Textured Co nanowire arrays with controlled magnetization direction

Article abstract of DOI:10.1016/j.jmmm.2005.01.010

Textured HCP Co nanowires with preferentially oriented (1 0 0) along the growth direction were fabricated by electrodeposition at high potential. Further increase of deposition potential results in the formation of twist wires with circumferential strain.

Full text of DOI:10.1016/j.jmmm.2005.01.010

ARTICLE IN PRESS
Journal of Magnetism and Magnetic Materials 295 (2005) 257–262
www.elsevier.com/locate/jmmm
Textured Co nanowire arrays withcontrolled
magnetization direction
Ã
H.N. Hua, , H.Y. Chena, S.Y. Yua, J.L. Chena, G.H. Wua, F.B. Mengb, J.P. Qub,
c
Y.X. Lib, H. Zh u, John Q. Xiaoc
aBeijing National Laboratory for Condensed Matter Physics, State Key Laboratory for Magnetism, Institute of Physics,
Chinese Academy of Sciences, Beijing 100080, China
bSchool of Material Sciences and Engineering, Hebei University of Technology, Tianjin 300130, People’s Republic of China
cDepartment of Physics and Astronomy, University of Delaware, Newark, DE 19716, USA
Received 11 November 2004; received in revised form 21 December 2004
Available online 5 February 2005
Abstract
Textured HCP Co nanowires withpreferentially oriented (1 0 0) along the growthdirection were fabricated by
electrodeposition at high potential. Further increase of deposition potential results in the formation of twist wires with
circumferential strain. These conclusions are unambiguously reached by combining TEM, XRD rocking curves and
pole figures. The magnetic properties are determined by the combination of magnetocrystalline, shape, and stress
anisotropies and magnetostatic interaction. Consequently, the magnetic easy axis can be tuned with structure and
temperature, thus paving the road for magnetic nanowire array use in applications where self-biasing of magnetization
is necessary.
r 2005 Elsevier B.V. All rights reserved.
PACS: 75.75; 81.07; 81.07.B
Keywords: Nanowire electrodeposition texture; Porous anodic aluminum oxide
1. Introduction
cording media and nanosensors [1–3]. More
recently, exciting developments in magnetic do-
main switching by spin-polarized currents seem to
promise breakthroughs in magnetic memory tech-
nology [4,5]. A spin-polarized current of a high
current density can be easily achieved in one-
dimensional structure. Precise structural control of
these one-dimensional structures allows us to
understand the mechanism of current-driven spin
One-dimensional nanostructures are of great
interest because of potential applications in areas
such as high-density perpendicular magnetic re-
Ã
Corresponding author. Tel.: +86 1 082 649 247;
fax: +86 1 082 649 485.
E-mail address: hnhu@aphy.iphy.ac.cn (H.N. Hu).
0304-8853/$ - see front matter r 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmmm.2005.01.010

ARTICLE IN PRESS
258
H.N. Hu et al. / Journal of Magnetism and Magnetic Materials 295 (2005) 257–262
dynamics. The synthesis and precise control of
sucha magnetic nanostructure on a large scale is a
challenging issue in materials science. One strategy
is to electrodeposit magnetic nanowires into
nanochannels of porous anodic aluminum oxide
(AAO) templates, which have been utilized by
many groups to prepare magnetic metals including
Ni, Co, and Fe [1–3,6–13]. In the case of Co
nanowires, almost all electrodeposited Co nano-
wires are polycrystalline. In this letter we report
the preparation of (1 0 0) oriented Co nanowires
withhexagon close packed (HCP) structure in a
controllable fashion. The structure has been
verified withconventional X-ray diffraction
(XRD), scanning electron microscopy (SEM) and
transmission electron microscopy (TEM). Detailed
X-ray rocking curve and pole figure measurements
reveal the formation of twisted HCP structure by
controlling the fabrication parameters. The cir-
cumferential tension in twisted HCP structures
significantly influences the magnetic anisotropy
and can be used to control the magnetization
direction in Co nanowire array.
figure measurements were carried out in a Philips
X’Pert MPD instrument using the Cu Ka radia-
tion. The magnetic measurements were performed
using a superconducting quantum interference
device (SQUID) magnetometer.
3. Results and discussion
Fig. 1 shows the SEM and TEM results on the
Co nanowire arrays and an isolated nanowire. The
lengthof Co nanowires is about 40 mm, consistent
with AAO template thickness. The wires show
uniform diameter of about 60 nm, resulting in a
high aspect ratio of about 700. Selected-
area electron diffraction (SAED) pattern has
been performed on different Co nanowires and
different positions on individual wires. All
2. Experiment procedure
Porous anodic aluminum oxide (AAO) template
was prepared by a two-step anodizing process on
aluminum foils [14,15]. Al foil withpurity of
99.999% was anodized in 0.3 M under a constant
voltage of 40 V at 12 1C. After anodization, the
remaining aluminium substrate and its barrier
layer at the bottom of the AAO template were
removed to obtain the through holes AAO
template withpore lengthof 40 mm and uniform
diameters of 60 nm. A Cu layer was sputter
deposited on one side of the AAO template to
serve as the working electrode in a two-electrode
electrochemical cell. An aqueous bath (pH ꢀ 3:0)
containing 0.2 M Co2+ and H3BO3 (30 g/l) was
used to prepare Co nanowire at room temperature
withconstant potential of ꢁ2 V or higher. Co-1
and Co-2 samples were deposited from the same
electrolyte withpotential of ꢁ1.95 and ꢁ2.15 V,
respectively. SEM and TEM were used to char-
acterize the morphology and the structure of the
nanowires. The X-ray rocking curve and pole
Fig. 1. (a) SEM images of cross-section view of Co nanowire
arrays after removing the AAO template; (b) typical bright field
TEM image of a single-crystal Co nanowire and electron
diffraction from the same sample (inset) showing the HCP
structure with c-axis aligned perpendicular to the wire.

ARTICLE IN PRESS
H.N. Hu et al. / Journal of Magnetism and Magnetic Materials 295 (2005) 257–262
259
obtained diffraction patterns show hexagonal
symmetry (inset in Fig. 1b), indicating that wires
are single crystal HCP structure with c-axis
perpendicular to the wire.
eachindividual nanowire is single crystal and has
[1 0 0] preferred orientation along the wires. The
nanowires array has random c-axis distribution in
the AAO template plane. The sharp (1 0 0) peak in
XRD and the narrow FWHM of rocking curves of
2.381 indicates the excellent pore parallelity in
AAO templates as well as a near-perfect texture in
our nanowire arrays. This clearly indicates that
optimum deposition condition can be achieved to
fabricate well-textured nanowires array over a
large area. It is important to emphasize that the
unusual twin-peak in rocking curves of Co-2
samples cannot be explained by the parallelity
deviation in pore structure, nor can it be explained
with the deviation from the overall orientation. A
poor parallelity only increases the FWHM of the
rocking curves, and orientation deviation increases
the FWHM and shifts the peak position.
To explain the double peaks in rocking curves
for Co-2 samples, we performed pole figure
measurement and (1 0 0) and (1 0 1) pole figures
for Co-1 and Co-2 samples as shown in Fig. 3. Th e
tilt angle (a) was varied from 0–851 in steps of 51.
The azimuth angle (b) was varied from 0–3551 in
steps of 51. Fig. 3(a) shows a well-defined HCP
(1 0 0) sharp peak at a ¼ 01 for Co-1, further
confirming the high degree of (1 0 0) alignment of
Fig. 2 shows XRD spectra of Co nanowire
arrays embedded in templates. All samples show
only a sharp (1 0 0) peak of HCP Co, indicating the
preferential orientation of whole single-crystalline
wires. The inset of Fig. 2 shows the results from
X-ray rocking curve of (1 0 0) reflection. Most of
samples, denoted as Co-1, show one sharp peak
centered at the specular angle of 20.81 withnarrow
peak widths at half-maximum (FWHM) in the
range of 2.38–5.511. Assuming a Gaussian dis-
tribution, these results indicate that at least 75% of
the wires have a preferential orientation within
1.2–2.81 of the [1 0 0] direction. Similar analyses
for polycrystalline samples show a larger distribu-
tion angle up to 91 [8]. Samples electrodeposited at
the potential of ꢁ2.15 V (denoted as Co-2) show
very different rocking curves, indicated in the inset
of Fig. 2. Although they have the same sharp
(1 0 0) XRD peak as Co-1 samples, there is no
peak at 20.81 in rocking curves. Interestingly, there
are two twin-peaks withthe same intensity at 15.7 1
and 26.11, respectively (i.e., about 751 mirror
imaging the [1 0 0] characteristic angle of 20.81).
Combining the results from SEM, TEM, XRD
and X-ray rocking curves, we can conclude that
8000
Co-1
[100]
6000
Co-2
4000
2000
10
20
30
θ (deg.)
0
20
40
60
80
2θ (deg.)
Fig. 2. X-ray diffraction (XRD) spectrum of Co nanowire
arrays embedded in porous alumina template. Inset is the (1 0 0)
rocking curve for two types of nanowire arrays.
Fig. 3. Three-dimensional (1 0 0) (a) and (1 0 1) (b) pole figures
of Co-1 samples and three-dimensional (1 0 0) (c) and (1 0 1) (d)
pole figures of Co-2 samples, respectively.

ARTICLE IN PRESS
260
H.N. Hu et al. / Journal of Magnetism and Magnetic Materials 295 (2005) 257–262
nanowires along the wire direction in Co-1
samples. The HCP (1 0 1) peak in Fig. 3(b) shows
the diffraction maximum as a crateriform ring
located at a ¼ 301; corresponding to the angle
between the neighboring [1 0 0] and [1 0 1] planes.
This means that the h1 0 1i axis of nanowires are
randomly distributed in the array. Based on this
observation, one may easily conclude that the
c-axis of single-crystalline wires also distribute
randomly in the array plane.
For Co-2 samples, the (1 0 0) peak has a volcano
shape located at about a ¼ 51 surrounding the
center of the pole figure (Fig. 3c). Its (1 0 1) pole
figure shows two crateriform rings located at
about a ¼ 25o and 35o, respectively. These shifted
angles are consistent withtwin-peak pattern
observed in (1 0 0) rocking curve (inset of Fig. 2).
If twin peak pattern in rocking curve suggests
lattice structure shown in Fig. 4d, then the pole
figure suggests that such distortion have circular
symmetry. Here, we propose a model of twisted
wire shown in Fig. 4 which will result in this cone
structure, this model is also consistent with
observed magnetic properties to be discussed later.
If the wire is twisted as it grows (Fig. 4b), the
atoms on the second plane will not be directly
above these on the first plane, but shifted along the
twisted direction. Consequently, the distance
between the first and second planes is reduced.
The reduction is most on the outmost radius and
zero along the center line since the twist creates the
largest displacement at the outmost radius. It is
clear the formation of this structure must be
accompanied by dislocations or defects. The cross-
sectional view of sucha distortion is shown in
Fig. 4d. The angle between wire axis and (1 0 0)
orientation for the twisted wire, in our case, should
be about 51. This effect gives rise to the twin peaks
shown in the inset of Fig. 2, and in the pole figure
volcano-like (1 0 0) peak at a ¼ Æ51 in Fig. 3c.
Similar analysis shows that double crateriform
rings should be observed for (1 0 1) orientation
located at a ¼ 30 Æ 51: Experimental pole figure
of (1 0 1) for twisted nanowire array is shown in
Fig. 3d. This is the only structure that explains
consistently the results shown in the rocking
curves and pole figures. This model has never
been proposed before, since no pole figure studies
have been done in nanowire array.
Magnetic properties were measured by
a
SQUID magnetometer from 5 to 300 K. Fig. 5
shows the hysteresis loops of Co-1 (a, b), Co-2
(c, d), and Co-2 withtemplate partially etched
away (e, f) at 5 and 300 K, respectively. The
H// wire
H wire
Co-1
5K
Co-2 dis.
5K
Co-2
5K
(e)
(f)
(c)
(d)
(a)
(b)
0.5
0.0
-0.5
Co-2
300K
Co-1
300K
Co-2 dis.
300K
0.5
0.0
-0.5
-5000 0 5000 -5000 0 5000 -5000 0 5000
H (Oe) H (Oe) H (Oe)
Fig. 4. The twisted nanowire model. The cross-section of both
normal and twisted nanowires are shown below the three-
dimensional diagram. (1 0 0) and (0 0 1) orientations are labeled
in the figure. (See text for detailed explanation).
Fig. 5. Hysteresis loops of sample Co-1 (a, b), Co-2 (c, d) and
Co-2 withtemplate partially etched away (e, f). Field is applied
perpendicular (- Á -) or parallel (ꢁ) to the nanowires.

ARTICLE IN PRESS
H.N. Hu et al. / Journal of Magnetism and Magnetic Materials 295 (2005) 257–262
261
control of anisotropy by structure and tempera-
ture is clear: for untwisted Co-1 wire array, the
easy axis is perpendicular to the wires at low
temperature and gradually decreases withincreas-
ing temperatures, while twisted Co-2 wire array
shows an easy axis along the wires at high
temperatures and decreases withdecreasing tem-
peratures. The total anisotropy represents the
competition among magnetocrystalline, shape,
and stress anisotropies and dipolar interaction
[7,9]. For HCP Co, the magnetocrystalline aniso-
tropy creates an easy magnetization direction
along the c-axis, which is reduced due to the
random c-axis orientations in bothCo-1 and Co-2
samples. The dipolar interaction, which is the same
for bothsamples because of the same diameter and
length, leads to an easy magnetization direction
perpendicular to the nanowire array. If the stress
effect in the Co-1 samples is neglected, then the
sum of magnetocrystalline anisotropy and dipolar
interaction is larger than the shape anisotropy.
This results in an easy magnetization direction
perpendicular to the nanowires. The magnetocrys-
talline anisotropy decreases faster than shape
anisotropy and dipolar interaction withincreasing
temperature, leading to the temperature-depen-
dent behaviors observed in sample Co-1. In sample
Co-2 , in addition to crystalline and shape
anisotropies and dipolar interaction, there is a
stress anisotropy, Eme ¼ 32ls sin2 y; where l is the
magnetostriction coefficient, s the stress, and y
the angle between Ms and s: In Co-2, the positive
circumferential tensile stress s and negative l for
Co along the c-axis gives rise to a negative stress
anisotropy, implying that along the wire direction
is an easy axis [16]. At low temperature, the stress
and shape anisotropy balances the magnetocrys-
talline anisotropy and dipolar interaction, result-
ing in similar magnetic properties in two
directions. For the same reason as in Co-1, at
the room temperature, the magnetization prefers
to be along the wire direction. It should be noted
that the coercivity of Co-2 samples is always larger
than that of Co-1 samples because of this
additional stress anisotropy. To confirm the stress
anisotropy in Co-2 sample, the Al2O3 template was
partially removed to release portion of stress, and
the magnetic properties at 5 and 300 K are shown
in Fig. 4(e,f). As anticipate the hysteresis loops are
between those non-stressed Co-1 nanowire arrays
and stressed (twisted) Co-2 nanowire array (Co-2),
clearly demonstrating that the stress anisotropy
plays a crucial role here.
4. Conclusions
Textured Co nanowire withpreferred growth
orientation of (1 0 0) has been fabricated by
electrodeposition at high potential. Further in-
crease of deposition potential results in the forma-
tion of twist wires. These conclusions are
unambiguously reached by combining, TEM,
XRD rocking curve and pole figure measurements.
The magnetic properties are determined by the
combination of magnetocrystalline, shape, and
strain anisotropies and dipolar interaction. Conse-
quently, the magnetic easy axis can be tuned with
structure and temperature, thus paving the road for
magnetic nanowire array use in applications where
self-biasing of magnetization is necessary.
Acknowledgements
This work was supported by National Natural
Science Foundation of China Grant no. 50371101.
JQX would also like to acknowledge the support
from Grant AFOSR F49620-03-1-0351 and Amer-
ican Chemical Society Petroleum 40057-AC5M.
References
[1] T.M. Whitney, J.S. Jiang, P.C. Searson, C.L. Chien,
Science 261 (1993) 1316.
[2] D. Almawlawi, N. Coombs, M. Moskovits, J. Appl. Phys.
70 (1991) 4421.
[3] T. Thurn-Albrecht, J. Schotter, G.A. Kastle, N. Emley, T.
Shibauchi, L. Krusin-Elbaum, K. Guarini, C.T. Black,
M.T. Tuominen, T.P. Russell, Science 290 (2000) 2126.
[4] M. Klaui, C.A.F. Vaz, J.A.C. Bland, W. Wernsdorfer, G.
Faini, E. Cambril, L.J. Heyderman, APL 83 (2003) 105.
[5] T.Y. Chen, Y. Ji, C.L. Chien, APL 84 (2004) 380.
[6] P.M. Paulus, F. Luis, M. KroK II, G. Schmid, L.J. de
Jong, J. Magn. Magn. Mater. 224 (2001) 180.
[7] R. Ferre0, K. Ounadjela, J.M. George, L. Piraux, S.
Dubois, Phys. Rev. B 56 (1997) 14066.

ARTICLE IN PRESS
262
H.N. Hu et al. / Journal of Magnetism and Magnetic Materials 295 (2005) 257–262
[8] J.-L. Maurice, D. Imhoff, P. Etienne, O. Durand, S.
Dubois, L. Piraux, J.-M. George, P. Galtier, A. Fert,
[12] L. Piraux, S. Dubois, J.L. Duvai, K. Ounadjela, A. Fert,
J. Magn. Magn. Mater. 175 (1997) 127.
J. Magn. Magn. Mater. 184 (1998) 1.
[13] Y. Henry, K. Ounadjela, L. Piraux, S. Dubois, J.-M.
George, J.-L. Duvail, Eur. Phys. J. B 20 (2001) 35.
[14] H. Masuda, K. Fukuda, Science 268 (1995) 1466.
[15] Y.W. Wang, L.D. Zhang, G.W. Meng, X.S. Peng, Y.X.
Jin, J. Zhang, J. Phys. Chem. B 106 (2002) 2502.
[16] B.D. Cullity, Introduction to Magnetic Materials,
Addison-Wesley, MA, 1972, pp. 260–264.
[9] K. Ounadjela, R. Ferre0, L. Louail, J.M. George, J.L.
Maurice, L. Piraux, S. Dubois, J. Appl. Ph ys. 81 (1997) 5455.
[10] H. Zeng, R. Skomski, L. Menon, Y. Liu, S. Bandyopad-
hyay, D.J. Sellmyer, Phys. Rev. B 65 (2002) 134426.
[11] Shihui Ge, Xiao Ma, Chao Li, Wei Li, J. Magn. Magn.
Mater. 226–230 (2001) 1867.