A study of magnetic properties: FexCo1-x alloy nanowire arrays

Article abstract of DOI:10.1016/S0009-2614(03)00809-1

Highly ordered arrays of Fe_xCo_1-x (x=0-1.0, nominally) alloy nanowire were prepared by electrodepositing Fe²⁺ and Co²⁺ into the porous anodic aluminum oxide (PAO) templates. XRD experiments proved that the crystal structure of alloy nanowire was bcc. The magnetic properties of Fe_xCo_1-x nanowire arrays showed strong magnetic anisotropy, whose easy axis was parallel to the nanowires. Suitable annealing temperature caused coercivity and squareness of sample to increase rapidly, which was higher than those of as-prepared samples. The change in coercivity and squareness of all samples was discussed in detail. The magnetic reversal mechanism may be attributed to localized nucleation based on Zeng's theory [J. Appl. Phys. 87 (2000) 4718].

Full text of DOI:10.1016/S0009-2614(03)00809-1

Chemical Physics Letters 374 (2003) 661–666
www.elsevier.com/locate/cplett
A study of magnetic properties: Fe Co alloy
x
1ꢀx
nanowire arrays
a
b
a
Dong-Huan Qin , Yong Peng , Lin Cao , Hu-Lin Li
a,
*
a
Department of Chemistry, Lanzhou University, Lanzhou, Gansu 730000, PR China
b
Joule Physics Laboratory, Institute for Materials Research, The University of Salford, Greater Manchester M5 4WT, UK
Received 19 March 2003; in final form 8 May 2003
Abstract
2þ
Highly ordered arrays of Fe
x
Co1ꢀx (x ¼ 0–1:0, nominally) alloy nanowire were prepared by electrodepositing Fe
and Co into the porous anodic aluminum oxide (PAO) templates. XRD experiments proved that the crystal structure
of alloy nanowire was bcc. The magnetic properties of Fe Co1ꢀx nanowire arrays showed strong magnetic anisotropy,
2
þ
x
whose easy axis was parallel to the nanowires. Suitable annealing temperature caused coercivity and squareness of
sample to increase rapidly, which was higher than those of as-prepared samples. The change in coercivity and
squareness of all samples was discussed in detail. The magnetic reversal mechanism may be attributed to localized
nucleation based on ZengÕs theory [J. Appl. Phys. 87 (2000) 4718].
Ó 2003 Elsevier Science B.V. All rights reserved.
1
. Introduction
purpose of application in ultra-high density per-
pendicular magnetic recording devices [2–4]. Dif-
ferent from conventional magnetic recording
materials, the nanowires are embedded in non-
magnetic media and there are only two stable
opposite magnetization directions due to the high
shape anisotropy of the nanowires. Such arrays
are also referred to as quantized magnetic disks
(QMD) [5]. Chou [6] obtained a recording density
about of 400 Gbits/in [1] in Ti/Au dots with 40-nm
pitch fabricated using nanoimprint lithography
and lift-off.
Perpendicular magnetic recording system is one
of the most promising candidates to realize the
extremely high-density recording (HDR) due to its
superior thermal stability and high track density
[
1]. One such technology to prepare sample with
high perpendicular magnetic anisotropy is based
on template method. There have been consider-
able efforts in the investigation of large area
highly ordered ferromagnetic nanowire for the
PAO has a highly oriented porous structure
with very uniform and nearly parallel pores that
can be organized in an almost precise hexagonal
structure. These merits make it become an ideal
template for preparing metal or semiconductor
*
Corresponding author. Fax: +86-931-891-2582.
E-mail addresses: qindonghuan@sohu.com (D.-H. Qin),
lihl@lzu.edu.cn (H.-L. Li).
0
009-2614/03/$ - see front matter Ó 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0009-2614(03)00809-1

662
D.-H. Qin et al. / Chemical Physics Letters 374 (2003) 661–666
nanowires. To date, semiconductors such as CdS
7] and GaAs [8] arrays have been prepared by
using PAO as a template successfully. For mag-
netic arrays, many early studies were mainly con-
cerned with single metal arrays such as Co [9], Fe
first process was carried out at a constant voltage
of 40 V in 0.2 M oxalic acid solution at 0 °C
for 3 h. Secondly, the oxide film was dissolved in
[
0.2 M H
2
Cr
2
O
4
, 0.4 M H
3
PO
4
at 60 °C, then, the
patterned aluminum substrate was anodized again
like the first process. The preparing process can be
seen in our previous work [16] and other groupÕs
work [17]. The diameter is about 50 nm and pore
[
10], and Ni [11] nanowires. However, recently,
attention has been shifted towards preparing alloy
arrays such as FeCo [12] or FeNi [13] and under-
standing of magnetization process. It is easy to
change the coercivity and squareness of recording
media to reach the requirements of real device by
adjusting the component of alloy.
10
ꢀ2
density about 10 cm in the resulting PAO
template prepared by two-step anodization (see
Fig. 1).
The electro-deposition was carried out by AC
voltage with graphite rod as counter-electrode and
PAO template with aluminum plate as working
electrode at room temperature. The electrolyte
solution consisted of 1.2–11.8 g/l CoSO Á 7H O,
In contrast to Fe, Co or Ni, Co
x
Fe1ꢀx alloy has
high saturation magnetization, low magnetic
crystalline anisotropy K (the fist magnetic crys-
1
talline anisotropy constant) and high Curie tem-
perature, which is more suitable for high
temperature applications. In the range of
4
2
1.2–11.8 g/l FeSO
4
Á 7H
2
O, 30 g/l boric acid and 1.5
ascorbic acid. The voltage and the frequency of
sine wave used in electro-deposition were 11.5 V
and 200 Hz, respectively. We found that the elec-
0
:3 < x < 0:7, Fe Co1ꢀx alloys in bulk material or
x
thin film undergo a phase transformation from a
disordered bcc structure to the ordered CsCl
structure below 730 °C. The ordered alloy shows
excellent soft magnetic properties with negligible
magnetocrystalline anisotropy K1 [14]. So the
2þ
tro-deposition velocity of Fe is almost the same
as that of Co during AC electrodeposition pro-
2þ
2þ
cess and slower than that of Co in the case of
DC electro-deposition. Based on these values, the
magnetic properties of Fe
are mainly predominated by shape anisotropy of
the nanowires. The shape anisotropy of the mag-
netic nanowires can be estimated by the simplest
x
Co1ꢀx nanowire arrays
component of Fe
x
Co1ꢀx alloy nanowires can be
easily controlled by adjusting the relative concen-
2
þ
2þ
tration of Co and Fe in electrolyte.
2
s
Stoner–Wohlfarth model [15] K
the corresponding anisotropy field H
s
¼ ðl =4ÞM , and
2.1. Instruments
0
c
ꢁ H ¼
A
2
K =l M ¼ l M =2, which shows that H is linear
Transmission electron microscopy (TEM) and
scanning electron microscopy (SEM) were used to
s
0
s
0
s
c
with M . Therefore, if we want to improve the
s
coercivity of materials, the good way is to improve
of materials or use materials with large M
M
s
s
and
weak crystalline anisotropy, such as Fe
x
Co1ꢀx
alloy. In this Letter, we report our work on fab-
rication and annealing effects on magnetic char-
acterization of Fe
x
Co1ꢀx nanowire arrays. Further,
localized nucleation model is adopted to explain
the reversal process in alloy arrays.
2
. Experimental
Highly ordered PAO templates with pore
diameter of about 50nm were prepared by anodic
oxidation of 99.999% pure Al sheet in oxalic acid
solution under two-step anodizing process. The
Fig. 1. AFM image of PAO template.

D.-H. Qin et al. / Chemical Physics Letters 374 (2003) 661–666
663
characterize the morphology of nanowire arrays,
while atomic absorbed spectrum device, X-ray
diffraction (XRD, Rigaku, Model D/max 2400; Cu
ꢀ
K
a
radiation, k ¼ 1:5418 A) was used to investi-
gate the content and crystalline structure of the
samples, respectively. Magnetic properties of the
as-prepared and annealing samples were tested by
a vibrating sample magnetometer.
To obtain morphology structure of alloy arrays,
the sample was dipped into a solution of 3.5 vol%
3 4 3
H PO and 45 g/l CrO for different times. For
SEM images it was 30 min, while for TEM images
it took about 3 h to dissolve the alumina matrix
completely and washed by distilled water several
times.
3
. Results and discussion
Fig. 2a shows the SEM images of arrays of
Fe0:3Co0:7 nanowires in PAO template. One can
see that a high degree of pore filling rate (up to
0
.9) is obtained from the electro-deposition. TEM
image (Fig. 2b) also shows that the nanowires
prepared by PAO template synthesis are of regu-
lar size and are continuous and parallel to each
other. The diameter is about 50 nm while length is
up to 1 lm, so the aspect ratio (length to diame-
ter) is up to 50. Hence, the aspect ratio of which is
Fig. 2. (a) SEM image of Fe0:4Co0:6 nanowire arrays. The
bright dots are the filled pores. EDAS of the surface of the
samples shows 40 at.%; 60 at.% Fe; (b) TEM image of 50 nm
FeCo alloy nanowires arrays with PAO templates removed.
about 80, the shape anisotropy of Fe
x
Co1ꢀx
nanowire arrays is very high in our case. Selected-
area electron diffraction (SAED) reveals that
Fe
talline structure.
x
Co1ꢀx alloy arrays are composed of polycrys-
The X-ray diffraction (XRD) spectra of the
Fe
arrays are shown in Fig. 3. From this figure, the
x
Co1ꢀx (with x range from 0.1 to 0.9) nanowire
main diffraction peaks can be assigned to FeCo
(
1 1 0), FeCo (2 1 1), the same as that of FeCo thin
film, and the other diffraction peaks of FeCo alloy
cannot be found in most of our spectra. The other
peaks (2H < 40) can be assigned to the peaks of
alumina.
Fig. 4 shows dependence of the coercivity H
and squareness (M =M ) on the Fe composition of
c
r
s
the alloy arrays before annealing. It can be seen
that with an increase of x (from 0 to 0.2), the co-
ercivity (the applied field parallel to nanowire ar-
Fig. 3. X-ray spectra of Fe
ent Fe components.
x
Co1ꢀx nanowire arrays with differ-

664
D.-H. Qin et al. / Chemical Physics Letters 374 (2003) 661–666
ropy along nanowire arrays. So coercivity in
Fe Co1ꢀx arrays drops down in the case of high Fe
content.
In order to study the annealing effects on
Co1ꢀx alloy nanowire arrays, all samples are
annealed at different temperatures 300, 400, 450,
00, 550 and 600 °C, respectively, with low vacuum
x
Fe
x
5
2
pressure of 10 Pa in Ar atmosphere. It is found
that the coercivities of all samples have increased
after annealing (Fig. 5a). The coercivity increases
linearly when annealing temperature ranges from
300 to 500 °C, however, too high annealing tem-
perature (above 550 °C) will deteriorate the mag-
netic properties of the samples. At low annealing
temperature and low Fe content, the M =M of all
Fig. 4. Coercivity (H
deposited samples: (//) external field parallel to nanowire axis;
?) external field perpendicular to nanowire axis.
c
) and squareness vs. Fe component in as-
(
r
s
samples (Fig. 5b) does not change a lot and
maintain at a very high level (at least 0.75), which
shows that the samples could keep their good
perpendicular recording properties. When Fe
content increases, the squareness drops down
rapidly to a very low value of 0.6. The samples
rays) firstly increases from 1200 to 2300 Oe, then
drops down almost linearly with increasing x (from
0
.2 to 0.9). Finally it drops down to 1600 Oe in the
case of pure Fe. However, the coercivity is less
than 600 Oe and squareness is less than 0.2 when
the applied field is perpendicular to the nanowire
arrays, which shows that such Fe
have high perpendicular anisotropy with their easy
x
Co1ꢀx arrays
axis being parallel to nanowire arrays. The change
of the coercivity H with the Fe content is closely
related to the variation of microstructure factors.
c
2þ
2þ
Due to the rapid deposition of Fe and Co into
the PAO template, intrinsic stress and defects are
high and a disordered bcc structure are found in
the samples. Disordered bcc structure and small
grain size in Fe
large negative K
x
Co1ꢀx alloy exhibits a relatively
, causes a high H due to the
magneto-crystallize anisotropy from the disor-
1
c
dered sample. However, the value of K in these
1
samples is still much lower than that of their shape
anisotropy, so that the coercivity is mainly domi-
nated by the shape anisotropy (K
in Fe Co1ꢀx alloy, large coercivity is expected in
FeCo alloy nanowire. The drop down of H when x
s
). With high M
s
x
c
is up to 0.2 can be explained by high stress tension
in the as-prepared sample and the decrease of Ms
in the material. In addition, Co tends to form hcp
2þ
structure when the concentration of Co is low
during electro-deposition, which had been con-
formed by Scarani [18] who prepared CoCu alloy
nanowire arrays. The hcp Co has its easy axis
perpendicular to nanowires and decreases anisot-
Fig. 5. (a) Coercivity vs. annealing temperature; (b) squareness
vs. annealing temperature, whose applied field is 16 kOe.

D.-H. Qin et al. / Chemical Physics Letters 374 (2003) 661–666
665
with different Fe contents annealing at high tem-
perature (about 600 °C) get the lowest squareness
Recently, Skomski et al. [22] pointed out that the
reversal process in thin magnetic nanowire with
small diameter should not be simply considered as
coherent rotation of Stoner–Wohlfarth and local-
ized reversal process was proposed due to the
polycrystalline and imperfection of the wire. Based
on H. ZengÕs research [10], the corresponding co-
ercivity could be expressed as
(
below 0.7) and coercivity. In addition, we have
found that coercivity as high as 3000 Oe and
squareness of about 0.91 can be obtained in
Fe0:3Co0:7 annealing at 550 °C compared to 2200
Oe for a non-annealed sample. We propose that
the change of magnetic properties of Fe Co
x
1ꢀx
after annealing is related to microstructure change
during annealing process. Firstly, thermal anneal-
ing relieves the internal stress in the samples in-
2
2
H
C
¼ 2K
0
=l M
s
ꢀ a DK =ð2Al M
s
Þ
0
0
2
2
¼ l M =2 ꢀ a DK =ð2Al M Þ;
0
s
0
s
2þ
2þ
duced by rapid deposition of Fe and Co and a
high degree of crystallinity is obtained. So the M
of annealed samples is higher than its as-deposited
state. As the M increases in Fe Co alloy arrays
in which A denotes the exchange stiffness, K is
0
s
effective uniaxial anisotropy, a determines the de-
fectÕs volume and DK is inhomogenity accompa-
nied by an easy-axis misalignment. From this
formula one can see that high H
at high M . H value is about 20–30% of H
¼ 2K
isotropy field, H =M ), as Sellmyer pointed
out. In the case of Fe Co , M is about 1.5–2.0 T,
s
x
1ꢀx
after annealing at some temperature, high H is
c
c
can be obtained
expected. Secondly, there is a large mismatch be-
tween the thermal expansion coefficients (a) of
s
c
A
(an-
Fe
1
x
Co1ꢀx alloy and alumina. The a (about
A
U
s
ꢀ
6
ꢀ1
4.0 ꢂ 10 K ) of FeCo alloy is much higher
x
1ꢀx
s
ꢀ
6
ꢀ1
so the H is estimated to be about 1800–3800 Oe,
than that of PAO (about 6.0 ꢂ 10 K ). So, it is
not hard to imagine that FeCo alloy tends to ex-
pand along the wire axis during annealing and
form column structure with easy axis along
nanowire arrays, which will improve shape an-
isotropy at local region. On the other hand, some
of the Fe or Co atoms does not form fcc FeCo
alloy during annealing, which may act as defects
and prevent the movement of domain wall, then
the coercivity will be improved in the samples.
Certainly, there are also microstructure changes
c
which is well confirmed by our experimental re-
sults. From this point, we suggest that localized
reversal model is very appropriate to explain the
reversal process in our Fe
To summarize, we have investigated the mag-
netic properties of Fe Co1ꢀx nanowire arrays in-
x
Co1ꢀx nanowires.
x
fluenced on Fe component and annealing. The
values of coercivity are found to increase roughly
with Fe component and decrease slowly at higher
Fe component peak position x ¼ 0:2. Annealing
effects cause the coercivity increase of Fe
x
Co1ꢀx
such as the change of K (the first magneto-crys-
1
nanowires and is found not to deteriorate their
squareness when the annealing temperature was
low. Microstructure change during annealing
process is proposed to explain the magnetic
properties change of samples. In addition, The
model of localized reversal model is proved to be
suitable to describe our nanowires.
talline anisotropy) and phase segregation due to
Fe/Co short-range order, which had been found in
FeCo thin film [19]. While annealing at high tem-
perature (up to 500 °C), internal stress will increase
and the alumina distort. The pore of PAO tem-
plate will deviate from its original place and the
shape anisotropy will drop down. Furthermore,
unavoidably, Fe will react with O existed in an-
2
odic aluminum oxide at high temperature as the Fe
content is high in the samples, which will deterio-
Acknowledgements
s
rate M and decrease perpendicular anisotropy in
the alloy arrays.
This work is supported by the National Natural
Science Foundation of China (No. 60171004). The
authors would like to express their appreciation to
Mr. Nigel Mellors for revising this paper more
carefully.
As for the reversal process in thin magnetic
nanowire arrays, many early work [20,21] involved
competition between shape anisotropy and mag-
neto-crystalline anisotropy in nanowire arrays.

6
66
D.-H. Qin et al. / Chemical Physics Letters 374 (2003) 661–666
[11] K. Nielsch, R.B. Wehrspohn, J. Barthel, Appl. Phys. Lett.
References
7
12] H.R. Khan, K. Petrikowski, J. Magn. Magn. Mater. 215
9 (2001) 1360.
[
[
[
[
[
[
[
1] S. Yamamoto, Y. Nakamura, S. Iwasaki, IEEE Trans.
Magn. 23 (1987) 2070.
(
2000) 526.
13] H.R. Khan, K. Petrikowski, Mater. Sci. Eng. C 19 (2002)
45.
2] S. Manalis, K. Babcock, J. Massie, V. Elings, M. Dugas,
Appl. Phys. Lett. 66 (1995) 2585.
3
[
[
14] R.M. Bozorth, Ferromagnetism. IEEE, New York, 1991.
15] E.C. Stoner, E.P. Wohlfarth, Philos. Trans. R. Soc. A 240
3] R.L. White, R.M.H. New, R.F.W. Pease, IEEE Trans.
Magn. 33 (1997) 990.
(
1948) 599.
4] M. Sun, Giovanni, M. Shamsuzzoha, R.M. Metzger,
Appl. Phys. Lett. 78 (2001) 2964.
[
[
[
16] D.H. Qin, M. Lu, Chem. Phys. Lett. 51 (2001) 350.
17] H. Masuda, K. Fukuda, Science 268 (1995) 1466.
18] V. Scarani, B. Doundin, J. Philippe, J. Magn. Magn.
Mater. 205 (1999) 241.
5] S.Y. Chou, M.S. Wei, P.R. Krauss, P.B. Fischer, J. Appl.
Phys. 76 (1994) 6673.
[
[
6] S.Y. Chou, IEEE Trans. Magn. 85 (1997) 652.
7] D. Routketvitch, A.A. Tager, et al., IEEE Trans. Electron.
Dev. 43 (1996) 1646.
[
[
[
[
19] R.H. Yu, S. Basu, L. Ren, et al., IEEE Trans. Magn. 36
(
2000) 3388.
20] A. Fert, L. Piraux, J. Magn. Magn. Mater. 200 (1999)
38.
21] G.T.A. Huysmans, J.C. Lodder, J. Appl. Phys. 64 (1988)
016.
[
[
8] X. Mei, D. Kim, H.E. Ruda, Appl. Phys. Lett. 81 (2002)
3
3
61.
9] Y. Peng, H.L. Zhang, S.L. Pan, H.L. Li, J. Appl. Phys. 87
2000) 7405.
2
(
22] R. Skomski, H. Zeng, D.J. Sellmyer, J. Magn. Magn.
Mater. 249 (2002) 175.
[
10] H. Zeng, M. Zheng, R. Skomski, D.J. Sellmyer, J. Appl.
Phys. 87 (2000) 4718.