masters thesis help

Colloquium: Opportunities in nanomagnetism

S. D. Bader Materials Science Division and Center for Nanoscale Materials, Argonne National Laboratory, Argonne, Illinois 60439, USA

!Published 3 January 2006"

Nanomagnetism is the discipline dealing with magnetic phenomena specific to structures having dimensions in the submicron range. This Colloquium addresses the challenges and scientific problems in this emerging area, including its fabrication strategies, and describes experiments that explore new spin-related behaviors in metallic systems as well as theoretical efforts to understand the observed phenomena. As a subfield of nanoscience, nanomagnetism shares many of the same basic organizing principles such as geometric confinement, physical proximity, and chemical self-organization. These principles are illustrated by means of several examples drawn from the quests for ultrastrong permanent magnets, ultra-high-density magnetic recording media, and nanobiomagnetic sensing strategies. As a final example showing the synergetic relationships to other fields of science, this Colloquium discusses the manipulation of viruses to fabricate magnetic nanoparticles.

DOI: 10.1103/RevModPhys.78.1 PACS number!s": 75.75.!a, 61.46."w

CONTENTS

I. Introduction 1 A. History 1 B. Challenges 1 C. Fabrication approaches 3

II. Specific Examples 3 A. Spin injection 3 B. Magnetic vortices 5 C. Bottom-up strategies 8

1. Use of spring magnets 8 2. Hierarchical assembly 10 3. Use of magnetic viruses 12

III. Summary and Conclusions 13 Acknowledgments 13 References 13

I. INTRODUCTION

A. History

Magnetism is one of the oldest scientific disciplines, but one also at the forefront of the emerging nanotech- nology era. Figure 1 illustrates this interplay of the past and the present. On the left is a reproduction of a wood- cut illustration from the book De Magnete written by William Gilbert and published in Latin in 1600. This book is regarded as the first to embrace the scientific method of inquiry. William Gilbert was a gentleman sci- entist whose day job was as a medical doctor, eventually as the personal physician to Queen Elizabeth I !Chap- man, 1944". Juxtaposed to the right in Fig. 1 is a modern simulation of a magnetic vortex structure. The two im- ages share the similarity of being circles that contain arrows. So what has changed in the 400+ years that have elapsed? The crucial difference is that Gilbert was con- cerned with geomagnetism; his image represents the planet Earth. The modern vortex structure to the right is

for a submicron ferromagnetic dot. The difference in di- ameters is upwards of 14 orders of magnitude. Herein lies the advance: in the present work, we will be con- cerned with the opportunities associated with the nanoscale.

B. Challenges

The scientific quest in nanomagnetism can be framed quite succinctly. The goal is to !i" create, !ii" explore, and !iii" understand new nanomagnetic materials and phe- nomena. Throughout this paper, examples will be pre- sented to highlight these three components of an inte- grated approach to nanomagnetism research. We create new materials via synthesis and fabrication routes. We explore them using major facilities such as synchrotron light sources, neutron scattering, and electron mi- croscopies, as well as via bench-top and conventional

FIG. 1. !Color" Two pictures of circles with arrows inside: !a" from De Magnete, published in 1600 by William Gilbert. The image is a woodcut representation of the Earth, including a mountainous topology, in an effort to understand geomag- netism. !b" A generic modern micromagnetic simulation of a submicrometer magnetic vortex structure in Permalloy, such as is discussed by Novosad et al. !2002". It includes a central core where the spins point out of plane.

REVIEWS OF MODERN PHYSICS, VOLUME 78, JANUARY 2006

0034-6861/2006/78!1"/1!15"/$50.00 ©2006 The American Physical Society1

laboratory facilities such as scanning probes, magnetom- etry, magneto-optics, and magnetotransport probes. To understand the materials we create, and the phenomena they exhibit, we rely on theory and simulation.

Figure 2 provides a sunburst representation of grand challenges in the field based on today’s perspective. The important feature to keep in mind is that, while the chal- lenges will be framed in terms of their relationships to societal benefits, the underlying basic science involves creating new materials and exploring and understanding issues in spin dynamics and spin transport. Starting at the top, we will go in clockwise fashion around the sun- burst.

Ultrastrong Permanent Magnets. The first bubble is the quest for ultrastrong permanent magnets. This quest addresses the national need for energy efficiency, energy conservation, and energy self-sufficiency. The energy is- sue could very well be the most important challenge the world faces. Its implications are broad and pervasive. Nanoscience offers the possibility to create and deliver ultrastrong permanent magnets that could, for example, lead to more efficient and compact motors. Lighter weight motors could save fossil fuels in auto and air transportation, since motors are ubiquitous in these sys- tems. Also, electric motors for hybrid automotive ve- hicles might become a major market soon.

Ultra-High-Density Media. Next, ultra-high-density media for magnetic recording is needed to reach the na- tion’s goals for information storage. Today’s media stores almost 100 Gbits / in.2. In order to advance to Tbits / in.2 and beyond, new approaches are required, and nanomagnetism might provide what is needed. The field is presently at a crossroads, and will soon no longer be able to incrementally improve on a technology origi- nally introduced in 1956. Thermal stability of small

structures, as covered by Weller and Moser !1999", is one of the major issues. The design and scaling of sensi- tive read head sensors, based on the giant magnetoresis- tive effect, pose other major issues !Parkin et al., 2003".

Spin Transistor with Gain. While the hard drive in one’s laptop computer is a well-known component, the future might include additional magnetic subsystems. The basis for today’s electronics is the semiconductor transistor #see Bardeen and Brattain !1948"$, perfected over the years, which has replaced vacuum-tube technol- ogy. Semiconductor electronics utilize the charge of the electron flowing in its circuitry, but the electron also has a spin, the basis for magnetism. A spin transistor could add value to present-day electronics, as highlighted by Wolf et al. !2001", and as we will see in examples that follow. The challenge is not only to create a spin transis- tor, but also to realize one with gain. Creating a spin gain transistor that not only utilizes but amplifies the spin signal has presented so far insurmountable prob- lems, as has been discussed by Nikonov and Bourianoff !2005", who propose a possible scheme. The fundamen- tals and applications within the field of magnetic elec- tronics, or spintronics, have been reviewed by Zutic et al. !2004".

Nearly 100% Spin-Polarized Materials. To create a cir- cuit where the flow of spins takes place, rather than, or in addition to, the flow of charge, requires nearly 100% spin-polarized materials. These can be found in exotic half-metallic ferromagnets, such as exist in some com- plex oxides and Heusler alloys !de Groot et al., 1983". Half-metallic refers to the Fermi level intersecting the density of states within a gap for one spin subband but not the other.

Room-Temperature Magnetic Semiconductors. A re- lated quest is for magnetic semiconductors, needed in order to interface the new functionalities of magnetic electronics to mainstream semiconductor circuitry !Ohno, 1998". Such magnetic semiconductors need to have their ferromagnetic ordering temperatures, or Cu- rie temperatures, well above room temperature in order to integrate with existing technologies.

Instant Boot-Up Computer. One new spintronic device is the magnetic random access memory !MRAM" chip presently under commercial development !Tehrani et al., 2003". MRAM with enough memory capacity could eventually enable the advent of the instant boot-up com- puter. Today’s laptops have semiconductor charge-based RAM that is volatile, losing its stored-up charge, and hence its stored information, when the power needed for its periodic refresh cycles is removed. MRAM is non- volatile. Once one of its memory elements is magne- tized, it retains its magnetization direction !and there- fore its binary coding" without the need for additional power consumption.

Magnetic Logic. Another new spintronic device con- cept also potentially useful in computer technologies would be magnetic logic, based on spin transistor archi- tectures, as in the proposals of Sugahara and Tanaka !2004" and Matsuno et al. !2004". The added value would be, for example, to create a magnetic central processor

FIG. 2. !Color online" Grand challenges in nanomagnetism, emphasizing those related to strategic national goals, such as stimulating the economy, energy efficiency, homeland security, and defense. The underlying basic research needed involves creating new materials and understanding their spin-transport and spin dynamic properties.

2 S. D. Bader: Colloquium: Opportunities in nanomagnetism

Rev. Mod. Phys., Vol. 78, No. 1, January 2006

unit !MCPU" that would have a soft architecture. It could be reconfigured to match the task at hand at any instant in time. It might sound far-fetched, but we are examining grand challenges, so we can think !and dream" outside the box. Today’s high-performance com- puters utilize parallel processors. If each processor were also dynamically optimized, performance would be ad- ditionally enhanced. Why optimize only for graphics or for number crunching if you can achieve each goal when needed? Initial steps toward the exploration of magnetic quantum cellular automata at room temperature have been demonstrated by Cowburn and Welland !2000", and programmable logic gates using giant magnetoresis- tance devices are discussed by Hassoun et al. !1997".

Spin-Based Qubits. Since we are moving away from mainstream approaches, we can proceed to the next bubble of spin-based qubits. Here we are driven by the quest to utilize electronic or nuclear spin systems to implement quantum-computing paradigms and to tran- scend binary logic completely. This would, for example, permit certain classes of problems that are computation- ally off limits at present to become tractable. DiVin- cenzo and Loss !1999" discuss quantum computers and quantum coherence. Proposals that focus on spin chains and molecular magnets include the work of Leuenberger and Loss !1996" and Meier et al. !2003". Experimental systems of interest include single-molecule magnets that exhibit quantum tunneling of the magnetization, such as is highlighted in the work of Wernsdorfer et al. !2002".

Hierarchically Assembled Media. Next, we must find new ways to hierarchically assemble systems, such as magnetic recording media. Good ideas about new phe- nomena are important to develop, but we have to be able to implement them as well. In the quest to address society’s issues !such as are associated with energy, trans- portation, stimulating the economy, enabling new ap- proaches to healthcare, homeland security, defense, etc.", we must create new pathways to solutions such that they can be manufactured cost effectively. This means that we need to embrace the goal of creating en- tire hierarchical subsystems and systems instead of merely assembling individual materials and components. We can ponder that ultimately the goal could be, for example, to pull a fully assembled computer out of a test tube.

Nanobiomagnetic Sensors. As we consider the level of self-organization needed to achieve such a lofty, futuris- tic goal as just proposed, we look to the world of biology and biomimetic approaches to assembly. The intersec- tion of biology and nanomagnetism has an immediate goal of giving rise to new concepts in sensing, such as are needed for homeland security and medical applications. A recent example can be found in the work of Hoff- mann et al. !2005". Applications of magnetic nanopar- ticles to biomedicine have been reviewed by Pankhurst et al. !2003". This has taken us full circle, emphasizing a bridging of disciplines, and motivated by a quest to serve strategic national needs and goals. Now we highlight some examples limited as they might be. The examples

will illustrate approaches to create, explore, and under- stand nanomagnetic systems.

C. Fabrication approaches

Nanoscience offers three major means to create new nanomagnetic materials, as reviewed by Martín et al. !2003". These are !i" top-down, !ii" bottom-up, and !iii" virtual fab. Top-down has been the traditional approach to miniaturization by sculpting via lithographic tools. But lithography needs enhancements to get to smaller length scales and to do so cost effectively. This is where the bottom-up approach enters centerstage. Bottom-up is self-assembly from molecular-precursor building blocks, as described by Whitesides and Grzybowski !2002". As such it encompasses the domains of chemistry and biology. However, self-assembly is not limited to these domains or length scales, because as we know on a majestic length scale, the entire Universe has self- organized into granular patterns of galaxies. Finally, we consider a third approach: virtual fab. This is the path taken by theorists and computationalists who create their new materials and properties in a computer simu- lation or analytically. For example, the virtual approach offers the best control of feature definition because the system is precisely specified at the outset. But the main advantage of virtual fab is to have the ability to obtain a fundamental understanding and to elucidate guiding principles. One can frame this, for example, as a quest to understand the rules that govern self-assembly. We know that the answer encompasses both thermodynam- ics and kinetics, and, hence, electronic structure. The main challenge is to seamlessly weave together almost every known computational tool and approach that ad- dresses multiple time and length scales into an as yet unrealized hierarchical computation package.

II. SPECIFIC EXAMPLES

A. Spin injection

This example concerns the all-metallic lateral spin valve. Although magnetic semiconductors have many virtues, all-metallic spin-transport systems offer their own opportunity to explore elementary spin-transport concepts of injection, diffusion, and detection.

Figure 3 illustrates the example under consideration. Metallic spin injection in a lateral spin-valve prototype structure was first explored in the pioneering work of Johnson and Silsbee !1985", and more recently by groups in Groningen !Jedema et al., 2001, 2002", Japan !Kimura et al., 2004", and the United States !Valenzuela and Tinkham, 2004". Figure 3 is taken from Ji et al. !2004". It shows a spin valve having two ferromagnetic electrodes separated by a nonferromagnetic metallic spacer. The magnetic electrodes are patterned via electron-beam li- thography from Permalloy !a Ni-Fe alloy", while the spacer is a 200-nm-wide gold stripe. Attached to these three elements in a rather nontraditional manner are current and voltage leads. In a traditional electronic cir-

3S. D. Bader: Colloquium: Opportunities in nanomagnetism

Rev. Mod. Phys., Vol. 78, No. 1, January 2006

Contents lists available at ScienceDirect

Journal of Magnetism and Magnetic Materials

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

Shape anisotropy and hybridization enhanced magnetization in nanowires of Fe/MgO/Fe encapsulated in carbon nanotubes

Dennis Aryeea,b, Dereje Seifub, ⁎

a Army Research Laboratory, Weapons and Materials Research Directorate, Aberdeen Proving Ground, MD 21005, USA b Department of Physics and Engineering Physics, Morgan State University, Baltimore, MD 21251, USA

A R T I C L E I N F O

Keywords: Nanowires Fe Magnetoresistance Anisotropy Sputtering

A B S T R A C T

Arrays of tunneling magnetoresistance (TMR) nanowires were synthesized for the first time by filling Fe/MgO/ Fe inside vertically grown and substrate supported carbon nanotubes. The magnetic properties of nanowires and planar nanoscale thin films of Fe/MgO/Fe showed several similarities, such as two-fold magnetic symmetry and ratio of orbital moment to spin moment. Nanowires of Fe/MgO/Fe showed higher saturation magnetization by a factor of 2.7 compared to planar thin films of Fe/MgO/Fe at 1.5 kOe. The enhanced magnetic properties likely resulted from shape anisotropy of the nanowires and as well as the hybridization that occur between the π- electronic states of carbon and 3d-bands of the Fe-surface.

1. Introduction

Tunneling magnetoresistance (TMR) is a macroscopic quantum phenomenon that allows electrical current to flow across an insulator with the application of an external magnetic field. It has attracted enormous attention because of important applications in non-volatile magnetoresistive random access memories (MRAM) and high-sensi- tivity field sensors [1]. Fe/MgO/Fe TMR based magnetic tunnel junctions (MTJs) are of great interest as witnessed by many reports on structural, magnetic, and electronic properties of the multilayer interfaces [2–7]. First principle modeling of TMR Fe/MgO/Fe tri- layers predicted zero bias magnetoresistance (MR) ratio several thousand percent on the basis of model structures involving abrupt Fe-MgO interfaces [3]. However, the MR ratio reduced to 1000% when the Fe/MgO interface oxidation is considered [2]. Theoretical calcula- tion of TMR with a disordered Fe/MgO/Fe junction showed the intermixing of Fe and Mg atoms at the interface decreases the MR ratio rapidly and when about 16% of interfacial Fe atoms are substituted by Mg the calculated MR saturates with increasing MgO thickness in good agreement with experiment [4]. The atomic moments at the interface are non-collinear with the bulk magnetization which may affect the net anisotropy or serve as spin scattering sites [5]. An MR ratio of 220% was achieved in MTJs with a crystalline MgO(001) barrier [7]. An MR ratio of 180% at room temperature and 247% at 20 K were also observed in single-crystal Fe/MgO/Fe MTJs [6]. The origin of this enormous TMR effect is coherent spin-polarized tunnel- ing, where the symmetry of electron wave function plays an important

role. Coherent TMR effect is key to making spintronic devices. Changing the geometry from planar tri-layer nanometric thin film to nanowire cylindrical geometry with nanometric diameter introduces shape anisotropy which can play an important role in coherence [8,9]. This report examine the similarities and differences in the magnetic properties of the two geometries, nanowires with cylindrical geometry and planar thin films, of Fe/MgO/Fe to shed light in understanding the effect of shape in magnetic properties of this ferromagnet/insulator/ ferromagnet tri-layer structure.

2. Experiment

Vertically aligned multi-wall carbon nanotubes (CNTs) were grown using a thermal chemical vapor deposition method [10] on SiO2 substrate. This method involves exposing silica structures to a mixture of ferrocene and xylene at 770 °C for 10 min. The furnace is pumped down to ≈200 mTorr in argon bleed and then heated to a temperature of 770 °C. The solution of ferrocene dissolved in xylene (≈0.01 g/ml) is pre-heated in a bubbler to 175 °C and then passed through the tube furnace. After returning to room temperature, the CNT-covered SiO2 substrate is transferred to an AXXIS sputtering tool (Kurt J. Lesker Company). The open ended CNTs tips were filled with Fe/MgO/Fe using DC magnetron and RF sputtering method at a substrate temperature of 100 °C. In planar Fe/MgO/Fe films, this substrate temperature yielded the highest value of coercive field compared to several other synthesis substrate temperatures as shown in Fig. 3. By using the combination of DC and RF sputtering, we were able to fill the

http://dx.doi.org/10.1016/j.jmmm.2017.01.014 Received 1 December 2016; Accepted 5 January 2017

⁎ Corresponding author. E-mail address: [email protected] (D. Seifu).

Journal of Magnetism and Magnetic Materials 429 (2017) 161–165

Available online 08 January 2017 0304-8853/ © 2017 Elsevier B.V. All rights reserved.

MARK

lumens (inner cylindrical volume) of carbon nanotubes. Planar thin films of Fe/MgO/Fe were epitaxially grown on MgO

(100) substrate of dimensions (5 mm X 5 mm X 0.5 mm) using magnetron DC and RF sputtering at several temperatures. Prior to deposition, the substrates were degassed at 350 °C in vacuum of 0.1 μTorr for 30 min. The source substrate distance was kept fixed at 30 cm and the substrate surface normal was kept at 45° with a line connecting the center of the sample to the center of the target, while being rotated at a constant rate of 20 rpm for uniform deposition. The deposition rate for Fe was 0.17 nm/s as calibrated by the deposition time versus thickness measurements for Fe films several hundred nm thick. With these conditions, epitaxial Fe grows on MgO (100) due to a good lattice match of MgO and Fe interaction [11,12]. Five planar samples of Fe/ MgO/Fe were synthesized at several substrate temperatures RT, 50 °C, 100 °C, 200 °C and 300 °C. All samples were annealed post-growth for 30 min under the same vacuum conditions.

3. Structural and spectroscopic characterization

Surface morphologies of nanowires of Fe/MgO/Fe grown in the lumens of CNTs were characterized by a Hitachi S-5500 field emission SEM/STEM. For STEM imaging, some nanotubes were scraped off SiO2 substrate and dispersed in dimethylformamide, the resulting solution was dripped on holey carbon coated TEM grid. Fig. 1(b) is a STEM micrograph of nanowires of Fe/MgO/Fe depicts a uniform composition.

X-ray absorption spectroscopy (XAS) measurements on thin films and nanowires of Fe/MgO/Fe were carried out at beamline 4UB at National Synchrotron Light Source (NSLS) in Brookhaven National Laboratory (BNL). XAS of thin film and nanowires synthesized at 100 °C is shown in Fig. 2(a). The XAS spectra reveals at 717 eV the existence of a missing shoulder at Fe L3 edge in the nanowires spectrum. At the Fe L2 edge the double peaks at 723.5 eV are of equal height for the nano-wires and a leading peak in the film spectra as shown in the inset in Fig. 2(a).

4. Results from magnetic measurements

4.1. Results on Fe/MgO/Fe nanowires

Magnetic force microscopy (MFM) was carried out using TT-AFM- MFM (AFM-Workshop Corporation), Fig. 1(a). Nanowires of Fe/MgO/ Fe depicts a uniform composition and Fig. 1(a) is MFM of nanowires of Fe/MgO/Fe. The MFM scans revealed the presence of stripe domains in the nanowire samples.

X-ray magnetic circular dichroism (XMCD) measurements of nanowires of Fe/MgO/Fe were carried out at beamline U4B at NSLS

in BNL, Fig. 2(b). XMCD of nanowires synthesized at 100 °C is shown in Fig. 2(b). Background corrected XMCD signal in Fig. 2(b) shows that the nano-wires XMCD signal at the Fe L3 and L2 edges is larger. Moreover, a switching between the two edges occur. At around 712 eV photon energy the intensity of the nanowires XMCD signal is smaller.

Vibrating Sample Magnetometer (VSM) measurements were car- ried out using Vector Magnetometer Model 10 VSM system (MicroSense) equipped with 3 T electromagnet, Fig. 3. The figure shows the coercive field of each sample (left axis) as well as the saturation magnetization (Ms) and remanent magnetization (Mr) (right axis). The solid lines connect the measurements of the Fe/MgO/Fe planar films deposited at different substrate temperatures. Data from the nanowires of Fe/MgO/Fe inside the CNTs are shown by an asterisk (‘*’).

Fig. 4(a) shows the hysteresis loops when H is applied parallel to the substrate plane (in-plane configuration) that is at 0°, at 45°, and perpendicular to the substrate, that is at 90°, to the sample surface (out-of-plane configuration) of nanowires of Fe/MgO/Fe synthesized at 100 °C. Note that the applied magnetic field is perpendicular to the nanowires for the in-plane configuration and parallel to the nanowires for the out-of-plane geometry.

Magnetic torque measurements were carried out using EV7 torque magnetometer (TMM) system equipped with a 2 T electromagnet (MicroSense), Fig. 5(a). Torque magnetometer measurements of nanowires of Fe/MgO/Fe synthesized at 100 °C for several applied fields is shown in Fig. 5(a). Fig. 5(d) depicts coercive field measured using VSM at several angles with respect to an applied field for nanowires of Fe/MgO/Fe synthesized at 100 °C.

4.2. Results on Fe/MgO/Fe thin films

XMCD measurements of thin films of Fe/MgO/Fe were carried out at beamline U4B at National Synchrotron Light Source (NSLS) in Brookhaven National Laboratory (BNL), Fig. 2(b). The XMCD shown in Fig. 2(b) is for a thin film of Fe/MgO/Fe synthesized at 100 °C.

Fig. 4 (b) shows hysteresis loops when H is applied parallel (in- plane), at 45°, and perpendicular to the sample surface (out-of-plane) of planar nanometric thin films and nanowires of Fe/MgO/Fe both synthesized at 100 °C.

TMM measurements of planar nanometric thin films of Fe/MgO/Fe synthesized at 100 °C for several applied fields is shown in Fig. 5(b). Magneto-optic Kerr effect (MOKE) measurements were carried out with an in-house built MOKE instrument in longitudinal symmetry, Fig. 5(c). Longitudinal MOKE measurements of planar thin film samples synthesized at several substrate temperatures is depicted in Fig. 5(c). External magnetic field applied along equivalent crystal- lographic directions did not produce equivalent hysteresis loops. This is

Fig. 1. (a) MFM of nanowires of Fe/MgO/Fe synthesized at 100 °C. (b) STEM image of nanowires of Fe/MgO/Fe synthesized at 100 °C.

D. Aryee, D. Seifu Journal of Magnetism and Magnetic Materials 429 (2017) 161–165

162

obvious when comparing Mr/Ms ratio obtained for symmetric orienta- tions, every 10° from 0° to 360°. Fig. 5(d) depicts coercive field measured using VSM at several angles with respect to an applied field for thin films of Fe/MgO/Fe synthesized at 100 °C. The coercive field of thin film showed strong dependence on angle than nanowires.

4.3. Comparison of magnetic properties of nanowires and thin films of Fe/MgO/Fe

The in-plane value of coercive field (Hc) of nanowires of Fe/MgO/ Fe is 54% higher than thin film of Fe/MgO/Fe both synthesized at a substrate temperature of 100 °C surprisingly, the in-plane saturation magnetization (Ms) of nanowires is 173% higher than thin film. The in- plane and out- of-plane values of Ms for nanowires is 25.5% whereas for films it is 76.5%. As the orientation of magnetic field with respect to the sample surface changes the variation in films is three times greater than in nanowires.

5. Discussion

MFM of nanowires of Fe/MgO/Fe is shown in Fig. 1(a). The stipe domains shown in the MFM image of Fe/MgO/Fe nanowires in Fig. 1(a) indicates the perpendicular anisotropy responsible for the formation of stripe domains results from strain magnetostriction and microstructure shape effect in the nanowires. As stripe domains were not observed in the planar films (MFM not shown here), this indicates that the shape anisotropy in nanowires alters the magnetic properties of the Fe/MgO/Fe system. Interestingly, the STEM micrograph of the nanowires of Fe/MgO/Fe shown in Fig. 1(b) depicts a uniform composition.

In the XAS spectra of Fig. 2(a) the leading peak in the film?s spectra

Fig. 2. (a) XAS of thin film and nanowires synthesized at 100 °C. The inset show a shoulder at Fe L3 edge and a leading peak at Fe L2 edge in the film spectrum. The lines are drawn connecting the points as a guide. (b) XMCD of thin film and nanowires synthesized at 100 °C. The lines are drawn connecting the points as a guide.

Fig. 3. VSM measurements of Hc, Ms, and Mr of thin films of Fe/MgO/Fe synthesized at several substrate temperatures. The lines are drawn connecting the points as a guide. (’*’) shows values for nanowires synthesized at substrate temperature of 100 °C. The lines are drawn connecting the points as a guide.

Fig. 4. (a) B-H loop of nanowires synthesized at 100 °C at several angles with the applied field. (b) B-H loop of thin film synthesized at 100 °C at several angles with the applied field.

D. Aryee, D. Seifu Journal of Magnetism and Magnetic Materials 429 (2017) 161–165

163

shown in the inset indicates Fe ions are not bonded with oxygen or any other atoms in the bulk indicating the metallic nature of Fe-layers. The Fe-layers on top and below the MgO layer are metallic, and we speculate that the Fe at the interface with the MgO may be partially oxidized while the bulk of the Fe remains metallic. The interface oxidation also causes spin moments per Fe atoms to be less than bulk Fe while anomalously large spin moment (L) is observed [13]. Oxidation effect at the interfaces of Fe and MgO is similar in both films and nanowires as observed in the XAS spectra both spectra possess a leading peak at 723.5 eV.

The well-defined XMCD spectra shown in Fig. 2(b) indicates the Fe film is ferromagnetic. Previous predictions of magnetically dead layers were based on the study of Fe films that were grown at room temperature [14]. Using sum rules spin moment ms is 1.62 μB for the film and 1.22 μB for the nanowires, and the spin moments of both the planar films and the nanotubes is smaller than that of bulk Fe (1.98 μB) [15]. The orbital moment mo is 0.056 μB for the film and 0.44 μB for the nanowires from XAS and XMCD spectra with applica- tions of the Eqs. (1) and (2) [15]. The ratios mo/ms in the in-plane direction of the present films are 3.5% for the film and 3.6% for the nanowires are similar to that observed for bulk Fe (4%) [15].

m n q r

= − (10 − ) 4 3o d3 (1)

m n p q r

T S

= − 6(10 − ) 6 − 4 (1 + 7 < > 2 < >

)s d z

z 3

−1 (2)

In the above sum rule q is integral over L3+L2 XMCD, r is integral over L3+L2 XAS, p is integral over the L3 edge of the XMCD spectrum, n3d is number of electrons in the Fe 3d orbital and we assume a value of 6.61 [15], T< >z is expectation value of the magnetic dipole operator, and T S< > / < >z z is taken −0.38% from 1st principle calculations [16].

At around 711 eV photon energy the intensity of nanowires XMCD signal is smaller and at 712 eV switching between the two edges occurred.

The XMCD signal of the nanowires at the Fe L3 and L2 edges is larger. Turning to the bulk magnetic properties shown in Fig. 3, the data

indicate that Hc, Ms, and Mr are higher in the nanowire samples than the Fe/Mgo/Fe samples. As shown in Fig. 3 the coercive field (Hc) is maximum at substrate synthesis temperature of 100 °C for the planar samples. The values of Hc, Mr, and Ms of the nanowires are higher than planar films, Hc by 37%, Mr by 55%, and Ms by 63%. We attribute these higher values to the shape anisotropy of the nanowires. According to DFT calculation magneto crystalline anisotropy is predicted to be caused by the change in the relative occupancy of the 3d-orbitals of Fe atoms at the Fe/MgO interface [17]. This higher magnetic property can also be attributed to the significant hybridization that may occur between the π-electronic states of carbon and 3d-bands of the Fe- surface since the trilayer Fe/MgO/Fe/MWCNTs nanowires are synthe- sized encapsulated inside carbon nanotubes. This work is useful since hybrid interface between ferromagnetic surfaces and carbon-based molecules play an important role in organic spintronics [18].

As shown in Fig. 4(a) and (b) the hysteresis loops of planar film of Fe/MgO/Fe show higher dependence on the angle between the field direction and the normal to the surface. A more pronounced depen- dence is seen in the films than nanowires samples. The hysteresis loops depend on the angle between lattice orientation and the applied magnetic field. The variation is more pronounced for the film as shown in Fig. 4(b) than for the nanowires shown in Fig. 4(a). The variation of the hysteresis loop in Fig. 4(b) for film implies that the easy axis lies in plane and the hard axis is perpendicular to the surface. The hysteresis loop width does not show significant change when the external magnetic field angle to the axis change, Fig. 4(a). This is due to the dense compaction (possibly close to a hexagonal close packing) of the nanowires in this study, and the presence of strong dipolar interactions among the nanowires. Previous results found in α-Fe nanowires fabricated with alumina templates and single nanowires inside dense nickel nanowire arrays [19,20], in which the dipole interactions are small due to a large separation between the nanowires, indicate that

Fig. 5. (a) TMM measurements of nanowires of Fe/MgO/Fe at several fields. The lines are drawn connecting the points as a guide. (b) TMM measurements of thin films Fe/MgO/Fe at several fields. (c) MOKE measurements of thin films at several synthesis temperatures. The lines are drawn connecting the points as a guide. (d) Coercive field measured using VSM at several angles with respect to the applied field for thin films (black) and nano wires (red) of Fe/MgO/Fe synthesized at 100 °C. The lines are drawn connecting the points as a guide.

D. Aryee, D. Seifu Journal of Magnetism and Magnetic Materials 429 (2017) 161–165

164