Electrochemical hydrogenation of crystalline Co powder

Article abstract of DOI:10.1149/1.2189978

The electrochemical hydrogenation properties of two boron-free crystalline Co powder samples were investigated. They exhibited almost the same maximum discharge capacities with 2-h charging, 415 and 428 mAhg (corresponding to Co H0.91 and Co H0.93) for samples A and B, respectively. The high discharge capacity of Co powder was attributed to hydrogenation of Co and phase transition between the phases of hexagonal close-packed and face centered cubic.

Full text of DOI:10.1149/1.2189978

A1128
Journal of The Electrochemical Society, 153 ͑6͒ A1128-A1131 ͑2006͒
0013-4651/2006/153͑6͒/A1128/4/$20.00 © The Electrochemical Society
Electrochemical Hydrogenation of Crystalline Co Powder
*
S.-R. Chung, K.-W. Wang, and T.-P. Perng
Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu 30043, Taiwan
The electrochemical hydrogenation properties of two boron-free crystalline Co powder samples were investigated. They exhibited
almost the same maximum discharge capacities with 2-h charging, 415 and 428 mAh/g ͑corresponding to CoH0.91 and CoH0.93͒ for
samples A and B, respectively. The high discharge capacity of Co powder was attributed to hydrogenation of Co and phase
transition between the phases of hexagonal close-packed and face centered cubic.
© 2006 The Electrochemical Society. ͓DOI: 10.1149/1.2189978͔ All rights reserved.
Manuscript received November 21, 2005; revised manuscript received February 6, 2006. Available electronically April 19, 2006.
Hydrogen can be stored in various forms such as high pressure
Experimental
gas, metal hydrides, adsorption on carbon materials, liquid hydro-
gen, etc. Many types of hydrogen storage alloy have been developed
in the past few decades, including AB-type intermetallic
compounds,1 AB2-type Laves phase alloys,2 Mg-based alloys,3 and
AB5-type rare-earth alloys.4 For battery application, lower plateau
pressure and good kinetics are required. For AB2-type alloys, al-
though the theoretical discharge capacities are higher than those of
AB5-type alloys, they are more difficult to activate to be used as a
negative electrode material in batteries. For Mg-based alloys, their
high discharge capacities are attractive, but the low cycling stability
and poor kinetics are not suitable for practical application. The
AB5-type hydrogen storage alloys are more often used in commer-
cial batteries because of their suitable plateau pressure and fast ac-
tivation. However, the discharge capacities of AB5-type alloys are
decreased by pulverization and oxidation of the alloys. Willems5 and
Chartouni et al.6 pointed out that the cycle life of electrode could be
improved by partial substitution of Ni by Co that lowers the hard-
ness of the alloys. Mixing hydrogen storage alloy with Co powder
might result in formation of a conductive layer of precipitate which
ensures good electrical contact of the alloy powder, in addition to a
slight increase in electrochemical capacity due to the reaction of
cobalt in the charge and discharge cycling.7 Durairajan et al. pointed
out that the discharge capacity of pure Co was only 50 mAh/g,8 that
might be due to the faradaic reaction of Co/Co͑OH͒2.
Recently, it has been reported that amorphous Co nanoparticles
prepared by reduction with NaBH4 can absorb a large amount of
hydrogen.9-12 When ultrafine Co–B amorphous alloy particles were
tested by electrochemical charging/discharging, the reversible maxi-
mum discharge capacity could be as high as 300 mAh/g.12 In the
cyclic voltammetry ͑CV͒ test, Mitov et al. proposed that the ca-
thodic and anodic peaks are due to electrochemical adsorption and
desorption of hydrogen, respectively. The residual B in the sample
may play an important role by exchanging position with hydrogen
atom in the metallic host lattice.9 Therefore, it would be possible to
use ultrafine Co–B amorphous powder as a new hydrogen storage
material.
Two Co powder samples were used for the electrochemical hy-
drogenation test. Sample A was a commercial powder purchased
from OM Group. Sample B was prepared by a precipitation method.
A precursor solution of 0.5 M of Co͑NO3͒2 dissolved in C2H5OH
was prepared, and the pH value was controlled at 12–13 by addition
of NaOH. N2H4 was then used as a reducing agent at 70°C. A black
Co powder was formed, which was separated from the solution by a
magnetic rod and dried at 50°C. Because N2H4 was used as the
reducing agent, no boron was present in sample B. The structures of
the samples were examined by X-ray diffraction ͑XRD͒ with a Cu
K␣ radiation.
The charge and discharge curves were measured in an electro-
chemical test cell, which contained one piece of positive electrode,
one piece of Co powder ͑0.02 g͒ as the negative electrode, and
polypropylene as the separator. The electrolyte was 6 M KOH
+ 1 wt % LiOH. The positive electrode material consisted of nickel
hydroxide, 5 wt % Co, and 5 wt % CoO. Each of the positive and
negative electrode materials was mixed with 3 wt % poly͑tetraflo-
roethylene͒ to form a paste and coated on a piece of Ni foam. The
electrode plates were cold pressed at a pressure of 50 kgf/cm2 for
30 s. Because the diffraction peaks of face-centered cubic ͑fcc͒
phase Co overlap with those of Ni foam, some pellets made of Co
and Cr powder at a mass ratio 1 to 4 were also tested to identify the
presence of fcc-phase Co during the charge–discharge process. The
charge and discharge currents were both set at 10 mA ͑equivalent to
500 mA/g͒, and the cutoff voltage was 900 mV. The charging time
was 1 or 2 h, and all experiments were conducted at 25 C.
Results and Discussion
Phase characterization.— The XRD patterns for the two
samples are shown in Fig. 1. In general, Co has two structures, fcc
͑␥-phase͒ and hexagonal close-packed ͑hcp͒ ͑␧-phase͒. The former
is metastable and the latter is stable at room temperature and ambi-
ent pressure.13 Sample A shows almost a pure hcp-Co phase. The
particle size is calculated to be ϳ31 nm by Scherrer’s equation. For
sample B, the relative intensities of the three major peaks are differ-
ent from those of sample A. Based on the standard pattern of the hcp
phase, the ͑101͒ peak at 47.57° is the strongest, followed by the
͑002͒ peak at 44.76°. For sample B, it is noted that the ͑002͒ peak is
stronger, presumably because of the presence of a small amount of
fcc phase, whose strongest ͑111͒ peak is located at 44.22°, and the
second strongest peak ͑200͒ is at 51.52°. The particle size of sample
B is approximately 46 nm.
During the charge process the negative electrode undergoes re-
duction reaction, pure cobalt cannot be reduced to other form be-
cause it is in the zero valence. Therefore, the decomposition of water
might be progressed. The hydrogen fugacity in electrochemical sys-
tems could be as high as several GPas. This makes it possible to
form hydrides for group VIII metals that can be identified by neu-
tron scattering. The electroreduction of water allows a higher hydro-
gen fugacity to be reached than in conventional gas-phase hydroge-
nation technique. This paper is devoted to the study of
electrochemical hydrogenation properties of B-free crystalline Co
powder. The hydrogenation process is elaborated. The discharge per-
formance and phase transformation of Co powder during repeated
charging and discharging are discussed.
Electrochemical hydrogenation properties.— Initially, without
charging, a blank discharge test with a fresh commercial Co powder
sample used as the negative electrode and NiOOH as the positive
electrode was made. The maximum discharge capacity was only
80 mAh/g. This means that the discharge capacity from the
Co/Co͑OH͒ faradaic reaction is low. Figure 2 shows the cyclic
2
discharge capacities of the two Co powder samples with two differ-
ent charge times. The activation rates for both samples were very
fast compared with those of typical hydrogen storage alloys. The
*
E-mail: tpperng@mx.nthu.edu.tw
Downloaded on 2015-06-06 to IP 134.129.182.74 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 153 ͑6͒ A1128-A1131 ͑2006͒
A1129
Figure 3. Electrochemical charge and discharge curves of the Co samples
for the second cycle. Both samples are charged for 60 min, rest for 10 min,
and then discharged to a cutoff potential of 0.9 V.
Figure 1. XRD patterns of the Co powder samples.
maximum discharge capacity could be obtained in the first few
cycles. When the charging time is 1 h, the maximum discharge ca-
pacities are 379 and 399 mAh/g for samples A and B, respectively.
As the charge time is extended to 2 h, the maximum discharge ca-
pacities are increased to 415 and 428 mAh/g, respectively. Both
samples have almost the same discharge capacity, near 350 mAh/g,
after 30 cycles of test. The maximum discharge capacities and
equivalent hydride formula ͑CoHx͒ of the two samples are summa-
rized in Table I.
In gas-phase hydrogenation, Co reacts with H2 to form CoH0.6
and CoH1.0 for the hcp and fcc structures, respectively, as hydrogen
pressure increases.13 By theoretical calculation, the maximum dis-
charge capacities of Co are 459 mAh/g for fcc-Co and 275 mA/g
for hcp-Co, based on the compositions of CoH1.0 and CoH0.6, re-
spectively. Since the discharge capacity due to the faradaic reaction
tial curves of the two samples at a constant current of 500 mA/g
during the second cycle. The cell voltage is between 1.3 and 1.7 V
in charging and between 1.3 and 1.1 V in discharging. It is noted
that there are two stages in the curve during the charging process.
The first stage has a higher slope which might be ascribed to easier
hydrogenation of the hcp phase, while the second stage with a
smaller slope is due to more difficult hydrogenation of the fcc phase.
The discharge curves of both A and B are almost the same. The
discharge plateaus of both samples can be compared with those of
AB2-type and AB5-type alloys. In general, the discharge plateaus
are between 1.3 and 1.2 V and between 1.3 and 1.1 V for AB2- and
AB5-type alloys, respectively.14 Therefore, the discharge capacities
of samples A and B can be attributed to the desorption of hydrogen
from Co.
Phase transition.— The XRD patterns of the samples after charge
and discharge are shown in Fig. 4 and 5, respectively. After charge,
the hcp-Co phase can be observed in both samples and no secondary
phase appears, as shown in Fig. 4. Based on the peak positions, the
unit cell volumes of samples A and B are calculated to be 22.16 and
22.11 Å3, respectively. Compared with the standard unit cell volume
hcp-Co ͑22.03 Å3͒, the unit cell volumes of both samples have been
expanded after charging, presumably because hydrogen atoms could
occupy the interstitial sites in the hcp-Co lattice. After discharge, as
shown in Fig. 5, the peak intensity of hcp-Co has decreased and
Co/Co͑OH͒ is only 80 mAh/g, the cyclic discharge capacities ob-
2
served for this electrochemical test are attributed to the CoHx/Co
reaction rather than from the faradaic reaction. With the above dis-
charge capacities, the maximum hydrogen storage capacity of
sample A is equivalent to 1.54 wt %, which corresponds to a com-
position of CoH0.91 in between CoH0.6 and CoH1.0. Therefore, it is
believed that both hcp-Co and fcc-Co coexist during the electro-
chemical test. If the samples behave like gas-phase hydrogenation
during electrochemical charging, the hcp-Co in sample A starts to
absorb hydrogen until saturation has been reached. Some of the
hcp-Co phase then transforms to fcc-Co. From Fig. 1, although there
is some fcc-Co present in sample B, the hcp phase is the main phase.
Therefore, the discharge behavior of sample B is almost the same as
that of sample A.
more Co͑OH͒ appears in both samples. This means that the fara-
2
daic reaction Co/Co͑OH͒ might have also occurred. The unit cell
2
volumes of samples A and B have been reduced to 22.09 and
22.07 Å3, respectively, because of release of the hydrogen atoms
during discharge. However, they are still slightly larger than that of
standard hcp-Co, implying that there is some residual hydrogen in
the Co lattice. Based on this result, it proves that hydrogen is
absorbed/desorbed by hcp-Co during the charge/discharge process.
Since Ni foam was used as electron collector in the present elec-
trochemical measurement, the peaks of fcc-Co, if any, would have
overlapped with those of Ni foam. Therefore, it cannot distinguish
the phase transformation between hcp-Co and fcc-Co during the
electrochemical charge/discharge cycling by using Ni foam. To
Figure 3 shows the electrochemical charge and discharge poten-
Table I. Electrochemical hydrogenation capacities of two Co
powder samples.
Discharge capacity Storage capacity Equivalent formula
͑mAh/g͒
͑wt %͒
͑x in CoHx͒
Sample
1 h
2 h
1 h
2 h
1 h
2 h
A
B
379
399
415
428
1.40
1.48
1.54
1.58
0.83
0.87
0.91
0.93
Figure 2. Discharge capacity curves of the two Co powder samples.
Downloaded on 2015-06-06 to IP 134.129.182.74 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1130
Journal of The Electrochemical Society, 153 ͑6͒ A1128-A1131 ͑2006͒
Figure 6. XRD pattern of the pellet consisting of sample A and Cr after
charge at the seventh cycle.
Figure 4. XRD patterns of the samples after charge.
check the possibility of the presence of fcc-Co, non-fcc metallic
binders have to be used. Furthermore, the powder has to be easy to
deform, stable in KOH solution, and not absorbing hydrogen.
Among common metallic powders that meet these requirements, Cr
was selected for this purpose. Sample A was mixed with Cr powder
to form a pellet and then tested in the same electrochemical system.
The XRD pattern after charging at the seventh cycle is shown in Fig.
6. In addition to Co and a small amount of Co͑OH͒2, both hcp-Co
and fcc-Co phases can be observed. This demonstrates that hcp-Co
has indeed, been partially transformed to fcc-Co, and this phase
transformation results in increased discharge capacity for sample A.
The standard unit cell volume of fcc-Co is 44.512 Å3. This was
increased to 44.663 Å3 after charging, based on the peak positions
shown in Fig. 6. It is also believed that similar phase transition
between hcp and fcc takes place in sample B.
used as an alloying element to improve the cycle life of MH elec-
trodes. It is also used in positive electrode to improve conductivity
and to reduce degradation of the electrode. Is it possible to obtain
CoHx by electrochemical charging shown in Reaction 4? When Co
was immersed in the KOH solution, it was observed that a thin
cobalt hydroxide layer was formed on the surface. It also covered
the cobalt electrode surface during cyclic charge/discharge. If the
surface of cobalt electrode is covered by cobalt hydroxide, the Re-
action 3 might occur. As mentioned above, with fresh commercial
Co powder used as the negative electrode and NiOOH as the posi-
tive electrode, the maximum discharge capacity is only 80 mAh/g
without prior charging. Therefore, the discharge capacity due to the
Co͑OH͒2/Co faradaic reaction is low. The discharge capacities ob-
tained from the present study should be attributed primarily to the
dehydrogenation of CoHx. Although the electrochemical hydrogena-
tion and dehydrogenation of Co are similar to those of conventional
hydrogen storage alloys ͑i.e., AB2, AB5, etc.͒, Co must undergo
phase transition during the charging process in order to reach the
maximum discharge capacity. It is also believed that the electro-
chemical hydrogenation of Co is an intrinsic property, regardless of
the synthesis process and crystallinity. Furthermore, alloying with
boron is not a prerequisite for hydrogenation of Co.
Hydrogenation mechanism.— For a nickel–metal hydride ͑Ni–MH͒
battery, the reactions of positive and negative electrodes during the
charge process can be written as
Ni͑OH͒ + OH− → NiOOH + H2O + e−
͓1͔
2
M + H2O + e− → MH + OH−
͓2͔
If the hydrogen storage alloy electrode is replaced by pure cobalt
powder, its charge reaction may be written as either of the following
two reactions
Conclusion
Co͑OH͒ + 2e− → Co + 2OH−
Co + xH2O + xe− → CoHx + xOH−
It is well known that Co is difficult to form hydride by gas-phase
hydrogenation at ambient pressures. Therefore, Co is commonly
͓3͔
In this paper, we report the electrochemical hydrogenation of two
crystalline Co powders. The electrochemical hydrogenation proper-
ties were measured in KOH solution. Although Co is difficult to
form hydride by gas-phase hydrogenation, high discharge capacities
by electrochemical hydrogenation were observed. For samples A
and B, the maximum discharge capacities with 2 h charging are
415 mAh/g ͑1.54 wt %͒ and 428 mAh/g ͑1.58 wt %͒, correspond-
ing to compositions of CoH0.91 and CoH0.93, respectively. The phase
transition between hcp-Co and fcc-Co takes place during the charge/
discharge process, resulting in high hydrogenation capacities.
2
͓4͔
Acknowledgments
This work was supported by the Ministry of Education of Taiwan
under contract no. A-91E-FA04-1-4 and the Ministry of Economic
Affairs under contract no. 93-EC-17-A-08-S1-0003.
National Tsing Hua University assisted in meeting the publication costs
of this article.
References
1. C. H. Chiang, Z. H. Chin, and T. P. Perng, J. Alloys Compd., 307, 259 ͑2000͒.
2. S. S. Huang, H. J. Chuang, and S. L. I. Chan, J. Alloys Compd., 330–332, 617
͑2002͒.
3. G. Liang, J. Alloys Compd., 370, 123 ͑2004͒.
4. S. R. Chung and T. P. Perng, J. Alloys Compd., 356–357, 768 ͑2003͒.
Figure 5. XRD patterns of the samples after discharge.
Downloaded on 2015-06-06 to IP 134.129.182.74 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 153 ͑6͒ A1128-A1131 ͑2006͒
A1131
5. J. J. G. Willems, Philips J. Res., 39, 1 ͑1984͒.
6. D. Chartouni, F. Meli, A. Züttel, K. Gross, and L. Schlapbach, J. Alloys Compd.,
10. M. Mitov, A. Popov, and I. Dragieva, Colloids Surf., A, 149, 413 ͑1999͒.
11. J. X. Yu, L. Wang, Y. D. Wang, H. Dong, and H. X. Yang, J. Electrochem. Soc.,
241, 160 ͑1996͒.
151, A1124 ͑2004͒.
7. T. Weizhong, G. Yingxin, and Z. Haoyu, J. Appl. Electrochem., 25, 874 ͑1995͒.
8. A. Durairajan, B. S. Haran, B. N. Popov, and R. E. White, J. Power Sources, 83,
114 ͑1999͒.
12. Y. D. Wang, X. P. Ai, and H. X. Yang, Chem. Mater., 16, 5194 ͑2004͒.
13. V. E. Antonov, T. E. Antonova, M. Baier, G. Grosse, and F. E. Wagner, J. Alloys
Compd., 239, 198 ͑1996͒.
9. M. Mitov, A. Popov, and I. Dragieva, J. Appl. Electrochem., 29, 59 ͑1999͒.
14. K. Hong, J. Alloys Compd., 321, 307 ͑2001͒.
Downloaded on 2015-06-06 to IP 134.129.182.74 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).