)
240
220
200
160
140
120
3.2% H2O
10% H2O
40% H2O
0.020
0.015
0.010
Proton conduction in Sn0.9In0.1P2O7
H2O͑g͒ + V•• → 2H•i + OOX and/or H O͑g͒ + 2h• → 2H• + 1 O
͑
͒
2
2
O
i
2
͓1͔
͓2͔
͓3͔
Anode reaction in Sn0.9In0.1P2O7 fuel cells
1.9
2.0
2.1
2.2
2.3
2.4
2.5
1000/T /K-1
t /oC
H2 → 2H•i + 2e−
Cathode reaction in Sn0.9In0.1P2O7 fuel cells
(b)
260
250
240
230
220
210
200
190
1 O2 + 2H•i + 2e− → H2O
2
0.01
1E-3
1E-4
1E-5
Oxide ion conduction in Sn0.9In0.1P2O7
1 O2 + V•O• + 2e− → OXO
͓4͔
͓5͔
͓6͔
͓7͔
2
Direct CO and methanol oxidation by oxide ion in anode
OXO + ͓R͔ → V•O• + ͓P͔ + 2e−
CO oxidation in anode through water–gas-shift reaction
H2O + CO → CO2 + H2
CH3OH oxidation in anode through reforming
CH3OH + H2O → CO2 + 3H2
Here, h•, H•, V•O•, OOϫ, ͓R͔, ͓P͔, and e− denote an electron hole, a
proton, an oxygen vacancy, a lattice oxide ion, reductant ͑CO and
CH3OH͒, reduction product, and an electron, respectively.
1.90
1.95
2.00
1000/T /K-1
2.05
2.10
2.15
Another possibility for proton conduction in Sn0.9In0.1P2O7 is the
residual phosphoric acid on the surface of material. If phosphoric
acid exists on the surface of material, it should be dissolved into
water when the Sn0.9In0.1P2O7 particles are soaked in water; thus,
the washed Sn0.9In0.1P2O7 should have a lower conductivity than
that of unwashed Sn0.9In0.1P2O7. However, the conductivity of
washed Sn0.9In0.1P2O7 shows a similar conductivity with the un-
washed one, indicating that proton conductivity in Sn0.9In0.1P2O7 is
mainly through a water–oxygen vacancy reaction. More experiments
are still needed to elucidate the exact role of residual phosphoric
acid in proton conduction of Sn0.9In0.1P2O7 if it exists.
To check the conductivity of Sn0.9In0.1P2O7 electrolyte in fuel
cells, the impedances of Sn0.9In0.1P2O7 membranes were measured
in O2 and N2 with and without the presence of water ͑Fig. 4͒. Com-
pare the impedances of Sn0.9In0.1P2O7 measured in dry N2 and dry
O2 ͑Fig. 4a͒. The low-frequency line did not intersect the real axis,
which indicates that Sn0.9In0.1P2O7 is an electronic insulator. Adding
water to oxygen gas results in an even lower electronic conductivity,
as evidenced by an almost straight line in the low-frequency region
͑Fig. 4b͒. Figure 4 clearly demonstrates that Sn0.9In0.1P2O7 is an
electronic insulator at temperatures below 200°C whether in a dry or
wet atmosphere.
Figure 3. Conductivities of ͑a͒ Sn0.9In0.1P2O7 in wet nitrogen with various
water vapor pressures, and ͑b͒ CsH2PO4 in 40% H2O/N2.
CsH2PO4, ͑Fig. 2a and b͒, the porosity of the former is higher than
the latter, i.e., the density of the CsH2PO4 pellet is higher than that
of Sn0.9In0.1P2O7 ͑Fig. 2c and d͒.
Figure
3 shows the conductivities of Sn0.9In0.1P2O7 and
CsH2PO4 at different temperatures in humidified nitrogen. The con-
ductivity of Sn0.9In0.1P2O7 continuously increased with increasing
water vapor pressure from 3.2 to 40% ͑Fig. 3a͒. The conduction
behavior of Sn0.9In0.1P2O7 at different temperatures is different from
the superprotonic character of CsH2PO4, which shows a sharp in-
crease in the conductivity of a few orders of magnitude by a struc-
tural transition from a low- to a high-temperature phase ͑Fig. 3͒. The
maximum protonic conductivities of Sn0.9In0.1P2O7, achieved at
around 200°C in 40% H2O, is 0.019 S/cm, which is similar to that
͑0.02 S/cm͒ of CsH2PO4 at 240°C. The thermal variations of proton
conductivities for Sn0.9In0.1P2O7 and CsH2PO4 were consistent with
reported results,4,14,27 although the conductivity of Sn0.9In0.1P2O7 is
slightly lower than reported values.14 The reason why the conduc-
tivity decreases after 200°C is not clear. It cannot be attributed to
the escape of water produced by the reversal of Reaction 1 because
the evolution of water vapor from Sn0.9In0.1P2O7 starts above
260°C.16 If it is due to the escape of water, the peak position should
not be fixed at 200°C but move to a higher temperature with higher
water partial pressure.
The maximum protonic conductivities of 0.019 S/cm are
achieved at around 200°C in 40% H2O. To avoid the conductivity
drop, the Sn0.9In0.1P2O7 membranes and Sn0.9In0.1P2O7 fuel cells
were tested at 170°C ͑i.e., below 200°C͒. Figure 5 shows the time
dependence of conductivity of Sn0.9In0.1P2O7 at 170°C and
CsH2PO4 at 235°C in wet nitrogen with 40% water vapor. It is clear
that the conductivities of both Sn0.9In0.1P2O7 and CsH2PO4 are very
Downloaded on 2013-02-22 to IP 132.203.235.189 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 155 ͑12͒ B1264-B1269 ͑2008͒
B1267
(a) 0.04
(a)
In dry N2
15000
10000
5000
0
In dry O2
100
50
0
0.03
106Hz
4000Hz
3.2*105Hz
0.02
1mHz
0
50
100
150
200
0.01
1mHz
0.00
0
5000
10000
15000
20000
0
5
10
15
20
25
Real Part /Ω
time /h
(b) 0.04
(b)
In wet nitrogen (40% H2O)
In wet oxygen (40% H2O)
800
600
400
200
0
0.03
0.02
0.01
0.00
1mHz
0.08Hz
1259Hz
106Hz
200
0
10
20
30
40
time /h
0
400
600
800
1000
1200
Real Part /Ω
Figure 5. ͑a͒ Time dependence of conductivities of Sn0.9In0.1P2O7 at 170°C
in 40% H2O/N2 and ͑b͒ that of CsH2PO4 at 235°C in 40% H2O/N2.
Figure 4. Impedance plots of Sn0.9Ino.1P2O7 in ͑a͒ dry and ͑b͒ wet condition
at 170°C, electrolyte thickness: 0.79 mm for dry oxygen, and 0.80 mm for
dry nitrogen, 0.88 mm for wet oxygen, 0.80 mm for wet nitrogen.
can be expressed as E0 = −tH+͑RT/2F͒ln͑PIHI O/PHI O͒, and the pro-
2
ton transport number, tH+, can be calculated from t2he measured E0,
tH+ = E0/Eth.29 Here, Eth = ͑RT/2F͒ln͑PIHI O/PHI O͒ is defined as the
2
theoretical emf. As shown in Fig. 6a, the2proton transport numbers
are 0.7–0.8 ͑average 0.76͒, depending on the temperatures. After
switching the carrying gas from oxygen to hydrogen, the oxygen-ion
transport number tO2− at different temperatures can be calculated
from Eq. 9 using the measured E0 values and calculated proton
transport numbers at the same temperatures. As shown in Fig. 6b,
the oxygen-ion transport numbers are around 0.12. The total ionic
transport number ͑0.88͒ of Sn0.9In0.1P2O7 is less than unity, which is
probably attributed to gas leakage in the steam concentration cells
due to the low density of the membrane ͑Fig. 2͒. The gas leakage
will result in a lower measured emf than the theoretical value,
and less than unity for the total ionic transport numbers. The con-
ductivity of oxide ions in Sn0.9In0.1P2O7 at 200°C is around
10−3 S/cm, which is close to the oxide-ion conductivities of
stable. Therefore, both Sn0.9In0.1P2O7 and CsH2PO4 are suitable
electrolytes for intermediate temperature fuel cells.
Based on Reactions 1 and 4, Sn0.9In0.1P2O7 should be a mixed
proton and oxide ion conductor in humidified gas. In order to
confirm Sn0.9In0.1P2O7 is a mixed-conducting conductor under
humidified conditions, the transport numbers of protons and
oxide ions were determined by measuring the electromotive force
emf of steam concentration cells H2/O2 ͑PH O = 0.6 K/Pa͒,
2
Pt/C͉Sn0.9In0.1P2O7͉Pt/C, H2/O2 ͑PH O = 3.2 K/Pa͒. The Nernst
2
potential of the steam concentration cell can be determined by the
following equations28
E0 = ͑RT/4F͓͒tion ln͑PIOI /POI ͒ − 2tH
+
ln͑P /PIH O͔͒
͓8͔
II
H O
2
2
2
2
where tion = tH+ + tO2− is the ionic transport number and the carrying
30
31
gas is oxygen. Or
La0.8Sr0.2Ga0.8Mg0.2 and Bi4Ti0.2V1.8O10.7 at 250°C.
E0 = ͑RT/2F͓͒− tion ln͑PIHI /PHI ͒ + tO2− ln͑P /PHI O͔͒
͓9͔
II
The carbon monoxide generated by the reforming reaction would
dilute the hydrogen fuel and/or poison the catalyst in the anode even
at an intermediate temperature and make the fuel cell performance
worse. Because Sn0.9In0.1P2O7 is a proton and oxide ion co-ion con-
ductor and the oxide ion will oxidize the CO ͑Reaction 5 can be
particularized as OOX + CO → VO•• + CO2 + 2e−͒, the exhaust CO
can become as a fuel in the direct methanol Sn0.9In0.1P2O7 fuel cells,
H O
2
2
2
2
where the carrying gas is hydrogen.
Because the pressures of the carrying gases ͑oxygen͒ at both
sides of the steam concentration cell are very close, and the transport
number for protons is larger than that for oxide ions ͑i.e, 2tH
ӷ tH+ + tO2−͒, the first parts of Eq. 8 are negligible. Therefore, Eq. 8
+
Downloaded on 2013-02-22 to IP 132.203.235.189 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

B1268
Journal of The Electrochemical Society, 155 ͑12͒ B1264-B1269 ͑2008͒
60
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
(a)
130oC
150oC
170oC
20
15
10
5
50
40
30
20
10
0
Theoretical
EMF
Proton transport number
0
130
140
150
160
170
180
190
0
10
20
30
40
50
60
70
Temperature /oC
Current Density /mAcm-2
8
0.5
0.4
0.3
0.2
0.1
0.0
(b)
Figure 7. ͑Color online͒ Fuel cell performance at different temperature using
Sn0.9In0.1P2O7 as electrolyte and hydrogen as fuel; electrolyte thickness:
0.78 mm.
EMF
Oxygen-ion transport number
6
4
2
0
switched from the H2–CH3OH mixture to pure methanol, the power
density for the CsH2PO4 fuel cell had an 85% drop, from
13.7 to 2.1 mW/cm2, while the drop for the Sn0.9In0.1P2O7 fuel cell
(a) 1.0
20
60% H2+40% H2O (Fuel)
45% H2+15% CH3OH+40% H2O
45% N2+15% CH3OH+40% H2O
0.8
15
0.6
130
140
150
160
170
180
190
Temperature /oC
10
0.4
Figure 6. ͑Color online͒ The emf values of Sn0.9In0.1P2O7 membrane gal-
vanic cells at different atmospheres. ͑a͒ Anode gas = O2 ͑PH = 3.2 kPa͒;
O
2
5
cathode gas = O2 ͑PH = 0.6 kPa͒. ͑b͒ Anode gas = H2 ͑PH = 0.6 kPa͒;
O
O
2
cathode gas = H2 ͑PH = 3.2 kPa͒; electrolyte thickness: 1.42mm.
0.2
O
2
0.0
0
0
10
20
30
40
50
60
70
and thus the direct methanol Sn0.9In0.1P2O7 fuel cells should per-
form better than direct methanol CsH2PO4 fuel cells.
Current Density /mAcm-2
To test the capability of the direct oxidation of CO by oxide ions,
the performance of a Pt/C fuel cell using a mixed proton/oxide ion
conducting Sn0.9In0.1P2O7 electrolyte was investigated in the tem-
perature range 130–170°C and compared with a fuel cell using a
pure proton conductor CsH2PO4 electrolyte at 235°C. Cell voltages
and power densities of two cells are shown in Fig. 7 and 8 as a
function of current density. As expected, the power density of the
Sn0.9In0.1P2O7 fuel cell increased with temperature and the maxi-
mum power densities at 170°C were 14.7 mW/cm2 for hydrogen,
14.9 mW/cm2 for a mixture of hydrogen and methanol, and
4.0 mW/cm2 for methanol. Due to the fast reaction kinetics at a
comparatively high operating temperature ͑235°C͒ for the CsH2PO4
fuel cell, the maximum power density ͑17.3 mW/cm2͒ of the H2/O2
CsH2PO4 fuel cell was slightly higher than the H2/O2 Sn0.9In0.1P2O7
fuel cell, even though the ionic conductivities of the two electrolytes
are very close. However, the power densities of the CsH2PO4 fuel
cell at 235°C using hydrogen–methanol ͑13.7 mW/cm2͒ and metha-
nol ͑2.1 mW/cm2͒ were much lower than that of the Sn0.9In0.1P2O7
fuel cell at 170°C. When 15% methanol was added into hydrogen,
the power density of the CsH2PO4 fuel cell lost 21% power density,
decreasing from 17.3 to 13.7 mW/cm2, but the power density of the
Sn0.9In0.1P2O7 fuel cell remained the same. When the fuel was
(b)
60% H2+40% H2O (Fuel)
45% H2+15% CH3OH+40% H2O
45% N2+15% CH3OH+40% H2O
1.0
0.8
0.6
0.4
0.2
0.0
20
15
10
5
0
0
10
20
30
40
50
60
70
Current density /mAcm-2
Figure 8. ͑Color online͒ Fuel cell performance with various fuels, electro-
lytes: ͑a͒ Sn0.9In0.1P2O7, thickness: 0.78 mm, 170°C. ͑b͒ CsH2PO4, thick-
ness: 0.89 mm, 235°C.
Downloaded on 2013-02-22 to IP 132.203.235.189 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 155 ͑12͒ B1264-B1269 ͑2008͒
B1269
was 75%, from 14.9 to 4.0 mW/cm2. Because both the amount of
platinum loading in the two fuel cells and the ionic conductivity of
two electrolytes ͑CsH2PO4 and Sn0.9In0.1P2O7͒ were the same, even
the CsH2PO4 cell has a faster electrochemical reaction kinetics than
that of the Sn0.9In0.1P2O7 cell because of its higher operation tem-
perature; the difference in performance must come from the electro-
lytes. The higher performance of methanol fuel cells using
Sn0.9In0.1P2O7 as an electrolyte at 170°C compared to using a
CsH2PO4 electrolyte at 235°C strongly indicated that CO was ef-
fectively oxidized by the oxide ions, resulting in an enhancement of
the fuel cell performance.
Although the CsH2PO4 fuel cell has a lower power density than
the Sn0.9In0.1P2O7 fuel cell when methanol is present in the fuel, its
open-circuit voltage is higher. This is due to the higher density of the
CsH2PO4 membrane as shown in Fig. 2, and the higher density may
be another reason that the power density of CsH2PO4 is higher than
Sn0.9In0.1P2O7 in a H2/O2 fuel cell, even though they have very
similar conductivities.
Office of FreedomCAR and Vehicle Technologies, as part of the
High Temperature Materials Laboratory User Program, Oak Ridge
National Laboratory, managed by UT-Battelle, LLC, for the U.S.
Department of Energy under contract no. DE-AC05-00OR22725
University of Maryland assisted in meeting the publication costs of this
article.
References
1. T. Norby, Solid State Ionics, 125, 1 ͑1999͒.
2. T. Norby, Nature (London), 410, 877 ͑2001͒.
3. S. M. Haile, D. A. Boysen, C. R. I. Chisholm, and R. B. Merle, Nature (London),
410, 910 ͑2001͒.
4. S. M. Haile, C. R. I. Chisholm, K. Sasaki, D. A. Boysen, and T. Uda, Faraday
Discuss., 134, 17 ͑2007͒.
5. T. Uda and S. M. Haile, Electrochem. Solid-State Lett., 8, A245–A246 ͑2005͒.
6. D. A. Boysen, T. Uda, C. R. I. Chisholm, and S. M. Haile, Science, 303, 68–70
͑2004͒.
7. T. Uda, D. A. Boysen, C. R. I. Chisholm, and S. M. Haile, Electrochem. Solid-State
Lett., 9, A261 ͑2006͒.
8. M. Cappadonia, O. Niemzig, and U. Stimming, Solid State Ionics, 125, 333
͑1999͒.
9. S. Haufe, D. Prochnow, D. Schneider, O. Geier, D. Freude, and U. Stiming, Solid
State Ionics, 176, 955 ͑2005͒.
10. T. Matsui, S. Takeshita, Y. Iriyama, T. Abe, and Z. Ogumi, J. Electrochem. Soc.,
Conclusions
A proton and oxide ion co-ion conductor, Sn0.9In0.1P2O7, was
synthesized and characterized. The conductivity of Sn0.9In0.1P2O7
and performance of Sn0.9In0.1P2O7 membrane fuel cell were inves-
tigated under various atmospheres in the temperature range of
130–230°C and compared with pure proton conductive CsH2PO4
and CsH2PO4 membrane fuel cells. The conductivity of
Sn0.9In0.1P2O7 reached 0.019 S/cm at 200°C in wet nitrogen, and
the transport numbers determined by steam concentration cells were
0.76 for protons and 0.12 for oxide ions. The higher performance of
a methanol Sn0.9In0.1P2O7 fuel cell at 170°C compared to a metha-
nol CsH2PO4 fuel cell at 235°C was attributed to the direct oxida-
tion of CO at the anode by the oxide ions transported through the
Sn0.9In0.1P2O7 electrolyte. Due to the unique concept of oxidation of
CO by oxide ion transported from the cathode, and preliminary re-
sults on a co-ion ͑proton/oxide ion͒ conductive fuel cell with metha-
nol fuel, the co-ion conducting membrane fuel cell, operating at
around 200°C, may be the next-generation methanol fuel cell and
reforming fuel cell because it combines the advantages of both
SOFCs and PEMFCs.
152, A167 ͑2005͒.
11. X. Chen, Z. Huang, and C. Xia, Solid State Ionics, 177, 2413 ͑2006͒.
12. X. Chen, X. Li, S. Jiang, C. Xia, and U. Stimming, Electrochim. Acta, 51, 6542
͑2006͒.
13. X. Chen, C. Xia, and U. Stimming, Electrochim. Acta, 52, 7835 ͑2007͒.
14. M. Nagao, A. Takeuchi, P. Heo, T. Hibino, M. Sano, and A. Tomita, Electrochem.
Solid-State Lett., 9, A105 ͑2006͒.
15. P. Heo, H. Shibata, M. Nagao, T. Hibino, and M. Sano, J. Electrochem. Soc., 153,
A897 ͑2006͒.
16. M. Nagao, T. Kamiya, P. Heo, A. Tomita, T. Hibino, and M. Sano, J. Electrochem.
Soc., 153, A1604 ͑2006͒.
17. E. P. Murray, T. Tsai, and S. A. Barnett, Nature (London), 400, 649 ͑1999͒.
18. S. Park, J. M. Vohs, and R. J. Gorte, Nature (London), 404, 265 ͑2000͒.
19. P. Mars and D. W. van Krevelen, Chem. Eng. Sci., 3, 41 ͑1954͒.
20. C. Doornkamp and V. Ponec, J. Mol. Catal. A: Chem., 162, 19 ͑2000͒.
21. F. Bertinchamps and E. M. Gaigneaux, Catal. Today, 91–91, 105–110 ͑2004͒.
22. H. Over, Y. D. Kim, A. P. Seitsonen, S. Wendt, E. Lundgren, M. Schmid, P. Varga,
A. Morgante, and G. Ertl, Science, 287, 1474 ͑2000͒.
23. H. Uchida, N. Maeda, and H. Iwahara, Solid State Ionics, 11, 117 ͑1983͒.
24. G. Ma, F. Zhang, J. Zhu, and G. Meng, Chem. Mater., 18, 6006 ͑2006͒.
25. B. Zhu, X. Liu, and T. Schober, Electrochem. Commun., 6, 378 ͑2004͒.
26. R. K. B. Gover, N. D. Withers, S. Allen, R. L. Withers, and J. S. O. Evans, J. Solid
State Chem., 166, 42 ͑2002͒.
27. J. Otomo, N. Minagawa, C. Wen, K. Eguchi, and H. Takahashi, Solid State Ionics,
Acknowledgment
156, 357 ͑2003͒.
This material is based upon work supported by the U.S. Army
Research Laboratory and the U.S. Army Research Office under
grant no. W911NF-06-1-0187. Partial work was done at Tennessee
Technological University. The XRD analysis is sponsored by the
Assistant Secretary for Energy Efficiency and Renewable Energy,
28. J. Guan, S. E. Dorris, U. Balachandran, and M. Liu, Solid State Ionics, 100, 45
͑1997͒.
29. D. P. Sutija, T. Norby, and P. Bjrönbom, Solid State Ionics, 77, 167 ͑1995͒.
30. J. W. Fergus, J. Power Sources, 162, 30 ͑2006͒.
31. M. J. Godinho, P. R. Bueno, M. O. Orlandi, and E. Longo, Mater. Lett., 57, 2540
͑2003͒.
Downloaded on 2013-02-22 to IP 132.203.235.189 address. Redistribution subject to ECS license or copyright; see www.esltbd.org