Preparation of a High Surface Area Nickel Electrode by Alloying and Dealloying in a ZnCl2-NaCl Melt

Article abstract of DOI:10.1149/1.1595662

An electrochemical method of preparing a high surface area nickel electrode by NiZn alloy formation and subsequent dealloying in ZnCl ₂-NaCl (60-40 mol %) at 450°C was studied. A nickel electrode was kept at 0.22 V vs. Zn in ZnCl₂-NaCl(sat) where γ-NiZn alloy was formed, and then at 0.5 V where partial dealloying occurred to yield α-NiZn alloy. Scanning electron microscope observation showed that the microporous structure was formed on the surface. The roughness factor of a sample was determined to be 240 from the double layer capacitance obtained in the ac impedance measurement. Thickness, porosity, and pore size of the surface layer were estimated.

Full text of DOI:10.1149/1.1595662

Journal of The Electrochemical Society, 150 ͑9͒ C585-C590 ͑2003͒
C585
0013-4651/2003/150͑9͒/C585/6/$7.00 © The Electrochemical Society, Inc.
Preparation of a High Surface Area Nickel Electrode by
Alloying and Dealloying in a ZnCl2-NaCl Melt
,z
*
A. Katagiri and M. Nakata
Electrochemistry Laboratory, Faculty of Integrated Human Studies and Graduate School of Energy Science,
Kyoto University, Sakyo-ku, Kyoto 606-8501, Japan
An electrochemical method of preparing a high surface area nickel electrode by NiZn alloy formation and subsequent dealloying
in ZnCl2-NaCl ͑60-40 mol %͒ at 450°C was studied. A nickel electrode was kept at 0.22 V vs. Zn in ZnCl2-NaCl͑sat͒ where
␥-NiZn alloy was formed, and then at 0.5 V where partial dealloying occurred to yield ␣-NiZn alloy. Scanning electron micro-
scope observation showed that the microporous structure was formed on the surface. The roughness factor of a sample was
determined to be 240 from the double layer capacitance obtained in the ac impedance measurement. Thickness, porosity, and pore
size of the surface layer were estimated.
© 2003 The Electrochemical Society. ͓DOI: 10.1149/1.1595662͔ All rights reserved.
Manuscript submitted September 24, 2002; revised manuscript received March 10, 2003. This was in part Paper 1465 presented
at the Philadelphia, Pennsylvania, Meeting of the Society, May 12-17, 2002. Available electronically July 17, 2003.
Zinc chloride-sodium chloride melts have relatively low liquids
temperatures and exhibit characteristic properties due to the Lewis
acidity of ZnCl2 . We studied the electrochemical behavior of the
nickel electrode in ZnCl2-NaCl ͑60-40 mol %͒ melt at 450°C by
cyclic voltammetry ͑CV͒ and constant potential electrolysis, and
reported that different types of NiZn alloys were formed depending
upon the potential.1 Equilibrium potentials of NiZn alloys of differ-
ent structures and compositions were calculated from the thermody-
namic data of zinc activity in alloys.2 CV revealed that the anodic
dissolution of zinc from NiZn alloys occurred on the potential scan
in the positive direction. Formation of crevices and pores, and there-
fore the increase of surface area, were expected in the dealloying
process. Preliminary experiments were performed in which ␥-NiZn
alloy was formed and then dealloyed to ␣-NiZn alloy, and the for-
mation of a porous surface was verified by a scanning electron mi-
croscope ͑SEM͒.3 The present paper describes an electrochemical
study of such alloying and dealloying, and presents some results of
X-ray diffraction ͑XRD͒, X-ray fluorescence analysis, SEM obser-
vation, and electrochemical impedance measurements. Properties of
the porous layer are also discussed.
High surface area nickel can be used as the cathode material for
hydrogen evolution in water electrolysis and in chlor-alkali
process.4-6 Several methods of preparing Raney nickel electrodes
from NiAl and NiZn alloys have been proposed.7-14 For example,
NiZn alloy coatings are made by electrodeposition in aqueous
solutions7-10 and by reaction of nickel with zinc vapor
͑sherardizing͒.11,12 Zinc is then leached out in hot concentrated al-
kali hydroxide solutions to form high surface area nickel electrodes.
The method proposed in this paper has an advantage; alloying and
dealloying can be made by controlling the potential in the same
electrolytic bath. Also the obtained product may have properties
different from those made in aqueous solutions.
were referred to this electrode. The equilibrium potential of zinc in
ZnCl2-NaCl ͑60-40 mol %͒ at 450°C was 0.23 V vs. RE.
Zinc chloride was made as an aqueous solution from zinc car-
bonate and hydrochloric acid, and freed from heavy metal impuri-
ties. The solution was concentrated by heating, and the solidified
material was vacuum dried at 200°C. The zinc chloride was further
purified by sublimation at 450°C under vacuum. Sodium chloride
was dried at 350°C under vacuum. ZnCl2-NaCl melt ͑60-40 mol %͒
was prepared from the thus obtained ZnCl2 and NaCl.
Characterization of produced materials was performed by using a
Shimadzu model XRD-6000 X-ray diffractometer with the Cu K␣
radiation, a Shimadzu model XRF-1500 X-ray fluorescence spec-
trometer. Microscopic observation was conducted with JEOL model
JSM-T100, Hitachi S-3500H, and Hitachi X-900 SEMs. Electron
probe microanalysis ͑EPMA͒ was carried out with a Shimadzu
model EPMA-1600 instrument. AC impedance was measured with
an NF Corporation model 5020 frequency response analyzer ͑FRA͒
and a Hokuto Denko HA-501 potentiostat.
Results and Discussion
Cathodic and anodic treatments of nickel electrodes.—In order
to establish appropriate conditions for cathodic and anodic treat-
ments, it is helpful to reinvestigate the voltammetric behavior of
nickel. Figure 1 reproduces a cyclic voltammogram of a nickel wire
electrode in ZnCl2-NaCl ͑60-40 mol %͒ melt at 450°C.1 The Figure
also shows the regions in which metallic zinc, ␥-, ␤1-, and ␣-NiZn
alloys are thermodynamically stable.1,2 A cathodic current observed
at potentials below 0.28 V on the leftward scan is due to the forma-
tion of ␥-NiZn alloy. An anodic current observed around 0.36 V on
the rightward scan corresponds to the dissolution of zinc from
␥-NiZn to yield ␤1-NiZn alloy. At potentials above 0.49 V an an-
odic reaction of ␤1-NiZn to form ␣-NiZn alloy occurs, and the
dissolution of zinc from ␣-NiZn follows. Although the deposition of
zinc can occur at potentials below the equilibrium potential of zinc
(E ϭ 0.23 V), the absence of the corresponding anodic current on
the rightward scan indicates that metallic zinc, if any, reacts imme-
diately with the nickel substrate to yield ␥-NiZn alloy.
Based on the voltammetric results, cathodic treatment at Ec
ϭ 0.22 V and anodic treatment at Ea ϭ 0.50 V were tried. The al-
loy phases expected were ␥-NiZn at 0.22 V and ␣-NiZn at 0.50 V.
Thus, a nickel plate ͑1 cm2͒, which had been electropolished in
advance, was cathodically treated at 0.22 V until a quantity of elec-
tricity Qc ϭ 35 C/cm2 passed. During this process ͑ca. 19 min͒, the
color of the electrode was silver white. As shown in Fig. 2, the
initial current of 70 mA/cm2 dropped quickly to 40 mA/cm2 after 2
min, and further decreased to 20 mA/cm2 after 19 min. When the
circuit was opened, the rest potential of the electrode was 0.27 V.
Subsequent anodic treatment was conducted at 0.50 V until the cur-
Experimental
Cathodic and anodic treatments of nickel were performed in a
stationary ZnCl2-NaCl ͑60-40 mol %͒ melt at 450°C in a Pyrex
glass cell. Nickel plates ͑Nilaco, 99.7%, 0.1 mm thick,
5
ϫ 10 mm, total area 1 cm2͒ were used as the starting material. Prior
to use they were electropolished in 1 M HCl at 0.5 A/cm2 for 30 s.
The counter electrode was a glassy carbon rod ͑Tokai Carbon, GC-
20, 3 mm diam͒ separated from the main compartment by a glass
frit. The reference electrode ͑RE͒ was molten zinc in ZnCl2-NaCl
͑saturated with NaCl͒ which was separated from the main compart-
ment by a small piece of sodium ␤ -alumina. All potential values
Љ
* Electrochemical Society Active Member.
z E-mail: katagiri.akira@a0016857.mbox.media.kyoto-u.ac.jp
Downloaded on 2014-07-01 to IP 128.252.67.66 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C586
Journal of The Electrochemical Society, 150 ͑9͒ C585-C590 ͑2003͒
Figure 1. Cyclic voltammogram of a nickel wire electrode in a ZnCl2-NaCl
͑60-40 mol %͒ melt at 450°C. Scan rate, 0.2 V/s. The alloy phases and
metallic zinc are thermodynamically stable in the regions divided by the
broken lines.
rent dropped to zero. The surface of the electrode turned black im-
mediately after the start of anodic treatment. As shown in Fig. 2, the
anodic current changed in two steps. Thus the initial current of 200
mA/cm2 decreased sharply, but a plateau of ca. 40 mA/cm2 appeared
after 2 min, which gradually decreased to zero. Such change in the
anodic current suggests that two different processes occurred in
dealloying. The quantity of electricity was Qa ϭ (26
Ϯ 0.1) C/cm2.
Figure 3. XRD patterns of ͑a͒ Sample I obtained by cathodic treatment
(Ec ϭ 0.22 V, Qc ϭ 35 C/cm2) and ͑b͒ Sample II obtained by cathodic
(Ec ϭ 0.22 V, Qc ϭ 35 C/cm2) and anodic (Ea ϭ 0.50 V, Qa
ϭ 26 C/cm2) treatments in ZnCl2-NaCl ͑60-40 mol %͒ at 450°C. ͑᭹͒ Ni,
͑᭺͒ ␣-NiZn alloy, and ͑ᮀ͒ ␥-NiZn alloy.
In order to investigate structural and property changes in ca-
thodic and anodic treatments, two samples were prepared. Sample I
was a nickel electrode treated only cathodically (Ec ϭ 0.22 V,Qc
ϭ 35 C/cm2); sample II was another one treated cathodically (Ec
ϭ 0.22 V,Qc ϭ 35 C/cm2) and then anodically (Ea ϭ 0.50 V,Qa
ϭ 26 C/cm2). Both samples were rinsed with water and dried.
XRD.—XRD measurements were performed to investigate the
crystallographic change in cathodic and anodic treatments. Figure 3a
and b show the results of sample I and sample II, respectively.
Sample I exhibits the diffraction patterns of ␥-NiZn alloy and nickel,
whereas sample II exhibits those of ␣-NiZn alloy and nickel. These
results confirm that ␥-NiZn alloy is formed by the cathodic treat-
ment at 0.22 V, and anodic dissolution of zinc from ␥-NiZn alloy
occurs at 0.50 V leaving ␣-NiZn alloy. Table I summarizes the ob-
tained lattice parameters which are in good agreement with the cor-
responding literature values.15,16
In our previous work,1 ␥-NiZn alloy was formed on a nickel
electrode in the constant potential electrolysis at 0.20 and 0.25 V,
whereas the formation of ␣-NiZn alloy was not confirmed at 0.50 V.
It seems that the formation of ␣-NiZn alloy occurs more easily in
dealloying of ␥-NiZn than in alloying of nickel at the same poten-
tial. As discussed previously,1 the diffusion coefficient of zinc is
larger in alloys of higher zinc concentration,17 and therefore zinc
atoms diffuse out from the ␥-NiZn alloy more easily than they dif-
fuse into the nickel metal. Another point is the difference in the
surface area. In the case of dealloying of ␥-NiZn, crevices and pores
Figure 2. Variation of current during cathodic and anodic treatments of a Ni
electrode in ZnCl2-NaCl ͑60-40 mol %͒ at 450°C. Cathodic treatment, Ec
ϭ 0.22 V, Qc ϭ 35 C/cm2; anodic treatment, Ea ϭ 0.50 V, Qa
ϭ 26 C/cm2.
Downloaded on 2014-07-01 to IP 128.252.67.66 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, 150 ͑9͒ C585-C590 ͑2003͒
C587
Table I. Lattice parameters of Ni, ␣-NiZn, and ␥-NiZn alloys.
a0
͑present work͒
a0
Crystal
structure
͑Ref. 15, 16͒
͑Å͒
Substance
͑Å͒
␥-NiZn
␣-NiZn
Ni
bcca
fccb
fcc
8.931
3.574
3.529
8.919
3.551
3.527
a bcc is body-centered cubic.
b fcc is face-centered cubic.
develop due to the shrinking, and therefore the surface area in-
creases, which accelerates the dealloying process. In the case of
alloying of nickel, such mechanism does not work.
X-ray fluorescence (XRF) analysis.—Table II shows the results
of XRF analysis of samples I and II. The intensity of the Zn K␣
signal was much larger in sample I than in sample II, whereas in-
tensities of Ni K␣ and Ni K␤ signals were larger in sample II than
in sample I. It is apparent that the zinc content of sample I was much
larger than that of sample II, although quantitative interpretation is
difficult for these layered materials because the observed X-ray in-
tensity is the sum of signals from different depths. The results are
consistent with the XRD data that sample I contains ␥-NiZn alloy
and sample II contains ␣-NiZn alloy.
SEM observation and EPMA.—In order to reveal microscopic
morphology of samples I and II, SEM observation was performed.
Figures 4 and 5 show SEM photographs of the surfaces of sample I
and sample II. While the surface of sample I seems compact with
sparse microcracks, the surface of sample II has many crevices and
pores, suggesting that the surface has a large internal area. Figure 6
shows SEM images of the surface of sample II in higher magnifica-
tions. The pore structure consists of interpenetrating phases of solid
and void. Figure 7 shows the cross section of sample II. A porous
layer of the thickness of about 10 ␮m can be noticed on the top of
the nickel substrate.
Distributions of nickel and zinc contents in the direction perpen-
dicular to the electrode surface were measured by EPMA using
cross-sectional samples. Figure 8 shows a backscattering electron
image ͑BEI͒ of a cross section of sample I, together with the inten-
sity profiles of Ni K␣ and Zn K␣ X-rays along a selected line
indicated on BEI. X-ray signals provide information on local nickel
and zinc contents with a resolution of 1 ␮m. BEI of sample I exhib-
its a surface layer of 10-13 ␮m in which intensity ratio of the Zn K␣
to Ni K␣ signal is about 2. Judging from XRD and XRF data, the
surface layer is considered to be ␥-NiZn alloy. It is noticed that
another type of NiZn alloy ͑such as ␤1 and ␣͒ does not exists in a
thickness larger than the limit of resolution of EPMA.
Figure 4. SEM photographs of the surface of sample I. See the captions of
Fig. 3.
Figure 9 shows EPMA results of a cross section of sample II. A
porous layer is recognized in which the intensity ratio of the Zn K␣
to Ni K␣ signal is about 1/8. This porous layer apparently consists
of ␣-NiZn alloy. Beneath this porous layer, there is no other alloy
layer such as ␤1- and ␥-NiZn. Comparing the surface layers of
sample I and sample II, the intensity ratio of Zn K␣ to Ni K␣ signal
has decreased by a factor of ca. 16. This change is due to the trans-
formation of ␥-NiZn to ␣-NiZn alloy.
Impedance measurements.—The roughness factor, which is the
ratio of the true surface area to the geometric surface area, was
determined from the double layer ͑dl͒ capacitance. Thus a piece of
sample was soldered to a copper wire, and fabricated to the working
electrode using a glass tube and Teflon seal tape. AC impedance of
the electrode was measured in 1 M NaOH solution at 30 °C at the
potential of Ϫ0.1 V vs. the reversible hydrogen electrode. The am-
plitude of AC signal was 10 mV, and the frequency was varied from
20 kHz to 100 mHz. The dl capacitance Cd was determined from the
Cole-Cole plot of impedance data. Table III summarizes the ob-
tained results. Assuming that the dl capacitance is proportional to
the true surface area and taking the smooth nickel plate ͑not elec-
tropolished͒ as the reference, the roughness factor of sample II is
calculated to be 240.
Table II. Results of XRF analysis.a
Specimen
Sample I
Fluorescent X-ray
Intensity ͑kcps͒
Ni K␣
Ni K␤
Zn K␣
10.9
2.0
21.0
Sample II
Ni K␣
Ni K␤
Zn K␣
30.5
5.1
2.1
a Target ϭ Rh; crystal ϭ LiF.
Downloaded on 2014-07-01 to IP 128.252.67.66 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C588
Journal of The Electrochemical Society, 150 ͑9͒ C585-C590 ͑2003͒
Figure 5. SEM photographs of the surface of sample II. See the captions of
Fig. 3.
Figure 6. SEM photographs of the surface of sample II in high magnifica-
tions. See the captions of Fig. 3.
Discussion on the characteristics of porous layer.—The results
obtained so far indicate that a single layer of ␥-NiZn alloy is formed
by the cathodic treatment at 0.22 V and subsequently a porous layer
of ␣-NiZn alloy is formed by the anodic treatment at 0.50 V. Thus,
the two-layer ͑␣- and ␥-NiZn͒ model proposed in a previous report18
is inadequate at least under stationary conditions. Now, processes of
alloying and dealloying, and characteristics of the porous layer are
discussed with the single layer model. Since the rest potential (E
ϭ 0.27 V) observed after the cathodic treatment is regarded as the
equilibrium potential of the ␥-NiZn alloy, its composition can be
estimated from Nernst equation
According to Faraday’s law, the experimental data of Qc
ϭ 35 C/cm2 gives the thickness of 18.5 ␮m of the ␥-NiZn alloy.
With this model, however, the percentage of dissolved zinc calcu-
lated from the compositions of ␥- and ␣-NiZn alloys does not agree
with that obtained from the quantities of electricity in cathodic and
anodic treatments. Thus, on the one hand, the percentage of dis-
solved zinc, A, is expressed in terms of the mole ratios of zinc to
nickel in ␥- and ␣-alloys as
E ϭ E0 Ϫ RT/2F͒ln a
E0 ϭ 0.23 V
͓1͔
XZn ␥͒/X ␥͒ Ϫ X ␣͒/X ␣͒
͑
͑
Ni
͑
Zn
͑
Ni
͑
Zn
A ϭ
ϫ 100 %͒
͓2͔
͑
XZn ␥͒/X ␥͒
͑
͑
Ni
Using the zinc activity data of the ␥-NiZn alloy,2 the mole fraction
of zinc is obtained as XZn(␥) ϭ 0.81: therefore the mole fraction of
nickel is XNi(␥) ϭ 0.19. Similarly, mole fractions of Zn and Ni in
the ␣-NiZn alloy after the anodic treatment at 0.50 V are estimated
to be XZn(␣) ϭ 0.12 and XNi(␣) ϭ 0.88. Table IV summarizes
compositions of the ␥- and ␣-NiZn alloys and related quantities.
Introducing the above-mentioned values yields A ϭ 96.8%. On the
other hand, the percentage of dissolved zinc should be expressed as
Qa
A ϭ
ϫ 100 %͒
͓3͔
͑
Qc
Downloaded on 2014-07-01 to IP 128.252.67.66 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, 150 ͑9͒ C585-C590 ͑2003͒
C589
Figure 9. BEI of the cross section of sample II and Ni K␣ and Zn K␣
signals as functions of depth in EPMA.
Figure 7. SEM photograph of the cross section of sample II. See the caption
of Fig. 3.
where p is the porosity, WNi(␥) and WNi(␣) are the weight fractions
of nickel in ␥- and ␣-NiZn alloys, respectively, and d(␥) and d(␣)
are the corresponding crystallographic densities. Equation 4 gives
Introducing Qc ϭ 35 C/cm2 and Qa ϭ 26 C/cm2 yields A ϭ 74%.
In order to clarify the reason for such disagreement, experimental
data are needed on the composition of alloys formed and on the
current efficiencies for cathodic and anodic processes.
p ϭ 1 Ϫ WNi ␥͒d ␥͒/W ␥͒d ␣͒
͓5͔
͑
͑
͑
Ni
͑
Although there are some uncertainties, it seems worth doing to
estimate the porosity and pore size in the porous layer. Assuming
that the thickness of the surface alloy layer does not change during
dealloying, simple mass balance calculation yields the porosity of
the ␣-NiZn alloy layer. Thus the nickel content in a unit volume of
the original ␥-NiZn alloy layer must be equal to that in the unit
apparent volume of the porous ␣-NiZn alloy layer. Therefore,
Introducing the values of Table IV, the porosity is calculated to be
p ϭ 0.82, that is, 82%. This calculation seems fair judged by the
SEM images in Fig. 6.
Assuming a simple model of cylindrical pores of diameter D and
length L and perpendicular to the geometric surface of the electrode,
the pore size can be estimated from the porosity and the roughness
factor. If the number of pores per unit geometric area is n, then the
porosity p and the roughness factor R are expressed as
WNi ␥͒d ␥͒ ϭ 1 Ϫ p͒W ␣͒d ␣͒
͓4͔
͑
͑
͑
͑
Ni
͑
p ϭ ␲ D/2͒2n
͓6͔
͓7͔
͑
R ϭ ␲DLn
Eliminating n from Eq. 6 and 7
D ϭ 4Lp/R
͓8͔
Introducing L ϭ 18.5 ␮m, p ϭ 0.82, and R ϭ 240, the diameter of
pores is calculated to be D ϭ 0.25 ␮m. Although actual pores are
not cylindrical, this estimation seems reasonable in comparison with
the SEM photographs of Fig. 6.
Table III. Double layer capacitance Cd and estimated roughness
factor R.
Cd
Specimen
͑␮F/cm2͒
Roughness factor, R
Ni ͑not electropolished͒
Ni ͑electropolished͒
Sample II
30
39
1
1.3
240
Figure 8. BEI of the cross section of sample I and Ni K␣ and Zn K␣ signals
as functions of depth in EPMA.
7.3 ϫ 103
Downloaded on 2014-07-01 to IP 128.252.67.66 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C590
Journal of The Electrochemical Society, 150 ͑9͒ C585-C590 ͑2003͒
Acknowledgment
Table IV. Compositions and densities of ␥- and ␣-NiZn alloys.
We would like to thank Professor Y. Awakura, T. Hirato, Dr. K.
Murase, and T. Tanaka of Kyoto University for conducting XRF
analysis, and T. Unesaki of Kyoto University for taking SEM pho-
tographs. EPMA measurements were performed by Mr. N. Tsumaki
of Shimadzu Techno-Research, Inc.
Alloy composition
Crystallographic
densityb
d
Mole fraction
of nickel,a
Weight fraction
Alloy
phase
XNi
of nickel WNi
͑g/cm3͒
␥-NiZn
␣-NiZn
0.19c
0.88d
0.174
0.868
7.78
8.66
Kyoto University assisted in meeting the publication costs of this article.
References
a Estimated from the Nernst equation and the Zn activity data.1
b Calculated from alloy compositions and lattice parameters (a0
ϭ 8.931 Å for ␥-NiZn and a0 ϭ 3.574 Å for ␣-NiZn; see Table I͒.
c Equilibrium composition at 0.27 V vs. RE.
1. Y. Okano and A. Katagiri, J. Electrochem. Soc., 144, 1927 ͑1997͒.
2. R. P. Anantatmula and D. B. Masson, Metall. Trans., 5, 605 ͑1974͒.
3. A. Yoshida, H. Takenishi, and A. Katagiri, Abstract 2B10, 31st Symposium on the
Molten Salt Chemistry, The Electrochemical Society of Japan ͑1999͒.
4. F. Hine, Electrode Processes and Electrochemical Engineering, Plenum Press, New
York ͑1985͒.
d Equilibrium composition at 0.50 V vs. RE.
5. Y. Ogata, H. Hori, M. Yasuda, and F. Hine, J. Electrochem. Soc., 135, 76 ͑1988͒.
6. D. Pletcher and F. C. Walsh, Industrial Electrochemistry, 2nd ed., Chapman and
Hall, London ͑1990͒.
Conclusions
7. L. Chen and A. Lasia, J. Electrochem. Soc., 138, 3321 ͑1991͒.
8. A. Rami and A. Lasia, J. Appl. Electrochem., 22, 376 ͑1992͒.
9. J. Balej, J. Divisek, H. Schmitz, and J. Mergel, J. Appl. Electrochem., 22, 705
͑1992͒.
10. J. Balej, J. Divisek, H. Schmitz, and J. Mergel, J. Appl. Electrochem., 22, 711
͑1992͒.
11. S. Rausch and H. Wendt, J. Electrochem. Soc., 143, 2852 ͑1996͒.
12. Th. Borucinsky, S. Rausch, and H. Wendt, J. Appl. Electrochem., 27, 762 ͑1997͒.
13. S. Tanaka, N. Hirose, and T. Tanaki, Denki Kagaku oyobi Kogyo Butsuri Kagaku,
65, 1044 ͑1997͒.
14. S. Tanaka, N. Hirose, and T. Tanaki, J. Electrochem. Soc., 145, 2477 ͑1999͒.
15. Joint Commitee on Powder Diffraction Standards, Powder Diffraction Files, no.
4-0850 ͑Ni͒, and no. 6-0653 (Ni5Zn21).
In order to obtain a high surface area nickel electrode, cathodic
and subsequent anodic treatments of nickel were conducted in a
ZnCl2-NaCl ͑60-40 mol %͒ melt at 450°C, and the following con-
clusions have been derived.
1. The cathodic treatment at 0.22 V vs. Zn in ZnCl2-NaCl(sat)
yields ␥-NiZn alloy, which is dealloyed in the anodic treatment at
0.50 V to give ␣-NiZn alloy.
2. The obtained material exhibits a porous surface with rough-
ness factor of e.g., 240.
3. The thickness L, porosity p, and pore size ͑diameter͒ D of the
porous layer can be estimated from the compositions of relevant
alloys and the quantities of electricity in cathodic and anodic treat-
ments. The values obtained under the conditions of the present study
are: L ϭ 18.5 ␮m; p ϭ 0.82; and D ϭ 0.25 ␮m.
4. The present method may be useful for producing high surface
area electrodes, although further study is necessary on their perfor-
mance in specified applications.
16. Gmelins Handbuch der anorganischen Chemie, Vol. B1, System Nr. 57, Nickel, p.
32, Verlag Chemie, Weinheim ͑1965͒.
¨
17. Diffusion in Solid Metals and Alloys, in Landolt-Bornstein Numerical Data and
Functional Relationships in Science and Technology, New Series, Group III, Vol.
26, Fig. 102, p. 363, Fig. 103, p. 364, Springer-Verlag, Berlin ͑1990͒.
18. A. Katagiri and M. Nakata, Abstract 1465, the Electrochemical Society Meeting
Abstracts, Vol. 2002-1, Philadelphia, Pennsylvania, May 12-17, 2002.
Downloaded on 2014-07-01 to IP 128.252.67.66 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).