arXiv:cond-mat/0604266v1 [cond-mat.mtrl-sci] 11 Apr 2006

arXiv:cond-mat/0604266v1 [cond-mat.mtrl-sci] 11 Apr 2006

all-ferromagneti superlatties

P. Padhan and W. Prellier

Laboratoire CRISMAT, UMR CNRS 6508, 6,

Bld. du Maréhal Juin, F-14050 Caen, Frane

R.C. Budhani

Department of Physis, Indian Institute of Tehnology, Kanpur 208016, India

(Dated: July 18,2018)

Abstrat

Thestruturalandmagnetipropertiesofaseriesofsuperlattiesonsistingoftwoferromagneti

metals La

0.7

^Sr

0.3

^MnO

3

^{(LSMO) and SrRuO}

3

^{(SRO)grown on (001) orientedSrTiO}

3

^arestudied.

Superlatties with a xed LSMO layer thikness of 20 unit ells (u..) and varying SRO layer

thiknessshowasuddendropinmagnetizationonoolingthroughtemperaturewhere both LSMO

and SROlayers are ferromagneti. Thisbehaviorsuggests an antiferromagneti oupling between

the layers. In addition, the samples having thinner SRO layers (n

<

6) exhibit enhaned satura-

tion magnetization at 10 K. Theseobservations areattributed to thepossible modiation in the

stereohemistryof the Ruand Mnions inthe interfaial region.

PACSnumbers:

bilayerongurations onsisting of ferromagneti(FM) - antiferromagneti (AFM), AFM -

AFM,andFM-FMthin lms. On several ombinationsof FM-AFMbilayersthe magneti

behavior exhibit horizontal and vertial hysteresis loop shifts, 1,2,3,4

while some AFM-AFM

bilayersystems show unexpeted ferromagneti behavior.

5,6,7

Ke et. al.

8

have observed ex-

hangebiaseetsinmultilayersonsistingofFM-FMbilayersomposedofLa

0.67

^Sr

0.33

^MnO

3

and SrRuO

3

^{. WhileUozu et. al.}

9

have observed anantiferromagnetiexhange ouplingin

theFMtrilayeronsistingofSr

0.7

^Ca

0.3

^RuO

3

^(SCRO)andLa

0.6

^Sr

0.4

^MnO

3

^{. inordertounder-}

stand the ouplingbetween the Ru and Mnions atthe interfaes of the two FMmedia,we

have synthesized aseries of superlatties onsisting of FM-FM bilayers of La

0.67

^Sr

0.33

^MnO

3

(LSMO) and SrRuO

3

^{(SRO), where the LSMO layer thikness (d}

LSM O

^{) is xed (~20 u..)}

and the SRO layer thikness (d

SRO

^{) is varied. We note that the} temperature dependene

of magnetization of these superlatties strongly varies with the SRO layer thikness when

the d

SRO

^{is less than 6 unit ells. In addition, these} superlatties show a sudden drop

in magnetization at a temperature where both LSMO and SRO layers are ferromagneti.

The reproduible zero-eld-ooled (ZFC) magneti minor hysteresis loopshapeinthe eld-

ooled(FC)stateofthesesuperlattiessuggests theexisteneofantiferromagnetiexhange

ouplingbut not the presene of aninterfaial layerof antiferromagnetiharater.

ThinlmsofLSMOandSROandtheirsuperlattiesweregrownon(001)orientedSrTiO

3

(STO)substrates at720

◦

Cinoxygenambientof300 mTorrusing amultitargetpulsedlaser

deposition tehnique. The deposition rates (typially ~0.26 Å/pulse and ~0.31 Å/pulse)

of SRO and LSMO respetively were alibrated individually for eah laser pulse of energy

density ~3 J/m

2

. The hamber was lled with oxygen of 300 Torr after ompletion of

depositionandthen thesampleswere ooledtoroomtemperatureattherate of15

◦

C/min.

The superlattiestrutures were synthesized by repeating15 timesthe bilayeronsisting of

a 20- unit ells thik LSMO layer and n- unit ells thik SRO layer, with n taking integer

values from 1 to 12. In all multilayer samples, the bottom and top layers are of LSMO.

Charaterizationofthestrutureandepitaxialnatureofthemultilayerandsinglelayerlms

wereperformedusingx-raydiration(XRD).Forthemagnetization(M)measurements,we

have used a superonduting quantum interferene devie based magnetometer (Quantum

DesignMPMS-5). Thesemeasurementswerearriedoutbyoolingthesampletothedesired

temperaturein the presene/absene of a magneti elds applied along the [100℄ and [001℄

measurements remained the same.

The lattie parameter of ubi STO (3.905 Å) is smaller than that of the pseudoubi

lattie parameter of SRO (3.93 Å) but larger than that of the LSMO (3.88 Å). Thus, STO

providesin-planetensilestress forthe epitaxialgrowth ofLSMO with alattiemismathof

-0.64%. Similarly,itisexpetedthatthe LSMOwouldprovidesin-planeompressivestress

for the epitaxial growth of SROwith a lattiemismath1.28 %. However, the LSMO-SRO

superlattie stabilizes pseudoubi phases of these perovskites. In the onventional

Θ

^-2

Θ

sans of the superlatties, no peaks were observed other than the (00l) Bragg reetions of

the onstituents,the substrate and the satellitesdue tohemialmodulationpresent inthe

multilayer. To evaluate the lm thikness of LSMO and SRO layers, we have arried out

quantitativerenementofthe

Θ

^-2

Θ

^{sanofthetrilayerstruturesusingDIFFaXprogram.10}

Theexperimental

Θ

^-2

Θ

^{sansandthesimulatedprolesofthetwosamplesreordedaround}

the (001) reetion of STO are shown in Fig. 1. The simulatedprole using the alibrated

thikness is in good agreement with the position of the Kiessig fringes and their relative

intensity ratio onrming the quality of the layers.

LSMO exhibits positive spin polarization and ferromagneti ordering with a relatively

high Curie temperature (T

C

^{~360 K), small magneti anisotropy and low oerive eld}

(H

C

^).

11

While SRO shows negative spin polarization with a relatively small T

C

^{(~150 K),}

stronguniaxial rystallineanisotropyand relativelylarge H

C

^.

12

In the magnetimultilayers

disussed here the thikness of the LSMO layer is xed at 20 u.. A typial eld-ooled

temperature-dependene magnetization M(T) for ~ 20 u.. thik LSMO and SRO lms is

shown in the Fig. 2a and 2b, respetively. These relatively thin lms of LSMO and SRO

show aredued T

C

^{of ~330 K and ~ 145 K}respetively. The redued T

C

^{seen in both the}

ases is due tonite size eet resultingfrom the strains.

13

Fig. 3 displays the eld-ooled (FC) magnetization of the superlatties with n=2, 4, 5,

6 and 7. These measurements were performed at 0.01 tesla eld applied along the out-

of-plane diretion of the STO. From this gure it appears that the alternative staking of

LSMO and SROin a superlattiehanges the M(T) signiantly ompared tothe M(T) of

theonstituents(e.g. LSMO andSRO). Similarbehaviorofmagnetizationisobtainedwhen

the magneti eld is applied in the plane (not shown), but the magnetization is larger in

the ase of the out-of-plane magneti eld. The onset of spontaneous magnetization in all

superlattiesoursatT=340K.ThebehaviorofM(T)atT

<

^{340 Kissigniantlydierent}

from that of the pure LSMO. For example, the FC magnetization of the sample with n=2

dereasesgraduallyasthetemperatureisraisedfrom10Kto33K.Thistrendisfollowedby

an inrease of magnetization till a maximum value is reahed at 100 K. A further inrease

in temperature leads to a monotoni drops in magnetization (see Fig. 3a). As the SRO

layer thikness is inreased up to 4 u.., the magnetization inreases slowly above 10 K,

beomesmaximumat60K,and then above60Kdereases slowly. Between 160 Kand 200

K, the magnetization shows a plateau. Two harateristi temperatures an be identied

in Fig. 3 in addition to the T

C

^{(350 K) above whih the sample beomes} paramagneti.

Therst one isthetemperature(T

C

^{*)atwhihthe SROlayerbeomes}ferromagnetisine

the magnetization rises sharply below this temperature (150 K). The seond one, is the

temperature(T

N

^{) below whih the} magnetizationdereases althoughboth the omponents

are inthe ferromagnetistate. To larifythis pointwe haveenlarge the M(T) urve inthis

regionand shown them inFig.3,panelf, g,h, i,and j forsuperlattiewith n=2,4, 5,6and

7,respetively.

The distint usp in magnetizationbelow the T

C

^{of SRO is an indiationof an antifer-}

romagneti behavior. Thus, we denote the temperature assoiated to this feature as the

Néeltemperatures (T

N

^{). However, whenthe SROlayerthiknessisinreasedbeyond7u..,}

this usp-likefeature below the Curietemperatureof SROis suppressed. TheT

C

^(LSMO),

T

C

^{* and T}

N

^{of the} superlatties extrated from Fig. 2 are plotted as a funtion of d

SRO

in Fig. 4. As disussed before, the T

C

^{of the} superlatties is nearly independent of d

SRO

^,

and islose tothe T

C

^{of LSMO(~340K), whilethe T}

C

^{* inreaseswith d}

SRO

^{.However, this}

T

C

^{* is not} distinguishable for the superlatties with d

SRO >

^{8 u.. On the ontrary, for the}

sampleswith n

<

^{5 u..,the T}

N

^{shows adistintinrease with the SROlayerthikness. The}

presene of a distint T

C

^{(SRO) at 150 K in the M(T) data of the} superlattie with

n > 4

indiatesthe formationofastoihiometriSROlayer. This observationsuggeststhatinthis

magneti system the interfae roughness aused by the magneti and strutural disorder is

small (~2 u..). The reproduible ZFC in-plane and out-of-plane minor hysteresis loops of

these samples in their orresponding FC state indiate the existene of an antiferromag-

neti exhange oupling, but do not show the presene of an interfaial antiferromagneti

layer.

8

The drop in magnetization, at a temperature whih we have marked T

N

^{, ould be}

due to this disordered interfae. A similar AFM exhange oupling has been observed in

LSMO/SCRO/LSMO trilayers by Uozu et. al. , and in LSMO/SRO bilayers by Ke and

oworkers 8

. While the former group has attributed the AF oupling to superexhange in-

teration, Ke et. al. attribute ittointerfaialhargetransfer. However, the AFMexhange

oupling with the SRO layer thikness as seen in the present ase suggests the possibility

of other physial proesses. Some well established soures aeting the magneti oupling

in superlatties are the substrate-indued strains, interfaial stress and interlayer exhange

oupling.

13,14

The larger value of T

C

^{*ompare tothe transition} temperature of SROould

also be due to the high paramagneti suseptibility 15

of SRO above the T

C

^{(SRO) and/or}

the reonstrution ofthe spin state ofRuand Mnions at/losetothe interfaes. The value

ofthe T

N

^{inreasesand saturatestoa}temperatureT

C

^{(SRO)athigherd}

SRO

^{,indiatingthe}

inuene of the size eet of the SRO layer.

In order to study the possible magneti onguration of Ru and Mn ions at 10 K, we

have measured the ZFC magneti hysteresis loop of these samples with the magneti eld

orientedalong the[100℄ and[001℄ diretionsof thesubstrate. The in-planeandout-of-plane

ZFC magneti hysteresis loops of two samples are shown in Fig. 5a. The in-plane and

out-of-plane saturation magnetizations (M

S

^{) of all samples are the same though their in-}

plane and out-of-plane saturation magneti elds are dierent. The M

S

^{of some samples,}

extrated from their ZFC in-plane and out-of-planehysteresis loopafter orreting for the

weak diamagneti response of the substrate, are shown in the Fig. 5b. This gure also

inludes the theoretial value of M

S

^{alulated from the spin-only M}

S

^{of the LSMO (3.34}

µ B

^/Mn)

16

and SRO (1.6

µ B

^/Ru)

17

. The higher value of the measured M

S

^{, as ompared}

to the theoretial one for the sample with

n = 2

^{to 7, indiates an enhanement of the}

magnetization.

The enhane magnetization ould be due to the modiation of the harge states of the

Ruand Mnions 18

attheinterfaes,and thereby aninrease inthe eetivethikness of the

interfaial layer in superlattie with the lower d

SRO

⁽

<

^{4 u..). In samples with d}

SRO > 4

^,

the eetivethikness ofthe interfaiallayerdereasesasthe stoihiometriSROlayerstart

to form. The interfaial magneti roughness seems to suppressed ompletely by the strong

long range ordering of the ferromagneti moments of SRO and LSMO at still larger d

SRO

^.

However, measurementsusing eletron energy lossspetrosopy shouldbe performtoverify

the valeny assumptions,in partiular the stabilization of the Ru

5+

spin state.

We have demonstrated that in La

0.7

^Sr

0.3

^MnO

3

^{and SrRuO}

3

superlatties an antiferro-

thikness. We attribute these hanges to interfaial magneti and eletri disorder, whih

appearstohealasthe SROlayerthiknessinreases andlongrangeorderingofthemagneti

moment assoiatedwith Ru ions beomes dominant.

We thank the nanial support from the Centre Frano-Indien pour la Promotion de

la Reherhe Avanee/Indo-Frenh Centre for the Promotion of Advane Researh (CE-

FIPRA/IFCPAR) under Projet No 2808-1. Partial support from the European Union

under the STREP researh projet CoMePhS (N

◦ 517039

^{)is also}aknowledged.

J.Noguesand I. K.Shuller,J. Magn.Magn.Mater. 192,203 (1999).

2

I. Panagiotopoulos, C. Christides,M. Pissas andD. Niarhos, Phys.Rev. B60, 485(1999).

3

P.Padhan andW. Prellier, Phys.Rev. B72,104416(2005).

4

P.Padhan andW. Prellier, Eur.Phys. J.B45,169 (2005).

5

K. Ueda, H.Tabataand T. Kawai, Siene 280,1064 (1998).

6

K.S.Takahashi,M. Kawasaki andY.Tokura,Appl. Phys.Lett. 79,1324 (2001).

7

P.Padhan,P.Murugaveland W.Prellier, Appl. Phys.Lett. 87,022506(2005).

8

X.Keand M.S.Rzhowskia,J. Belenkyand C.B. Eom,Appl. Phys.Lett. 84,5458 (2004).

9

Y.Uozu, T. Nakajima,and M. Nakamura, Y. Ogimoto,M. Izumi and K. Miyano,Appl. Phys.

Lett.85, 2875 (2004).

10

seeat: http://p14.sims.nr.a/p/p14/ftp-m irror/ diax/pub /trea y/DIFFaX

\

^_v1807/

11

J.-H. Park, E. Vesovo, H.-J. Kim, C. Kwon, R. Ramesh and T. Venkatesan, Nature 392, 794

(1998).

12

G.Cao, S.MCall, M. Shepard, J.E. Crowand R.P.Guertin,Phys.Rev. B56,321 (1997).

13

K.R.Nikolaev, A.Dobin,I.N. Krivorotov, W.K.Cooly, A.Bhattaharya, A..L.Kobrinskii,L.I.

Glazman,R.M.Wentzovith., E. DanDahlbergandA.M. Goldman, Phys. Rev.Lett., 85,3728

(2000).

14

P.Padhan,R.C.Budhani and R.P.S.M. Lobo, Europhys. Lett. 63,771 (2003).

15

G.A.Mulhollan, R.L.Fink, J.L.Erskine and G.K. Walters, Phys.Rev.B43, 13645 (1991).

16

G.Banah,R.Tyerand W.M. Temmerman,J.Magn. Magn.Mater. 272,1963 (2004).

17

D.Singh, J.Appl. Phys.79,4818 (1996).

18

C. Martin,A. Maignan, M. Hervieu, C. Autret, B. Raveau and D. I. Khomskii, Phys. Rev. B,

63,174402(2001).

Figure1: The experimental (solid line) and simulated (DIFFaX) (dotted line) - 2 x-ray

dirationproles of the superlattie withn=4 and 8. The (001)Bragg's reetion ofSTO

and several orders

(0, ±1, ±2)

^{of satellite peaksare indexed.}

Figure 2: Temperature dependene of the out-of-plane magnetization under 0.01 tesla

eld for a 20u.. thik lmof LSMO (panel a)and SRO(panel b).

Figure 3: Temperature dependene of the out-of-plane magnetization under 0.01 tesla

eld of the superlatties with n=2, 4, 5, 6 and 7 (panels a, b, , d and e, respetively, for

0-300K and panels f, g, h, i, and j, respetively, for the enlarge part orresponding to the

temperature range where the LSMO and SRO are ferromagneti (panel f, g, h, i, and j

respetively). The arrows indiatethe T

C

^{and T}

N

^.

Figure 4: Evolution of the T

C

^{, T}

C

^{* and T}

N

^{(panel a, b and} respetively) of several

superlatties as a funtion of the SRO layer thikness. In panel a and b the solid lines

representtheT

C

^{ofbulkLSMOand SRO,}respetively,whileinpanelthesolidlineisonly

a guide to the eyes. Both T

C

^{and T}

C

^{* have been alulated from the} intersetion of the

slopearound the transitionof magnetization.

Figure 5: (a) ZFC magnetization loop for the (20 u..)LSMO/(n u..)SRO superlattie

withn=4and12. Magnetieldisorientedalongthe[100℄and[001℄diretionsofSTO.(b)

: Experimental and theoretial saturation magnetizationfor the superlatties as afuntion

of the SrRuO

3

^{layer thikness. The solid lines are guide tothe eyes.}

18 20 22 24 26 28 -3

-2 -1

+2 +1

I nt e ns i t y ( a br . uni t s )

Two theta (degree) n = 8

n = 4

0 0

+5 -5

-4 -3

-2 -1

+4 +3 +2

+1 (001) SrTiO

3

Exp.

Fit

FIG. 1: Padhan et. al.

1

0 200 400 0

40 80

0 150 300

0 80 160

Temperature (K)

Magnetization (emu/cm

3 )

T

C

(a)

20 u.c. LSMO

at 0.01 tesla

T

C

Temperature (K)

Magnetization (emu/cm

3 )

(b)

20 u.c. SRO

at 0.01 tesla

FIG. 2: Padhan et. al.

2

0 80

T

C

(a) n = 2

120 130

T

N

(f) n = 2

0 80

T

C

(b) n = 4 90

100

T

N

(g) n = 4

0 80

T

C

Magnetization (emu/cm

3 )

(c) n = 5

60 90

T

N

Magnetization (emu/cm

3 )

(h) n = 5

0 80

T

C

(d) n = 6

40 80

T

N

(i) n = 6

0 150 300

0 80

T

C

Temperature (K) (e) n = 7

0 100 200

75 90

T

N

(j) n = 7

FIG. 3: Padhan et. al.

3

0 4 8 12 0

150 0 150 0 300

SRO layer thickness (unit cell) (c) T

N

(b) T

C

*

T e m pe r a t ur e ( K ) (a) T

C

(LSMO)

FIG. 4: Padhan et. al.

4

0 2 4 6 8 10 12 0

2 4

M S

( B

/unitcell)

SRO layer thickness (unit cell) (b)

experimental

theoretical

-4 -2 0 2 4

-4 -2 0 2 4

n = 4

n = 12

M( B

/unitcell)

Magnetic field (tesla) (a)

H//[100]

H//[001]

FIG. 5: Padhan et. al.

5