ApplPhysLett_88_062502

Combining half-metals and multiferroics into epitaxial heterostructures for spintronics

H. Béa, M. Bibes, M. Sirena, G. Herranz, K. Bouzehouane et al.

Citation: Appl. Phys. Lett. 88, 062502 (2006); doi: 10.1063/1.2170432

View online: https://www.sodocs.net/doc/e55604294.html,/10.1063/1.2170432

View Table of Contents: https://www.sodocs.net/doc/e55604294.html,/resource/1/APPLAB/v88/i6

Published by the American Institute of Physics.

Additional information on Appl. Phys. Lett.

Journal Homepage: https://www.sodocs.net/doc/e55604294.html,/

Journal Information: https://www.sodocs.net/doc/e55604294.html,/about/about_the_journal

Top downloads: https://www.sodocs.net/doc/e55604294.html,/features/most_downloaded

Information for Authors: https://www.sodocs.net/doc/e55604294.html,/authors

Combining half-metals and multiferroics into epitaxial heterostructures for spintronics

H.Béa

UnitéMixte de Physique CNRS-Thales,Route Départementale128,91767Palaiseau,France

M.Bibes a?

Institut d’Electronique Fondamentale,UniversitéParis-Sud,91405Orsay,France

M.Sirena,G.Herranz,K.Bouzehouane,and E.Jacquet

UnitéMixte de Physique CNRS-Thales,Route Départementale128,91767Palaiseau,France

S.Fusil

Universitéd’Evry,Batiment des Sciences,rue du père Jarlan,91025Evry,France

P.Paruch and M.Dawber

DPMC,University of Geneva,24Quai E.Ansermet,1211Geneva4,Switzerland

J.-P.Contour and A.Barthélémy

UnitéMixte de Physique CNRS-Thales,Route Départementale128,91767Palaiseau,France

?Received4August2005;accepted9December2005;published online6February2006?

We report on the growth of epitaxial bilayers of the La2/3Sr1/3MnO3?LSMO?half-metallic ferromagnet and the BiFeO3?BFO?multiferroic,on SrTiO3?001?by pulsed laser deposition.The growth mode of both layers is two dimensional,which results in unit-cell smooth surfaces.We show that both materials keep their properties inside the heterostructures,i.e.,the LSMO layer?11nm thick?is ferromagnetic with a Curie temperature of?330K,while the BFO?lms shows ferroelectricity down to very low thicknesses?5nm?.Conductive-tip atomic force microscope mappings of BFO/LSMO bilayers for different BFO thicknesses reveal a high and homogeneous resistive state for the BFO?lm that can thus be used as a ferroelectric tunnel barrier in tunnel junctions based on a half-metal.?2006American Institute of Physics.?DOI:10.1063/1.2170432?

To a great extent,today’s research in spintronics focuses on the development of materials and devices concepts.1In many cases,the former are requisites to the latter.Recently, several families of materials have been developed in this sense,a typical example being that of diluted magnetic semiconductors.2As spintronics effects rely on the spin polarization of the electrical current,materials with large, ideally total,spin-polarization have also been extensively investigated.With these so-called half-metals,3a consider-able increase in the tunnel magnetoresistance of magnetic tunnel junctions4has indeed been achieved,at least at low

temperatures.5

Another emerging family of magnetic materials are

multiferroics.6In these compounds,several ferroic orders ?e.g.,magnetic and electric?coexist,with some coupling between them?the magnetoelectric effect?.7Most are

ferroelectric and antiferromagnetic,a notable exception

being BiMnO3that is ferromagnetic.8A propotypical

multiferroic that has attracted a lot of attention lately is

BiFeO3?BFO?.9,10It is a ferroelectric and weakly ferromagnetic rhombohedral perovskite with order tem-peratures far above room temperature?T C=1043K,11 T N=647K?.12BFO thus crystallizes in the same structure as several known half-metallic ferromagnets?such as La2/3Sr1/3MnO3,La2/3Ca1/3MnO3,or Sr2FeMoO6?,which makes it possible to combine it with these materials in mul-tifunctional epitaxial heterostructures.Several promising types of devices can be imagined from this combination,as discussed,for instance,by Binek et al.13,14In particular,one can think of using very thin layers of BiFeO3as multiferroic tunnel barriers.If ferroelectric,these layers should have the same functionalities as those of recently developed15and modeled16,17ferroelectric tunnel barriers,combined with a magnetic ordering and a possible magnetoelectric coupling.

In this letter,we describe the growth and properties of bilayers of the La2/3Sr1/3MnO3?LSMO?half-metal combined with BFO and epitaxially grown onto SrTiO3?001?substrates by pulsed laser deposition.We study the morphological, structural,electrical,and magnetic properties of BFO/LSMO bilayers.We show that the magnetic properties of LSMO are preserved and that the BFO layers are insulating and ferro-electric,down to thicknesses of t BFO=5nm.BFO ultrathin layers thus ful?ll some important criteria for being used as ferroelectric tunnel barriers.

The samples used in this study have been grown by pulsed laser deposition using a Nd:yttrium–aluminium–garnet?YAG?laser,at a frequency of2.5Hz.The LSMO target was stoichiometric while for BFO,a target with nomi-nal composition Bi1.15FeO3was used,in order to compensate for the high volatility of Bi.?001?-oriented SrTiO3?STO?substrates were used.In the pseudo-cubic representation, BFO has a unit-cell parameter of3.96?,LSMO of3.88?, and STO of3.905?,so that the STO substrate induces a tensile strain on LSMO and a compressive strain on BFO. The optimal growth conditions for LSMO were previously determined to be of a deposition temperature T dep=720°C and an oxygen pressure of P dep=0.41mbar,18with a postan-nealing at300mbar of O2.We recently determined the

a?Electronic mail:manuel.bibes@ief.u-psud.fr

APPLIED PHYSICS LETTERS88,062502?2006?

0003-6951/2006/88?6?/062502/3/$23.00?2006American Institute of Physics

88,062502-1

P dep -T dep phase stability diagram for BFO ?lms and found that optimal conditions are around T dep =580°C and P dep =6?10?3mbar.19In the present samples,the LSMO layer ?11nm thick for all samples ?was grown ?rst,and the BFO ?lm ?t BFO =1–70nm ?afterward.To limit a possible deoxygenation of the manganite,after the growth of the LSMO ?lm,the pressure was kept at the LSMO deposition pressure while decreasing T dep to 580°C.Then,the pressure was rapidly decreased to 6?10?3mbar,to grow the BFO layer.The sample was ?nally cooled to room temperature in 300mbar of oxygen.

Re?ection high-energy electron diffraction ?RHEED ?patterns were collected ?at an acceleration voltage of 25kV ?before growth,and after depositing each layer.Images for the ?100?direction are shown in Figs.1?a ?–1?c ?,and indicate a two-dimensional growth for the LSMO and the BFO layer ?up to at least t BFO =35nm ?.An azimuthal analysis revealed an in-plane epitaxy for both layers.

X-ray diffraction ?-2?spectra ?in the 15°–115°2?range ?were collected for all samples and only showed peaks corresponding to ?00??re?ections ?pseudo-cubic indexation ?of STO,LSMO,and BFO.Figure 1?d ?shows the spectra for t BFO =5nm and 70nm.From the angular position of the ?003?re?ections of BFO,we calculated the out-of-plane pa-rameter c BFO ?4.10?that was found to vary only slightly with thickness.A reciprocal space map ?RSM ?of the ?103?re?ections ?see Fig.1?e ??for a 70nm ?lm shows that both the LSMO and BFO layers are very heavily strained on the STO substrate.Note that no splitting of the BFO ?103?peak is detected,suggesting a tetragonal rather than monoclinic or rhombohedral symmetry for the BFO layer,in contrast to the results of Xu et al.20or Qi et al.,21respectively.

Figure 2?a ?shows a magnetization hysteresis cycle of a BFO ?5nm ?/LSMO bilayer measured at 10K with the ?eld applied in-plane along ?100?after zero-?eld cooling,mea-sured in a superconducting quantum interference device ?SQUID ?.The saturation magnetization ?M S ?for t BFO =5nm is about 580emu cm ?3,close to the magnetization of

bulk LSMO ?590emu cm ?3?.22Even for larger BFO thick-ness values,the contribution of the BFO layer to the magne-tization was not visible,as expected from the very small magnetization values obtained for optimized BFO single ?lms grown in the same conditions.19We measured the evo-lution of the magnetization with the temperature in order to check the quality of the LSMO layer ?see Fig.2?b ??.The Curie temperature ?T C ?is 330K,somehow smaller than the bulk T C =370K,22but in good agreement with the T C of thin single ?lms 23of similar thickness.

The BFO/LSMO bilayers were imaged with a conductive-tip atomic force microscope ?CTAFM ?.In these type of experiments,the morphology of the sample surface and the resistance between the bottom electrode and the tip are measured simultaneously,and coupled height-resistance maps are recorded.24Examples of such maps 3?3?m 2are shown in Figs.3?a ?and 3?b ?for a BFO ?5nm ?/LSMO bi-layer.The left image reveals a very ?at surface ?in agreement with the two-dimensional growth mode inferred from the RHEED patterns ?with terraces separated by ?4?high steps,i.e.,a perovskite unit cell.The corresponding resis-tance map ?Fig.3?b ??shows a high resistance level ?average value ?1G ??,with a remarkable homogeneity.Similar coupled maps were collected for samples with different BFO thicknesses.Identical unit-cell steps were observed up to t BFO =20nm.Resistance maps could be collected for the t BFO =1,2,and 5nm samples,without saturating the capabil-ity of the CTAFM electronics.The average resistance of the maps is plotted as a function of the BFO thickness in Fig.3?c ?.The data at t BFO =0correspond to a single LSMO ?lm.An exponential increase of the resistance with t BFO is ob-tained,which indicates that transport occurs through the BFO ?lm by tunneling.This observation,together with the ?atness and the homogeneity of the BFO on the LSMO buff-ers,quali?es BFO very thin ?lms as potential barriers in tunnel junctions.

Piezoelectric atomic force microscopy ?PFM ?was used to probe the ferroelectricity of the BFO ?lms in BFO/LSMO bilayers,25for several values of t BFO .An alternatively posi-tive and negative voltage was applied between the conduc-tive tip of the AFM and the bottom electrode ?LSMO ?to pole the BFO layer into “up”and “down”stripes.PFM was then used to detect the domain structure.Figure 4?a ?shows the topography,and Figs.4?b ?and 4?c ?the PFM images of the 5nm ?lm after writing.We observe a clear contrast in Fig.4?b ?that reveals the presence of up and down ferroelectric do-mains in this ?lm.Note that this pattern is not observed

in

FIG.1.RHEED patterns in the ?100?direction of the sample with t BFO =20nm ?a ?of the STO substrate before deposition,?b ?after deposition of the LSMO layer,and ?c ?after deposition of the BFO.The pattern is more diffuse in ?b ?because the pressure of measurement was 6?10?3mbar com-pared to 10?6mbar for ?a ?and ?c ?.?d ?X-ray diffraction spectra of BFO/LSMO//STO bilayers with t LSMO =11nm and t BFO =5nm and 70nm.S,L,and B label peaks of STO,LSMO,and BFO,respectively.?e ?RSM of the ?103?re?ections of the BFO ?70nm ?/LSMO//STO bilayer;r.l.u.is for re-ciprocal space

units.

FIG.2.?a ?Magnetic hysteresis loops of the BFO/LSMO//STO bilayer with t BFO =5nm measured by SQUID at 10K.?b ?Temperature dependence of the magnetization in a ?eld of 1kOe,normalized to the magnetization at 10K ?M 10K ?for the same sample.

the topography.Figure 4?c ?,collected after writing two wider stripes perpendicularly to the ?rst pattern of Fig.4?b ?,shows that the PFM response can be switched.These PFM measurements are not quantitative;so that we cannot conclude on a possible decrease of the polarization of the BFO ?lm when thickness decreases,as reported for PbTiO 3,for instance.26However,they demonstrate that the ferroelec-tric character is preserved in BFO layers down to a thickness of 5nm.

In summary,we have successfully grown epitaxial het-erostructures integrating a LSMO bottom electrode and a BFO layer.The growth of both layers is two dimensional;they are heavily strained and the surface of the structure is unit-cell ?at.SQUID measurements indicate that the LSMO layer is ferromagnetic with a T C of 330K,while the BFO

?lms give virtually no signal,as expected for a weak ferro-magnet.Through a combined CTAFM and PFM study on these BFO/LSMO heterostructures,we have shown that BFO layers as thin as 5nm are ferroelectric and can be used as tunnel barriers.This opens the way for the realization of several types of devices,such as ferroelectric tunnel junctions 15or magnetic tunnel junctions 4with ferroelectric tunnel barriers.In this latter type of structure,a control of the ferroelectric polarization by an external magnetic ?eld can be envisaged,via the magnetoelectric effect.

The authors are very grateful to M.Viret and D.Colson for providing the BFO target and to J.-M.Triscone for fruit-ful discussions.Two of the authors ?G.H.and M.S.?acknowledge ?nancial support from the Ministère de l’Education Nationale,de l’Enseignement Supérieur et de la Recherche ?France ?.Another author ?H.B.?acknowledges ?nancial support from the Conseil Général de l’Essone ?France ?.

1

S.A.Wolf,D.D.Awschalom,R.A.Buhrman,J.M.Daughton,S.von Molnar,M.L.Roukes,A.Y .Chtchelkanova,and D.M.Treger,Science 294,1488?2001?.2

A.H.MacDonald,P.Schiffer,and M.Samarth,Nat.Mater.4,195?2005?.3

J.M.D.Coey and M.Venkatesan,J.Appl.Phys.91,8345?2002?.4

J.S.Moodera,L.R.Kinder,T.M.Wong,and R.Meservey,Phys.Rev.Lett.74,3273?1995?.5

M.Bowen,M.Bibes, A.Barthélémy,J.-P.Contour, A.Anane,Y .Lema?tre,and A.Fert,Appl.Phys.Lett.82,233?2003?.6

N.A.Hill,J.Phys.Chem.B 104,6694?2000?.7

M.Fiebig,J.Phys.D 38,R123?2005?.8

N.A.Hill and K.M.Rabe,Phys.Rev.B 59,8759?1999?.9

J.Wang,J.B.Neaton,H.Zheng,V .Nagarajan,S.B.Ogale,B.Liu,D.Viehland,V .Vaithyanathan,D.G.Schlom,U.V .Waghmare,N.A.Spaldin,K.M.Rabe,M.Wuttig,and R.Ramesh,Science 299,1719?2003?.10

W.Eerenstein,F.D.Morrison,J.Dho,M.G.Blamire,J.F.Scott,and N.D.Mathur,Science 307,1203a ?2005?.11

G.A.Smolenskii,V .M.Yudin,E.S.Sher,and Y .E.Stolypin,Sov.Phys.JETP 16,622?1963?.12

P.Fischer,M.Polomska,I.Sosnowska,and M.Szymanski,J.Phys.C 13,1931?1980?.13

C.Binek and B.Doudin,J.Phys.:Condens.Matter 17,L39?2005?.14

C.Binek, A.Hochstrat,X.Chen,P.Borisov,W.Kleemann,and B.Doudin,J.Appl.Phys.97,10C514?2005?.15

J.Rodríguez Contreras,H.Kohlstedt,U.Poppe,R.Waser,C.Buchal,and N.A.Pertsev,Appl.Phys.Lett.83,4595?2003?.16

H.Kohlstedt,N. A.Pertsev,J.Rodríguez Contreras,and R.Waser Phys.Rev.B 72,125341?2005???17

M.Ye.Zhuravlev,R.F.Sabirianov,S.S.Jaswal,and E.Y .Tsymbal,Phys.Rev.Lett.94,246802?2005?.18

J.-L.Maurice,R.Lyonnet,and J.-P.Contour,J.Magn.Magn.Mater.211,91?2000?.19

H.Béa,M.Bibes, A.Barthélémy,K.Bouzehouane, E.Jacquet,A.Khodan,J.-P.Contour,S.Fusil,F.Wyczisk,A.Forget,D.Lebeugle,D.Colson,and M.Viret,Appl.Phys.Lett.87,072508?2005?20

G.Xu,H.Hiraka,G.Shirane,J.Li,J.Wang,and D.Viehland,Appl.Phys.Lett.86,182905?2005?.21

X.Qi,M.Wei,Y .Lin,Q.Jia,D.Zhi,J.Dho,M.G.Blamire,and J.L.MacManus-Driscoll,Appl.Phys.Lett.86,071913?2005?.22

A.Urushibara,Y .Moritomo,T.Arima, A.Asamitsu,G.Kido,and Y .Tokura,Phys.Rev.B 51,14103?1995?.23

J.-L.Maurice,F.Pailloux,A.Barthélémy,A.Rocher,O.Durand,R.Lyonnet,and J.-P.Contour,Appl.Surf.Sci.188,176?2002?.24

F.Houzé,R.Meyer,O.Schneegans,and L.Boyer,Appl.Phys.Lett.69,1975?1996?.25

P.Paruch,T.Tybell,and J.-M.Triscone,Appl.Phys.Lett.79,530?2001?.26

C.Lichtensteiger and J.-M.Triscone,Phys.Rev.Lett.94,047603?2005?

.

FIG.3.3?3?m 2morphology ?a ?and resistance ?b ?maps of the BFO ?5nm ?/LSMO bilayer measured simultaneously by a CTAFM.?c ?Logarithm of the average resistance of these maps for different BFO thicknesses.The error bars correspond to the full width at half maximum of the resistance

distribution.

FIG.4.Topography ?a ?and PFM image ?b ?after writing stripe domains using the AFM tip as a top electrode of the 5nm BFO layer.PFM image ?c ?after writing two wider stripes perpendicularly to the pattern seen in ?b ?.