Improvement of in-plane alignment for surface oxidized NiO layer on textured Ni substrate by two-step heat-treatment

Article abstract of DOI:10.1016/j.physc.2004.01.106

Epitaxial growth of NiO on a textured Ni substrate as a template for an REBa₂Cu₃O_y coated conductor was investigated. Highly in-plane aligned NiO layers were successfully fabricated using a new process of a two-step heat-treatment for oxidation. In the first-step, a highly in-plane aligned thin NiO layer was formed on a textured Ni substrate under a low driving force of oxidation. Then, in the second-step, a thick NiO layer was grown at a higher rate with maintaining its high in-plane grain alignment, as if the first NiO layer acts as a seed crystal layer. Further, growth rates and microstructures of the NiO layers were studied comparatively in the cases with and without the first layer. It was found that the oxidation rate in the case with the first layer was lower than that without the first layer. The microstructure observation revealed that the NiO without the first layer was poly-crystalline with many grain-boundaries. On the other hand, in the case with the first layer, grain-boundaries of the NiO were hardly observed. Hence, the reason for this difference of the growth rate and the microstructure of the NiO layers were discussed in view of a diffusivity path.

Full text of DOI:10.1016/j.physc.2004.01.106

Physica C 412–414 (2004) 864–870
www.elsevier.com/locate/physc
Improvement of in-plane alignment for surface oxidized
NiO layer on textured Ni substrate by two-step heat-treatment
a,
a
a
Katsuya Hasegawa *, Toru Izumi , Teruo Izumi ,
a
Yuh Shiohara , Toshihiko Maeda
b
a
Superconductivity Research Laboratory, International Superconductivity Technology Center, 1-10-13 Shinonome,
Koto-ku, Tokyo, 135-0062 Japan
b
The Furukawa Electric Co., Ltd., Nikko 321-1493, Japan
Received 29 October 2003; accepted 13 January 2004
Available online 7 May 2004
Abstract
Epitaxial growth of NiO on a textured Ni substrate as a template for an REBa2Cu3Oy coated conductor was
investigated. Highly in-plane aligned NiO layers were successfully fabricated using a new process of a two-step heat-
treatment for oxidation. In the first-step, a highly in-plane aligned thin NiO layer was formed on a textured Ni substrate
under a low driving force of oxidation. Then, in the second-step, a thick NiO layer was grown at a higher rate with
maintaining its high in-plane grain alignment, as if the first NiO layer acts as a seed crystal layer. Further, growth rates
and microstructures of the NiO layers were studied comparatively in the cases with and without the first layer. It was
found that the oxidation rate in the case with the first layer was lower than that without the first layer. The micro-
structure observation revealed that the NiO without the first layer was poly-crystalline with many grain-boundaries. On
the other hand, in the case with the first layer, grain-boundaries of the NiO were hardly observed. Hence, the reason for
this difference of the growth rate and the microstructure of the NiO layers were discussed in view of a diffusivity path.
ꢀ 2004 Elsevier B.V. All rights reserved.
PACS: 74.72.Bk; 74.76.Bz
Keywords: Surface oxidation; NiO; In-plane crystal alignment
1. Introduction
oxidized NiO layer on a textured Ni by the so-
called SOE (surface oxidation epitaxy) method is
one of the promising candidates because of its
simple process which is suitable for low-cost fab-
rication [1]. However, deterioration of the in-plane
alignment of the NiO layer was inevitable, i.e., in-
plane FWHM (full width at half maximum) values
evaluated from /-scan measurement of X-ray
diffraction (XRD) for the NiO layers were worse
than those of the textured Ni substrate. Indeed, the
In-plane grain crystal alignments of RE-
Ba2Cu3Oy (RE123: RE ¼ Y, Sm, Nd. . .) on metal
substrates for coated conductors are introduced by
textured buffer layers or substrates. A surface
* Corresponding author. Tel.: +81-3-3536-5711; fax: +81-3-
3536-5714.
E-mail address: khasegawa@istec.or.jp (K. Hasegawa).
0921-4534/$ - see front matter ꢀ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.physc.2004.01.106

K. Hasegawa et al. / Physica C 412–414 (2004) 864–870
865
deterioration of the NiO prepared by the conven-
tional SOE (cSOE) method was confirmed in our
previous study [2]. In consequence, degrees of in-
plane grain alignment of thereon BaZrO3 (BZO)
buffer and SmBa2Cu3Oy (Sm123) films were lim-
ited, whose in-plane FWHM values ranged be-
tween 12ꢁ and 28ꢁ which are too large to realize
higher Jc values than 1 · 106 A/cm2.
In this study, improvement of in-plane grain
alignment for the surface oxidized NiO layer as a
template of an RE123 coated conductor was
investigated. A highly in-plane aligned NiO layer
was realized by this modified SOE process
(mSOE). Applicability of the newly developed,
mSOE-grown NiO layer was confirmed by high-Jc
Sm123 coated conductors [3].
them employed the conditions of high-tempera-
tures over 1000 ꢁC and high P(O2) near the
ambient pressure, as a heat-treatment for oxida-
tion. Those conditions have an advantage of rel-
atively higher growth rates than 1 lm/min [5,6]. In
view of crystal growth, however, a higher growth
rate due to higher supersaturation occasionally
induces a lower crystallinity such as unfavorable
oriented or non-epitaxial nucleation. In the SOE-
NiO growth, the supersaturation ratio ðrÞ could be
defined as,
r ¼ PðO2Þ=PðO2ÞÃ
ð1Þ
where P(O2)Ã is the equilibrium pressure of O2 at a
certain temperature. P(O2)Ã values for the NiO/Ni
system calculated from thermodynamical data [7]
were as low as 10À10 Pa at 750 ꢁC. Therefore,
oxidation of Ni may proceed under the much
lower P(O2) atmosphere than the above-men-
tioned conditions.
2. Experimental
Fig. 1 shows P(O2) dependences of in-plane
FWHM, degree of (1 0 0) orientation, and the
surface roughness. The P(O2) was ranged from
5 · 10À3 to 50 Pa using a tube furnace under an O2/
Ar mixture gas flow with vacuum pumping. The
annealing temperature and the time were em-
ployed as following two conditions: (a) 750 ꢁC · 10
h, (b) 1100 ꢁC · 1 h. All the data except for 1100 ꢁC
and P(O2) of 50 Pa show the same tendency that
the in-plane alignment was improved as the P(O2)
decreased, accompanying both the higher degrees
of (1 0 0) orientation and a lower average surface
roughness ðRaÞ. In-plane FWHM values were
reached about 7ꢁ in the low P(O2) conditions in
this study, which is close to that of the Ni sub-
strate.
Not only in-plane alignment but also out-of-
plane alignment of NiO was improved by main-
taining the low P(O2) atmosphere. Fig. 2 shows
XRD h–2h profiles of the Ni substrate and thereon
NiO: (a) Ni substrate, (b) mSOE-NiO (after the
first-step heat-treatment), (c) mSOE-NiO (after the
second-step one), (d) cSOE-NiO. It should be no-
ted that the vertical axis is a logarithmic scale. The
second-step heat-treatment was 1200 ꢁC · 2 h in
air, and the same heat-treatment was conducted
for cSOE-NiO growth as shown in Fig. 2(d). It was
clearly observed that the NiO layer grown by the
Specimens were cut from a {1 0 0} Æ0 0 1æ cube-
textured Ni tape of 10 mm in width and 70 lm in
thickness. The textured Ni substrate was con-
firmed to be highly out-of-plane and in-plane
aligned from an XRD measurement, which shows
in-plane FWHM values about 7ꢁ in /-scan of
Ni(1 1 1) and a strong (2 0 0) peak and no other
diffractions in the h–2h profiles. A heat-treatment
for oxidation was conducted under the following
conditions: temperature was ranged from 750 to
1200 ꢁC, time from 10 min to 5 h, Oxygen partial
pressure (P(O2)) from 5 · 10À3 to 2 · 104 Pa.
Crystal orientation was characterized by XRD and
the pole-figure measurements. Its surface rough-
ness was measured by using a stylus profiler with a
scan-length of 500 lm. Thicknesses of the NiO
layer were measured directly from the cross-sec-
tional optical microscopy and scanning electron
microscopy (SEM) images. Further, the grain
structure of the NiO layer was investigated after
electrolytically etching with CH3COOH and HF in
H2O to distinguish grain-boundaries [4].
3. Results and discussion
So far, growth conditions of SOE-NiO were
reported from several groups [5]. Almost all of

866
K. Hasegawa et al. / Physica C 412–414 (2004) 864–870
Fig. 2. XRD h–2h profiles of Ni substrate and SOE-grown
NiO: (a) Ni substrate, (b) modified SOE (after the first-step), (c)
modified SOE (after the second-step), (d) conventional SOE.
Note that the vertical axis was a logarithmic scale.
about 98%. In contrast, the NiO layer grown by
the mSOE method shows only (2 0 0) diffraction.
After the second-step of mSOE, intensities of
NiO(2 0 0) became stronger than those after the
first-step because of the increased NiO thickness.
XRD pole figures of the SOE-grown NiO(1 1 1)
were summarized in Fig. 3: (a) mSOE (after the
first-step), (b) mSOE (after the second-step), (c)
cSOE. The mSOE-NiO, both after the first-step
and after the second-step, showed a four-fold
symmetry with sharp peaks. The cSOE-NiO also
showed a four-fold symmetry, but their peaks were
much broader than those of mSOE-NiO.
Fig. 1. Oxygen partial-pressure dependence of in-plane
FWHM, degree of (1 0 0) orientation, and surface roughness.
Annealing temperature and time were as follows: (a) 750 ꢁC ·
10 h, (b) 1100 ꢁC · 1 h.
cSOE method shows several diffractions other
than (2 0 0) plane. The index of the (1 0 0) orien-
Fig. 4 shows thickness dependences of in-plane
FWHM of NiO(1 1 1); solid symbols represent the
P
tation fraction defined as Ið200Þ= IðhklÞ was

K. Hasegawa et al. / Physica C 412–414 (2004) 864–870
867
Fig. 3. XRD pole figures of SOE-grown NiO(1 1 1): (a) modified SOE (after the first-step), (b) modified SOE (after the second-step), (c)
conventional SOE. In-plane FWHM values were 7.2ꢁ, 7.5ꢁ, and 14.0ꢁ, respectively.
of the relationship between in-plane grain align-
ment and its thickness. That is, lower in-plane
FWHM values with almost completely flat trends
were recognized at any annealing temperatures and
growth thicknesses in the case of mSOE-NiO. On
the other hand, in the case of cSOE-NiO, in-plane
FWHM values were worse than those of mSOE-
NiO and increased with increasing its thickness.
Fig. 5 shows thicknesses of the SOE-grown NiO
layers as a function of the annealing time: (a)
cSOE and (b) mSOE method. For the case of
mSOE, the annealing time was those of the sec-
Fig. 4. Thickness dependence of in-plane FWHM of
NiO(1 1 1); solid symbols represent the mSOE-grown NiO lay-
ond-step. All data of the NiO thickness ðxÞ and the
time for oxidation ðtÞ was fitted to the well-known
ers and open symbols for the cSOE-grown NiO layers. (d)
mSOE at 1000 ꢁC, (m) mSOE at 1100 ꢁC, (r) mSOE at 1200
parabolic growth equation, which is defined as a
solution of the differential equation dx=dt ¼ 2kp=t,
ꢁC, ( ) cSOE at 1000 ꢁC.
ꢀ
x2 ¼ kpt þ c
ð2Þ
where kp is a so-called parabolic rate-constant and c
is an integration constant. The physical meaning of
kp was explained by Wagner [8] where the material
mSOE-grown NiO layers and open symbols for the
cSOE-grown NiO layers. Significant difference be-
tween mSOE and cSOE was observed in the trends
Fig. 5. Thicknesses of the SOE-grown NiO layer as a function of the annealing time: (a) conventional SOE, (b) modified SOE. The
growth rate of the NiO by the modified SOE was always lower than that of the conventional SOE method as compared at the same
annealing temperature.

868
K. Hasegawa et al. / Physica C 412–414 (2004) 864–870
is transported by diffusion across a single and
homogeneous layer of the reaction product. The
constant, c, is estimated by the boundary condition
of the thickness x at t ¼ 0. In the case of cSOE,
c ¼ 0 was obtained, since oxidation started form a
bare-Ni surface. While in the case of mSOE, c ¼ 0:7
lm, which is the thickness of the NiO layer by the
first-step annealing, was employed. The values of kp
were calculated by a least-squares approximation
method. Then, x À t curves using the obtained kp
values at different annealing temperatures were also
plotted in Fig. 5. The kp, and hence, the growth rate
of the NiO layer by mSOE was always lower than
that of the cSOE method when compared at the
same annealing temperature.
Fig. 6 shows the Arrhenius plot for kp of the
SOE-grown NiO layers at the temperature range
from 1000 to 1200 ꢁC. The broken-line shows the
experimental results of single-crystalline NiO/Ni
by Graham et al. [9]. In the case of mSOE-NiO,
the kp and the activation energy were both close to
those of Graham et al. While, in the case of cSOE-
NiO, the kp values were larger and the activation
energy was smaller than the others. The calculated
kp;l value by Rhines and Connell [10] using a vol-
ume diffusion coefficient, Dl, of Ni in NiO was also
plotted for comparison. The reason for the differ-
ent growth rates between mSOE and cSOE is
discussed at the later part of this section.
Fig. 6. Arrhenius plot for parabolic rate-constant (kp) of NiO
layers grown by mSOE ( ) and cSOE (4). Temperature ranged
ꢁ
from 1000 to 1200 ꢁC. The values of kp were obtained by a least-
squares method from data of Fig. 5. Broken-line shows the
experimental results of single-crystalline NiO/Ni [9]. (+): cal-
culated kp;l value by Rhines [10] using the volume diffusion
coefficient, Dl, for Ni in NiO was also plotted for comparison.
consists of poly-crystalline grains with many grain-
boundaries. On the other hand, in the case of
mSOE, grain-boundaries of the NiO were hardly
observed even after electrolytical etching. Consid-
ering the facts that the underlying Ni substrate
consists of relatively large grains whose grain-
boundaries were clearly observed by optical
microscopy and that in-plane FWHM values of
mSOE-NiO were the same as the Ni substrate,
mSOE-NiO may also have grain-boundaries due
to misorientation angles between grains, even
though the degree of epitaxy of the NiO on Ni was
ideally proceeded. Possible explanations for
this observation of few grain-boundaries in
Fig. 7 shows surface optical microscopy images
of the NiO layers grown by (a) cSOE and (b)
mSOE methods. The surfaces of the NiO layers
were mechanically polished and etched electrolyt-
ically to distinguish grain-boundaries. It can be
found that the NiO layer without the first layer
Fig. 7. Surface optical microscopy images of NiO layer grown by (a) conventional SOE, (b) modified SOE. The surfaces of the NiO
layer were polished mechanically and etched electrolytically to distinguish grain-boundaries.

K. Hasegawa et al. / Physica C 412–414 (2004) 864–870
869
Table 1
Comparison of density of grain-boundary, kp and apparent diffusivity for the NiO layers grown by cSOE and mSOE methods
Method
Density of grain-
boundary (mÀ1
Density of dislocation-
pipe (mÀ2
)
kp=kp;l (1000 ꢁC)
Dapp=D1
)
cSOE
mSOE
4.7 · 105
4.4 · 104
2.0 · 1014
1.1 · 1013
99
7.9
390
23
Density of grain-boundary was evaluated from surface optical microscopy images and the ratio of kp values were from the data in Fig.
6. The contribution of grain-boundaries to the apparent diffusivity was estimated based on the Hart’s model [12].
mSOE-NiO are as follows: If the mSOE-NiO
consists of continuous domains though small-an-
gle grain boundary are connected strongly each
other, the grain-boundaries are hardly etched
similarly to the single crystals of NiO.
Ni substrate under the low driving force of oxi-
dation. Then, in the second-step, a highly in-plane
aligned thick NiO layer was grown at higher rates.
In addition, it was revealed that the modified
SOE-grown NiO layer shows a single-crystalline
microstructure, which is much different from that
of the conventional SOE-grown NiO layer. Fur-
ther, the correlation between the oxidation rate
and the microstructure was explained in view of
grain-boundary diffusion at least qualitatively.
Finally, the above-mentioned different kp values
and microstructures between mSOE and cSOE
methods were reasonably explained by grain-
boundary diffusions. The apparent diffusivity of Ni
in NiO, DNi, was measured through the oxidation
study of Ni by Fueki and Wagner [11]. In their
proposed model, the value of DNi was evaluated
experimentally using P(O2) dependences of kp. In
this study, the kp values were obtained only at the
fixed P(O2) of 2 · 104 Pa. However, Fueki’s data
show that DNi is proportional to the kp values, at
least in the heat-treatment conditions employed in
this study. Then, in the followings, we assume the
ratio of kp was the same as that of the ratio of DNi.
The contribution of grain-boundaries was esti-
mated based on the Hart’s model [12], in which
grain-boundaries consist of ‘‘dislocation-pipes’’
with the specific distance derived from misoriented
angles. Detailed treatments for a high-diffusivity
path of the NiO will be described elsewhere [13].
Here the experimental data and theoretical values
were summarized in Table 1. The tendency of the
experimental data coincide well with the theoreti-
cal value of the Hart’s model. Thus, the reason for
the difference in the oxidation-rates was correlated
to microstructures at least qualitatively.
Acknowledgements
This work was supported by the New Energy
and Industrial Technology Development Organi-
zation (NEDO) as Collaborative Research and
Development of Fundamental Technologies for
Superconductivity Application.
References
[1] K. Matsumoto, Y. Niiori, I. Hirabayashi, N. Koshizuka,
T. Watanabe, Y. Tanaka, M. Ikeda, in: K. Osamura, I.
Hirabayashi (Eds.), Advances in Superconductivity, vol.
10, Springer-Verlag, Tokyo, 1998, pp. 611–614.
[2] M. Kai, N. Hobara, K. Hasegawa, T. Izumi, S. Asada, Y.
Nakamura, T. Izumi, T. Watanabe, Y. Shiohara, Physica
C 378–381 (2002) 998.
[3] T. Izumi, K. Hasegawa, S. Asada, T. Izumi, Y. Shiohara,
T. Maeda, paper in preparation.
[4] F.N. Rhines, R.G. Connell Jr., M.S. Choi, J. Electoro-
chem. Soc.: Solid-State Sci. Technol. 126 (1979) 1061.
[5] Z. Lockman, A. Berenov, W. Goldacker, R. Nast, B.
deBoer, B. Holzapfel, J.L. MacManus-Driscoll, J. Mater.
Res. 18 (2003) 327.
4. Conclusion
[6] K. Matsumoto, T. Watanabe, T. Tanigawa, T. Maeda,
S.B. Kim, I. Hirabayashi, IEEE Trans. Appl. Supercond.
11 (2001) 3138.
A new process of the two-step heat-treatment
for surface oxidation of textured Ni substrates was
developed. In the first-step, a highly in-plane
aligned thin NiO layer was formed on the textured
[7] O. Kubaschewski, C.B. Alcock, Metallurgical Thermo-
chemistry, fifth ed., revised and enlarged, Pergamon Press,
1979, p. 382.

870
K. Hasegawa et al. / Physica C 412–414 (2004) 864–870
[8] C. Wagner, Z. Physik. Chem. 21B (1933) 25.
at the 1974 TMS-AIME fall meeting, Metallurgical Society
of AIME, 1975, pp. 94–113.
[9] M.J. Graham, D. Caplan, M. Cohen, J. Electoro-
chem. Soc.: Solid-State Sci. Technol. 119 (1972)
1265.
[11] K. Fueki, J.B. Wagner Jr., J. Electorochem. Soc.: Solid-
State Sci. Technol. 112 (1965) 384.
[12] E. Hart, Acta Met. 5 (1957) 597.
[10] F.N. Rhines, R.G. Connell Jr. Stress effects and the
oxidation of metals. in: Proceedings of the Symposium held
[13] K. Hasegawa et al., paper in preparation.