Multi-cracking and magnetic behavior of Ni80Fe20 nanowires

Apr 18, 2018 - These measurements show, on the one hand, a delay in crack initiation relative to the non-patterned thin film, and, on the other hand, ...
0 downloads 0 Views 6MB Size
Subscriber access provided by Kaohsiung Medical University 80

20

Multi-cracking and magnetic behavior of Ni Fe nanowires deposited onto a polymer substrate Skander Merabtine, Fatih Zighem, Damien Faurie, Alexis GarciaSanchez, Pierpaolo Lupo, and Adekunle Olusola Adeyeye Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b00922 • Publication Date (Web): 18 Apr 2018 Downloaded from http://pubs.acs.org on April 18, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Multi-cracking and magnetic behavior of

Ni80 Fe20

nanowires deposited onto a polymer

substrate

S. Merabtine1 , F. Zighem1 ,∗ D. Faurie1 ,† and A. Garcia-Sanchez1 1

Laboratoire des Sciences des Procédés et des Matériaux, CNRS,

Université Paris 13-Sorbonne Paris Cité, 93430, Villetaneuse, France

2

P. Lupo2 and A. O. Adeyeye2‡ Information Storage Materials Laboratory, Department of Electrical and Computer Engineering, National University of Singapore - 117576 Singapore

(Dated: April 04

®

th

2018)

This work presents the eect of large strains (up to 20%) on the behavior of magnetic nanowires (Ni80 Fe20 ) deposited on a Kapton

substrate. The multi-cracking phenomenon was followed by in

situ tensile tests combined with atomic force microscopy measurements. These measurements show,

on the one hand, a delay in crack initiation relative to the non-patterned thin lm, and, on the other hand, a saturation of the length of the nanowire fragments. The latter makes it possible to retain the initial magnetic anisotropy measured after deformation by ferromagnetic resonance. In addition, the ferromagnetic resonance line prole (intensity, width) is little aected by the numerous cracks, which is explained by the small variation in magnetic anistropy and the low magnetostriction coecient of

Ni80 Fe20 .

Keywords: Ferromagnetic nanowires, exible systems, nanomechanics, nanomagnetism

Stretchable/exible electronics has been a growing eld for several years [15]. Particularly, magnetic systems made on exible substrates are of growing interest [611] because of the potential applications of exible/ stretchable magneto-electronics include the detection of magnetic elds in media with complex geometries and especially in/on human tissues [12]. In addition, the increasingly complex forms of consumer electronics and intelligent textiles will benet from the magnetic functionality of these systems due to their adaptability to non-planar surfaces (conformability) [13]. Other factors contributing to the growth of this market are the weight savings and the low cost of production compared to rigid substrates such as silicon. From a general point of view, the main limitation of these exible/stretchable systems is their often still too low durability [14]. While substrates generally made of polymers are suitable for large deformations, the inorganic nanolms and nanostructures that carry the magnetic functionality are intrinsically much more brittle. Indeed, when a metal is deposited on a exible substrate, it is subjected to high mechanical stresses due to the stretching or the curvature of the system and to the metal/polymer interfacial adhesion [1517]. Thus, there are still major challenges in this eld and in particular to understand the phenomena of multi-cracking which are usual when large strains are applied [8, 1821]. Moreover, it is important to check the impact of this phenomenon on the magnetic properties. It is thus crucial to study the eects of lateral nanostructuring of thin lms on coupled mechanical and

∗ Corresponding

author. E-mail: [email protected] author. E-mail: [email protected] ‡ Corresponding author. E-mail: [email protected] † Corresponding

magnetic properties. This kind of studies is lacking but is indispensable for applications requiring magnetic nanoobjects on polymer substrates such as devices based on magnonic crystals [22]. In this paper, we present a systematic study of the eects of multi-cracking on the magnetic behaviour of 20 nm thick Ni80 Fe20 arrays of nanowires fabricated on Kapton (polymide). For comparison, lms of same composition and thickness have been also fabricated and studied. Uniaxial tests along the nanowires were carried out with in situ atomic force microscopy measurements. These experiments made it possible to demonstrate the evolution of the crack density as well as the length of the residual magnetic fragments. Ferromagnetic resonance (FMR) measurements performed after 20% of strain show that the multi-cracking of nanowires does not aect the probed magnetic properties.

®

The arrays of nanowires have been fabricated using interference nanolithography technique [23, 24]. The main advantage of interference lithography techniques, compared with electron beam lithography for instance, is that they allow the fabrication of large arrays of nanostructures (0.5 × 0.5 cm2 in our case) which leads to a sufciently large signal-to-noise ratio in ferromagnetic resonance measurements [24]. Prior to nanowires fabrication, rectangular pieces (width: ∼ 0.6 cm and length: ∼ 4 cm) of Kapton substrates have been cut in order to be used as tensile specimens thereafter. Using the protocol presented in ref. [24], 0.5 × 0.5 cm2 arrays of Ni80 Fe20 nanowires have been fabricated on 125 μm substrates. thick-rectangular (0.6 × 4.0 cm2 ) Kapton Note that no adhesion layer (Cr, Ti, ...) has been deposited on the substrate in order to avoid their inuence on the possible initiation of cracks in the nanowires [25]. The arrays of nanowires were fabricated at the center of

ACS Paragon Plus Environment

®

®

2 5 mm

b)

y 5 mm

x

Tensile specimen

a)

Arrays of nanowires

y x c)

Figure 1: a) Photography of typical tensile specimens. The

®

arrays of nanowires have been fabricated at the center of rectangular Kapton

substrates. The colorations are due to day-

light diraction by nanowires which depends on slight samples misorientations. b) Zoom-in sketch of the arrays showing the direction of the nanowires as compared to the tensile speci-

TM tensile machine

mens. c) Photography of the 300 N-Deben

“cracks + buckling”

tensile axis

Figure 2:

“cracks”

tensile axis

“absence of cracks”

used in this study.

Crack density evolution as a function of the ap-

µ

plied strain for the non-patterned lm. Two typical AFM 2 images (15 × 15 m ) highlighting the two regimes (cracks and cracks+buckling) are inserted in the graph. A zoom-in of the second image shows typical buckling at the frontier of the cracks.

the tensile specimens as shown in gure 1-a). The arrays of nanowires correspond to colored squares at the center the tensile specimens which is due to daylight diraction. Indeed, the nanowires periodicity (width: 400 nm and spacing: 200 nm) is within the visible spectrum. Although not a result in itself, this coloring is a simple way to verify that the nanostructuration has been carried out over the whole region. Finally, as schematically shows by gure 1-b), the nanowires are aligned along the x direction (along the tensile direction). Following, on these tensile specimens, we have measured the mechanical behavior using a tensile machine in situ combined with AFM observations [26]. The tensile

loads were applied to specimens by means of a DebenTM tensile module. This tensile tester is equipped with a 300 N load cell enabling the force measurement with a precision of 0.1 N. The strain rate has been kept constant, equal to 3 × 10−4 s−1 during loading ramps. As shown in gure 1-c), samples are mounted horizontally, clamped to a pair of jaws and supported on stainless steel sliding bearings. A dual threaded lead screw drives the jaws symmetrically in opposite directions, keeping the sample centered in the eld of view. A beforehand characterization of the 20 nm Ni80 Fe20 (non-patterned) thin lm is rst presented. Figure 2 presents the crack density variation as a function of the strain. It has been determined thanks to in situ imaging of the lm surface. Two typical images are presented in this gure as insert. As usual, we observe straight cracks which appear at a critical strain of ∼ 2%. They propagate perpendicular to the stress direction and cross the entire width of the sample. Beyond 2%, we observe a sequential multissuration. The cracks propagate roughly in the middle of each lm fragment formed by the previously propagated cracks. Finally, from 7 − 8%, we observe the appearance of buckles (local decohesion between lm and substrate) with a stop of the cracking process and a saturation of the distance between the cracks [16, 27]. Similar measurements and analysis have been performed on the Ni80 Fe20 array of nanowires. Figure 3 shows typical AFM images of an array of nanowires submitted to strains ranging from 0 to 20%. We recorded AFM images at every 1% of strain. First cracks appear at a critical strain of ∼ 7% as compared to 2% for the lm. We can explain this by the fact that the strain is more concentrated in the areas of the substrate devoid of wires. As in lm case, these cracks are quite straight within the nanowires where they appear. Nevertheless, unlike the lm case, these cracks do not cross the entire sample, but are initially limited to a single nanowire. Then, by increasing the strain, new cracks appear, rather like the lm case. Note that the new cracks appear not far from the rst, this phenomenon is particularly put into evidence at 20% of applied strain (see gure 3). Indeed, these cracks propagate step by step (from nanowire to nanowire). It is clear here that the initiation of the rst cracks played the role of nucleation point for the subsequent cracks. Actually, it is probable that a rst crack generates a strain eld in the substrate which induces a secondary crack in a neighboring wire but often with a small shift compared to the initial one (see the enlarged area in gure 3). Finally, we did not observe the presence of buckling during these tests. Compared to continuous lms, the wires are less subject to compression stresses causing buckling. In fact, the lateral size of the wires is below a subcritical size allowing buckling. Figure 4 shows the mean residual fragment length (= 5 × 103 m at 0%) variation of the nanowires and of the lm as function of the applied strain. As mentioned above, the cracks in the nanowires are initiated at higher strains (∼ 7% instead of ∼ 2%) as it is high-

µ

3

tensile axis

tensile axis

tensile axis

ε=8%

Figure 3: Typical AFM images (25

× 25

ε=12%

µ

ε=20%

−1 m ) highlighting the multissuration of a 20 nm thick Ni80 Fe20 array of nanowires

40 30 20 10 0

Nanowires Continuous film

1st NWs cracks

1st film cracks

Fragment length (µm)

at dierent strain states (8%, 12% and 20%). The images are taken at almost the same area.

5

10

15

20

Strain (%) Figure 4: Residual fragments length as a function of the applied strains of Ni80 Fe20 lm and arrays of nanowires.

lighted in gure 4. We immediately see the interest that this delay can represent from a technological point of view. Indeed, the geometric optimization of 2D arrays of nanostructures could further delay or even avoid cracking for usual strains of stretchable exible systems (several tens of percents). After the initiation of rst cracks, the fragment length decreases quickly and tends to a nal value of ∼ 3.22 m for a strain of 20%. This is relatively close to the value found for continuous thin lm (∼ 2.4 μm for a strain of 13%). Actually, the behavior shown in this gure is asymptotic and it is likely that the fragment length of the wires can reach values less than 3 microns. This is also an important point for the magnetic properties, because given their width (0.4 microns), the shape of the fragments will remain elongated despite the high crack density. FMR measurements have been carried out before and after damaging (strain = 20%). For an array of

µ

nanowires, we expect the presence of a strong initial uniaxial anisotropy (in absence of strains) which is due to the shape anisotropy of these nanowires. The presence of this shape anisotropy is highlighted in gure 5-a) which presents the frequency dependencies of all observable modes as function of the applied magnetic eld along and perpendicular to the nanowires. Indeed, we note the presence of one mode when the applied eld is along the nanowires and at least two modes when the applied eld is perpendicular to the nanowires (see gure 5-c)). As previously discussed in ref. [24], this is due to lateral size reduction that leads to wave-vector quantization of the propagating spin wave. The full lines of gure 5-a) correspond to ts of the rst mode by introducing a shape anisotropy eld Hshape = 270 Oe into the frequency dependence formula which can be written as:  γ 2 f2 = (H cos(ϕM − ϕH ) + Hshape cos 2ϕM ) 2π  H cos(ϕM − ϕH ) + Hshape cos2 ϕ + 4πMs (1) γ = where γ and Ms are the gyromagnetic ratio ( 2π −1 0, 295 GHz.Oe ) and the saturation magnetization (Ms = 720 emu.cm−3 ). ϕM (resp. ϕH ) is the angle between the magnetization (resp. applied magnetic eld) and x-direction. The presence of the shape anisotropy leads to the splitting of the frequency dependence along and perpendicular to the nanowires. Good agreements between experimental data and ts are found. Figure 5b) shows similar frequency dependencies obtained after the application of strains (20%). The continuous lines correspond to the same ts than in gure 5-a). We clearly note that these frequency dependencies are almost unmodied despite the presence of cracks and certainly to residual stresses inside the nanowires. This absence of

Nano Letters

Page 4 of 6

4 15

a)

Frequency (GHz)

Frequency (GHz)

15

10

0%:

5

0° 90° 0

0

500

1000

1500

b)

10

After 20%: 0° 90°

5

0

2000

0

Applied field (Oe) 300

c)

120

After 20%

H

0%

200

500

1000

1500

2000

Applied field (Oe) d)

∆Hpp (Oe)

Intensity (µV)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

90 60

100

H

After 20%

0% After 20% Linear fit

30

0

0% 500

1000

1500

0

0

Applied field (Oe)

5

10

15

Frequency (GHz)

Figure 5: a-b) Frequency dependencies as a function of applied magnetic eld along and perpendicular to the nanowires for a Ni80 Fe20 array of nanowires. The dependencies are presented for two stresses states: at 0% (a) and at 20% (b). c) Typical FMR spectra of the NWs for magnetic eld applied along and perpendicular to the NWs at 0% of strain and after 20% of strain. d) Peak to peak FMR linewidth

∆Hpp

varia-

tions as a function of the frequency obtained from the spectra recorded with a magnetic eld applied along the NWs. The

since it governs the speed at which the magnetization can be reversed or reoriented. We note that the large strains have only a few inuences on the ∆Hpp variation. Indeed, one can see on gure 5-c that resonance spectra are very slightly aected by the high applied strain. This is quite remarkable when one puts these results in relation to the multiple cracks present within the nanowires. Indeed, the low magnetostriction makes this material insensitive from a magnetic point of view to the strains heterogeneities. To conclude, we have shown that the phenomenon of multi-cracking in arrays of Ni80 Fe20 nanowires is delayed in comparison with thin lms of the same thickness. Once this is initiated, the length of residual fragments reaches an asymptotic value of about 3 µm. This value is sucient to maintain a magnetic shape anisotropy comparable to the initial anisotropy of innite nanowires (width: 0.4 µm). Thus, due to the very low magnetostricition coecient of Ni80 Fe20 , the numerous cracks probably associated with signicant heterogeneous residual strains have no noticeable eect on magnetic anisotropy as well as FMR linewidth that is directly related to magnetic damping. This is an important result for applications in exible spintronics because it shows that despite the numerous cracks due to large strains, the array of nanowires retains its initial static and dynamicalproperties.

dashed line is a linear t and serves as a guide for the eyes.

Acknowledgments

inuence may be due to the almost zero magnetostriction coecient of Ni80 Fe20 material [28]. One can deduce that the shape anisotropy of the remaining fragments is equivalent to that of innite nanowires. Finally, we have used the spectra recorded along the nanowires axis to extract the peak to peak FMR linewidth (∆Hpp ) variation as function of the frequency which is of strong importance in spintronic applications

Authors would like to thank Université Sorbonne Paris Cité and the National University of Singapore for their support through the USPC-NUS program (MagnoFlex project, ANR- 11-IDEX-0005-02 and Nanoex project, 2016-04-R/USPC-NUS). DF, FZ and AGS would like to thank Labex SEAM for nancial support (MECAMAG project, ANR 10 LABX 96).

[1] M. Kaltenbrunner, M. S. White, E. D. Gªowacki, T. Seki-

[10] A. Bedoya-Pinto, M. Donolato, M. Gobbi, L.E. Hueso,

tani, T. Someya, N. S. Sariciftci and S. Bauer, Nat. Com-

and P. Vavassori, Applied Physics Letters 104, 062412

mun. 3, 770 (2012)

(2014)

[2] L. Li, J. Liang, S.-Y. Chou, X. Zhu, X. Niu, Z. Yu, Q. Pei, Sci. Rep. 4, 4307 (2014)

[11] M. Gueye, B. M. Wague, F. Zighem, M. Belmeguenai, M. S. Gabor, T. Petrisor Jr, C. Tiusan, S. Mercone, D.

[3] S. Das, R. Gulotty, A. V. Sumant, A. Roelofs, Nano Lett. 14, 2861 (2014)

Faurie, Applied Physics Letters 105, 062409 (2014) [12] D.

[4] Q. Hua, J. Sun, H. Liu, R. Bao, R. Yu, J. Zhai, C. Pan, Z.L. Wang, Nature Communications 9, 244 (2018) [5] S. Wang et al., J. B.-H. Tok and Z. Bao, Nature 555, 83 (2018)

Makarov,

M.

Melzer,

D.

Karnaushenko,

O.

G.

Schmidt, Applied Physics Review 3, 011101 (2016) [13] G. S. C. Bermúdez,

D. D. Karnaushenko,

D. Kar-

naushenko, A. Lebanov, L. Bischo, M. Kaltenbrunner, J. Fassbender, O. G. Schmidt and D. Makarov, Sci. Adv.

[6] G. Lin, D. Makarov, M. Melzer, W. Si, C. Yan and O. G. Schmidt, Lab Chip 14, 4050 (2014)

4: eaao2623 (2018) [14] A. Kleinbichler, M. Bartosik, B. Völker, M.J. Cordill,

[7] S. Kim, H. Y. Jeong, S. K. Kim, S.-Y. Choi, and K. J. Lee, Nano Lett. 11, 5438 (2011)

Adv. Eng. Mater. 19, 1 (2016) [15] J. W. Hutchinson and Z. Suo, Advances in Applied Me-

[8] M. Donolato, C. Tollan , J. M. Porro , A. Berger and P. Vavassori, Adv. Mater. 25, 623 (2013)

chanics 29, 63 (1991) [16] M. J. Cordill, F.D. Fischer , F.G. Rammerstorfer, G.

[9] H. Li, Q. Zhan, Y. Liu, L. Liu, H. Yang, Z. Zuo, T. Shang, B. Wang, and R.-W. Li, ACS Nano 10, 4403 (2016)

Dehm, Acta Mater. 58, 5520 (2010) [17] H.-J. Kim, M.-W. Moon, D.-Ik Kim, K.-R. Leeb and K.

ACS Paragon Plus Environment

Page 5 of 6

Nano Letters 5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

H. Oh, Scr. Mater. 57, 1016 (2007)

Chong, J. Alloys. Compd. 449, 261 (2008)

[18] J. Marthelot, B. Roman, J. Bico, J. Teisseire, D. Dalmas and F. Melo, Phys. Rev. Lett. 113, 085502 (2014) 540 (2013) Melzer,

D. Faurie,

M. Belmeguenai,

A. Garcia-

Sanchez, P. Lupo and A. O. Adeyeye, Appl. Phys. Lett.

[19] N. Lambricht, T. Pardoen and S. Yunus, Acta Mater. 61, [20] M.

[24] F. Zighem,

111, 052408 (2017) [25] S. Olliges, P. A. Gruber, S. Orso, V. Auzelyte, Y. Ekinci,

D.

Makarov,

A.

Calvimontes,

D.

Kar-

naushenko, S. Baunack, R. Kaltofen,Y. Mei and O. G. Schmidt, Nano Lett. 11, 2522 (2011)

H. H. Solak, R. Spolenak, Scr. Mater. 58, 175 (2008) [26] D. Faurie, F. Zighem, A. Garcia-Sanchez, P. Lupo, A.O. Adeyeye, Appl. Phys. Lett. 110 (9), 091904 (2017)

[21] M. Melzer, G. Lin , D. Makarov and O. G. Schmidt, Adv. Mater. 24, 6468 (2012)

[27] S. Frank, U.A. Handge, S. Olliges, R. Spolenak, Acta Mater. 57, 1442 (2009)

[22] V. V. Kruglyak, S. O. Demokritov and D. Grundler, J. Phys. D: Appl. Phys. 43, 264001 (2010)

[28] M. Gueye, P. Lupo, F. Zighem, D. Faurie, M. Belmeguenai, A. O. Adeyeye, EPL 114, 17003 (2016)

[23] Q. Xie, M.H. Hong, H.L. Tan, G.X. Chen, L.P. Shi, T.C.

ACS Paragon Plus Environment

Applied field (Oe)

1 2 3 4

Page 6After 20% of 6

H

0%

200

Intensity (µV)

Nano Letters 300

100

H

After 20%

ACS Paragon Plus Environment 0 0%

ε=20%

500

1000

Applied field (Oe)

1500