Deconvolution of mixed magnetism in multilayer grapheneAkshaya Kumar Swain and Dhirendra Bahadur Citation: Applied Physics Letters 104, 242413 (2014); doi: 10.1063/1.4884426 View online: http://dx.doi.org/10.1063/1.4884426 View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/104/24?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Multilayer nanogranular films (Co40Fe40B20)50(SiO2)50/α-Si:H and (Co40Fe40B20)50(SiO2)50/SiO2: Magneticproperties J. Appl. Phys. 113, 17C105 (2013); 10.1063/1.4794361 Magnetization reversal and spintronics of Ni/Graphene/Co induced by doped graphene Appl. Phys. Lett. 102, 112403 (2013); 10.1063/1.4795764 Quantifying interlayer exchange coupling via layer-resolved hysteresis loops in antiferromagnetically coupledmanganite/nickelate superlattices Appl. Phys. Lett. 95, 102504 (2009); 10.1063/1.3222944 Competition between interlayer exchange and Zeeman energies on the way to saturation of magnetization inFe/Cr multilayers J. Appl. Phys. 105, 013920 (2009); 10.1063/1.3057512 Magnetic patterning of exchange-coupled multilayers Appl. Phys. Lett. 84, 2853 (2004); 10.1063/1.1699475

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

137.30.242.61 On: Mon, 08 Dec 2014 21:33:23

Deconvolution of mixed magnetism in multilayer graphene

Akshaya Kumar Swain1 and Dhirendra Bahadur2,a)

1IITB-Monash Research Academy, Department of Metallurgical Engineering and Materials Science,IIT Bombay, Mumbai 400076, India2Department of Metallurgical Engineering and Materials Science, IIT Bombay, Mumbai 400076, India

(Received 11 March 2014; accepted 9 June 2014; published online 19 June 2014)

Magnetic properties of graphite modified at the edges by KCl and exfoliated graphite in the form of

twisted multilayered graphene (<4 layers) are analyzed to understand the evolution of magnetic

behavior in the absence of any magnetic impurities. The mixed magnetism in multilayer graphene

is deconvoluted using Low field-high field hysteresis loops at different temperatures. In addition to

temperature and the applied magnetic field, the density of edge state spins and the interaction

between them decides the nature of the magnetic state. By virtue of magnetometry and electron

spin resonance studies, we demonstrate that ferromagnetism is intrinsic and is due to the interac-

tions among various paramagnetic centers. The strength of these magnetic correlations can be con-

trolled by modifying the structure. VC 2014 AIP Publishing LLC.

[http://dx.doi.org/10.1063/1.4884426]

A weak spin-orbit coupling, a longer spin-lattice relaxa-

tion time, high electron mobility, and the 2-dimensional

geometry of graphene along with the ease to induce para-

magnetic centers in graphene have made it a potential mate-

rial for spintronics applications.1,2 The carrier density can

also be tuned accordingly with a desired bias voltage, by

which its electronic, magnetic, and optical properties are

modified.3–5 Thus, a deep understanding of the magnetism of

graphene has become necessary for further progress in this

field. The ferromagnetism of graphene in the absence of any

d or f-shell impurities is very intriguing and is a subject of

debate.6–9 There have been several reports on graphene exhib-

iting ferromagnetic (FM), antiferromagnetic (AFM), paramag-

netic, or diamagnetic behavior.10–16 In the absence of any

magnetic impurities, the dominating factor that decides its

magnetic state is the density of edge states/paramagnetic cen-

ters and any interaction between them. In order to observe

these interesting phenomena, several techniques have been

designed by experimentalists to examine these propositions.17

In this report, we show the effect of edge modification on the

magnetic states of exfoliated graphite (EG) which mainly

consists of twisted bi and tri-layered graphene. The different

magnetic contributions in graphene have been separated using

low field (LF)-high field (HF) hysteresis loops so as to empha-

size the effect of interactions that induce FM behavior in

it. We speculate a defect mediated mechanism to be suitable

to explain the observed FM behavior of EG where the

p-electrons (conduction carriers) interact with edge-state

spins. We believe that the current study would add more

insight to the origin of mixed magnetism and its deconvolu-

tion, in graphitic samples. The material chosen for this study

is a potential candidate for various electronic (spintronics) de-

vice applications due to its properties.18 Various composites

based on twisted graphene can easily be prepared due to the

presence of KCl in graphite-KCl compound (GKC) which

will be reported in near future.

Natural graphite powder (size< 45 lm, purity> 99.99%)

was chosen to be the starting material to prepare GKC.

Further, KCl was washed off from GKC to produce EG. We

have reported the synthesis and characterization of these mate-

rials recently.18 By virtue of experimental conditions, GKC is

turbostatic due to which the graphene planes are misoriented

to attain a low energy state. However, EG is free of KCl

and contains primarily a mixture of twisted bi and tri-layer

graphene samples.18 Figures S1 and S2 in supplementary

material19 show the XRD and Raman spectra of graphite

(GRT), GKC, and EG, respectively. GKC has an ordered crys-

tal structure similar to GRT while, EG has more dangling

bonds at the edges and also contains more surfaces due to

exfoliation of the GRT. These changes are also confirmed

through nature and shape of the 2D-bands. More discussion

on GKC and EG can be seen in our previous report.18

GRT is made up of stacked graphene layers along the

c-axis, held by weak van-der Waals interactions. The carbon

atoms in the graphene plane are sp2 hybridized, leaving a

delocalized p-electron per carbon atom. This kind of structure

of GRT makes it convenient to have a substantial interaction

between the guest and the host forming donor-acceptor com-

pounds. Hence, the graphene layers could be either positively

or negatively charged depending upon their surroundings. For

example, the intercalation of halogens would make the guest

positively charged. The donors (acceptors) facilitate transfer

of electrons (holes) to the GRT lattice. Thus, a maximum

charge density would be at the interface which would decay

with distance. Hence, an electrostatic potential is produced

between the surface and the bulk. Also, due to chemical inter-

action between the guest and the host, the Fermi level should

shift accordingly. As a result, the properties of the host would

get modified.20 Hence, we notice quite interesting magnetic

properties in GKC and EG, which are very different from that

of GRT.21

The plots of zero field cooled (ZFC)-field cooled (FC)

data of magnetization (M) v/s temperature for GKC after the

background (BG) subtraction and EG without any BG subtrac-

tion measured at 50 Oe are presented in Figures 1(a) and 1(b),

a)Author to whom correspondence should be addressed. Electronic mail:

dhirenb@iitb.ac.in. Tel.: þ91 22 2576 7632. Fax: þ91 22 2572 6975.

0003-6951/2014/104(24)/242413/4/$30.00 VC 2014 AIP Publishing LLC104, 242413-1

APPLIED PHYSICS LETTERS 104, 242413 (2014)

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

137.30.242.61 On: Mon, 08 Dec 2014 21:33:23

respectively. From the ZFC data, it can be seen that the M of

GKC increases if the temperature is increased and attains a

maximum value at the Neel temperature, TN, after which it

decreases with further increase in temperature.22 GKC shows

similar behavior even when a different field is applied (Figure

S3 in supplementary material).19 Hence, it is indicated that

GKC has an AFM order until the transition temperature, TN

above which paramagnetic response is seen. Eventually, the

moment becomes negative, signifying a diamagnetic state

(above 25 K). This is due to the presence of the diamagnetic

GRT and KCl. Recently, Grujic et al. have predicted the AFM

nature of graphene rings using the mean-filed Hubbard

model.23 The edges and the surfaces of the GKC are modified

due to bonding of graphene with KCl. This results in a strong

dependence of the magnetic state of GKC on the distribution

of the moments at the outer edges of it.24 Also, the surface of

the nanoparticles has a strong effect on the nature of the spin

canting giving rise to magnetic anomalies.25 The AFM phase

was found to be preferred by the mean-field Hubbard model

which we noticed experimentally in GKC. Even, first princi-

ples density functional calculations have also predicted the

AFM phase of long zigzag edges in graphene nanoribbons.26

FM behavior is seen in EG throughout temperature range of

2–330 K except an AFM transition at around 60 and 155 K

(Figure S4 in supplementary material).19 The diamagnetic

contribution from the bulk has been suppressed due to the

presence of higher density of defects in EG resulting in a FM

state. This kind of behavior has been reported for several non-

magnetic systems.27–31

M v/s applied magnetic field (H) loops would add more

insight to the simultaneous presence of different magnetic

states in the samples. EG and GKC consist of a mixture of

both diamagnetic (from the bulk) and paramagnetic centers

(from the defect sites). Also, FM and AFM states were noticed

in the samples. We use LF-HF techniques to separate different

magnetic contributions. It is well known that the spin and the

orbital angular momentum of the electron about the nucleus in

an atom is responsible for a paramagnetic state while diamag-

netic behavior is due to the change in orbital moment induced

by H. The FM and AFM states support the presence of an

exchange field that keeps the electron spins arranged in an or-

dered manner. It is the strength of the interactions of spins

with the external field that helps in separating the individual

contributions. At a given temperature, the magnetic suscepti-

bility (v) of a sample is dependent on its density of defect

states and the interactions among them. Thus, v would be a

sum of all the individual contributions (BG of the instrument,

diamagnetic, paramagnetic, FM, and AFM)

v ¼ vBG þ v ðdiamagneticÞ þ v ðparamagneticÞ þ vAFM = FM:

The BG contribution can be easily separated by measuring the

v in the absence of the material. The FM contribution would

be saturating at a particular field above which only paramag-

netic contributions will be present along with diamagnetic in-

terference from the bulk. Both diamagnetic and paramagnetic

contributions can be seen only in the presence of H and is pro-

portional to the strength of the H. v (paramagnetic) is directly

proportional while v (diamagnetic) is inversely proportional to

the H. Thus, a linear behavior in M v/s H loop would be seen

at fields above the saturation magnetic field (Hs).

Furthermore, the slope of the line (in M v/s H loop) at HFs

above Hs would represent the sum of diamagnetic and para-

magnetic contributions. Therefore, the MH loop should show

its saturation behavior after diamagnetic and paramagnetic

parts are subtracted.32

The insets of Figures 1(a) and 1(b) represent the M v/s H

loops of GKC and EG (before and after BG/paramagnetic sub-

traction), respectively, at 2 K. It can be seen that, both GKC

and EG get saturated eventually at HFs after the BG/paramag-

netic contributions (linear responses) have been deducted.

Thus, both GKC and EG show FM behavior.33 At low temper-

atures (<5 K), GKC is found to be AFM. Hubbard model and

density functional theory calculations support the formation of

spin polarized edge states.17 There could be an intra or inter-

zigzag interaction resulting in a FM or AFM state.17,26 The

AFM state is due to the conduction p-electrons mediating the

spin polarized zigzag edge state spins.17,24,34 It can be also

seen that, the coercive field (Hc) of GKC is larger than that of

EG while, EG has a higher M-value than that of GKC. The

high Hc value of GKC is a direct consequence of its turbo-

static effect producing lattice strains.35–37 Also, GKC is

expected to have larger number of nearest neighboring edge

states by virtue of its graphitic structure. Thus, the edge states

in GKC have a stronger exchange interaction between them.

The higher M-value of EG in comparison to that of GKC is

due to the presence of greater density of defect states in it.38,39

Recently, a very similar nature of M v/s H loop (with similar

magnitudes in M) was observed in graphene by Chen et al.40

The increase in the M-value comes due to a strong internal

molecular field (Hm) which is at least 1000 times that of H.

FIG. 1. M v/s temperature plots for (a) GKC and (b) EG at H¼ 50 Oe. The

inset figures represent the corresponding hysteresis loops at 2 K. M v/s H

plots signify the occurrence of a FM state in both the samples. AFM state

was also observed at certain temperatures.

242413-2 A. K. Swain and D. Bahadur Appl. Phys. Lett. 104, 242413 (2014)

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

137.30.242.61 On: Mon, 08 Dec 2014 21:33:23

Thus the effective field (He) experienced by the material is a

sum of H and Hm.

The M v/s H loop of EG (at 5 K) presented in Figure 2

indicates a decrease in the value of M with that measured at

2 K. As Hm is proportional to M, we could infer that the

strength of the Hm and, hence, the interaction has decreased

due to the randomness created by the temperature rise.

However, the FM behavior of EG is still seen after the sub-

traction of BG and paramagnetic contributions. This can be

easily seen as a consequence of remanence in EG (inset in

Figure 2). To understand the temperature effect on magnetic

interactions, we plot the maximum moment (after BG and

paramagnetic subtraction) of the sample (Ms) and Hc v/s

temperature in Figure S5 (in supplementary material).19 The

three inset figures depict the MH loops of EG at three different

temperatures (40, 100, and 300 K). At this stage, we omit any

measurement of GKC as it is diamagnetic at these tempera-

tures. The Ms for EG varies from 0.864 (at 2 K)–0.02 emu/g

(at 300 K). The value of Ms and Hc decreases with increase in

temperature as expected for FM materials (Figure S5). It is

interesting to see that the curve of MH loop at 300 K (above

100 K) lies in the negative axis. This is due to the dominant

diamagnetic contribution at HF at 300 K. While at LF, the

moment is still positive as expected in case of a FM material.

GKC was also found to be FM at 5 K (Figure S6 in supple-

mentary material)19 after the subtraction of BG and paramag-

netic contributions.

The samples were analyzed by inductively coupled

plasma-atomic emission microscopy (ICP-AES) to deter-

mine the magnetic impurity concentration. Assuming the

iron content is pure, it would require 3927 and 91 ppm of Fe

to produce the observed values of moment at 2 at 300 K,

respectively.41 But the measured Fe content in EG (by

ICP-AES) was found to be 4.72 ppm. Thus, it is obvious that

the FM interaction in EG is not due to any magnetic impur-

ities. Rather, it is the interaction of the paramagnetic centers

which results in a FM state. Hence, it is certain that the ferro-

magnetism is intrinsic to EG.

To further verify this point, we performed electron spin

resonance (ESR) measurements of EG at 300 and 77 K

(Figure 3(a)). There is essentially no change in the line

width (dH) and g-value of EG measured at 77 and 300 K (at

77 K: dH� 14.5 G, g� 1.9899, ESR microwave frequency:

t77 K � 9154 MHz; and at 300 K: dH� 14.4 G, g� 1.9896,

t300 K� 9448 MHz).42,43 It has been reported that, quasi-two-

dimensional (Q2D) model proposed for graphene sheet gives

a weak or no temperature dependence of the dH with a value

around 10 G that arise due to the localized spins.44,45 The

measured dH values of EG are close to the values predicted

by the Q2D model of graphene sheet. Hence, it is the local-

ized spins in EG which are responsible for the observed mag-

netic properties. The lattice vibrations would increase with

increase in temperature. Thus the electron-phonon interac-

tion at 300 K is stronger than that of at 77 K. As a result, the

g-value decreases with increase in temperature due to pho-

nons mediating the lattice spin-orbit interaction.46 EG was

found to contain �2.89� 1019 spins/g at 300 K. In contrast,

the spin density from superconducting quantum interference

device (SQUID) measurements at 300 K was estimated to be

�2.15� 1018 spins/g. The spin density calculation details

are explained in the supplementary material.19 The lower

estimate of spin density from SQUID measurements may be

due to the decrease in the value of Ms (due to subtraction of

BG/paramagnetic contribution). The nature of the ESR spec-

tra (temperature independent and absence of any magnetic

impurities in full spectrum; given in Figures S7 and S8 in

supplementary material)19 and the g-value (approximate to

the g-value of free electron� 2.0023) does not support the

presence of any ferromagnetic impurities in the sample.34 In

addition, the ESR spectra are not absolutely Lorentzian in

shape, although a single ESR line is obtained. Rao et al.have observed two distinct ESR signals in carbon.47 The

ESR signal in graphene and its derivatives (in the absence of

metallic impurities) mainly arise due to the p-electrons and

non-bonding localized states in the graphene sheet. There are

reports based on theoretical calculations and experimentalFIG. 2. M v/s H loop for EG at 5 K. The inset figure shows the presence of

remanence.

FIG. 3. (a) ESR spectra of EG at 77 and 300 K. (b) Schematic model to

understand the exchange interactions between the edge-state spins in an arbi-

trary shaped graphene sheet.

242413-3 A. K. Swain and D. Bahadur Appl. Phys. Lett. 104, 242413 (2014)

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

137.30.242.61 On: Mon, 08 Dec 2014 21:33:23

observations to support the presence of electron spin magne-

tism at the zigzag edges of graphene.48 Thus, a possible long

range exchange interaction between the conduction electrons

and the paramagnetic centers can be a dominant factor in

obtaining FM in EG.49

A schematic model is presented in Figure 3(b) to explain

the possible interaction between various edges in an arbitrary

shaped graphene sheet. The zigzag edge-state spins are local-

ized and exhibit a strong FM intra-zigzag edge exchange

interaction (J0� 103 K).17 Further, the conduction carriers

(p-electrons) can mediate an inter-zigzag-edge interaction

(J1) between the individual FM coupled zigzag edge-state

spins separated by armchair edges. Based on the sign of J1,

this could generate either a FM (þJ1) or AFM (�J1) state.

Thus, the edge structure and the shape of the graphene sheets

in EG play an important role in deciding the dominating

behavior of the material giving rise to a resultant non-zero

magnetic moment. These interactions can also propagate

along the c-axis of EG to yield higher values of M. The FM

interaction could also be due to RKKY-type of interaction

where the conduction carriers in the graphene lattice interact

with the local magnetic moments. Thus, we find the defect

mediated mechanism to be suitable for the magnetic behav-

ior of EG.

In summary, EG exhibits a state of mixed magnetism

where the temperature and field play important roles in

deciding the final nature of the magnetic state. At low tem-

peratures (<5 K), the main contribution to the magnetic

moment comes from the internal molecular fields resulting

in a FM state, while diamagnetic state dominates at room

temperatures. The nature of exchange interaction between

the edge sates (inter and intra zigzag) in the multilayer gra-

phene (in EG) is responsible for its AFM/FM state. FM in

multilayer graphene is intrinsic, and is a result of a strong

correlation between various defects.

The authors would like to thank IITB-Monash Research

Academy and DST-Nanomission for their financial support.

1R. R. Nair, I. L. Tsai, M. Sepioni, O. Lehtinen, J. Keinonen, A. V.

Krasheninnikov, A. H. C. Neto, M. I. Katsnelson, A. K. Geim, and I. V.

Grigorieva, Nat. Commun. 4, 2010 (2013).2F. Simon, F. Mur�anyi, and B. D�ora, Phys. Status Solidi B 248(11), 2631

(2011).3A. H. C. Neto, N. M. R. Peres, K. S. Novoselov, and A. K. Geim, Rev.

Mod. Phys. 81(1), 109 (2009).4X. H. Hu, L. T. Sun, and A. V. Krasheninnikov, Appl. Phys. Lett. 100(26),

263115 (2012).5A. Zhou, W. Sheng, and S. J. Xu, Appl. Phys. Lett. 103(13), 133103

(2013).6T. Enoki and K. Takai, Solid State Commun. 149(27–28), 1144 (2009).7P. Esquinazi, A. Setzer, R. Hohne, C. Semmelhack, Y. Kopelevich, D.

Spemann, T. Butz, B. Kohlstrunk, and M. Losche, Phys. Rev. B 66(2),

024429 (2002).8M. Sepioni, R. R. Nair, S. Rablen, J. Narayanan, F. Tuna, R. Winpenny, A.

K. Geim, and I. V. Grigorieva, Phys. Rev. Lett. 105(20), 207205 (2010).9R. R. Nair, M. Sepioni, I. L. Tsai, O. Lehtinen, J. Keinonen, A. V.

Krasheninnikov, T. Thomson, A. K. Geim, and I. V. Grigorieva, Nat.

Phys. 8(3), 199 (2012).10R. Hohne and P. Esquinazi, Adv. Mater. 14(10), 753 (2002).11H. Ohldag, T. Tyliszczak, R. Hohne, D. Spemann, P. Esquinazi, M.

Ungureanu, and T. Butz, Phys. Rev. Lett. 98(18), 187204 (2007).

12C. N. R. Rao, H. S. S. R. Matte, K. S. Subrahmanyam, and U. Maitra,

Chem. Sci. 3(1), 45 (2012).13O. V. Yazyev, Rep. Prog. Phys. 73(5), 056501 (2010).14L. Brey, H. A. Fertig, and S. Das Sarma, Phys. Rev. Lett. 99(11), 116802

(2007).15M. Sherafati and S. Satpathy, Phys. Rev. B 83(16), 165425 (2011).16Y. Zhang, S. Talapatra, S. Kar, R. Vajtai, S. K. Nayak, and P. M. Ajayan,

Phys. Rev. Lett. 99(10), 107201 (2007).17T. Enoki, “Magnetism of nanographene,” in Graphene: Synthesis,

Properties, and Phenomena, edited by C. N. R. Rao and A. K. Sood

(Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2012), p.

131.18A. K. Swain and D. Bahadur, RSC Adv. 3(42), 19243 (2013).19See supplementary material at http://dx.doi.org/10.1063/1.4884426 for

XRD and Raman spectra for GRT, and EG; M v/s T for GKC and EG; Ms

and Hc v/s T for EG; M v/s H for EG at various temperatures; M v/s H at

5 K for GKC; Full range ESR spectrum for EG at 300 K with measuring

parameters; Computation formulae for spin density; Computation details

of spin density from SQUID and ESR data; and SQUID measurement

procedures.20A. K. Swain and D. Bahadur, J. Phys. Chem. C 118(18), 9450 (2014).21J. Zhou, M. M. Wu, X. Zhou, and Q. Sun, Appl. Phys. Lett. 95(10),

103108 (2009).22C. M. Julien, A. Ait-Salah, A. Mauger, and F. Gendron, Ionics 12(1), 21

(2006).23M. Grujic, M. Tadic, and F. Peeters, Phys. Rev. B 87(8), 085434 (2013).24J.-H. Lee and J. C. Grossman, Appl. Phys. Lett. 97(13), 133102 (2010).25N. K. Sahu and D. Bahadur, J. Appl. Phys. 113(13), 134303 (2013).26D. E. Jiang, B. G. Sumpter, and S. Dai, J. Chem. Phys. 127(12), 124703

(2007).27A. K. Swain, D. Li, and D. Bahadur, Carbon 57, 346 (2013).28D. Dutta and D. Bahadur, J. Mater. Chem. 22(47), 24545 (2012).29A. Prakash, S. K. Misra, and D. Bahadur, Nanotechnology 24(9), 095705

(2013).30P. Esquinazi, W. Hergert, D. Spemann, A. Setzer, and A. Ernst, IEEE

Trans. Magn. 49(8), 4668 (2013).31S. Tongay, S. S. Varnoosfaderani, B. R. Appleton, J. Wu, and A. F.

Hebard, Appl. Phys. Lett. 101(12), 123105 (2012).32C. Richter and B. A. Vanderpluijm, Phys. Earth Planet. Inter. 82(2), 113

(1994).33M. Kan, J. Zhou, Y. Li, and Q. Sun, Appl. Phys. Lett. 100(17), 173106

(2012).34S. S. Rao, S. N. Jammalamadaka, A. Stesmans, V. V. Moshchalkov, J. van

Tol, D. V. Kosynkin, A. Higginbotham-Duque, and J. M. Tour, Nano.

Lett. 12(3), 1210 (2012).35Y. C. Wang, J. Ding, J. B. Yi, B. H. Liu, T. Yu, and Z. X. Shen, Appl.

Phys. Lett. 84(14), 2596 (2004).36F. Zhai and L. Yang, Appl. Phys. Lett. 98(6), 062101 (2011).37N. Levy, S. A. Burke, K. L. Meaker, M. Panlasigui, A. Zettl, F. Guinea, A.

H. C. Neto, and M. F. Crommie, Science 329(5991), 544 (2010).38R. Faccio and A. W. Mombr�u, J. Phys.: Condens. Matter 24(37), 375304

(2012).39B. Panigrahy, M. Aslam, and D. Bahadur, Appl. Phys. Lett. 98(18),

183109 (2011).40L. Chen, L. Guo, Z. Li, H. Zhang, J. Lin, J. Huang, S. Jin, and X. Chen,

Sci. Rep. 3, 2599 (2013).41Y. Wang, Y. Huang, Y. Song, X. Y. Zhang, Y. F. Ma, J. J. Liang, and Y.

S. Chen, Nano. Lett. 9(1), 220 (2009).42L. �Ciric, A. Sienkiewicz, B. N�afr�adi, M. Mionic, A. Magrez, and L. Forr�o,

Phys. Status Solidi B 246(11–12), 2558 (2009).43M. Drescher and E. Dormann, Europhys. Lett. 67(5), 847 (2004).44V. Likodimos, S. Glenis, N. Guskos, and C. Lin, Phys. Rev. B 68(4),

045417 (2003).45V. Sitaram, A. Sharma, S. Bhat, K. Mizoguchi, and R. Menon, Phys. Rev.

B 72(3), 035209 (2005).46A. A. Konakov, A. A. Ezhevskii, A. V. Soukhorukov, D. V. Guseinov, S.

A. Popkov, and V. A. Burdov, J. Phys.: Conf. Ser. 324, 012027 (2011).47S. S. Rao, A. Stesmans, Y. Wang, and Y. Chen, Physica E 44(6), 1036

(2012).48K. Xu and P. D. Ye, Appl. Phys. Lett. 104(16), 163104 (2014).49L. �Ciric, A. Sienkiewicz, D. M. Djokic, R. Smajda, A. Magrez, T. Kaspar,

R. Nesper, and L. Forr�o, Phys. Status Solidi B 247(11–12), 2958 (2010).

242413-4 A. K. Swain and D. Bahadur Appl. Phys. Lett. 104, 242413 (2014)

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

137.30.242.61 On: Mon, 08 Dec 2014 21:33:23