Invariance of the magnetic behavior and AMI in ferromagnetic biphase films with distinct non-magnetic metallic spacers

Contents lists available atScienceDirect

Physica B

journal homepage:www.elsevier.com/locate/physb

Invariance of the magnetic behavior and AMI in ferromagnetic biphase

films with distinct non-magnetic metallic spacers

E.F. Silva

, M. Gamino

, A.M.H. de Andrade

, M. Vázquez

, M.A. Correa

, F. Bohn

b_{Departamento de Física, Universidade Federal de Pernambuco, 50670-901 Recife, PE, Brazil} c_{Instituto de Física, Universidade Federal do Rio Grande de Sul, 91501-970 Porto Alegre, RS, Brazil} d_{Instituto de Ciencia de Materiales de Madrid, CSIC, 28049 Madrid, Spain}

A R T I C L E I N F O

A B S T R A C T

We investigate the quasi-static magnetic, magnetotransport, and dynamic magnetic properties in ferromagnetic biphasefilms with distinct non-magnetic metallic spacer layers. We observe that the nature of the non-magnetic metallic spacer material does not have significant influence on the overall biphase magnetic behavior, and, consequently, on the magnetotransport and dynamic magnetic responses. We focus on the magnetoimpedance effect and verify that the films present asymmetric magnetoimpedance effect. Moreover, we explore the possibility of tuning the linear region of the magnetoimpedance curves around zero magneticfield by varying the probe current frequency in order to achieve higher sensitivity values. The invariance of the magnetic behavior and the asymmetric magnetoimpedance effect in ferromagnetic biphase films with distinct non-magnetic metallic spacers place them as promising candidates for probe element and open possibilities to the development of lower-cost high sensitivity linear magneticfield sensor devices.

1. Introduction

Advances in thefield of the magnetization dynamics during the last decades have contributed to the comprehension in the context of fundamental physics on how the magnetization responds to external magneticfields and electric currents, and also have led to enormous progress in the development of sophisticated technological devices [1,2]. In this direction, the magnetoimpedance effect (MI) appears an important tool to investigate nanostructured magnetic materials, revealing the magnetic properties at different frequency ranges and magneticfield values, at saturated and unsaturated states, as well as at resonant and non-resonant regimes [3]. Moreover, the effect is of technological interest due to the potential of application of materials exhibiting magnetoimpedance as probe element in magnetic sensor devices for low-field detection[4].

The magnetoimpedance effect corresponds to the change of the real and imaginary components of electrical impedance Z=R+iX of a ferromagnetic material caused by the action of an external static magnetic field. In recent years, the interest on materials exhibiting MI to act as probe element in technological sensor devices has grown significantly[5–8], justifying the increasing number of reports addres-sing different magnetic structures[9–15]. The sensitivity and linearity

as a function of the magneticfield are the most important parameters for practical applications of the magnetoimpedance effect. However, although most of the soft magnetic materials are highly sensitive to small field variations at low magnetic fields, they use to show a nonlinear MI behavior around zerofield[16].

For this reason, the design of novel magnetic materials with optimized properties and specific dynamic magnetic response is fundamental, making possible the observation of new phenomena and their use to the development of a next generation of technological devices. In this sense, several studies have been carried out to improve the linear features of the MI. This explains the recent interest in materials with asymmetric magnetoimpedance (AMI), which can present linear response around zero magneticfield. These materials are obtained by inducing an asymmetric magnetic configuration, usually done by magnetostatic interactions or exchange bias. The primary AMI results have been measured in wires[17–19], amorphous ribbons [20,21], and exchange biased multilayered films[16,22,23], opening possibilities for the use of this kind of material for the development of auto-biased linear magneticfield sensors[22].

Beyond the aforementioned magnetic systems, recently, it has been shown that AMI can be obtained in NiFe/Cu/Co films presenting biphase magnetic behavior and which consist of hard and soft

http://dx.doi.org/10.1016/j.physb.2016.11.009

Received 31 August 2016; Received in revised form 3 November 2016; Accepted 4 November 2016

⁎_{Corresponding author.}

E-mail address:felipebohn@fisica.ufrn.br(F. Bohn).

Available online 15 November 2016

0921-4526/ © 2016 Elsevier B.V. All rights reserved.

ferromagnetic phases intermediated by a non-magnetic metallic layer acting together[24]. In this case, the linear region of the AMI curves has been tuned by varying the thickness of the Cu spacer layer, and probe current frequency. Moreover, as an improvement with respect to this structure, it has been already verified that the AMI sensitivity at lowfields can be amplified if ferromagnetic multilayered biphase films are considered[25].

In this work, we investigate the quasi-static magnetic, magnetotran-sport, and dynamic magnetic properties in ferromagnetic biphasefilms with distinct non-magnetic metallic spacers. By analyzing the magne-tization dynamics through the magnetoimpedance effect, we verify that thefilms present asymmetric magnetoimpedance. We confirm that the linear region of the AMI curves around zero magnetic field can be tuned by the probe current frequency. We show that the nature of the non-magnetic metallic spacer material does not have significant influence on the overall biphase magnetic behavior and, consequently, on the magnetotransport and dynamic magnetic responses. The invariance of the magnetic behavior and the AMI in ferromagnetic biphasefilms with distinct non-magnetic metallic spacers place them as promising candidates for probe element and represent an important step to the development of lower-cost high sensitivity linear magnetic field sensor devices.

2. Experiment

For this study, we produce a set of ferromagnetic biphasefilms with distinct non-magnetic metallic spacers, i.e. Ni81Fe19(25 nm)/ NM(7 nm)/Co(50 nm) films, where NM=Cu, Ta, and Au. The films are deposited by magnetron sputtering onto glass substrates, with dimensions of 8 × 4 mm2_{. For all, a 10 nm-thick Ta buffer layer is}

considered in order to reduce irregularities of the substrate and enhance the adherence of the magnetic film, as well as a cap layer with similar thickness is employed to avoid oxidation. The deposition process is carried out with the parameters presented inTable 1. During the deposition, a constant 2 kOe magneticfield is applied perpendicu-larly to the main axis of the substrate in order to induce magnetic anisotropy.

The deposition rates, also presented in Table 1, are obtained through low angle X-ray diffractometry, while the structural properties of thefilms are investigated by high-angle X-ray diffraction measure-ments. The latter, not shown here, reveals the crystalline structural character of the ferromagnetic layers and indicates the fcc cubic Co (111) and NiFe (111) preferential growth for all films [24,25], irrespective on the non-magnetic metallic spacer material.

The quasi-static magnetic properties are investigated through magnetization curves obtained using a vibrating sample magnetometer. The curves are acquired at room temperature, with in-plane magnetic field of maximum amplitude of ± 300 Oe applied in two directions, along (ϕ = 0°) and perpendicular (ϕ = 90°) to the main axis of thefilm. Magnetotransport properties are obtained through magnetoresis-tance (MR) measurements performed using the four probe method. The experiments are carried out for the ϕ = 90° configuration, with in-plane magnetic field of maximum amplitude of ± 300 Oe, and dc electric current of 0.2 mA applied along the main axis of thefilm. The MR results are presented in terms of the magnetoresistance ratio,

defined by R R R H R H R H Δ (%) = [ ( ) − ( )] ( ) × 100%, max max (1)

where R(H) is the electric resistance at a given magneticfield value and R H( max)is the resistance at the maximum applied magneticfield, in which thefilm is magnetically saturated.

The magnetization dynamics is investigated through longitudinal MI effect, where magnetic field and current are applied along the direction of the main axis of thefilm, ϕ = 0°. The measurements are obtained using a RF-impedance analyzer Agilent model E4991, with E4991Atest head connected to a microstrip, in which the sample is the central conductor. To avoid propagative effects and acquire just the sample contribution to MI, the analyzer is calibrated at the end of the test head connector by performing open, short, and load (50Ω) measurements using reference standards. The curves are taken by acquiring the real R and imaginary X components of the impedance Z in a wide range of frequencies from 0.1 GHz up to 3.0 GHz, in a linear regime with 1 mW (0 dBm) constant power, and with in-plane magneticfield varying between ± 300 Oe. To quantify the sensitivity as a function of the frequency, we calculate the magnitude of the impedance change at lowfields through[24,25]

Z H Z H Z H |Δ | |Δ |= ( = 6 Oe) − ( = −6 Oe) 12 . (2)

Here, we consider the absolute value of|Δ |Z since the impedance around zerofield can present positive or negative slopes, depending on the sample and frequency. It is verified that |Δ |/|Δ |Z H is roughly constant at least for a reasonable lowfield range.

3. Results and discussion

First all, we address the experimental results associated to the quasi-static magnetic and magnetotransport properties, which provide us information on the magnetic anisotropy, the orientation of the magnetization in each ferromagnetic layer, and the magnetization reversal process.

Fig. 1shows the normalized magnetization curves and MR ones for the NiFe/NM/Cofilms with NM = Cu, Ta and Au.

The angular dependence of the magnetization curves,Fig. 1(a,c,e), indicates the existence of a magnetic anisotropy, with the easy magnetization axis oriented perpendicularly to the main axis, induced by the magneticfield applied during deposition. The films exhibit a biphase magnetic behavior, as clearly verified from the magnetization curves obtained for ϕ = 90°. The two-stage magnetization process is characterized by magnetization reversal of the soft NiFe layer at low magneticfield, followed by the reversal of the hard Co layer at higher field.

The orientations between the magnetizations of the ferromagnetic layers at different magnetic field values, as well as the magnetization reversal process, previously discussed can also be identified through the MR curves. The largest MR ratio values are found when an antiparallel alignment between the magnetizations of the soft and hard ferromagnetic layers is established, while the smaller MR ratios occur in the case of the parallel alignment [26–29]. Thus, from the MR curves,Fig. 1(b,d,f), starting from the maximum negativefield value, as the magneticfield increases, the abrupt rise in the MR curve occurs at the switching field of the soft NiFe layer. As the magnetic field continuously increases, the sudden drop in the MR curve in located at the switchingfield of the hard Co layer. Similar behavior is verified in the MR curves along the other branch of the hysteresis loop, with decreasingfield starting from the maximum positive field value. These features correspond to a clear signature that the shape of the MR curve is directly related to the magnetization reversal process and to the biphase magnetic behavior observed through the magnetization curves. The biphase magnetic behavior verified through both the

magne-Table 1

Parameters employed for the deposition of the films.

Parameters NiFe Co Cu Ta Au

Base pressure (Torr) 10−8 10−8 10−8 10−8 10−8 Deposition pressure (mTorr) 2.0 2.0 2.0 2.0 2.0

Arflow (sccm) 32 32 32 32 20

Power (W) 150 150 100 200 60

Source DC DC RF DCP RF

tization and MR curves suggests that the ferromagnetic layers are essentially uncoupled. This behavior is dependent on the relationship between the thickness and magnetic anisotropy constant of the soft and hard ferromagnetic layers, as well as on the thickness of the spacer material [24]. Small variations in the switching fields, area of the magnetization loops, and values of the MR ratios observed with the change of the non-magnetic metallic spacer material may be possibly associated to the quality of the ferromagnetic/spacer and spacer/ ferromagnetic interfaces and electric properties of the spacer materials. In particular, none significant influence of the spacer on the biphase magnetic behavior is verified, suggesting that this behavior does not dependent specifically on the nature of the metallic spacer material, at least for considered elements and thickness.

On the dynamic magnetic behavior, it is well-known that the quasi-static magnetic properties play a fundamental role in the dynamic magnetic response and MI behavior. The intensity of the effective magnetic anisotropy, associated to its orientation with respect to the magnetic field and ac probe current, determines the MI response [15,16,30,31].

Fig. 2shows the evolution of the MI curves as a function of the magneticfield, for selected frequency values, for the NiFe/Au/Co film, as a representative example of the general MI behavior verified for the studied samples. It is important to notice that, although the MI curves are acquired over a complete magnetization loop and present hysteretic behavior, here we show just part of the curve, when thefield goes from negative to positive values. In particular, when the field goes from positive to negative values, the MI behavior is reversed, a feature due to the opposite sense of the magnetizations of the NiFe and Co layers [24,25].

The curves have similar features for thefilms with distinct spacers. In particular, they present a double peak structure for the whole frequency range, as expected, since that the external magneticfield and ac probe current are perpendicular to the easy magnetization axis[32]. On the position of the peaks, for the allfilms, the peaks position

remains unchanged for frequencies below∼0.5 GHz. For the film with Au as spacer layer, the peak at negativefield is located at∼−6Oe, while the peak at positivefield is at ∼+26 Oe. Similar behavior is verified for thefilms with Cu and Ta, whose peaks are located at ∼−4 and∼+30Oe, and ∼−8 and ∼+16 Oe, respectively. It worth to notice that, for each sample, these magneticfield values in which the impedance peaks are located, in modulus, correspond to the switchingfields of the NiFe and Co ferromagnetic layers, respectively, identified through the magneti-zation and MR curves. At this low frequency range, the invariance of the position of the peaks is a clear indication that the skin effect is the main responsible for the MI variations[33]. However, above∼0.5 GHz, besides the skin effect, the FMR effect also becomes an important mechanism responsible by the MI variations, a fact evidenced by the displacement of the peaks position toward higherfield as the frequency is increased[33,34]. Moreover, above∼1.0 GHz, as a signature of the higher contribution of the FMR effect to the dynamic magnetic behavior, the displacement of the peaks toward higherfields suppresses the peak asymmetry resulting in symmetric peaks around zerofield with same amplitude.

An interesting issue on the results resides in the asymmetry in the MI curves measured at the low and intermediate frequency ranges, below∼1.0 GHz.Fig. 3shows the MI curves as a function of magnetic field, at the selected frequency of 0.5 GHz, for the NiFe/NM/Co films with NM=Cu, Ta, and Au.

The AMI behavior is evidenced by the asymmetric position of the peaks and the difference of their amplitude. If the ferromagnetic layers were completely uncoupled, one could expect multiple peaks MI behavior, associated to the anisotropyfields of the layers. This is not verified here, indicating that the AMI cannot be explained assuming independent reversal of the soft NiFe and hard Co layers. Thus, the asymmetric behavior is a result of both the orientation of the magnetization of the NiFe and Co layers, as well as the kind of magnetic interaction between the ferromagnetic layers, which in this case is characterized by a magnetostatic coupling[18,20,24,25,35,36].

Fig. 1. (a,c,e) Normalized magnetization and (b,d,f) MR curves for the NiFe/NM/Co films with NM=Cu, Ta, and Au. The magnetization curves are measured with in-plane magneticfield applied along (ϕ = 0°) and perpendicular (ϕ = 90°) to the main axis. The MR curves are acquired for the ϕ = 90° configuration, with the electric current applied along the main axis of thefilms.

Fig. 2. MI curves, ϕ = 0°, at selected frequencies for the NiFe/Au/Cofilm. Similar behavior is obtained for thefilms with the Cu and Ta spacers. Although the curves are acquired over a complete magnetization loop, here we show just part of the curve, when thefield goes from negative to positive values, to make easier the visualization of the whole MI behavior.

It is worth noting that our MI experimental results are in qualitative agreement with the theoretical predictions for a tri-layered bimagnetic system recently reported by Buznikov and Antonov[36].

From the technological point of view, the MI sensitivity and linearity are the main parameters to be controlled [20]. Here, it is important to notice that for all the investigatedfilms, irrespective on the non-magnetic metallic spacer layer, the AMI provides a nearly linear response around zerofield.Fig. 4shows the frequency spectrum of impedance variations, defined by(2), indicating the sensitivity at low fields. The highest sensitivity values are verified at ∼0.75 GHz, reaching ∼10.4 mΩ/Oe for film with Au spacer. The sample with Cu spacer reaches a maximum of∼8.0 mΩ/Oe at ∼0.75 GHz, while the film with Ta has maximum of∼7.2 mΩ/Oe at ∼0.67 GHz. In particular, the small differences of sensitivity values may be devoted to the quality of interfaces and electric properties of the spacer materials, as previously discussed.

After all, as an issue of fundamental interest for technological applications, the most strikingfinding here resides in the invariance of the magnetic behavior and AMI in ferromagnetic biphasefilms with distinct non-magnetic metallic spacers. The fact that ferromagnetic biphasefilms with distinct non-magnetic metallic spacers share similar features and magnetic responses represents an important step to the development of lower-cost high sensitivity linear magneticfield sensor devices.

4. Conclusion

In summary, we have investigated the quasi-static magnetic, magnetotransport, and dynamic magnetic properties in ferromagnetic biphase films with distinct non-magnetic metallic spacers. We have observed that the nature of the non-magnetic metallic spacer material does not have significant influence on the overall biphase magnetic behavior and, consequently, on the magnetotransport and dynamic magnetic responses. The magnetoimpedance results strictly depend on

the structure of the biphasefilms and probe current frequency, and can be understood in terms of the orientation of the ferromagnetic layers and kind of magnetic interaction between the ferromagnetic layers. The small differences in the behavior of distinct samples may be associated to the quality of interfaces and electric properties of the non-magnetic metallic spacer materials. We have verified that the films present AMI behavior and have tuned the linear region of the MI curves around zero magneticfield by varying the probe current frequency. In particular, we have estimated the MI sensitivity at lowfield and have verified that significant sensitivity values can be achieved. It is known that the employment of multilayers enables us to reach higher sensitivity values. In this sense, to improve the response of the ferromagnetic biphasefilms with distinct non-magnetic metallic spacers, experiments in multilayered ferromagnetic biphasefilms are currently in progress. These results and the invariance of the magnetic behavior and AMI in ferromagnetic biphase films with distinct non-magnetic metallic spacers represent an important step from the technological point of view, placing them as promising candidates for probe element, and open possibilities to the development of lower-cost high sensitivity linear magneticfield sensor devices.

Acknowledgments

The research is supported by the Brazilian agencies CNPq (Grants nos. 306423/2014-6, 471302/2013-9, 306362/2014-7, 152650/ 2016-4, and 441760/2014-7), CAPES, and FAPERN (Pronem no. 03/ 2012).

References

[1] G. Durin, S. Zapperi. The science of hysteresis: physical modeling, micromagnetics and magnetization dynamics, vol. II, ch. III, The Barkhausen Effect, Cambridge University Press, Amsterdam, 2006.

[2] S.M. Rezende, Materiais e dispositivos eletônicos, Livraria da Física, São Paulo, 2004.

[3] R.B. da Silva, et al., Appl. Phys. Lett. 104 (2014) 102405.

[4] H. Hauser, L. Kraus, P. Ripka, IEEE Instrum. Meas. Mag. 4 (2001) 28. [5] M. Knobel, et al., J. Appl. Phys. 79 (1996) 1646.

[6] C. Morón, A. García, J. Magn. Magn. Mater. 290–291 (2005) 1085.

[7] H. Chiriac, M. Tibu, A.E. Moga, D.D. Herea, J. Magn. Magn. Mater. 293 (2005) 671.

[8] L.V. Panina, K. Mohri, Appl. Phys. Lett. 65 (1994) 1189.

[9] J. Velázquez, M. Vázquez, D.-X. Chen, A. Hernando, Phys. Rev. B 50 (1994) 16737. [10] K.R. Pirota, L. Kraus, M. Knobel, P.G. Pagliuso, C. Rettori, Phys. Rev. B 60 (1999)

6685.

[11] S.-Q. Xiao, et al., Phys. Rev. B 61 (2000) 5734.

[12] M. Corrêa, A. Viegas, R. da Silva, A. de Andrade, R. Sommer, Phys. B Condens. Matter 384 (2006) 155.

[13] G.V. Kurlyandskaya, J. Magn. Magn. Mater. 321 (2009) 659.

[14] L. Chen, Y. Zhou, C. Lei, Z.M. Zhou, W. Ding, J. Magn. Magn. Mater. 322 (2010) 2834.

[15] K. Agra, et al., J. Magn. Magn. Mater. 393 (2015) 593. [16] R.B. da Silva, et al., Appl. Phys. Lett. 104 (2014) 102405. [17] G.V. Kurlyandskaya, et al., J. Appl. Phys. 87 (2000) 4822.

[18] J. Torrejón, M. Vázquez, L.V. Panina, J. Appl. Phys. 105 (2009) 033911. [19] J. Sinnecker, A. de Araujo, R. Piccin, M. Knobel, M. Vázquez, J. Magn. Magn.

Mater. 295 (2005) 121.

[20] C.G. Kim, K.J. Jang, D.Y. Kim, S.S. Yoon, Appl. Phys. Lett. 77 (2000) 1730. Fig. 3. MI curves, ϕ = 0°, at the frequency of 0.5 GHz measured for the NiFe/NM/Cofilms with NM=Cu, Ta, and Au.

Fig. 4. Frequency spectrum of impedance variations between ± 6 Oe for the NiFe/NM/ Cofilms with NM=Cu, Ta, and Au, indicating the sensitivity around zero field.

[21] N.A. Buznikov, C. Kim, C.O. Kim, S.S. Yoon, Appl. Phys. Lett. 85 (2004) 3507. [22] R.B. da Silva, et al., Appl. Phys. Lett. 94 (2009) 042501.

[23] C. García, J.M. Florez, P. Vargas, C.A. Ross, Appl. Phys. Lett. 96 (2010) 232501. [24] E.F. Silva, et al., Appl. Phys. Lett. 105 (2014) 102409.

[25] E. Silva, et al., J. Magn. Magn. Mater. 393 (2015) 260.

[26] S. Tsunashima, M. Jimbo, Y. Imada, K. Komiyama, J. Magn. Magn. Mater. 165 (1997) 111.

[27] T. Luciński, et al., J. Magn. Magn. Mater. 269 (2004) 78.

[28] L. Kong, Q. Pan, B. Cui, M. Li, S.Y. Chou, J. Appl. Phys. 85 (1999) 5492. [29] B. Dieny, et al., Phys. Rev. B 43 (1991) 1297.

[30] A. Zhukov, M. Ipatov, V. Zhukova, Handbook of magnetic materials, Vol. 24, Ch. 2, Advances in Giant Magnetoimpedance of Materials, Elsevier, Amsterdam, 2008. [31] N.A. Usov, A.S. Antonov, A.N. Lagar'kov, J. Magn. Magn. Mater. 185 (1998) 159. [32] R.L. Sommer, C.L. Chien, Appl. Phys. Lett. 67 (1995) 3346.

[33] M.A. Corrêa, F. Bohn, R.B. da Silva, R.L. Sommer, J. Appl. Phys. 116 (2014) 243904.

[34] D. de Cos, A. García-Arribas, J.M. Barandiaran, J. Magn. Magn. Mater. 304 (2006) 218.

[35] C.G. Kim, K.J. Jang, D.Y. Kim, S.S. Yoon, Appl. Phys. Lett. 77 (2000) 1730. [36] N. Buznikov, A. Antonov, J. Magn. Magn. Mater. 420 (2016) 51.