Giant magnetoresistance (GMR) and ferromagnetic properties of DC and pulse electrodeposited Cu–Co alloys

Journal of Magnetism and Magnetic Materials 324 (2012) 2983–2988

Contents lists available at SciVerse ScienceDirect

Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm

Giant magnetoresistance (GMR) and ferromagnetic properties of DC and pulse electrodeposited Cu–Co alloys N. Rajasekaran a, S. Mohan a, J. Arout chelvane b, R. Jagannathan a,n a b

Central Electrochemical Research Institute (CECRI-CSIR), Karaikudi-630006, India DMRL, Hyderabad-500058, India

a r t i c l e i n f o

abstract

Article history: Received 12 January 2012 Received in revised form 8 March 2012 Available online 11 May 2012

Cu–Co ferromagnetic alloys occurring as granular films and exhibiting giant magnetoresistance (GMR) property have been synthesized using both DC and pulse electrodeposition techniques. The growth process of these electrodeposits comprising multiple granules of disparate morphology, magnetic features exhibits critical dependence on electrodeposition conditions. Using ferromagnetic resonance and magnetic hysteresis data, we have attempted a correlation between evolution of these electrodeposits and their ferromagnetic properties with special emphasis on GMR property. & 2012 Elsevier B.V. All rights reserved.

Keywords: Electrodeposition Cu–Co alloys Giant magnetoresistance Ferromagnetic resonance

1. Introduction Fascinating success of Fe–Cr superlattice magnetic multilayers exhibiting giant magneto-resistance (GMR) properties and finding applications as high density recording media, magnetic sensors and reading head etc., is the driving force for the sustained interest in exploring newer materials, their magnetic characteristics [1–4]. Application of GMR systems in memory, sensor devices has stimulated more interest for augmenting scope for newer materials, properties and hence newer avenues for applications. These types of magnetic multilayer structures are normally prepared through electrodeposition methods. This conventional electrochemical route — as a strategic technique has gained much practical significance in preparing such magnetic systems in terms of ease of preparation, cost, and control on synthesis parameters [5–7]. Motivated by the success of this technique, in this investigation we discuss the results concerning the material, magnetic characteristics of Co–Cu films prepared using both DC and pulse electro deposition methods.

2. Experimental aspects Cu–Co alloys were electrodeposited on indium tin oxide coated glass substrate with resistivity in the range of 8 to 12 O cm  2 using the electrolyte composition comprising 0.04 M n

Corresponding author. E-mail addresses: [email protected] (S. Mohan), [email protected] (R. Jagannathan). 0304-8853/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jmmm.2012.04.027

CuSO4, 0.3 M CoSO4 and 0.05 M tri-sodium citrate as chelating agent at pH 4. The deposition was carried out in a three-electrode cell system under a galvanostatic control using Parstat 2273 model electrochemical analyzer. Co–Cu films deposited at different conditions served as the working electrode (cathode) while a platinum sheet was used as the counter-electrode (anode). Throughout the electrodeposition process a saturated calomel electrode (SCE) was used as the reference electrode. In this investigation, depositions of alloys were carried out using both direct current (DC) and pulse current (square wave from frequency range 10–100 Hz with 10% duty cycle) techniques with the average current density of 10 mA cm  2 (for a duration of 100 s in all cases). The samples are labeled as given in Fig. 1. The chemical purity and integrity of these electrodeposits were ascertained in terms of powder XRD pattern consistent with the standard file (Fig. 1) corresponding to the chemical composition Co0.48Cu0.52 (ICDD#50-1452) and their least squares refined crystallographic cell parameters (Table 1). Corresponding XRD line-width and intensity data were used to estimate the crystallite size (using Scherrer formula) and relative percentage of crystallinity. The magnetic hysteresis curves were recorded using a vibrating sample magnetometer (DMS-1600) at room temperature with applied magnetic field in the range of þ20 to 20 kOe. Ferromagnetic resonance (FMR) measurements were carried out using a Bruker X-band (9.83 GHz) electron paramagnetic resonance spectrometer. FMR spectra were recorded with the plane of electrodeposited films both parallel and perpendicular direction wrt the magnetic field axis of the electromagnet of the EPR system.

2984

N. Rajasekaran et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 2983–2988

1600

(nm)

Cu Kα (300K) S.G.: Fm 3m (225)

33

a (Å) ± 0.004

Intensity (arb.unit)

PED @

3.593

1200 28

100Hz (E)

3.585

28

50Hz (D)

3.586

800 33

25Hz (C)

3.591 10Hz (B)

51

400

(220)

(200)

(111)

3.597 DC (A) ICDD # 50-1452

0 40

50

60 2θ (deg)

70

80

Fig. 1. XRD pattern of Cu–Co electrodeposits obtained under different conditions as labeled and in comparison with ICDD standard for Cu0.52Co0.48 composition. Crystallite size F in nm and fitted crystallographic cell parameter a in A˚ values are given under respective column heads.

Table 1 Materials and magnetic characteristics of Cu–Co electrodeposits. Materials properties

Magnetic Properties Refined crystallographic cell parameter (a) d ¼ 7 0.004 A˚

Crystallite size (F) nm

A: DC; Cu0.94Co0.06 Pulsed @ B: 10 Hz; Cu0.91Co0.09 C: 25 Hz; Cu0.90Co0.10

3.5974

50.5

3.591 3.5839

D: 50 Hz; Cu0.88Co0.12 E: 100 Hz; Cu0.80Co0.20

3.5852 3.5926

Sample label: conditions; composition

Relative % of crystallinity

RH R0  100, R0

FMR

MR ¼ DR/ R0  100%

Coercivity (Hc) Oe

Remanence magnetization Mr X 10-3

Keff9  105

Han9 (KOe)

94.3

60.4

3.85

6.75

1.65

1.77

32.5 22.7

77.8 55.4

42.2 45.1

3.09 5.19

51.4 100

46.7 56.7

2.88 3.85

1.62 (i)1.64 (ii)0.76 1.34 0.89

0.84 1.1

22.7 32.5

6.51 (i)6.68 (ii)1.48 4.4 2.01

The magnetoresistance (MR) measurements were carried out using a Keithley Nanovoltmeter (6220) and power source (2182A) in four-probe method. During the MR measurements, direction of the magnetic field (0–6.3 kOe) was perpendicular to the applied current direction (transverse MR) which can be conveniently measured in current in plane configuration, commonly referred as CIP mode. That is in this CIP mode the current flow direction was parallel to the plane of the electrodeposited film(s). The percentage variation of MR values were calculated using the relation. MR % ¼

VSM

ð1Þ

with RH being the resistance with field and R0 being the resistance without field of the given film. Furthermore, the growth of these electrodeposits and their morphologies were followed-up with the help of scanning electron microscope images (Hitachi S3000H), its EDAX facility for compositional analysis and also atomic force microscope (Agilent technologies 5500).

2.7 1.1

3. Results and discussion 3.1. Material properties of Cu–Co electrodeposits The chemical integrity of the electrodeposits obtained using both DC and pulse deposited films yield nearly identical X-ray powder diffraction patterns consistent with the ICDD#50-1452 standard pattern corresponding to Cu0.52Co0.48 system (Fig. 1). Also the crystallographic unit cell parameter values of the different samples obtained using a standard least squares refinement procedure show good agreement with the standard values (Table 1) thereby confirming the chemical similarity and phase singularity of the electrodeposits being investigated. It is pertinent to note that in the Cu1 xCox alloy system, a wide variation (x¼ 0.06 to 0.20) in the relative composition is possible and these samples are indexable under one kind of XRD pattern [8]. However, the crystallite size (F) determining the growth rate and relative percentage of crystallinity (%C) show profound dependence on the preparation conditions of these electrodeposits. The crystallite size is the highest for the film

N. Rajasekaran et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 2983–2988

deposited using DC (sample A) while its relative degree of crystallinity is comparable to that of the film deposited using pulsed current with high frequency of 100 Hz (sample E). For other pulse plated samples at different frequencies the percentage of crystallinity showed a wide variation, which can be rationalized in terms of porosity versus film-growth conditions [9]. In this report, the possibility of oxide contamination obscuring the original results can be ruled-out for multiples reasons viz., (i) all the MR, FMR measurements were done for the well preserved electrodeposit films without much delay after the sample preparation (ii) also the EDAX data did not support the presence of any oxides. Between DC and pulsed electrodeposition of Cu–Co alloys, the former has been reported to yield a more Co rich composition when the off-time is high (  1000 m s) owing to pronounced galvanic exchange reaction during the off-time of the pulsed deposition process [10–12]. In the DC deposition there is no scope for change in set-current while in the pulse deposition the amplitude of the set-current is expected to be higher (peak current) during the peak time which may increase the propensity to accelerate the electrodeposition of less noble metals. Accordingly in the present investigation samples of pulsed deposition showed higher Co-content attributed to higher peak current than the average current density favoring the increased deposition of a less noble metal like Co in the alloy. During the course of pulseelectrodeposition, there can be process or secondary effects such as peak current deposition (during ON time), galvanic-exchange effect (during OFF time) determining the relative composition of the deposits. Although these processes are expected to occur simultaneously during the course of pulse electrodeposition, the impact of galvanic exchange is expected be less pronounced due to off-time period shortening. We note from the composition analysis that in the pulse deposition samples, Co content increases with frequency which can be explained through more occurrences of peak current and also decrease in off-time limiting the scope of galvanic exchange. 3.2. Film growth and giant magneto resistance Of different samples synthesized, sample D the pulse-deposited sample @50 Hz showed the highest magnetoresistance value of 2.7% as against the sample-B pulse deposited at 10 Hz showing the lowest value of 0.84% (Fig. 2a, Table 1). Whilst the sample-A obtained through the DC-electrodeposition showed a moderate magnetoresistance value of 1.77%. The percentage of crystallinity is the lowest for the sample-D showing the highest GMR (Table 1). Growth mechanism of these deposits can have critical dependence on the experimental parameters of electrodeposition, in particular the frequency of pulse deposition as can be seen from results given in the SEM images (Fig. 2b). Even on a cursory glance of these images, at least three kinds of granular growth can be figured-out and their relative abundance change with frequency of electrodeposition. Two types of granules appear prominently in the SEM images viz., as dark particles (size  0.2 to 0.3 m) and relatively bright white particles (size 0.5 m), the latter type shows a more pronounced growth behavior with size changing with pulse deposition frequency whilst the former type of particles does not show any substantial growth in size the third kind of particles interspersed in the backdrop of the brighter particles can be seen with less prominent features (size  0.1 m) fading in intensity upon increasing deposition frequency. The sample E deposited at high frequency of 100 Hz exhibiting pronounced agglomeration of these brighter particles started showing decrease in GMR which eventually suggest critical interdependence on growth versus GMR properties. As can be seen from AFM images and corresponding profile analysis traces given in Fig. 2c three kinds of particle growth in these

2985

electrodeposited films can be confirmed. Furthermore these three kinds of particles seem to posses different kinds magnetic characteristics which may either aid or hamper GMR properties according as whether they are soft or hard ferromagnetic species. 3.3. Ferromagnetic properties Ferromagnetic properties of these electrodeposits such as magnetic hysteresis curve, ferromagnetic resonance (FMR) spectra show subtle difference in terms of values and pattern (Fig. 3). Hystersis loop for various samples yield nearly comparable pattern with low coercivity and also remanence values (Fig. 4). But closer scrutiny and comparison of these data reveal subtle difference in terms of coercivity, remanence especially for sample C showing a two fold increase in remanance indicating a different origin or nature of the ferromagnetic phase as corroborated by the FMR results. Turning to coercivity value we can notice a trend showing shift towards origin with decrease in frequency of pulse deposition (indicated by a shady arrow in Fig. 4) with the sample A deposited using DC seems to be an exception. Moreover ferromagnetic properties of these granular films may have directional and also size dependence determined by the growth/ electrodeposition conditions which can be brought out using ferromagnetic resonance (FMR) studies. From the FMR spectra taken with magnetic field directions either parallel or perpendicular to the plane of these film samples (Fig. 3), drastic change in patterns taken in perpendicular (H?) configuration is so obvious as compared to that in parallel (H:) mode meriting a more significant consideration of the former. That is in the H? configuration, the FMR spectra show variety of signals with changing features with deposition conditions suggesting multiplicity of magnetic phases while such features are not so obvious in the H: mode. We can observe the emergence of intense Lorentizian band around 750 Oe gaining prominence with frequency at the expense of intensities of sharp weak feature (  1.6 kOe) and featureless band (  2.7 kOe) fading with frequency. Analyzing this data for sample C (pulse deposited at 25 Hz) we can see two sets of anisotropic factor [13] and field parameter values having a large (  3-fold) difference (Table 1). In this report, the effective anisotropic constant (Keff) and field (Han) have been calculated using the following formula [13].   o vðorÞs ¼ M H þ 4 p M =2, ð2Þ K ef ? R f

g

K ef f ? ¼ K vef f þ

2K Sef f , d

ð3Þ

Han? ¼ 4pM

2K ef f , M

ð4Þ

where M – magnetization o/2p – microwave frequency g – gyromagnetic ratio HR – resonance field Han? – annisotropy field K ef f ? – effective anisotropy constant. In order to fix the origin of these multiple magnetic phases, we compare this with evolution of different granular phases of these samples changing complexion with electrodeposition condition. Considering the relative intensity of FMR bands vis a vis the relative prominence of these three particle morphologies we could identify three kinds of magnetic phases as indicated in Fig. 3 inset. Also it should be noted that sample C showing the highest remanence value in the magnetic hysteresis data indicating a more ferromagnetic phase yet turned- out to be a low GMR yielding sample (Table 1). On the other hand the sample D pulse deposited at 50 Hz is having the lowest remanance, also the crystallinity turned out to be the highest GMR yielding system

2986

N. Rajasekaran et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 2983–2988

0.4

0.8

0.0 -0.4

Pulsed @

25 Hz (C) -1.2

100 Hz (E)

2.7

ΔR/R0 X 100

10 Hz (B) -0.8

-1.6

DC (A) -2.0 -2.4 50 Hz (D) -2.8 -8

-6

-4

2 -2 0 Magnetic Field kOe

4

6

8

Fig. 2. (a) GMR profile of DC, pulse electrodeposits with conditions as indicated sample labels in brackets, (b) corresponding SEM projections, (c) AFM images of representative samples with corresponding profile analysis curves.

whilst samples having higher crystallinity also the crystallite size yielded only lower GMR values. It seems reasonable to hypothesize that there exists a critical limit (in terms of magnetic remanance, crystallite size and crystallinity) for these granular electrodeposits to achieve good GMR results. GMR property based on spin-dependent scattering [14] might result in pronounced scattering losses when the critical limit is exceeded at the well defined, larger grain boundaries, interfaces. Also it is important to

note that there are already similar research concerning magnetoresistance, magneto-impedance properties and their correlation with ferromagnetic resonance properties of Cu–Co multilayers, Cox[Al2O3] (1  x) and electrodeposited Cu–Co–Ni alloys [15–20]. These studies also have highlighted the importance of local microstructure of these electrodeposits, concentration of magnetic ion(s) and the magnitude of effective anisotropy filed in influencing the magneto resistance properties.

N. Rajasekaran et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 2983–2988

1

2

2987

3

4000000 E 0

1500000 D

0

1

C 3

2

II

80000

E

4000000 0

D

2000000 0

C

2000000 0

B

3000000 0

A

2000000 0

Substrate

2000000 0 -2000000

B

0 80000

A 0

80000

Substrate

0 -80000 0

1

2

3

4 0 Magnetic Field kOe

1

2

3

Absorption Derivative (a.u)

40000

2768 Oe

1654 Oe

766 Oe

Absorption Derivative (a.u)

0

4

Magnetisation emu

Fig. 3. Ferromagnetic resonance spectra of these electrodeposits taken in parellel (:) and perpendicular (?) mode. 1, 2 and 3 correspond to three kinds of resonance band/ peaks. Inset: three kinds of particle morphologies assigned to three FMR peaks. FMR spectrum of bare substrate is also included.

5

A B C D E

-120

-60

60

Hc

120

to 2.8%.Granular growth of these electrodeposits stems from three kinds of disparate magnetic particles with their relative abundance changing with electrodeposition condition. Pulse deposition of these sample(s) show Co-content increasing with frequency that can be explained on the basis of peak-current and also shortening of off-time width. Ferro magnetic resonance spectra of these alloy films confirm the occurrence of three kinds of magnetic particles having different ferromagnetic characteristics. Morphological study in conjunction with GMR results suggests a critical size for these granular particles to achieve optimum GMR value.

Acknowledgment

-5

Magnetic Field Oe Fig. 4. Magnetic hysteresis curves for different electrodeposits as labeled. Shift in coercivity towards origin is indicated by the shady arrow.

Our Sincere thanks to CECRI, Karaikudi-CSIR, New Delhi for the support we received in this work. One of the authors (N.R) expresses his sincere thanks to the CSIR-New Delhi for the award of a Senior Research Fellowship.

References

4. Conclusions Cu–Co films synthesized through DC and pulse electrodeposition methods show giant magnetoresistance in the range of 0.8%

[1] M.N. Baibich, J.M. Broto, A. Fert, F. Nguyen Van Dau, F. Petroff, P. Eitenne, G. Creuzet, A. Frederich, J. Chazelas, Physical Review Letters 61 (1988) 2472. [2] G. Binasch, P Grunberg, F. Suarenbach, W. Zinn, Physical Review B 39 (1989) 4828.

2988

N. Rajasekaran et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 2983–2988

[3] Dali Sun, Lifeng Yin, Chengjun Sun, Hangwen Guo, Zheng Gai, X.G. Zhang, T.Z. Ward, Zhaohua Cheng, Jian Shen, Physical Review Letters 104 (2010) 236602 1. ¨ [4] A. Jesche, A. Gorbunoff, A. Mensch, H. Stocker,1, A.A. Levin, D.C. Meyer, Journal of Applied Physics 107 (2010) 023904 1. [5] D. Rafaja, C. Schimpf, T. Schucknecht, V. Klemm, L. Peter, I. Bakonyi, Acta Materialia 59 (2011) 2992. [6] S.K. Ghosh, A. Dogra, C. Srivastava, S.K. Gupta, Journal of Alloys and Compounds 504 (2010) 452. [7] Bakonyi, L. Peter, Progress in Materials Science 55 (2010) 107. [8] K. Miyazaki, S. Kainuma, K. Hisatake, T. Watanabe, N. Fukumuro, Electrochimica Acta 44 (1999) 3713. [9] J.C. Puippe and F. Leaman (Eds.) Theory and Practice of Pulse Plating p-11American Electroplaters and Surface Finishers Society (1986. [10] J.J. Kelly, P.E. Bradley, D. Loandolt, Journal of the Electrochemical Society J 147 (2000) 2975. [11] S. Roy, M. Matlosz, D. Landolt, Journal of the Electrochemical Society J 141 (1994) 1509. [12] P.E. Bradley, D. Landolt, Electrochimica Acta 45 (1999) 1077.

[13] B. Schulz, R. Schwarzwald, K. Baberschke, Surface Science 307–309 (1994) 1102. [14] W. Wang, F. Zhu, W. Lai, J-Q. Wang, G. Yang, J. Zhu, Z. Zhang, Journal of Physics D: Applied Physics 32 (1999) 1990. [15] Shihui Ge, Haohua Li, Chao Li, Li Xi, Wei Li, Junhang Chi, Journal of Physics: Condensed Matter 12 (2000) 5905. [16] D.P. Belozorov, V.N. Derkach, S.V. Nedukh, A.G. Ravlik, S.T. Roschenko, I.G. Shipkova, S.I. Tarapov, F. Yildiz, B. Aktas, Journal of Magnetism and Magnetic Materials 263 (2003) 315. [17] D.P. Belozorov, V.N. Derkach, S.V. Nedukh, A.G. Ravlik, S.T. Roschenko, I.G. Shipkova, S.I. Tarapov, F. Yildiz, International Journal of Infrared and Millimeter Waves 22 (2001) 1669. [18] S. Tarapov, T. Bagmut, V. Derkach, S. Nedukh, F. Yildiz, S. Kazan, Journal of Magnetism and Magnetic Materials 300 (2006) e78. [19] I.H. Karahan, O.F. Bakkaloglu, M. Bedir, Pramana Journal of Physics 68 (2007) 83. [20] W. Schindler, Th. Koop, J. Kirsehner, Zeitschrift fur Physikalische Chemie, Bd. 208 (1999) S.93.