JOURNAL OF APPLIED PHYSICS

VOLUME 89, NUMBER 3

1 FEBRUARY 2001

Magnetoresistance in manganiteÕalumina nanocrystalline composites L. E. Hueso and J. Rivasa) Departamento de Fı´sica Aplicada, Universidad de Santiago de Compostela, 15706 Santiago de Compostela, Spain

F. Rivadulla and M. A. Lo´pez-Quintela Departamento de Quı´mica-Fı´sica, Universidad de Santiago de Compostela, 15706 Santiago de Compostela, Spain

共Received 13 July 2000; accepted for publication 9 November 2000兲 Magnetotransport properties of manganite/insulator composites are studied in this article. By merging the half metallic character of La2/3Ca1/3MnO3 and the mixed composition with alumina grains dispersed in the structure we have been able to increase three times intergranular magnetoresistance around the percolation threshold. The transport properties have been studied employing a theoretical model for ferromagnetic–insulator systems and a two channel equation in order to reproduce the behavior of resistivity in the whole temperature range. The percolation theory is introduced to try to understand and improve the fascinating properties of these mixed systems. © 2001 American Institute of Physics. 关DOI: 10.1063/1.1338518兴

I. INTRODUCTION

article is to study the magnetotransport properties of fine particle composites designed with ferromagnetic La0.67Ca0.33MnO3 and insulating Al2O3 nanoparticles. We have chosen these compounds for different reasons: the manganite proposed has been studied for several years and its magnetotransport properties are well documented. It is also possible to obtain particles in the nanometric range using the sol–gel technique. On the other hand, alumina is a wellknown inert insulator, usually employed in junction magnetoresistive devices,19 which may be used in order to study the breakdown of the conductivity. Additionally, its particle size may be nanometric and following this protocol we have tried to enhance the properties of the composite by increasing the grain surface.

Magnetoresistance 共MR兲 effects have been one of the most frequently studied topics in solid state physics since the discovery of giant magnetoresistance in Fe/Cr multilayers,1 and a few years after in heterogeneous alloys with ferromagnetic granules embedded in a nonmagnetic metallic matrix.2,3 In this scenario, revisited early studies of the 1950s4 on mixed-valence manganese perovskites led to the fabrication of high quality thin films that had considerable magnetoresistive effects,5,6 the so-called colossal MR 共CMR兲. CMR is restricted to a narrow range of temperatures around the ferromagnetic–paramagnetic phase transition, and this makes it difficult to apply to electronic devices. However, much effort was made to understand the physical properties of this kind of compound, which led to the discovery of another type of MR, intergranular MR.7 It is present at low temperatures in polycrystalline materials, such as ceramics,7 fine particle systems,8 or granular films,9 and can be significantly larger than the intrinsic CMR of manganites. Its origin is attributed to spin tunneling trough grain boundaries, and the clue to the great values reported lies in the nearly full spin polarization of manganites.10 This property has been extensively explored in trilayers,11 artificial grain boundaries,12 spin injection devices,13 etc., and has also led us to consider other compounds that only share the halfmetallic character with manganese perovskites, such as Fe3O4, 14 CrO2, 15 or Sr2FeMoO6. 16 In the last year, several groups have tried to merge the high degree of spin polarization present in manganites with the advantages of a heterogeneous granular structure.17,18 The mixture of a metallic ferromagnetic perovskite with an insulator is followed by magnetoresistance enhancement, especially near the percolation threshold. The propose of this

II. EXPERIMENT

For this study, the sol–gel method was used to prepare La0.67Ca0.33MnO3 共LCMO兲 nanoparticles with urea as a gelifying agent. This procedure has been explained in detail in previous articles.20,21 The final sintering temperature was 900 °C, which led to a mean grain size of 150 nm. Commercial alumina nanoparticles with a mean grain size of 100 nm were provided by Goodfellow Cambridge Ltd. The combinations (1⫺x) La0.67Ca0.33MnO3 ⫹x Al2O3 共with x⫽0%, 5,5%, 8%, 15%, and 25% in volume兲 were mixed in a ballmilling agate mortar for 1 h. The resulting homogeneous powder was pelletized at a pressure of 3 tons/cm2, and then sintered rapidly for 1 h at 1100 °C. Resistivity was measured by the standard four-probe method in the range 77–300 K and in fields up to 10 kOe. Magnetization was measured with a vibrating sample magnetometer in the same temperature and field range. X-ray diffraction was used to examine the crystalline structure and the particle size and shape were investigated by means of scanning electronic microscopy.

a兲

Author to whom correspondence should be addressed; electronic mail: [email protected]

0021-8979/2001/89(3)/1746/5/$18.00

1746

© 2001 American Institute of Physics

Downloaded 17 Dec 2002 to 131.111.8.102. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/japo/japcr.jsp

Hueso et al.

J. Appl. Phys., Vol. 89, No. 3, 1 February 2001

1747

FIG. 1. SEM photograph of 15% alumina sample. The brightest regions indicate alumina grains, as was proved by energy dispersive analysis by xray microanalysis. Pores and empty regions are also observable.

III. RESULTS AND DISCUSSION

X-ray diffraction pattern analyses show that the reflection peaks of LCMO do not shift. The peaks corresponding to Al2O3 are also observable, and their intensity increases as alumina percentage does. This is an indication that no reaction between the two compounds occurred, without which it is impossible to draw conclusions about the composite properties. Grain size increased up to around 500 nm in the final sintering process. In the composite samples, alumina regions present a different contrast, and are seen as particles dispersed among the LCMO granules 共see Fig. 1兲. On the other hand, Curie temperature T C remains unchanged around 260 K. The general behavior of magnetization versus temperature curves indicates a long-range ferromagnetism in all the cases. The saturation magnetization value at 80 K is obtained from high-field magnetization curves, fitting these experimental data to the saturation approach law22

冉

M 共 T 兲 ⫽M S 1⫺

冊

b a ⫺ 2 , H H

共1兲

where a and b are suitable constants. The results can be seen in Fig. 2, and it is clear that the decrease in M S is due only to the alumina percentage present in the sample. The magnetic contribution is reduced by the same amount as the alumina percentage, which leads us to disregard any kind of extra magnetic disorder. The main point of this article is the discussion about magnetotransport properties. Resistivity increases in almost an exponential manner when it is plotted against alumina concentration 共see Fig. 3兲. The percolation point where resistivity verifies a huge upturn can be seen. The value obtained at 100 K is more than 3 orders of magnitude larger for a 25% sample than for the undoped one. It is also clear that at all the temperatures studied, the resistivity versus alumina percentage curves follow the same trend. Moreover, pure LCMO undergoes the characteristic metal–insulator transition at 270 K (T M – I ), which is also nearly coincident with Curie temperature. Doping depresses metallic conduction and induces a considerable decrease in T M – I , which is more than 140 K for 5.5% of alumina in our

FIG. 2. Saturation magnetization at 80 K for various composite compositions. We can see that the decrease in magnetization is proportional to the alumina percentage 共x兲 present in the sample. The line is a linear fit to the data.

sample. This transition point is completely suppressed in our experimental measurement range for doping levels higher than 8% 共Fig. 4兲. As a result, the sample is within a ferromagnetic–insulator regime for an extensive range of temperature. This behavior is clearly not homogeneous, as expected for such a distorted structure. To avoid any mistakes, we must point out that this increase in resistivity and the disappearance of the metal–insulator transition are not due to the doping inside LCMO particles, but exclusively to the increase in the electron scattering with alumina grains placed in the microstructure. It is not an intrinsic property but an extrinsic one to LCMO particles. Resistivity in the ferromagnetic–insulator region follows the law for conductivity in granular metals embedded in an insulator that was first proposed by Sheng et al.23

FIG. 3. Resistivity at two temperatures 共100 and 300 K兲 vs alumina percentage. A great upturn arises for values of x higher than 10%.

Downloaded 17 Dec 2002 to 131.111.8.102. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/japo/japcr.jsp

1748

Hueso et al.

J. Appl. Phys., Vol. 89, No. 3, 1 February 2001

FIG. 4. Temperature dependence of the resistivity for different percentages. Metal–insulator transition temperature (T M – I ) drops, and is also suppressed by alumina contribution.

再冑 冎

␳ 共 T 兲 ⫽a exp 2

E , kT

共2兲

where a is a constant and E is the activation energy, which is proportional to the tunnel-barrier thickness and the charging energy of the metallic grains. The data of the relation between ␳ and 1/T 1/2 are shown in Fig. 5. We can clearly see that in the ferromagnetic–insulator regime, between T C and T M – I the linear dependence is satisfactory, but not as good as one might expect. However, more sophisticated relations proposed by the same authors24 such as

再 冑冎

1 1 ⫽ exp ⫺2 ␳共 T 兲 ␳T

E 1 ⫹ kT ␳C

共3兲

FIG. 5. Resistivity plotted against T ⫺1/2 in order to indicate a temperature range where the samples follow the law predicted for granular metals in an insulator.

FIG. 6. Experimental resistivity temperature dependence 共solid lines兲 and fits to the two-channel model proposed in the text 关Eq. 共4兲兴 共dotted lines兲. Not all alumina compositions are plotted for clarity.

do not offer better results. We have reproduced the resistivity curve throughout the whole range measured for all the samples employing a two channel model, that is, two resistances in parallel, weighting each one with a different fitting factor: 1 A B ⫽ ⫹ , ␳ 共 T 兲 ␳ LCMO共 T 兲 ␳ E 共 T 兲

共4兲

where in this expression ␳ LCMO(T) indicates the resistivity inside the LCMO grains. As we have no available resistivity data on La0.67Ca0.33MnO3 single crystals, we have taken our own results from a high temperature sintered ceramic sample with grain size larger than 20 ␮m and residual resistivity ␳ 0 ⬇1 m⍀ cm. 25 ␳ E (T) represents an insulating channel created by the extra electron scattering with alumina regions. In the interest of simplicity, this term is represented as a semiconductor like resistivity: exp(E/kT), with the activation energy obtained from first-order high temperature fits (E ⫽100 meV兲. Fits to experimental results are satisfactory 共Fig. 6兲, although not completely perfect because of the lack of this simple model. In any case, these results indicate that the fundamental physics of the problem are taken into consideration. MR behaves in an interesting manner. The low temperature value (T⫽77 K) is increased nearly three times by adding 8% of alumina to pure LCMO 共Figs. 7 and 8兲. The low field sensitivity 共defined as the maximum slope of MR versus H curve兲 behaves in the same way. These results are very similar to those obtained by other groups17,18 employing related systems and they are associated with the extra disorder induced by the alumina regions present in the sample. At this point, we must adduce percolation as the underlying effect most likely to explain the magnetotransport results presented before. First, it is necessary to justify the percolation threshold value obtained, which is around 10%

Downloaded 17 Dec 2002 to 131.111.8.102. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/japo/japcr.jsp

Hueso et al.

J. Appl. Phys., Vol. 89, No. 3, 1 February 2001

1749

stay together through dipolar interactions. This increases the facility of alumina grains to form the percolation pathway. Such a scenario makes it easy to understand a reduction from 20% to 10%. The MR upturn around the percolation value in granular materials is a well known phenomena17,18,27 shown here again, and could be related to the formation of an infinite cluster that will enormously increase spin tunneling between ferromagnetic regions. The most common theoretical explanation for intergranular MR inserts an additional magnetic exchange energy arising when the magnetic moments of the neighboring grains are not parallel and the electron spin is conserved in tunneling.28 This assumption only takes into account linear terms in magnetic energy E m , and in the case of ferromagnetically coupled grains, the MR is given by

冉 冊

JP ⌬␳ ⫽⫺ 关 M 2 共 H,T 兲 ⫺M 2 共 0,T 兲兴 , ␳0 4kT FIG. 7. MR at T⫽77 K vs alumina percentage. The optimum doping value for MR enhancement is around 10%.

for our composites. The percolation value for a concrete structure is related to the packing fraction through the following relation: p C⫽

0.16 , ␯

共5兲

where p C is the critical percolation value, ␯ is the packing fraction, and 0.16 is an almost structure independent constant.26 Picnometry measurements have enabled us to estimate the packing fraction of our samples as a value around 0.8. This leads to a percolation value of 0.2 共that is, 20%兲, which is higher than 10%, the value obtained from our transport data. However, if we take into account interactions between grains 共as studied, for example, in Ref. 26兲, we may justify the reduction of the theoretical value to the experimental one. Grains of the magnetic material should tend to

共6兲

where J is the exchange coupling constant, P is electronic polarization, and M is magnetization. Within this model, it could be possible to predict both temperature and field MR dependences. Our results suggest that MR and magnetization at a fixed temperature do not follow this reliance. This contradicts previously published results,7,25,27 and the reason for that is the broken relation between transport and magnetic properties in the composite sample due to the alumina presence. New energy terms should be taken into account in the theoretical approach to explain MR behavior. IV. CONCLUSION

Manganite/insulator composite systems have been produced and carefully characterized by several experimental techniques. Through them we have detected a complex microstructure and a very rich phenomenology. Electrical transport is progressively destroyed as alumina concentration increases, and the extent can be qualitatively predicted by a two channel model presented in the text. We have been able to observe percolation threshold in this system for 10% of the insulating component, and around this critical point a great increase in intergranular MR has been detected. This is the most stimulating result obtained and could be the breakthrough needed to improve the MR response in granular systems. ACKNOWLEDGMENTS

˜ eiro for stimulating discussions. We want to thank Y. Pin Two of the authors 共L.E.H. and F.R.兲 want to thank M.E.C. of Spain for Ph.D. grants. This work was developed as a part of Project No. MAT98-0416. 1

FIG. 8. MR isotherms at 77 K for three different composites. The different form of the MR curves can be seen clearly.

M. N. Baibich, J. M. Broto, A. Fert, F. Nguyen Van Dau, F. Petroff, P. Etienne, G. Creuzet, A. Friederich, and J. Chazelas, Phys. Rev. Lett. 61, 2472 共1988兲. 2 A. E. Berkowitz, J. R. Mitchell, M. J. Carey, A. P. Young, S. Zhang, F. E. Spada, F. T. Parker, A. Hutten, and G. Thomas, Phys. Rev. Lett. 68, 3745 共1992兲. 3 J. Q. Xiao, J. S. Jiang, and C. L. Chien, Phys. Rev. Lett. 68, 3749 共1988兲. 4 G. H. Jonker and J. H. Van Santen, Physica 共Amsterdam兲 16, 337 共1950兲. 5 R. von Hemlolt, J. Wecker, B. Holzapfel, L. Schultz, and K. Samwer, Phys. Rev. Lett. 71, 2331 共1993兲.

Downloaded 17 Dec 2002 to 131.111.8.102. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/japo/japcr.jsp

1750 6

Hueso et al.

J. Appl. Phys., Vol. 89, No. 3, 1 February 2001

S. Jin, T. H. Tiefel, M. McCormack, R. A. Fastnach, R. Ramesh, and L. H. Chen, Science 264, 413 共1994兲. 7 H. Y. Hwang, S-W. Cheong, N. P. Ong, and B. Batlogg, Phys. Rev. Lett. 77, 2041 共1996兲. 8 R. D. Sa´nchez, J. Rivas, C. Va´zquez-Va´zquez, M. A. Lo´pez-Quintela, M. T. Causa, M. Tovar, and S. Oseroff, Appl. Phys. Lett. 68, 134 共1996兲. 9 A. Gupta, G. Q. Gong, G. Xiao, P. R. Duncombe, P. Lecoeur, P. Trouilloud, Y. Y. Wang, V. P. Dravid, and J. Z. Sun, Phys. Rev. B 54, 15629 共1996兲. 10 J.-H. Park, E. Vescovo, H.-J. Kim, C. Kwon, R. Ramesh, and T. Venkatesan, Nature 共London兲 392, 794 共1998兲. 11 M. Viret, M. Drouet, J. Nassar, J. P. Contour, C. Fermon, and A. Fert, Europhys. Lett. 39, 545 共1997兲. 12 N. D. Mathur, G. Burnell, S. P. Isaac, T. J. Jackson, B.-S. Teo, J. L. MacManus-Driscoll, L. F. Cohen, J. E. Evetts, and M. G. Blamire, Nature 共London兲 387, 266 共1997兲. 13 V. A. Was’ko, V. A. Larkin, P. A. Kraus, K. R. Nikolaev, D. E. Grup, C. A. Nordman, and A. M. Goldman, Phys. Rev. Lett. 78, 1134 共1997兲. 14 J. M. D. Coey, A. E. Berkowitz, Ll. Balcells, F. F. Putris, and F. T. Parker, Appl. Phys. Lett. 72, 734 共1998兲. 15 J. M. D. Coey, A. E. Berkowitz, Ll. Balcells, F. F. Putris, and A. Barry, Phys. Rev. Lett. 80, 3815 共1998兲. 16 K.-I. Kobayashi, T. Kimura, H. Sawada, K. Terakura, and Y. Tokura, Nature 共London兲 395, 677 共1998兲.

17

Ll. Balcells, A. E. Carrillo, B. Martı´nez, and J. Fontcuberta, Appl. Phys. Lett. 74, 4014 共1999兲. 18 D. K. Petrov, L. Krusin-Elbaum, J. Z. Sun, C. Feild, and P. R. Duncombe, Appl. Phys. Lett. 75, 995 共1999兲. 19 See, for a recent review, J. S. Moodera and G. Mathon, J. Magn. Magn. Mater. 200, 248 共1999兲. 20 C. Va´zquez-Va´zquez, M. C. Blanco, M. A. Lo´pez-Quintela, R. D. Sa´nchez, J. Rivas, and S. B. Oseroff, J. Mater. Chem. 8, 991 共1998兲. 21 L. E. Hueso, J. Rivas, F. Rivadulla, and M. A. Lo´pez-Quintela, J. Appl. Phys. 86, 3881 共1999兲. 22 A. H. Morrish, The Physical Properties of Magnetism 共Wiley, New York, 1965兲. 23 P. Sheng, B. Abeles, and Y. Arie, Phys. Rev. Lett. 31, 44 共1973兲. 24 B. Abeles, P. Sheng, M. D. Coutts, and Y. Arie, Adv. Phys. 24, 407 共1975兲. 25 L. E. Hueso, F. Rivadulla, R. D. Sa´nchez, D. Caeiro, C. Jardo´n, C. Va´zquez-Va´zquez, J. Rivas, and M. A. Lo´pez-Quintela, J. Magn. Magn. Mater. 189, 321 共1998兲. 26 R. Zallen, The Physics of Amorphous Solids 共Wiley, New York, 1983兲. 27 A. Milner, A. Gerber, B. Groisman, M. Karpovsky, and A. Gladkikh, Phys. Rev. Lett. 76, 475 共1996兲. 28 J. S. Helman and B. Abeles, Phys. Rev. Lett. 37, 1429 共1976兲.

Downloaded 17 Dec 2002 to 131.111.8.102. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/japo/japcr.jsp