Fabrication and magnetic properties of ultrathin Fe nanowire arrays

Article abstract of DOI:10.1063/1.1621459

The study of the fabrication process of ultrathin Fe nanowires arrays and their magnetic properties was presented. The fabrication procedure was based on the use of anodic porous alumina template for electrodeposition process. The ultrathin nanowires exhi

Full text of DOI:10.1063/1.1621459

Fabrication and magnetic properties of ultrathin Fe nanowire arrays
X. Y. Zhang, G. H. Wen, Y. F. Chan, R. K. Zheng, X. X. Zhang, and N. Wang
Citation: Applied Physics Letters 83, 3341 (2003); doi: 10.1063/1.1621459
View online: http://dx.doi.org/10.1063/1.1621459
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/83/16?ver=pdfcov
Published by the AIP Publishing
This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
128.143.23.241 On: Sun, 10 Nov 2013 19:42:06

APPLIED PHYSICS LETTERS
VOLUME 83, NUMBER 16
20 OCTOBER 2003
Fabrication and magnetic properties of ultrathin Fe nanowire arrays
X. Y. Zhang, G. H. Wen, Y. F. Chan, R. K. Zheng, X. X. Zhang, and N. Wanga)
Department of Physics, Institute of Nano Science and Technology, The Hong Kong University of Science
and Technology, Clear Water Bay, Hong Kong, China
͑
Received 2 June 2003; accepted 29 August 2003͒
Ultrathin Fe nanowire ͑about 5 nm in diameter͒ arrays have been fabricated by electrodeposition
using anodic porous alumina templates. These ultrathin nanowires exhibited uniaxial anisotropy and
a quite large coercivity ͑4190 Oe͒ at 5 K. In addition, the field needed to saturate the magnetization,
when the field was applied perpendicularly to the easy axis, was much larger than the shape
anisotropy field (2␲M ). This saturation field increased with decreasing temperature. We believed
S
that this enhanced saturation field was mainly due to the contribution of the surface spins. © 2003
American Institute of Physics. ͓DOI: 10.1063/1.1621459͔
In recent years, there has been increasing interest in the
fabrication of one-dimensional nanostructures because of
their potential utilization in many fields.1 Magnetic nano-
wire arrays are of great interest for both fundamental science
and potential applications in, such as, high-density perpen-
dicular magnetic recording media and nanosensors.5 The
synthesis and precise control of such a magnetic nanostruc-
ture on a large scale is a challenging issue in materials sci-
ence. One strategy is to electrodeposit the magnetic nanow-
ires into the nanochannels of anodic porous alumina
templates, which have been investigated previously by many
ing the typical microstructure of the porous alumina tem-
plates. The diameters and the separation distances of the
pores are about 5 and 15 nm, respectively. Figure 1͑c͒ is a
typical cross-sectional view of the template filled with Fe
nanowires whose diameters and lengths are about 5 nm and 2
␮m, respectively. Fe nanowires are well aligned indicating
the nanochannels in the substrate are uniformly distributed
through the entire substrates. Since the TEM sample was
sectioned along a certain direction, only some of the nanow-
ires were imaged in this figure. Figure 1͑d͒ shows an isolated
Fe nanowire after completely dissolving the alumina tem-
plate and dispersing the sample on a carbon supporting film.
Most Fe nanowires observed were single crystalline. The
selected-area electron diffraction pattern ͓inset in Fig. 1͑d͔͒
taken from this nanowire was indexed as the ͓110͔ zone axis
of bcc Fe structure. This result was consistent with that ob-
tained from the XRD measurement. Using similar method,
we have also synthesized Ni and Ni–Co alloy nanowires in
the templates. Their structure and morphology were very
–4
–8
6
–13
groups.
The porous alumina substrates were fabricated by anod-
izing pure aluminum foils ͑purity 99.999%͒ in a sulfuric acid
solution. Prior to anodizing, the aluminum foils were an-
nealed at 500 °C for 2 h in order to homogenize the micro-
structures and reduce the density of defects in the foils. This
was essential for the formation of high quality nanochannels
over large areas in the alumina substrates. Subsequently, the
foils were electropolished in the solution of C H OH mixed
2
5
with HClO ͑9:1 in volume͒ in order to obtain a high quality
4
of flat surfaces. Then, anodization was carried out under a
condition of constant cell voltage of 5.5 V in a 3.5 M H SO4
2
solution. The temperature was kept constant at 0 °C. The
preformed alumina layer was removed by phosphoric acid
and chromic acid, and then the aluminum foil with a fresh
and clean surface was anodized again under the same anod-
izing condition. After the anodization, Fe nanowires were
prepared by ac ͑14 V, 50 Hz͒ electrodeposition in the solu-
tion of FeSO •7H O (45 g/l)ϩH BO ͑50 g/l͒ with a pH
4
2
3
3
value of 3.5. The temperature for the electrodeposition was
25 °C.
Figure 1͑a͒ shows the x-ray diffraction ͑XRD͒ spectrum
of the Fe nanowire arrays after completely dissolving the
aluminum substrate in a HgCl saturated solution. The dif-
2
fraction peaks of ͑110͒, ͑200͒, and ͑211͒ are clearly distin-
guishable, and match perfectly with the bcc Fe structure with
a lattice constant of aϭ0.2866 nm. Figure 1͑b͒ is a plan-
view transmission electron microscope ͑TEM͒ image show-
FIG. 1. ͑a͒ XRD result taken from Fe nanowire arrays. ͑b͒ TEM image of
the alumina template with 5 nm diameter pores. ͑c͒ Cross-sectional view of
the template filled with Fe nanowire arrays. ͑d͒ An isolated Fe nanowire
with a diameter of about 5 nm and its selected-area electron diffraction
pattern ͑inset͒.
a͒Author to whom correspondence should be addressed; electronic mail:
phwang@ust.hk
This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
003-6951/2003/83(16)/3341/3/$20.00 3341 © 2003 American Institute of Physics
28.143.23.241 On: Sun, 10 Nov 2013 19:42:06
0
1

3342
Appl. Phys. Lett., Vol. 83, No. 16, 20 October 2003
Zhang et al.
similar to the Fe nanowires illustrated earlier.
To understand the magnetism in the magnetic nanowire
arrays, hysteresis loops have been measured in a wide tem-
perature range from 5 to 350 K. Because of the very larger
length-to-diameter ratio (ϳ200) of Fe nanowires, it is ex-
pected that the effective magnetic anisotropy is dominated by
the shape anisotropy 2␲M , where MS is the saturation
S
magnetization of the materials. This means that the easy axis
of magnetization is along the wire axis. Figure 2͑a͒ shows
the hysteresis loops obtained from Fe nanowire arrays with
the applied external magnetic field perpendicular to the
sample plane ͑parallel to the wire axis͒ and parallel to the
sample plane ͑perpendicular to the wire axis͒ at 300 K. It
should be pointed out that the magnetization data plotted
here are the values extracted by using MϭM -␹H, where
tot
Mtot is the measured value and ␹ is the diamagnetic suscep-
tibility of the templates, a small negative constant. It is
clearly seen from Fig. 2͑a͒ that the magnetic easy axis is
along the wire axis, which can be easily saturated as we
expected. From the hysteresis loop data, the parameters that
describe the hysteresis loop behavior are obtained: HcЌ
ϭ1470 Oe, Hc
ʈ
ϭ530 Oe, MrЌ /MsЌϭ0.88 and Mr
ʈ
/Ms
ʈ
ϭ0.10. Previous studies show that the coercivity of magnetic
nanowires depends strongly on their diameters.1 At 300 K,
1,12
there is a critical diameter (d ) for Fe wires at which the
C
maximum coercivety appears. The fact that coercivity de-
creases with decreasing the diameter of the wires, if the wire
diameter dϽd , is suggested to be due to the thermal
C
1
4
effects. The reported d for Fe nanowires is 13 nm, and the
C
1
2
largest value of coercivity is 2640 Oe. Since the diameters
of our Fe nanowires are much smaller than d , it is expected
C
that the value of coercivity obtained at 300 K is smaller than
that reported by the others.1
1–13
However, the diameter de-
pendence of the coercivity measured at a low temperature of
K shows a very different behavior because no maximum
has been observed, i.e., the coercivity increases monotoni-
FIG. 2. ͑a͒ The hysteresis loops of the Fe nanowires at 300 K. ‘‘Ќ’’ indicates
the magnetic field applied perpendicularly to the sample surface, i.e., paral-
ʈ
lel to the nanowire axes. ‘‘ ’’ denotes the magnetic field parallel to the
sample surface. ͑b͒ The temperature dependence of the coercivity of Fe
5
11,12
cally with decreasing the diameter.
The smallest diameter
nanowires for the field applied perpendicular to the sample plane. The inset
of Fe nanowire arrays reported by Menon et al. is 9 nm.12
These nanowires exhibit a coercive field of 3700 Oe at 5 K.
The diameter of Fe nanowires prepared in this work, which
show a higher coercive field of 4190 Oe at 5 K is about 5
nm, much smaller than 9 nm. From a simple extrapolation of
the data shown in Fig. 3͑b͒ in Ref. 12 ͑the coercivity as a
function of diameter measured at 5 K͒, we find that the co-
ercivity will be roughly about 4200 Oe for the 5 nm Fe
nanowires, indicating that our results are qualitatively in line
is the replot of HC as a function of T1
/2
.
anisotropy constant and volume of the wire; k is the Bolt-
B
9
zmann constant, f (ϳ10 Hz), and ␶ is the attempt fre-
0
quency and relaxation time. In the inset to Fig. 2͑b͒, the
coercivity data are replotted as a function of T1 , where the
line is the linear fit of the data. Obviously, the coercivity has
/2
1
/2
T
dependence, i.e., mϭ2. Actually, mϭ2 is corresponding
with what reported previously.1
1,12
to the case of aligned, noninteracting Stoner–Wohlfarth par-
ticles, where the reversal of magnetization is coherent. The
coherent reversal of magnetic moment in 5 nm nanowires
should be expected.11 Inspection of the high-field magnetiza-
tion data obtained at 300 K reveals that the magnetization
saturates at about 1.2 T for the field parallel to the sample
plane ͑or perpendicular to the wire axis͒. This saturation field
Figure 2͑b͒ shows the temperature dependence of the
coercivity with the field applied perpendicularly to the
sample plane ͑parallel to wire axis͒. It is manifest that HcЌ
decreases with increasing temperature, from 4190 Oe at 5 K
to 1470 Oe at 300 K, indicating a much stronger
temperature-dependent characteristic than that reported in
Ref. 11 for Fe nanowires with large diameters. By consider-
(
H ) is usually very close to the anisotropy field (H ) in
S
A
11
nanowires. But this value is larger than that reported for the
Fe wires with larger diameters and also lager than the theo-
ing the thermal effect in nanowire structure, the coercivity
can be described by11
retical value (2␲M ϭ1.073 T) of the shape anisotropy in
S
1
/m
k T
nanowire structure for Fe at 300 K. Figure 3͑a͒ shows mag-
netization curves for the applied field parallel to the sample
plane ͑or perpendicular to the wire axis͒ obtained at different
B
H ϭHA 1Ϫ
ͫ
ln͑ f ␶͒
ͬ
,
͑1͒
C
ͭ
0
ͮ
K V
u
This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
where H is the anisotropy field, K and V are the uniaxial
temperatures. It is obvious that the saturation field which is
A
u
1
28.143.23.241 On: Sun, 10 Nov 2013 19:42:06

Appl. Phys. Lett., Vol. 83, No. 16, 20 October 2003
Zhang et al.
3343
the surface spin is about two atomic layers, the percentage of
the surface spins will be ϳ20% for the Fe nanowire with a
diameter of 5 nm. Due to the reduction of the coordination
and the interface between Fe wire and the template, the mag-
netic configuration of the surface spins must be different
from that in the bulk, which must affect the magnetic behav-
ior of the wires. Surface-spin effect in nanostructured mag-
netic materials has been well known and studied
intensively.1
5,16
Surface spins are normally very difficult to
be aligned, even in a very larger magnetic field as observed
1
5,16
in magnetic nanoparticles.
In some cases, surface spins
With increasing the diam-
form a spin-glass-like layer.1
5,16
eters of the wires, the surface spin contribution becomes
weaker and weaker; the saturation of magnetization will be
dominated by the shape anisotropy (2␲M ) as observed by
S
Sellmyer et al.11
If the surface spins form a spin-glass-like layer in our
ultrathin Fe wires, it will be in a frozen state at low tempera-
tures and become superparamagnetic ͑or paramagnetic͒
above the freezing temperature. Therefore, it is difficult to
reach the saturation state in the whole temperature range.
The existence of the surface spin contribution can be ob-
served from the low temperature magnetization curves and
the temperature dependent magnetization. Figure 3͑b͒ re-
veals the temperature dependent magnetization measured
with 1 T magnetic field applied perpendicularly to the sample
plane ͑or parallel to the wire axis͒. Clearly, there is a strong
temperature-dependent magnetic contribution at low tem-
peratures superimposing to the weak temperature-dependent
magnetization resulted from the wire cores ͑bcc Fe͒. This can
also be observed from the field-dependent magnetization
measured at 5–50 K ͓inset to Fig. 3͑b͔͒.
This work was supported by grants from the Research
FIG. 3. ͑a͒ Plot of magnetization vs field obtained at different temperatures,
where the field is parallel to the sample plane ͑or perpendicular to the wire
axis͒. The inset shows the estimated saturation field vs temperature and the
saturation field plotted as a function of temperature. ͑b͒ The temperature-
dependent magnetization with 1 T magnetic field applied perpendicularly to
the sample plane ͑or parallel to the wire axis͒. The inset reveals the magne-
tization versus field obtained at 5 and 50 K. It is clear that the surface spins
are much difficult to be aligned at low temperatures.
Grant Council of Hong Kong ͑Project Nos. HKUST6151/
0
1P and 6165/01P͒.
1
2
S. Iijima, Nature ͑London͒ 354, 56 ͑1991͒.
M. H. Dvoret, D. Esteve, and C. Urbina, Nature ͑London͒ 360, 547
͑1992͒.
3
H. Dai, E. W. Wong, Y. Z. Lu, S. Fan, and C. M. Lieber, Nature ͑London͒
3
75, 769 ͑1995͒.
4
5
A. P. Alivisatos, Science ͑Washington, DC, U.S.͒ 271, 933 ͑1996͒.
K. Uosaki, K. Okazaki, H. Kita, and H. Takahashi, Anal. Chem. 62, 652
larger than 2␲MS increases with decreasing temperature.
The estimated saturation field decreases monotonically with
increasing temperature ͓inset to Fig. 3͑a͔͒.
From the strong temperature dependence of coercivity
and the saturation magnetic field ͓inset to Fig. 3͑a͔͒, the heu-
ristic understanding of the observation is that except the con-
͑
1990͒.
6
T. M. Whitney, J. S. Jiang, P. C. Searson, and C. L. Chien, Science ͑Wash-
ington, DC, U.S.͒ 261, 1316 ͑1993͒.
C. Z. Zhang and J. C. Lodder, J. Magn. Magn. Mater. 88, 236 ͑1990͒.
D. Almawlawi, N. Coombs, and M. Moskovits, J. Appl. Phys. 70, 4421
7
8
͑
1991͒.
9
0
tribution of the shape anisotropy (2␲M ) to the saturation
R. Ferr e´ , K. Ounadjela, J. M. George, L. Piraux, and S. Dubois, Phys. Rev.
B 56, 14 066 ͑1997͒.
M. Zheng, L. Menon, H. Zeng, Y. liu, S. Bandyopadhyay, R. D. Kirby, and
D. J. Sellmyer, Phys. Rev. B 62, 12 282 ͑2000͒.
D. J. Sellmyer, M. Zheng, and R. Skomski, J. Phys.: Condens. Matter 13,
R433 ͑2001͒.
L. Menon, M. Zheng, H. Zeng, S. Bandyopadhyay, and D. J. Sellmyer, J.
Electron. Mater. 29, 510 ͑2000͒.
F.-Y. Li, L. Zhang, and R. M. Metzger, Chem. Mater. 10, 2470 ͑1998͒.
B. D. Cullity, Introduction to Magnetic Materials ͑Addison-Wesley, New
York, 1972͒, pp. 386–387.
S
field, there must be another contribution that makes the mag-
netization difficult to be saturated. In the present case, the
diameters of the Fe wires are about 5 nm, much smaller than
the coherent diameter dcohϭ11 nm for Fe.11 Therefore, the
1
11
12
reversal of the magnetization should be coherent if the wire
is free of local imperfections.11 If the reversal of magnetiza-
13
tion is coherent, the saturation field should equal the anisot-
ropy field, which is dominated by the demagnetization field
14
15
R. H. Kodama, S. A. Makhlouf, and A. E. Berkowitz, Phys. Rev. Lett. 79,
1
Foner, ibid. 77, 394 ͑1996͒.
B. Mart ´ı nez, X. Obradors, L. Balcells, A. Rouanet, and C. Monty, Phys.
2
␲M in very long and narrow wires. Here, we observe that
S
393 ͑1997͒; R. H. Kodama, A. E. Berkowitz, E. J. McNiff, Jr., and S.
the saturation field is much larger than 2␲M in the whole
S
temperature range ͑5–350 K͒, which might be attributed to
16
This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
the surface spins of the very narrow wires. If we assume that
Rev. Lett. 80, 181 ͑1998͒, and references therein.
128.143.23.241 On: Sun, 10 Nov 2013 19:42:06