Direct observation of electronic inhomogeneities induced by point defect disorder inmanganite filmsM. Sirena, A. Zimmers, N. Haberkorn, E. Kaul, L. B. Steren, J. Lesueur, T. Wolf, Y. Le Gall, J.-J. Grob, and G.Faini Citation: Journal of Applied Physics 107, 113903 (2010); doi: 10.1063/1.3428458 View online: http://dx.doi.org/10.1063/1.3428458 View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/107/11?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Epitaxial stabilization of ultra thin films of electron doped manganites Appl. Phys. Lett. 104, 202409 (2014); 10.1063/1.4879456 The influence of point defects and inhomogeneous strain on the functional behavior of thin film ferroelectrics Appl. Phys. Lett. 94, 212905 (2009); 10.1063/1.3132583 Impact of near-surface native point defects, chemical reactions, and surface morphology on ZnO interfaces J. Vac. Sci. Technol. B 26, 1477 (2008); 10.1116/1.2919158 Lattice-strain control of magnetoresistivity and magnetoresistive anisotropy induced by ion-beam milling in La0.65 Ca 0.35 Mn O 3 films near the metal-insulator transition Appl. Phys. Lett. 92, 022501 (2008); 10.1063/1.2827192 Iodization of rf sputter induced disordered Ag thin films reveals volume plasmon-exciton “transition” J. Appl. Phys. 100, 064314 (2006); 10.1063/1.2353238

[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:

209.183.183.254 On: Wed, 10 Dec 2014 03:24:38

Direct observation of electronic inhomogeneities induced by point defectdisorder in manganite films

M. Sirena,1,a� A. Zimmers,3 N. Haberkorn,1 E. Kaul,1 L. B. Steren,2 J. Lesueur,3 T. Wolf,3

Y. Le Gall,4 J.-J. Grob,4 and G. Faini51Instituto Balseiro, Univ. Nac. de Cuyoand CNEA, Av. Bustillo 9510, Bariloche, Rio Negro 8400, Argentina2Centro Atómico Constituyentes, Av. Gral. Paz 1499, San Martín 1650, Buenos Aires, Argentina3UPR5-LPEM-CNRS, Physique Quantique, ESPCI, 10 Rue Vauquelin, 75231 Paris, France4Institut d’Électronique du Solide et des Systèmes, UMR 7163, 23 rue du Loess–BP20, F-67037 StrasbourgCedex 02, France5LPN-CNRS, Route de Nozay, 91460 Marcoussis, France

�Received 27 January 2010; accepted 15 April 2010; published online 1 June 2010�

We have investigated the influence of point defect disorder in the electronic properties of manganitefilms. Real-time mapping of ion irradiated samples conductivity was performed through conductiveatomic force microscopy �CAFM�. CAFM images show electronic inhomogeneities in the sampleswith different physical properties due to spatial fluctuations in the point defect distribution. Asdisorder increases, the distance between conducting regions increases and the metal-insulatortransition shifts to lower temperatures. Transport properties in these systems can be interpreted interms of a percolative model. The samples saturation magnetization decreases as the irradiation doseincreases whereas the Curie temperature remains unchanged. © 2010 American Institute of Physics.�doi:10.1063/1.3428458�

I. INTRODUCTION

Since their rediscovery in the past decade,1

A1−xAx�MnO3 �A: La, Pr; A�: Sr, Ba, and Ca� manganiteshave attracted a lot of attention in the scientific community.The renewed interest in these materials is due to their colos-sal magnetoresistance effect and its possible technologicalapplications. From early works it became clear that disorder,through changes in the Mn–O distance and the Mn–O–Mnangle, plays an important role in the transport and magneticproperties of bulk and thin manganite films.2–8 Many resultson this subject were explained in terms of phase-separationor phase coexistence9–11 in these systems. However, the ori-gin of this phenomenon is still a matter of discussion.Shenoy et al.12 recently proposed that Coulomb interactionsmay be responsible of electronic inhomogeneities in manga-nites rather than disorder or phase competition. Conductiveatomic force microscopy �CAFM�,13 through direct mappingof the system’s conductivity has proven to be a powerful toolto study this kind of problem. Last year, phase separation inpolycrystalline manganite films was established by this mea-surement method.14

In this work, vacancy-interstitial pairs were introducedby ion implantation into ferromagnetic La0.75Sr0.25MnO3 thinfilms in order to induce lattice disorder in a controlledway.15,16 Ion irradiation has recovered importance for the de-veloping of new technological devices like planar tunneljunctions. Typically ferromagnetic tunnel junctions present acurrent perpendicular to the plane configuration. However,their fabrication is complex, involves several lithographysteps, and suffers from pinholes in the thin insulating barrier.Irradiation has been successfully used by the authors to fab-

ricate planar high temperature superconductor �HTc� Joseph-son junctions and high transition temperature �Tc� supercon-ducting devices17 which present the same technologicalchallenges than oxide tunnel junctions, from the fabricationpoint of view. The irradiation can be used to render insulat-ing the manganite and draw the mesa circuits. Planar tunneljunctions can be fabricated creating a thin insulator barrierby ion irradiation through a photoresist mask with a smallaperture across, done with electronic lithography. The planartunnel junctions can be placed directly and easily in the cir-cuit opening the corresponding holes in the photoresist mask.Of course, studying the influence of ion irradiation in thephysical properties of manganite films becomes very impor-tant to achieve this objective. CAFM was used to obtaindirect images of the influence of point defect disorder in thephase-separation of manganite films. Transport and magneticproperties of the samples were analyzed in the frame of thismodel.

II. EXPERIMENTAL DETAILS

Thin 50 nm La0.75Sr0.25MnO3 films were grown onsingle-crystal SrTiO3 �100� substrates by dc magnetron sput-tering from a stoichiometric ceramic target. More detailsabout the films fabrication method and characterization aregiven elsewhere.18 The films were irradiated with oxygenions �O+� at 150 keV with different doses, ranging from �=1�1014 to �=10�1014 ion /cm2. The choice of the oxy-gen energy was made considering the authors previous expe-rience with oxygen implantation into HTc materials. The 150keV O+ ions are not expected to alter the material, eitherchemically or physically. The ions can be stopped by metal-lic or photoresist masks of appropriate thickness, which isimportant as mentioned before for the fabrication of differentdevices.17 The damage profile calculated by TRIMa�Electronic mail: martin.sirena@hotmail.com.

JOURNAL OF APPLIED PHYSICS 107, 113903 �2010�

0021-8979/2010/107�11�/113903/4/$30.00 © 2010 American Institute of Physics107, 113903-1

[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:

209.183.183.254 On: Wed, 10 Dec 2014 03:24:38

simulations19 is constant along the film depth for 50 nm filmsthick. The resistivity of the samples has been measured usinga standard four-probes configuration. The temperature andfield dependence of the magnetization have been studied us-ing a superconductor quantum interference device magneto-meter. CAFM measurements were performed in a Veeco Di-mension 3100® scanning probe microscope �SPM� with aCAFM module. The scans were performed using a diamonddoped with boron conductive tip in contact mode. The mea-surements have been made keeping the probe polarizationvoltage fixed to 0.5 V, with the same deflection setpoint �0.4V�. Tests, performed with different probe polarizations anddeflection setpoints have been done in order to assure thatCAFM results were independent of the measurements condi-tions.

III. RESULTS

Figure 1 presents the topographic �left� and CAFM cur-rent �right� images of irradiated La0.75Sr0.25MnO3 �LSMO�films. The pristine sample presents very low roughness��0.2 nm�. The images put in evidence the correlation be-tween the irradiation of the samples and their roughness: theincrease of the irradiation dose leads to an augmentation ofthe films roughness reaching 4 nm for the highest dose ��=10�1014 ion /cm2�. CAFM images show that even thenonirradiated sample presents a finite and spatially inhomo-geneous distribution of conductivity. We define the conduct-ing regions as those that show nonzero CAFM current whileinsulator zones represent those where no CAFM current isobserved for the standard measuring conditions. In the im-ages, the conducting islands are clear while the insulator re-gions are dark. The scale lengths of the conducting and in-sulator regions in the pristine sample are clearly differentfrom the ones measured in the irradiated samples, even forthe lowest irradiation dose. The origin of the inhomogene-ities in pristine thin films may be associated to the intrinsic Acation disorder and to inhomogeneous strain field due to thedifferent A cation radii in manganite compounds. On theother hand, the electronic inhomogeneity observed in the ir-radiated samples would be probably associated to the ran-dom distribution of point defect, strain field induced by theion bombardment or inhomogeneous changes of the Mn3+/4+

ratio to reduce the lattice strain in the irradiated areas.The influence of point defect disorder in the conductivity

of the irradiated samples can be clearly seen in a histogramof conducting surface as function of the CAFM current, i.e.,conductivity �shown in Fig. 2 for three irradiation doses�.The histograms present a single continuum distribution ofconductivity for each sample. Continuum conductivity histo-grams were also found by STM measurements onPr0.68Pb0.32MnO3 single crystals20 and seem to indicate thatthere is not a phase separation in the system. Instead, therewould be an arrangement of regions with proper characteris-tics and physical properties due to the inhomogeneous distri-bution of strain and point defects at the nanoscale. I�V�curves �Fig. 3� for the so-called conducting islands and insu-lator regions are nonlinear in agreement with recent resultsobtained on polycrystalline films.14 However, it is clearly

discernible from these curves that the energy gap in the con-ducting regions �Fig. 3�a�� is much smaller than the one mea-sured for the insulator regions �Fig. 3�b��, when the irradia-tion dose increases the gap also increases �Fig. 3�c��. Atroom temperature, the carries in the different regions are stilllocalized but the energy “gap” is smaller for reduced disor-der. This means that for the same applied voltage, the CAFMcurrent is higher in the “metallic” regions.

The pristine sample is ferromagnetic with a transitiontemperature, Tc, of 320 K and a total saturation magnetiza-tion of 545 emu /cm3 in agreement with the calculated valueconsidering the Mn3+ /Mn4+ ratio and its lattice parameter.The sample presents a metal-insulator transition at a charac-

a)

b)

c)

d)

FIG. 1. �Color online� Topographic �left� and CAFM �right� images �5�5 �m2� of irradiated LSMO films for different irradiation doses: �a� �=0 ion /cm2, �b� �=3�1014 ion /cm2, �c� �=7.5�1014 ion /cm2, and �d��=10�1014 ion /cm2.

113903-2 Sirena et al. J. Appl. Phys. 107, 113903 �2010�

[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:

209.183.183.254 On: Wed, 10 Dec 2014 03:24:38

teristic temperature �TMI� around the Tc, as expected fromthe double exchange model.21 As the irradiation dose in-creases the saturation magnetization decreases but the Tcdoesn’t change. However, the TMI systematically decreasesfor increasing irradiation doses and for the highest dose thesample becomes insulator in all the measured temperaturerange. This TMI decoupling from the magnetic transition tem-perature is in contradiction with the early believe that for ionirradiated films Tc follows the TMI,

15 but has already beenobserved in previous works as a result of increasingdisorder.22 A detailed study of the magnetic and transportproperties of the ion irradiated films will be presentedshortly.23

Figure 4 presents the saturation magnetization, measuredat 5 K and the magnetic transition temperature for thesamples with different irradiation doses, as function of thetotal conducting area obtained from the CAFM images. Thetotal conducting area is obtained integrating the surface ofthe areas presenting nonzero CAFM current. A clear correla-tion between the saturation magnetization and the total con-ducting area can be observed: as the irradiation dose in-creases, the total conducting area decreases and there is areduction in the saturation magnetization. This suggests thatthe conducting regions are related to the ferromagnetism inthe sample. The observed reduction in the magnetization

with increasing disorder, while the magnetic transition re-mains unchanged, is rather unexpected. We believe that themicroscopic origin of this behavior can be related to the ex-istence of conducting ferromagnetic untouched regions sepa-rated from insulating “damaged” regions. The number andvolume of the untouched regions is reduced dramatically asthe irradiation dose increases. This decrease in the number ofuntouched areas rapidly reduces the magnetization with in-creasing irradiation dose, but of course the measured Tcdoesn’t change since it is given by the magnetic transitionscorresponding to these undamaged regions.

Figure 5 shows the metal-insulator transition tempera-ture obtained from resistance versus temperature measure-ments as function of the mean distance between conductingregions, obtained graphically from the CAFM images. Wehave measured with the software the border to border dis-tance of around 30 random conducting areas of differentsizes and distances in order to estimate its mean value anddispersion. As the irradiation dose increases the distance be-tween conducting areas increases slowly. Figure 5 indicatesthat as the distance between the conducting islands increases,the films TMI decreases. The metal-insulator transition in

0.1 1 10 100

0.38

3.81

38.11

co

ve

red

su

rfac

e(%

)

CAFM current (nA)

Increasing dpa

Vpol = 500 mV

Non irradiated

FIG. 2. �Color online� Histograms of the covered surface for different val-ues of the CAFM current �i.e., conductivity� of the irradiated LSMOsamples at room temperature.

-4 -2 0 2 4-2

-1

0

1

2

(c)

(b)

I(n

A)

V (V)

T = 300 K(a)

FIG. 3. Typical I�V� curves for �a� the conducting regions corresponding tothe sample irradiated with a dose of 3�1014 ion /cm2, �b� insulating regionsof the same sample and �c� insulating regions of the sample irradiated with4.5�1014 ion /cm2 �c�.

10 20 30 40 50 60 7010

100

1000

10

100

1000

Tc

(K)

Ms

(a.u

)

conducting covered surface (%)

FIG. 4. �Color online� Saturation magnetization and magnetic transitiontemperature as function of the covered conducting surface at room tempera-ture. The inset shows a typical detail of the CAFM image with highlightedconducting and insulating regions. Lines are the best linear fit of the experi-mental data.

0 20 40 60 80 100 120 140 160

0

50

100

150

200

250

300

TM

I(K

)

<d> (nm)

FIG. 5. �Color online� Metal-insulator transition temperature as function ofthe mean distance between conducting regions. The line is a linear fit of thedata.

113903-3 Sirena et al. J. Appl. Phys. 107, 113903 �2010�

[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:

209.183.183.254 On: Wed, 10 Dec 2014 03:24:38

these inhomogeneous systems can be described with a per-colative model.24,25 When the temperature decreases the in-sulating regions gradually increase due to the decrease in thethermally activated conductivity. However, below a criticaltemperature these areas increase their conductivity and theirsize.14 At the macroscopic TMI a percolation of the conduct-ing islands occurs. As the point defect disorder increases,lower temperatures are required for the system to undergothe percolative transition. This explains the decoupling be-tween the magnetic transition and the metal-insulator transi-tion observed in these systems. Generally, ion irradiation cre-ates vacancy-interstitial �cation and oxygen� pairs, locallyinducing lattice strain enhancement26 and Mn3+ /Mn4+ ratiodepletion27 which are responsible for the observed results.

IV. CONCLUSIONS

We have found that point defect disorder influences in acomplex way the transport and magnetic properties of man-ganite films. Increasing the density of vacancy-interstitialpairs not only increases the systems resistance but also in-creases its electronic inhomogeneity. Transport properties inthese systems can be interpreted in terms of a percolativemodel, as disorder increases the distance between conductingregions increases, decreasing the macroscopic metal-insulation transition temperature. We have also found that asdisorder increases, the saturation magnetization of thesamples decreases. However, the magnetic transition tem-perature remains unchanged.

ACKNOWLEDGMENTS

The authors acknowledge K. Bouzehouane and S. Fusilfor the formation received in CAFM measurements; Micraand Veeco crew for extraordinary technical support. Thiswork was partially supported by the ANPCYT �Grant Nos.PICT 06 2092 and PICT 05 33304�. M.S., N.H., and L.B.S.are members of CONICET, Argentina. L.B.S is a fellow ofthe Guggenheim Foundation.

1R. von Helmolt, J. Wecker, B. Holzapfel, L. Schultz, and K. Samwer,Phys. Rev. Lett. 71, 2331 �1993�.

2J. Goodenough, Interscience Monographs on Chemistry �Wiley, NewYork, 1963�, Vol. I.

3A. Millis, T. Darling, and A. Migliori, J. Appl. Phys. 83, 1588 �1998�.4J. Fontcuberta, V. Laukhin, and X. Obradors, Appl. Phys. Lett. 72, 2607�1998�.

5L. Rodriguez-Martinez and J. P. Attfield, Phys. Rev. B 54, R15622 �1996�.6W. Zhang, X. Wang, M. Elliott, and I. Boyd, Phys. Rev. B 58, 14143�1998�.

7H. Ju, K. Krishnan, and D. Lederman, J. Appl. Phys. 83, 7073 �1998�.8J. Aarts, S. Freisen, R. Hendrikx, and H. Zandbergen, Appl. Phys. Lett. 72,2975 �1998�.

9T. Becker, C. Streng, Y. Luo, V. Moshnyaga, B. Damaschke, N. Shannon,and K. Samwer, Phys. Rev. Lett. 89, 237203 �2002�.

10A. Biswas, M. Rajeswari, R. C. Srivastava, T. Venkatesan, R. L. Greene,Q. Lu, A. L. de Lozanne, and J. Millis, Phys. Rev. B 63, 184424 �2001�.

11A. Tebano, C. Aruta, P. G. Medaglia, F. Tozzi, G. Balestrino, A. A. Si-dorenko, G. Alloni, R. De Renzi, G. Ghiringhelli, C. Dallera, and L. Bari-covich, Phys. Rev. B 74, 245116 �2006�.

12V. B. Shenoy, T. Gupta, H. R. Krishnamurthy, and T. V. Ramakrishnam,Phys. Rev. Lett. 98, 097201 �2007�.

13M. Bibes, M. Bowen, A. Barthelemy, A. Anane, K. Bouzehouane, C.Carretero, E. Jacket, J. P. Contour, and O. Durand, Appl. Phys. Lett. 82,3269 �2003�.

14Y. H. Chen and T. B. Wu, Appl. Phys. Lett. 93, 224104 �2008�.15C.-H. Chen, V. Talyansky, C. Kwon, M. Rajeswari, R. P. Sharma, R.

Ramesh, T. Venkatesan, J. Meingailis, Z. Zhang, and W. Chu, Appl. Phys.Lett. 69, 3089 �1996�.

16J. F. Gibbons, Proc. IEEE 60, 1062 �1972�.17N. Bergeal, J. Lesueur, M. Sirena, G. Faini, M. Aprili, J. P. Contour, and

B. Leridon, J. Appl. Phys. 102, 083903 �2007�.18M. Sirena, L. Steren, and J. Guimpel, Thin Solid Films 373, 102 �2000�.19J. F. Ziegler, J. P. Biersack, and U. Littmark, The Stopping and Range of

Ions in Solids, Pergamon Press, New York �1985�.20S. Rößler, S. Ernst, B. Padmanabhan, S. Elizabeth, H. L. Bhat, F. Steglich,

and S. Wirth, Europhys. Lett. 83, 17009 �2008�.21C. Zener, Phys. Rev. 81, 440 �1951�.22L. B. Steren, M. Sirena, and J. Guimpel, J. Magn. Magn. Mater. 211, 28

�2000�.23M. Sirena, A. Zimmers, N. Haberkorn, E. Kaul, L. B. Steren, J. Lesueur,

T. Wolf, Y. Le Gall, J.-J. Grob, and G. Faini, Phys. Rev. B 81, 134439�2010�.

24M. Uehara, S. Mori, C. H. Chen, and S.-W. Cheong, Nature �London� 399,560 �1999�.

25N. Mathur and P. Littlewood, Phys. Today 56, 25 �2003�.26R. Bathe, K. P. Adhi, S. I. Patil, G. Marest, B. Honneyer, and S. Ogale,

Appl. Phys. Lett. 76, 2104 �2000�.27R. Bathe, S. I. Patil, K. P. Adhi, B. Hannoyer, and G. Marest, J. Appl.

Phys. 93, 1127 �2003�.

113903-4 Sirena et al. J. Appl. Phys. 107, 113903 �2010�

[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:

209.183.183.254 On: Wed, 10 Dec 2014 03:24:38