Study of anode catalysts and fuel concentration on direct hydrazine alkaline anion-exchange membrane fuel cells

Article abstract of DOI:10.1149/1.3082129

A platinum-free fuel cell using liquid hydrazine hydrate (N ₂H₄·H₂O) as the fuel and comprised of a cobalt or nickel anode and a cobalt cathode exhibits high performance. In this study, the fuel cell performances using nic

Full text of DOI:10.1149/1.3082129

Journal of The Electrochemical Society, 156 ͑4͒ B509-B512 ͑2009͒
B509
0013-4651/2009/156͑4͒/B509/4/$23.00 © The Electrochemical Society
Study of Anode Catalysts and Fuel Concentration
on Direct Hydrazine Alkaline Anion-Exchange
Membrane Fuel Cells
Koichiro Asazawa,a,c Tomokazu Sakamoto,a Susumu Yamaguchi,a Koji Yamada,a,
*
Hirotoshi Fujikawa,a Hirohisa Tanaka,a,z and Keisuke Ogurob,c
aDaihatsu Motor Company, Limited, Frontier Technology Research and Development Division,
Shiga 520-2593, Japan
bNational Institute of Advanced Industrial Science and Technology, Osaka 563-8577, Japan
cKobe University, Kobe 657-8501, Japan
A platinum-free fuel cell using liquid hydrazine hydrate ͑N2H4·H2O͒ as the fuel and comprised of a cobalt or nickel anode and a
cobalt cathode exhibits high performance. In this study, the fuel cell performances using nickel, cobalt, and platinum as anode
catalysts are evaluated and compared. It is found that fuel cell performance in the case of nickel and cobalt is higher than that in
the case of platinum. Further, anodic reactions are discussed by comparing the hydrazine consumption and ammonia generation
when cobalt and nickel are used as anode catalyst. Cobalt exhibits a higher rate of decomposition than nickel. Nickel is found to
be the most suitable anode catalyst among the above mentioned anode catalysts for this fuel cell. The influence of hydrazine
hydrate and KOH concentrations in the fuel on cell performance is also discussed. Cell performance is the highest at a hydrazine
hydrate concentration of 4 M and a KOH concentration of 1 M. The maximum power density of the alkaline anion-exchange
membrane fuel cell, comprised of a nickel anode and a Co–PPY–C ͑cobalt polypyrrole carbon͒ cathode, is 617 mW cm−2
© 2009 The Electrochemical Society. ͓DOI: 10.1149/1.3082129͔ All rights reserved.
.
Manuscript submitted November 5, 2008; revised manuscript received January 23, 2009. Published February 20, 2009.
Proton-exchange membrane ͑PEM͒ fuel cells have been recog-
nized as the future power source for vehicles. These fuel cells do not
adversely affect air quality, health, and the environment, because
they do not emit toxic substances and carbon dioxide. However,
economic problems associated with the large-scale use of platinum
and many restrictions associated with the infrastructure and storage
of hydrogen have limited the widespread use of fuel cell vehicles.
In order to solve these problems, a platinum-free direct hydrazine
fuel cell has been developed.1 In conventional fuel cells, materials
with high corrosion resistance are required for the fabrication of
electrodes because the PEMs used in these cells are highly acidic in
nature. Thus far, metals other than platinum have not been used for
the fabrication of electrodes for fuel cells. In alkaline anion-
exchange membrane fuel cells, basic hydroxide anion ͑OH−͒ ex-
change polymer membranes are used. These fuel cells have a mildly
corrosive environment. Therefore, various types of metals can be
used in these fuel cells.2-6 A platinum-free fuel cell using liquid
hydrazine hydrate ͑N2H4·H2O͒ as the fuel and comprised of a cobalt
or nickel anode and a cobalt or silver cathode has been developed.
The power generation performance of this fuel cell is comparable to
that of hydrogen fuel cells, and this fuel cell is characterized by zero
CO2 emissions ͑Fig. 1͒.
bonyl groups in the polymer. The onboard readsorption of hydrazine
hydrate is possible; the polymer releases the required amount of
liquid hydrazine hydrate in a timely manner for generation and the
hydrazine hydrate supply in a filling stand can also be attained with-
out returning to a factory. Therefore, hydrazine hydrate can be re-
garded as a promising carrier for realizing a hydrogen society.
The new platinum-free anion-exchange membrane fuel cell can
be potentially used for a wide range of applications. Moreover, this
technology attracted considerable attention at the Environmental
Showcase of the G8 Hokkaido Toyako Summit held in July 2008.8
In this study, the power densities of the anion-exchange membrane
fuel cell using nickel, cobalt, and platinum as an anode catalyst are
compared. The influence of hydrazine and KOH concentrations in
the fuel on fuel cell performance is also investigated.
Experimental
Materials.— Three types of unsupported nanoparticles, Ni ͑Inco,
type 210H͒, Co ͑Junye Nano Material Co., Ltd.͒, and Pt ͑Tanaka
K.K.͒, were used as an anode catalyst in the anion-exchange mem-
brane fuel cell. The specific surface areas of Ni, Co, and Pt were 6.2,
33, and 6.9 m2 g−1
, respectively. In all three cases, Co–
polypyrrole͑PPY͒–C was used as the cathode catalyst. A Co–PPY–C
cathode was used in previous research by Bashyam and Zelenay.9
The concentration of Co in Co–PPY–C was 10 wt %. An anion-
exchange membrane ͑ion exchange capacity 1.7 meq g−1; thickness
30 ␮m; quaternary ammonium form͒ and an anion-exchange resin
solution ͑ion exchange capacity 2.0 meq g−1͒, both provided by
Hydrazine-air fuel cells were developed in the 1960s.7 A high-
concentration alkaline liquid ͑9 M KOH͒ is used as the electrolyte
in these cells. However, in these fuel cells, CO2 neutralization of the
electrolyte takes place. The CO2 tolerance of the anion-exchange
membrane fuel cells is discussed in Ref. 1. The power density of
these cells reduces slightly when the concentration of CO2 inside
them is approximately 50 times that in the atmosphere. However, the
cell performance reverts back to the original level shortly after the
removal of CO2. This result shows that CO2 neutralization is not a
serious problem for anion-exchange membrane fuel cells. Vorcoe et
al. have reported similar results in the case of hydrogen–oxygen fuel
cells comprising an anion-exchange membrane.2
High-concentration hydrazine hydrate is not only poisonous but
also mutagenic. Although the flash point of hydrazine hydrate
͑74°C͒ is higher than that of gasoline ͑Ͻ −40°C͒, hydrazine hy-
drate has to be handled carefully. A detoxification method has been
developed previously1 which repeatedly fixes hydrazine to the car-
Anion exchange membrane
fuel cell
Proton exchange membrane
fuel cell
Load
e-
Load
e-
O2
H2
O2
N2H4·H2O
OH-
H+
H2O
N2
H2O
H2O
Cathode electrode
(Cobalt, Silver)
Proton exchange
Membrane
Anion exchange
membrane
Electrode
(Platinum)
Anode electrode
(Nickel, Cobalt)
Acidic
Alkaline
*
Electrochemical Society Active Member.
z E-mail: hiroshisaគtanaka@mail.daihatsu.co.jp
Figure 1. Working principle of alkaline anion-exchange membrane fuel cell.

B510
Journal of The Electrochemical Society, 156 ͑4͒ B509-B512 ͑2009͒
0.9
700
600
500
400
300
200
100
0
Tokuyama Corp., were used to fabricate the membrane electrode
assembly ͑MEA͒. Hydrazine hydrate ͑60 wt %͒ was obtained from
Otsuka Chemicals, and potassium hydroxide ͑85 wt %͒ was pur-
chased from Hayashi Junyaku K.K.
a)
0.8
0.7
0.6
0.5
0.4
0.3
Methods.— Comparison of anode catalysts.— For preparation
of anode ink, 0.2 g of Ni and 2 g of a solvent ͑1-propanol 80:THF
20 by weight͒ were sonicated for 3 min. Then, 1.1 g of an ionomer
͑2 wt %͒ and 1.4 g of the solvent were added to the mixture. The
ink was dispersed by using a rotary shaker for 15 min. For the
preparation of cathode ink, 0.2 mg of Co–PPY–C and 1.3 g of the
solvent were sonicated for 3 min. Then, 1.9 g of an ionomer solu-
tion ͑2 wt %͒ was added to the mixture. Finally, the ink was dis-
persed by using a rotary shaker for 15 min.
0
500
1000
1500
The inks were splayed onto the MEA ͑geometric surface area of
the electrodes: 1.33 cm2͒. The spray-painted MEA was then pressed
at 110 MPa at room temperature for 30 s. The loading amounts of
Ni, Co, and Pt on the MEA were 2.7, 2.7, and 2.8 mg cm−2, respec-
tively. In all three cases, Co–PPY–C was used as the cathode cata-
lyst, and the loading amount of Co on the cathode was
Current density / mA cm-2
0.9
0.8
0.7
0.6
0.5
0.4
0.3
b)
0.24 mg cm−2
.
For the preparation of fuel, 5 M KOH aqueous solution was
prepared by diluting KOH ͑85 wt %͒ in ultrapure water. Hydrazine
solutions of the required concentration were prepared from a con-
centrated aqueous solution ͑60 wt %͒ and the 5 M KOH aqueous
solution by diluting these solutions in ultrapure water ͑Millipore
Milli-Q system, 18.2 M ⍀ cm͒.
The fabricated MEA ͑spherical shape; working electrode area
1.33 cm2͒ was inserted in a single cell to measure the cell perfor-
mance. An aqueous solution of 1 M hydrazine hydrate and 1 M
10
100
1000
10000
Current density / mA cm-2
KOH preheated at 55°C was supplied to the anode at 2 mL min−1
,
and oxygen humidified at 50°C was supplied to the cathode at
500 mL min−1. The cell temperature was controlled and maintained
at 80°C. The cell performance was evaluated after continuously op-
erating the cell for 10 h as a conditioning. Current–voltage ͑I-V͒
measurements were performed in the potentiostatic mode. The I-V
measurements were started only after voltage remained constant at
0.4 V for at least 15 min. Then, it was increased in steps of 20 mV
at time intervals of 6 s. The high-frequency resistance of the cell
was determined by ac impedance spectroscopy at 0.4 V.
In order to analyze the reaction during power generation by de-
termining the difference in the anode catalysts, the concentrations of
hydrazine and ammonia were measured by absorption spectropho-
tometry and ion chromatography, respectively. The schematic dia-
gram of the equipment used for exhaust composition analysis is
shown in Fig. 1 of Ref. 10.
Influence of hydrazine hydrate and KOH concentration on cell per-
formance.— The influence of hydrazine hydrate and KOH concen-
tration on fuel cell performance was investigated. During the mea-
surements, the hydrazine hydrate concentration in the aqueous solu-
tion was varied at 1, 2, 4, and 8 M, while the KOH concentration
was 1 M. Further, the influence of KOH concentration on fuel cell
performance was measured by using an aqueous solution of 4 M
hydrazine hydrate in the absence of KOH, and for 0.1 and 1 M
KOH. The cell performance was measured on the Ni anode
͑2.5 mg cm−2͒ and Co–PPY–C cathode ͑0.21 mg cm−2͒. Other ex-
perimental conditions were fundamentally the same as the above-
mentioned.
Figure 2. Influence of concentration of hydrazine hydrate on the cell volt-
ages and power density vs current density at ͑b͒ 1, ͑᭺͒ 2, ͑᭡͒ 4, and ͑᭿͒
8 M. The cell temperature was 80°C. An aqueous solution of hydrazine and
KOH ͑2 mL min−1, ␭
= 10, 20, 40, and 80 at i = 1 A cm−2͒ was supplied
stoic
to the anode, and humidified oxygen ͑55°C, relative humidity ͑RH͒ 27%,
500 mL min−1, ␭ = 23 at i = 1 A cm−2͒ was supplied to the cathode.
stoic
͑a͒ Electrochemical reaction
N2H4 + 4OH− → N2 + 4H2O + 4e− ͑four-electron reaction͒
͓1͔
N2H4 + 3OH− → N2 + ͑1/2͒H2 + 3H2O + 3e−
͑three-electron reaction͒
͓2͔
N2H4 + 2OH− → N2 + H2 + 2H2O + 2e− ͑two-electron reaction͒
͓3͔
N2H4 + OH− → N2 + ͑3/2͒H2 + H2O + e−
͑one-electron reaction͒
͓4͔
͓5͔
N2H4 + OH− → ͑1/2͒N2 + NH3 + H2O + e−
͑one-electron reaction͒
͑b͒ Chemical reaction
N2H4 → N2 + 2H2 ͑decomposition͒
3N2H4 → 4N2 + NH3 ͑decomposition͒
͓6͔
͓7͔
Results and Discussion
Comparison of anode catalysts.— The fuel cell performances
using Ni, Co, and Pt as an anode catalyst were measured. The I-V
performance and the power density performance of the cell using Ni
and Pt as an anode catalyst are shown in Fig. 2. It is observed that in
the anion-exchange membrane fuel cell using hydrazine as the fuel,
use of Ni and Co as anode catalysts results in a high power density
as compared to Pt. The anodic reaction of hydrazine in an alkaline
environment takes place through the following reactions
In the electro-oxidation reaction of hydrazine, the most desirable
reaction is the four-electron reaction given by Eq. 1.
In the graph shown in Fig. 3, Pt ͑829 mV͒ has a higher open-
circuit potential ͑OCP͒ than Ni ͑797 mV͒ and Co ͑792 mV͒. Ni
͑527 mW cm−2͒ and Co ͑517 mW cm−2͒ exhibit a higher power
density than Pt ͑389 mW cm−2͒. The difference between the I-V

Journal of The Electrochemical Society, 156 ͑4͒ B509-B512 ͑2009͒
B511
0.9
0.8
0.7
0.6
0.5
0.4
0.3
700
600
500
400
300
200
100
0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
700
600
500
400
300
200
100
0
0
500
1000
1500
0
500
1000
1500
Current density / mA cm-2
Current density / mA cm-2
Figure 3. Effect of anode catalysts on ͑a͒ cell voltage and power density vs
current density: ͑b͒ nickel, ͑᭺͒ cobalt, and ͑᭡͒ platinum. The loadings of
nickel, cobalt, and platinum were 2.7, 2.7, and 2.8 mg cm−2, respectively.
The cell temperature was 80°C. An aqueous solution of hydrazine and KOH
Figure 4. Influence of KOH concentration in fuel on the cell voltage and
power density vs current density: ͑b͒ without KOH, ͑᭺͒ 0.1 M KOH, and
͑᭡͒ 1 M and ͑᭿͒ 3 M KOH. The cell temperature was 80°C. An aqueous
solution of hydrazine and KOH ͑2 mL min−1, ␭
= 40 at i = 1 A cm−2
͒
stoic
͑2 mL min−1; ␭
= 10 at i = 1 A cm−2͒ was supplied to the anode, and
was supplied to the anode, and humidified oxygen ͑55°C, RH 27%,
stoic
humidified oxygen ͑55°C, RH 27%, 500 mL min−1
,
␭
= 23 at i
500 mL min−1, ␭ = 23 at i = 1 A cm−2͒ was supplied to the cathode.
stoic
stoic
= 1 A cm−2͒ was supplied to the cathode. ͑b͒ Tafel plots derived from I-V.
͑Same conditions as Fig. 2a.͒
suitable anode catalyst among Ni, Co, and Pt for use in the direct
hydrazine alkaline anion-exchange membrane fuel cell.
slopes of these anode catalysts does not influence the high-
frequency resistance of the cell because the resistance of these anode
catalysts is the same ͑0.15 ⍀ cm2͒. In the Tafel plot shown in Fig. 3,
although a clear difference is observed in the characteristics of these
anode catalysts in high-current regions, no clear difference in the
low-current regions below 100 mA cm−2 is observed. This plot
shows that the low ion-conductive anion-exchange ionomer does not
show a difference in catalytic activities of the above-mentioned an-
ode catalysts.
Because Ni and Co exhibited equivalent I-V characteristics, the
reaction at each anode was investigated. The concentrations of hy-
drazine and ammonia in the exhaust measured by absorption spec-
trophotometry and ion chromatography, respectively, are shown in
Table I. Ammonia is formed as a by-product of hydrazine decom-
position. However, ammonia does not deactivate the catalysts, un-
like carbon monoxide of direct methanol fuel cells.11 However, the
concentration of ammonia should be low to achieve high efficiency
and zero emissions. Hydrazine consumption and ammonia genera-
tion in the case of the Ni catalyst were less than that in the case of
Co catalyst. In the case of Co catalyst, the consumption of hydrazine
was considerably high, and the decomposition reaction given by
Eq.7 occurred at the OCP. There was no relationship between cur-
rent densities and ammonia concentrations, implying that ammonia
was generated by the chemical reaction.7 Moreover, this indicated
that Co was in the metal form at the OCP, because Co exhibited a
high rate of decomposition. An increase in the current density re-
sulted in the oxidation of Co and a decrease in the rate of decom-
position. From these results, it can be assumed that Ni is the most
Influence of hydrazine hydrate and KOH concentrations on fuel
cell performance.— The influence of the concentration of hydrazine
hydrate on fuel cell performance is discussed in this section. The
change in the I-V performance and power density performance with
a change in the hydrazine hydrate concentration from 1 to 8 M is
shown in Fig. 4. The cell performance improves until the hydrazine
hydrate concentration is 4 M. However, the cell performance dete-
riorates at a KOH concentration of 8 M. Crossover of hydrazine was
almost the same because the difference between the OCPs is small.
Although the high-frequency resistance of the cell remains almost
the same ͑0.15–0.16 ⍀ cm2͒ from 1 to 4 M, it increases rapidly
͑0.23 ⍀ cm2͒ at 8 M. The increase in the hydrazine concentration
from 1 to 4 M results in an increase in the reactivity. However, an
excess concentration of hydrazine ͑8 M͒ adversely affects the cell
performance because of a decrease in ion conductivity.
The anion-exchange membrane fuel cell, comprised of a Ni an-
ode and a Co–PPY–C cathode, exhibits the maximum power density
of 617 mW cm−2 at a hydrazine hydrate concentration of 4 M.
The influence of KOH concentration in the fuel on cell perfor-
mance was also investigated. The hydrazine hydrate concentration
was set to 4 M. The change in the I-V performance and power
density performance of the cell with a change in the KOH concen-
tration from 0 to 3 M is shown in Fig. 4. It was found that the
addition of KOH improved the cell performance remarkably. The
cell performance was very poor in the absence of KOH and for
0.1 M KOH. In particular, the cell exhibited a very low OCP of
0.282 V in the absence of KOH. Hydrazine is a weak base ͑pKa
= 7.96͒; at pH values well below pKa, hydrazine exists primarily as
+
hydrazinium ions ͑N2H5 ͒. This result indicates that the low OCP
Table I. The exhaust composition from the anode side of the fuel
cell during operation.
can be attributed to the restriction of the diffusion of hydrazinium
ions into the catalyst layer due to the inverse effect of an anion-
exchange resin of electrode. Further, it is considered that the addi-
tion of KOH improves the ion conductivity of anion-exchange mem-
branes. However, the cell performance decreases when the KOH
concentration is 3 M. This decrease in the cell performance is not
due to its high-frequency resistance ͑0.13 ⍀ cm2 at 3 M and
0.15 ⍀ cm2 at 1 M͒. Therefore, an increase in the KOH concentra-
tion results in a decrease in cell performance; this decrease can be
attributed to mass transportation that influences cathode
polarization.12,13 In alkaline fuel cells, alkaline liquid electrolytes
Current density Hydrazine hydrate Ammonia
Anode
Ni
Cathode
͑mA cm−2
͒
͑%͒
͑ppm͒
Co–PPY–C
0
4.98
4.53
4.34
4.77
4.12
3.99
33
38
75
1193
617
618
376
752
0
376
752
Co
Co–PPY–C

B512
Journal of The Electrochemical Society, 156 ͑4͒ B509-B512 ͑2009͒
Acknowledgments
with high concentration ͑9 M KOH͒ can be used. Electrolytes with
low concentration ͑1 M KOH͒ can be used in this anion-exchange
membrane fuel cell.
The authors thank Dr. T. Kobayashi ͑AIST͒ and Dr. M. Taniguchi
͑Otsuka Chemical͒ for their advice.
Conclusion
Daihatsu Motor Co., Ltd., assisted in meeting the publication costs of
this article.
1. In an alkaline anion-exchange membrane fuel cell using hy-
drazine as the fuel, the power density was higher in the case of Ni
and Co used as an anode catalyst than in the case of Pt used as an
anode catalyst.
2. Ni was found to be the most suitable anode catalyst among Co,
Ni, and Pt for use in the direct hydrazine anion-exchange membrane
fuel cell, because the rate of decomposition of Ni was lower than
that of Co, and Ni exhibits higher power density than Pt.
3. An increase in the hydrazine concentration from 1 to 4 M re-
sults in an increase in the reactivity. However, an excess concentra-
tion of hydrazine adversely affected the cell performance because of
a decrease in ion conductivity.
4. Addition of KOH can improve cell performance remarkably.
In the absence of KOH, hydrazine exists primarily as hydrazinium
ions ͑N2H+5͒. Low OCP can be attributed to prohibition of hy-
drazinium ion into the catalyst layer by an inverse affect of alkaline
anion-exchange resin of the electrode.
5. The anion-exchange membrane fuel cell, comprised of a Ni
anode and a Co–PPY–C cathode, exhibits a maximum power den-
sity of 617 mW cm−2 at a hydrazine hydrate concentration of 4 M.
References
1. K. Asazawa, K. Yamada, H. Tanaka, A. Oka, M. Taniguchi, and T. Kobayashi,
Angew. Chem., Int. Ed., 46, 8024 ͑2007͒.
2. L. A. Adams, S. D. Poynton, C. Tamain, R. C. T. Slade, and J. R. Varcoe, Chem.
Sus. Chem., 1.2, 79 ͑2008͒.
3. J. R. Varcoe and R. C. T. Slade, Fuel Cells, 2, 187 ͑2005͒.
4. J. R. Varcoe, R. C. T. Slade, and E. L. H. Yee, Chem. Commun. (Cambridge),
2006, 1428.
5. E. H. Yu and K. Scott, J. Appl. Electrochem., 35, 91 ͑2005͒.
6. K. Yamada, K. Yasuda, N. Fujiwara, Z. Siroma, H. Tanaka, Y. Miyazaki, and T.
Kobayashi, Electrochem. Commun., 5, 892 ͑2003͒.
7. G. E. Evans and K. V. Kordesch, Science, 158, 1148 ͑1967͒.
8. http://www.mofa.go.jp/Mofaj/Gaiko/summit/toyako08/showcase/pdf/E08.pdf, last
accessed Feb 18, 2009.
9. R. Bashyam and P. Zelenay, Nature, 443, 63 ͑2006͒.
10. K. Yamada, K. Asazawa, K. Yasuda, T. Ioroi, H. Tanaka, Y. Miyazaki, and T.
Kobayashi, J. Power Sources, 115, 236 ͑2003͒.
11. S. Gottesfeld and J. Pafford, J. Electrochem. Soc., 135, 2651 ͑1988͒.
12. Z. P. Li, B. H. Liy, K. Arai, A. Asaba, and S. Suda, J. Power Sources, 126, 28
͑2004͒.
13. W. X. Yin, Z. P. Li, J. K. Zhu, and H. Y. Qin, J. Power Sources, 182, 520 ͑2008͒.