Electrodeposition and characterization of Cu/Co multilayers: Effect of individual Co and Cu layers on GMR magnitude and behavior

Article abstract of DOI:10.1016/j.jmmm.2008.03.020

The present work discusses the successful electrodeposition of Cu/Co multilayers, exhibiting appreciable GMR of 12-14% at room temperature. The effect of individual Cu and Co layers on the magnitude and behavior of GMR has been studied. By varying the thi

Full text of DOI:10.1016/j.jmmm.2008.03.020

ARTICLE IN PRESS
Journal of Magnetism and Magnetic Materials 321 (2009) 974–978
www.elsevier.com/locate/jmmm
Electrodeposition and characterization of Cu/Co multilayers: Effect of
individual Co and Cu layers on GMR magnitude and behavior
Ã
D.K. Pandya, Priyanka Gupta, Subhash C. Kashyap , S. Chaudhary
Thin Film Laboratory, Department of Physics, Indian Institute of Technology Delhi, Hauz Khas, New Delhi-110 016, India
Available online 15 March 2008
Abstract
The present work discusses the successful electrodeposition of Cu/Co multilayers, exhibiting appreciable GMR of 12–14% at room
temperature. The effect of individual Cu and Co layers on the magnitude and behavior of GMR has been studied. By varying the
thickness of individual layers the field at which saturation in GMR is observed can be controlled. It was observed that for lower
thicknesses of Co layer, the saturation fields are reduced below 1 kOe. The Cu layer thickness seems to control the nature of magnetic
coupling and the saturation field, with the two showing a correlation.
r 2008 Elsevier B.V. All rights reserved.
PACS: 68.65.Ac; 75.47.De; 75.70.Cn
Keywords: Magnetic multilayer; Magnetoresistance; Electrodeposition
1
. Introduction
via control of the thickness of individual Co and Cu layers,
and understanding the role of layer thickness on the
behavior of GMR.
Electrodeposition in modern day technology has
emerged as a novel economically viable technique with
large-scale production capabilities [1]. The technique has
generated immense interest worldwide because of its
simplicity, low cost, easy deposition irrespective of surface
size and area, the deposition being possible even in
nanopores [2]. The observation of appreciable GMR in
electrodeposited Co/Cu and CoNi/Cu multilayers [3],
comparable to that grown by vacuum techniques, further
raised the interest in this technique. Co/Cu multilayers
exhibiting appreciable GMR have been studied widely and
there are some reports on the dependence of GMR on
factors like number of bilayers [4], thickness of individual
Cu-layers [5,6] and Co-layers [7] and interfacial roughness
The multilayers were deposited in a potentiostatic pulse
mode, using a computer-interfaced potentiostat (PAR,
model 263A) as a potential source. Single sulfate bath
comprising of CoSO ꢀ 7H O, CuSO ꢀ 5H O, Na C H O
4
2
4
2
3
6
5
7
and NaCl in de-ionized water was used for the electro-
deposition of multilayers. All the depositions were made in
a three-electrode cell containing Pt as a counter electrode,
saturated calomel electrode (SCE) as reference and ITO
coated glass as a working electrode. Cyclic voltammetry
(CV) experiments were carried out to optimize the
deposition potentials for individual Cu and Co layers.
The thicknesses of individual layers, tCu and tCo, were
estimated from in situ I–t curves using Faraday’s laws. The
details have been presented elsewhere [9]. The resistance of
the deposited films was measured at room temperature in
current in plane/field in plane (CIP/FIP) configuration by
making four linear contacts, current being perpendicular to
the applied magnetic field. The externally applied magnetic
field was varied upto 8 kOe. The GMR was calculated
using the usual definition, GMR% ¼ (R ꢁR )/R ꢂ 100,
[
8]. However, low sensitivity and high saturation fields have
been restricting the application potential of these multi-
layers. After obtaining high GMR values of 12.6% in
Co/Cu multilayer stacks [9,10], in the present paper we
report about our efforts of lowering GMR saturation field
Ã
Corresponding author. Tel.: +91 11 26591346; fax: +91 11 26581114.
E-mail address: skashyap@physics.iitd.ac.in (S.C. Kashyap).
H
0
sat
where R is the resistance in presence of applied field, R is
H 0
0304-8853/$ - see front matter r 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmmm.2008.03.020

ARTICLE IN PRESS
D.K. Pandya et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 974–978
975
the resistance in absence of magnetic field and Rsat is the
resistance at higher field where it almost saturates. The
magnetization studies were carried out using vibrating
sample magnetometer (VSM), at room temperature by
varying the external field up to 8 kOe.
4 kOe for the sake of clarity. Appreciable GMR values
have been obtained in both cases. For sample having
tCo ¼ 10 nm, Fig. 1(a), the GMR saturates to a value of
7.5% at a relatively low magnetic field value ofo1 kOe.
For t ¼ 22 nm sample shown in Fig. 1(b), the GMR
Co
obtained was 12.6% which does not show saturation up to
a magnetic field of 8 kOe. The measurements were done for
both longitudinal (B?I) transverse (BJI) magnetoresis-
tance. For both, the magnitudes were almost equal and
negative, which clearly indicated the presence of prominent
GMR component without any AMR component. The
attainment of appreciable GMR indicates the presence of
appropriate antiferromagnetic alignment between the Co
layers.
2
. Results and discussion
2
.1. Effect of Co layer thickness
Figs. 1(a) and (b) show the variation of giant magne-
toresistance with applied magnetic field for two different
samples having Co layer thickness tCo ¼ 10 and 22 nm,
respectively. Both the samples had tCu ¼ 6 nm and n
(
number of bilayers) ¼ 50. The GMR curves for both the
Figs. 2(a) and (b) show the variation in GMR value with
increase in Co layer thickness. The thickness of Co layer
samples have been obtained upto 8 kOe, but shown up to
1
0
8
6
4
2
0
0
n = 50
Co = 10nm
Cu = 6nm
-
-
-
3
6
9
Cu = 4nm
Cu = 2nm
-
12
0
-
4
-2
0
2
4
5
6
7
8
9
10
Co layer thickness (nm)
H (kOe)
14
12
10
8
6
4
n = 50
Co = 22nm
Cu = 6nm
Cu = 6nm
-
-
4
8
-
12
-
8
-4
0
4
8
8
10
12
14
16
18
20
22
H (kOe)
Co layer thicness (nm)
Fig. 1. Variation in magnitude of GMR with applied magnetic field for a
0 bilayer sample on ITO (a) tCo ¼ 10 nm and (b) tCo ¼ 22 nm. Both the
samples have tCu ¼ 6 nm and number of bilayers, n ¼ 50.
Fig. 2. Variation in magnitude of GMR with thickness of Co layer:
(a) tCo ¼ 5–10 nm and tCu ¼ 2 and 4 nm and (b) tCo ¼ 8–22 nm and
5
tCu ¼ 6 nm.

ARTICLE IN PRESS
D.K. Pandya et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 974–978
976
was varied in two different ranges, 5–10 nm for lower tCu
of 2 and 4 nm and 8–22 nm for higher tCu of 6 nm,
respectively. An increase in GMR with increasing thickness
of Co layer was observed in all the sets of multilayers,
where the thickness of Cu layer was kept fixed at 2, 4 and
Cu = 4nm
1
1
0
0
0
.2
.0
.8
.6
.4
6
nm. This achievement of appreciable GMR and its
increase with Co layer thickness indicates that there is no
bulk or FM contribution from the Co layers. Even for
thicker Cu layers there exists sufficient AF coupling
between the adjacent Co layers. It also indicates that the
intermediate Cu spacer layer is continuous to the extent of
giving high enough GMR. If the Cu layer was not uniform
or was discontinuous, Co layers would have exhibited bulk
behavior, with negligible or no GMR.
Though both the samples exhibited appreciable GMR, a
remarkable difference was observed in their behavior
regarding the achievement of saturation in magnetoresis-
tance and the saturation magnetic field. It is observed from
Fig. 2(a) that for the samples having tCu of 2 and 4 nm, the
GMR tends to saturate as tCo increases to about 10 nm,
whereas Fig. 2(b) shows no saturation tendency for GMR
even as the tCo values approach 22 nm and magnetic field
approaches till 8 kOe, though after 3 kOe the change in
GMR was quite gradual. Figs. 3(a) and (b) show the
variation in saturation field with Co layer thickness, for
two set of multilayers having tCu ¼ 4 and 6 nm. It is
observed that saturation magnetic field initially increases in
all cases with increasing Co layer thickness. But for thinner
Co (5–10 nm) and Cu (2–4 nm) layers the saturation fields
as low as 0.5–1.2 kOe are achieved. All the ML samples
having Co layer thickness in this range exhibited saturating
behavior of GMR. For the case of tCu ¼ 6 nm, the
saturation field continues to increase beyond 1 kOe with
Co layer thicknesses increasing beyond 10 nm. The ML
samples having thicker Co layers in the range 8–22 nm had
higher saturation fields and exhibited a non-saturating
behavior of GMR. The attainment of higher saturation
fields, with increase in Co layer thickness can be explained
on the basis of surface roughness of individual layers. It
was observed that with increase in layer thicknesses, the
roughness of individual layers and hence the average
roughness of ML increases. This surface roughness seems
to induce granular like behavior in the MLs, exhibiting
superparamagnetic (SPM) behavior and in turn causing
them to either show a non-saturating behavior or exhibit
higher saturation fields. Moreover, it can also be said that,
larger is the thickness of Co layer; stronger will be the
coupling between two adjacent Co layers. Thus higher
fields will be required to saturate the GMR values. But at
present we do not have any further support or explanation
for this reasoning.
5
6
7
8
9
10
Co layer thickness (nm)
4
3
2
1
Cu = 6nm
8
10
12
14
16
18
20
22
24
Co layer thickness (nm)
Fig. 3. Variation in the saturation field values for two sets of samples.
Co
(a) tCu ¼ 4 nm and t lying in the range of 5–10 nm and (b) tCu ¼ 6 nm
and tCo lying in the range of 8–22 nm.
the GMR magnitude will be low with lower saturation
fields. For MLs having thin Co layers, if the layers
are continuous, this is expected. Also because of lower
surface roughness, FM regions are more prominent
than SPM regions, giving rise to lower GMR and low
saturation fields. For the other two paths where SPM
regions are also involved, the MLs are expected to have
prominent granular behavior, thus exhibiting either a non-
saturating behavior or having high fields of saturation.
Thicker Co layers are expected to have some SPM regions
along with FM regions, because in electrodeposited MLs
due to dissolution and exchange reactions during deposi-
tion, we cannot expect to have very smooth and continuous
films as we go for higher thicknesses, i.e. deposition done
for larger times.
The Co layers can be considered to have both
ferromagnetic and SPM regions within it. So, a conduction
electron crossing through the Cu spacer layer can travel
from one FM region to another FM or SPM region or
between both SPM regions. In case, the conduction
electron encounters only FM regions, it is expected that

ARTICLE IN PRESS
D.K. Pandya et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 974–978
977
1
1
1
4
2
0
8
6
4
2
n = 50
2.0
1.6
1.2
0.8
2
4
6
8
10
0
2
4
6
8
10
Cu layer thickness (nm)
Cu layer thickness (nm)
Fig. 5. Variation in the saturation field H
n ¼ 50.
S
with thickness of Cu layer for
Fig. 4. Variation in the magnitude of GMR with thickness of Cu layer for
n ¼ 50.
2
.2. Effect of Cu layer thickness
value with changing Cu layer thickness. The oscillations in
the values of saturation fields indicate the switching in the
coupling from antiferro to ferromagnetic. It can be seen
that the oscillations in GMR and saturation field are
correlated.
Fig. 4 shows the variation in GMR as the thickness of
Cu layer is changed from 1 to 10 nm, keeping the Co layer
of approximately same thickness (20 nm). The number of
bilayers was kept 50. It was observed that GMR exhibits
oscillations in its magnitude as the thickness of Cu layer
was varied. The peak in GMR values were seen for Cu
layer thickness of about 2, 4, 6 and 8 nm; though at 4, 6 and
3. Conclusions
Electrodeposited stacks of alternate Co and Cu layers, of
8
nm the GMR values were high in the range of 12–14%
thicknesses ranging from 2–20 nm, comprising of 50
bilayers exhibited appreciable high GMR up to 12–14%.
GMR was found to increase with increase in Co layer
thickness in the range 5–20 nm. The saturation field values
were also found to increase with the increase in Co layer
thickness. However, stacks with thinner Co and Cu layers
exhibited low saturating fields that saturate to a value of
about 1 kOe as Co thickness reaches 10 nm. For thicker Co
and appeared to be saturated. At intermediate values of tCu
lower values of GMR were obtained, suggesting an
oscillatory behavior of GMR. In electrodeposited films,
much of the previous work shows absence of oscillatory
behavior, due to structural imperfections [4,11–13],
whereas some reports [14,15] show the presence of
oscillatory behavior. Even in our case, the observed
oscillatory trend is quite different from that observed by
other workers [5,15–17]. The bilayer period obtained in our
case is 2 nm, unlike to the 1 nm bilayer period obtained by
other workers. This can be attributed to the higher surface
roughness induced in the MLs by the initial surface
roughness of ITO glass substrate. Also, we were unable
to investigate the effect at intermediate Cu layer thick-
nesses. The oscillations obtained in GMR with Cu spacer
layer variation, confirms that even at higher Cu thicknesses
of the range of 10 nm, the Co layers are not entirely
decoupled. Parkin et al. [5] have stated that Co/Cu
structures show the evidence of AF coupling for thicker
Co layers ranging from 0.25 to more than 20 nm, and also
large GMR values are possible for Cu layers more than
and Cu layers a non-saturating behavior in H and GMR is
S
seen for fields up to 4 kOe. The GMR and saturation field
values were found to exhibit oscillatory behavior with Cu
layer thickness in a correlated manner. The M–H behavior
clearly shows antiferro coupling for Cu thicknesses of
4–8 nm. Thus it is successfully established that by control-
ling the thicknesses of individual Cu and Co layers, the
magnitude of GMR as well as saturation field values can be
tuned in a controlled manner.
Acknowledgment
The authors acknowledge the financial support by
Ministry of Information Technology for this work.
2
0 nm thick. Actually from our magnetization studies of
these samples we find that for Cu spacer layer thickness of
, 6 and 8 nm the M–H loop shows signatures of AF
4
References
coupling and at all other Cu layer thicknesses M–H loop is
FM in nature. Fig. 5 shows the variation in saturation field
[1] C.A. Ross, Annu. Rev. Mater. Sci. 24 (1994) 159.

ARTICLE IN PRESS
D.K. Pandya et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 974–978
978
[
[
[
2] T. Ohgai, X. Hoffer, L. Gravier, J.-E. Wegrowel, J.-P. Ansermet,
Nanotechnology 14 (2003) 978.
Z.Z. Bandic, M. Rooks, R. Berger, T. Ando (Eds.), Nanostructured
and patterned materials for information storage, Materials Research
Society Symposium Proceedings of the 961E, Warrendale, PA, 2007
(0961-O12-03).
3] I. Bakonyi, J. Toth-Padar, J. Toth, T. Becsei, T. Tarnoczi,
P. Kamasa, J. Phys. Condens. Matter 11 (1999) 963.
4] L. Peter, A. Cziraki, L. Pogany, Z. Kupay, I. Bakonyi, M. Uhlemann,
M. Herrich, B. Arnold, T. Bauer, K. Wetzig, J. Electrochem. Soc. 148
[10] P. Gupta, D.K. Pandya, S.C. Kashyap, S. Chaudhary, Phys. Status
Solidi (A) 1–8 (2007).
(
2001) C168.
5] S.S.P. Parkin, R. Bhadra, K.P. Roche, Phys. Rev. Lett. 66 (1991)
152.
6] Q.X. Liu, L. Peter, J. Padar, I. Bakonyi, J. Electrochem. Soc. 152
2005) C316.
[11] I. Bakonyi, L. Peter, Z. Rolik, K. Kiss-Szabo, Z. Kupay, J. Toth,
L.F. Kiss, J. Padar, J. Phys. Rev. B 70 (2004) 054427.
[12] M. Shima, L. Salamanca-Riba, T.P. Moffat, R.D. McMichael,
J. Electrochem. Soc. 148 (2001) C518.
[
[
[
[
2
(
[13] E.C. Camblong, P.M. Levy, Phys. Rev. Lett. 69 (1992) 2855.
[14] V. Weihnacht, L. Peter, J. Toth, J. Padar, Zs. Kerner, C.M.
Schneider, I. Bakonyi, J. Electrochem. Soc. 150 (2003) C507.
[15] S. Kashiwabara, Y. Jyoko, Y. Hayashi, Physica B 239 (1997) 47.
[16] D.H. Mosca, F. Petroff, A. Fert, P.A. Schroeder, W.P. Pratt,
R. Laloee, J. Magn. Magn. Mater. 94 (1990) L1.
7] M. Shima, L. Salamanca-Riba, T.P. Moffat, R.D. McMichael,
J. Magn. Magn. Mater. 198–199 (1999) 52.
8] A. Cziraki, M. Koteles, L. Peter, Z. Kupay, J. Padar, L. Pogany, I.
Bakonyi, M. Uhlemann, M. Herrich, B. Arnold, J. Thomas, H.D.
Bauer, K. Wetzig, Thin Solid Films 433 (2003) 237.
[
9] D.K. Pandya, P. Gupta, S.C. Kashyap, S. Chaudhary, MRS Fall
Meeting, Boston, MA, USA, November 27–December 1, 2006, in:
[17] H. Kubota, S. Ishio, T. Miyazaki, Z.M. Stadnik, J. Magn. Magn.
Mater. 129 (1994) 383.