J O U R N A L O F C H E M I S T R Y Materials EVect of compositional homogeneity on the magnetic properties of La0.7Ca0.3MnO3 Kumar P. S. Anil,a Joy P. Alias,b* and Sadgopal K. Dateb aCentre for Advanced Studies in Materials Science and Solid State Physics, Department of Physics, University of Pune, Pune 411007, India bPhysical andMaterials Chemistry Division, National Chemical L aboratory, Pune 411008, India Dc magnetization and ac magnetic susceptibility have been measured in the temperature range 80–300 K for La0.7Ca0.3MnO3, synthesized by the conventional ceramic and the low temperature combustion methods and annealed at diVerent temperatures.Temperature dependence of the dc magnetization and the ac susceptibility is strongly dependent on the processing conditions resulting in compositional inhomogeneity in the samples.A sharp ferromagnetic transition is observed only for a compositionally homogeneous sample. The discovery of unusually high magnetoresistance (known as ation measurements. The ACS measurements were pursued as giant magnetoresistance, GMR) in the ferromagnetic perovsk- this technique can provide more information about magnetic ite manganites, Ln1-xCaxMnO3 (Ln=La, Pr, Nd, etc.) has ordering and the presence of other magnetic impurities in stimulated the need for a detailed study of their electrical a sample.15 transport and magnetic properties.1 Jonker and Van Santen2 first reported the onset of ferromagnetism in the system La1-xCaxMnO3.Wollan and Koehler3 made a detailed investi- Experimental gation of the magnetic structure of this system from neutron La0.7Ca0.3MnO3 was prepared by the combustion synthesis diVraction studies.The double-exchange (DE) magnetic intermethod (hereafter referred to as the combustion sample) follow- action between Mn3+ and Mn4+, as proposed by Zener, ing the literature method,12 and was annealed in air at diVerent explained the origin of ferromagnetism in these compounds.4 temperatures (600–1300 °C).In the ceramic synthesis method Recently it has been shown that small magnetic polarons are (referred in the text as the ceramic sample), the corresponding responsible for the pronounced magnetoresistance of these oxides were mixed in the required molar ratio and heated compounds at the Curie temperature.5 The GMR eVect in this initially at 1000 °C for 72 h with six intermediate grindings, system has been studied extensively using single crystals, thin and subsequently annealed in air at higher temperatures films, and polycrystalline samples derived from diVerent pro- (1000–1300 °C). The samples were characterized by powder X- cessing routes.Large variations in the Curie temperature are ray diVraction (XRD) method using a Philips PW1730 powder reported for a single composition processed by diVerent X-ray diVractometer. The powder XRD patterns of all the methods.Various types of anomalies in their structural, electrisamples revealed their single phase nature without any second- cal, and magnetic properties are also reported.6–11 The Curie ary or impurity phases. The microstructural features were temperature12 and magnetic entropy change13 of low temperaobtained using a Leica-Cambridge 440 scanning electron ture synthesized compounds (by the combustion of a urea– microscope. The Mn4+ content of the samples was estimated nitrate mixture or by the sol–gel method) are shown to depend by redox titrations using potassium permanganate and iron(II) on particle size; maximum Tc and entropy changes are obammonium sulfate solutions.The low field (at 10 Oe and served for bigger particles obtained by heating the sample at 27 Hz) ac susceptibility (ACS) studies were performed using higher temperatures. Since the GMR eVect can be controlled the mutual inductance technique in an APD cryogenics closed- by manipulating the grains and grain boundaries,14 the low cycle helium cryostat (50–300 K).The dc magnetization temperature synthesized compounds, which always give finer measurements were carried out on a EG&G PAR vibrating particles, oVer control over grain/grain-boundary related elecsample magnetometer (VSM) model 4500 attached to a Janis tronic properties. As these properties depend on the extent of compositional, magnetic, and structural homogeneities, com- liquid nitrogen cryostat (80–300 K).positional homogeneity at a microscopic level is the most desired factor as it leads to a uniform distribution of Mn3+/Mn4+ in the compound. The extent of the DE inter- Results and Discussion action between Mn3+ and Mn4+ which determines the Tc as In Fig. 1 the temperature dependence of the dc magnetization well as the sharpness of the ferromagnetic transition, and the (measured at 5000 Oe) is shown for the samples prepared by degree of resistivity anomaly at Tc, thus, may depend purely the ceramic and combustion methods and annealed at diVerent on the compositional homogeneity.temperatures. The samples annealed below 1300 °C show an In order to understand the origin of the diVerent types of initial increase in the magnetization at ca. 270 K and show a anomalies reported for the magnetic properties of the manbroad magnetic transition with no well defined transition ganites, we have studied the magnetic behavior of temperature. The ceramic sample annealed at 1000 °C and the La0.7Ca0.3MnO3 synthesized by two diVerent methods; (i) by combustion sample annealed at 1200 °C show a second broad the conventional ceramic method and (ii) by combustion transition below ca. 240 K. After annealing at 1300 °C for 48 synthesis or the urea–nitrate method, and annealed at diVerent hours, the first magnetic transition becomes sharp and is temperatures. The magnetic properties were evaluated using shifted to ca. 245 K for the combustion sample with a small low field ac susceptibility (ACS), high field dc magnetization, and field cooled (FC) and zero field cooled (ZFC) magnetiz- step in the magnetization at ca. 230 K. On the other hand, a J. Mater. Chem., 1998, 8(5), 1219–1223 1219well defined Tc#245 K is obtained for the ceramic sample annealed at 1300 °C for 48 hours. Fig. 2 shows the temperature variation of the ac susceptibility measured on the combustion samples.The samples annealed up to 1200 °C show an initial increase in the susceptibility above 260 K. After annealing at 1200 °C, the susceptibility curve shows multiple steps at lower temperatures, as well as the initial increase at ca. 265 K. Even the sample annealed at 1300 °C for 48 h shows a small step at ca. 230 K apart from the sharp transition at Tc=246 K.The magnetic transition temperature (Tc), which corresponds to the maximum in the dxac/dT curve of each sample, is indicated in Table 1. Inset A Fig. 2 shows the derivative of the susceptibility, dxac/dT, of the samples annealed at 1000 °C (curve 1) and 1300 °C (curve 2). Curve 1 is very broad with a maximum at 266 K and shoulders at ca. 270 K and ca. 245 K, whereas curve 2 shows only a sharp maximum at 246 K.The ac susceptibility behavior of the ceramic (curve 2) and combustion (curve 1) samples annealed at 1200 °C is compared in inset B. Both curves show almost identical behavior. Fig. 1 Temperature dependence of the magnetization (HA=5000 Oe) The ACS curves of the ceramic samples are shown in Fig. 3. of La0.7Ca0.3MnO3 annealed at various temperatures: (a) prepared by The natures of the ac susceptibility curves of the samples the combustion method, and (b) prepared by the ceramic method annealed up to 1200 °C are identical.As observed in Fig. 2, an initial increase in the susceptibility is observed below ca. 270 K and a second rise at low temperature is observed at 185 K for the 1000 °C annealed sample. The temperature at which this second transition occurs is shifted to higher temperatures (see Table 1) as the annealing temperature is increased to 1200 °C (193 K for curve b and 208 K for curve c) with an increase in the magnitude of the transition below 270 K, and finally both the transitions are merged together to form a single broad Fig. 2 Temperature dependence of the ac susceptibility of La0.7Ca0.3MnO3 synthesized by the combustion method and annealed at various temperatures.Inset A: the dxac/dT curves of (1) 1000 °C and (2) 1300 °C annealed samples; inset B: ACS curves of the samples Fig. 3 Temperature dependence of the ac susceptibility of annealed at 1200 °C, prepared by (1) combustion and (2) ceramic La0.7Ca0.3MnO3 synthesized by the ceramic method and annealed at methods.various temperatures Table 1 Mn4+ content and the Curie temperature(s) of the samples annealed at various temperatures (Tc corresponds to the tempreature at which the dxac/dT shows a maximum value) annealing temperature/°C; sample duration/ha Mn4+ content (%) Tc/K La0.7Ca0.3MnO3 1000; 72 34 262, 185 (ceramic) 1100; 12 32 263, 193 1200; 12 32 263, 208 1300; 12 30 249 1300; 24 30 245 1300; 48 30 245 La0.7Ca0.3MnO3 600; 12 62 264 (combustion) 800; 12 42 270 1000; 12 34 266 1200; 12 32 265, 246, 195 1300; 48 30 246 La0.9Ca0.1MnO3 1000; 72 14 265, 216 (ceramic) 1100; 12 14 263, 164 1200; 12 12 150 1300; 12 10 155 1300; 48 10 155 aThe ceramic samples initially annealed at 1000 °C; 72 h are subsequently annealed at higher temperatures. 1220 J. Mater. Chem., 1998, 8(5), 1219–1223transition for the sample annealed at 1300 °C for 12 hours.inhomogeneity only and not due to the excess of Mn4+ in La0.7Ca0.3MnO3. This can be justified with the following This magnetic transition becomes sharp for the sample annealed at 1300 °C for longer duration (Tc#245 K). reasons. (1) The Curie temperature for the first magnetic transition at ca. 263 K remains unchanged for the ceramic Fig. 4 shows the FC and ZFC curves of the ceramic sample annealed at two diVerent temperatures. The FC and ZFC samples annealed up to 1200 °C, though the Mn4+ content is decreased slightly (see Table 1). (2) If the magnetic transition curves deviate at a temperature slightly below Tc#245 K for the high temperature annealed sample, whereas for the low at ca. 263 K is due to excess Mn4+, there is no reason for a second magnetic transition at a lower temperature (Fig. 3). temperature annealed sample this deviation starts at ca. 265 K, close to the first magnetic transition observed in the ac The Curie temperature for this second transition increases with increasing annealing temperature and this transition must be susceptibility curve (Table 1).Both the FC and ZFC curves show a second magnetic transition at ca. 185 K. For the due to the presence of another phase with a diVerent composition. (3) There is not much variation in the Tc of the combustion combustion samples also similar FC and ZFC curves were obtained. sample with annealing temperature up to 1200 °C, though the estimated Mn4+ content is continuously decreased.With In La1-xCaxMnO3, maximum Tc (ca. 270 K) is observed for x=0.33, and Tc decreases as x is decreased or increased from decreasing Mn4+ content, the transition temperature is expected to be decreased. (4) The combustion sample, when this value.16 The temperature at which the dc magnetization and the ac susceptibility of the low-temperature-annealed annealed below 1200 °C and containing excess Mn4+, showed a single (broad) magnetic transition above 260 K; the same samples show an initial increase is close to the Tc of La0.67Ca0.33MnO3, indicating the presence of small amounts sample when annealed at 1200 °C showed three magnetic transitions as evidenced from three maxima in the dxac/dT of the x=0.33 phase in the samples.The onset of a first magnetic transition below ca. 270 K with a second broad curve, though this sample contained less Mn4+ compared to the low-temperature-annealed samples. magnetic transition at a lower temperature, and the absence of a well defined magnetic ordering temperature for the samples The above facts indicate that the observed anomalous magnetic behavior of those samples processed below 1300 °C annealed below 1300 °C, indicate that the samples prepared by the ceramic method and the combustion method, annealed is due to the presence of phases with varying compositions in the La1-xCaxMnO3 system.Validity for the above arguments at low temperatures, may contain diVerent compositions (diVerent x values) in the series La1-xCaxMnO3. It is possible comes from the results of Baythoun and Sales who had earlier shown by careful EDAX analysis that the samples prepared that on initial heating of the mixture of oxides (ceramic method) and in the decomposed urea–nitrate mixture (combus- by a low temperature sol–gel process and annealed at a temperature as high as 1400 °C were not compositionally tion method), phases with diVerent compositions in the entire range 0x1 are formed.On further heating of this mixture homogeneous.18 The ideal composition La0.5Sr0.5MnO3 synthesized by them at low temperatures and annealed at 1200 °C at higher temperatures, the compositional range is narrowed, and finally after heating at a higher temperature for suYcient had only 50% of the total sample within a compositional band of x between 0.4 and 0.6, with the rest of the sample having duration, the required composition is obtained.That is, the samples annealed at low temperatures are compositionally other compositions in the range 0x1 in La1 -xSrxMnO3. Similarly, after annealing at 1400 °C, 64% of the sample had inhomogeneous. The increasing FC magnetization (Fig. 4) at low temperatures for the low-temperature-annealed sample the ideal composition but the rest of the phases were between x=0.3 and x=0.6.A similar distribution of the various com- indicates the presence of a paramagnetic phase along with the ferromagnetic phase(s) in the sample. positions in La1-xCaxMnO3 is responsible for the observed magnetic behavior in our experiments. The magnetic transition The Curie temperatures of the manganites containing excess oxygen are reported to be higher than those of the stoichio- (at 245 K) of the combustion sample annealed at 1300 °C is sharper (Fig. 2) than that of the sample prepared by the metric samples, and this eVect is more pronounced for LaMnO3 and low Ca doped samples.2,17 The higher Tc of the oxygen- ceramic method and annealed under similar conditions (Fig. 3). This implies that the ceramic sample annealed at 1300 °C for excess compositions is due to the presence of an excess of Mn4+ in the compounds and a maximum Tc of ca. 270 K is 48 h is still inhomogeneous but with the additional phases having compositions above and below, but very close to, x= observed for 33% Mn4+ containing compositions. This Mn4+ content is equivalent to that present in La0.67Ca0.33MnO3, 0.3.The combustion sample, on the other hand, has two, almost fixed, compositions, one with x=0.3 and another which shows a magnetic transition at ca. 270 K.16 In the present results, however, the origin of the first with x#0.25. In order to show that the x=0.33 phase is formed initially, magnetic transition at ca. 270 K is due to compositional for low Ca doped samples also, the ACS curves of the composition La0.9Ca0.1MnO3 prepared by the ceramic method are shown in Fig. 5. For the sample annealed at 1000 °C for 72 h, a small increase in the susceptibility is observed at ca. Fig. 4 Temperature dependence of FC and ZFC magnetization (HA= Fig. 5 Temperature dependence of the ac susceptibility of La0.9Ca0.1MnO3 synthesized by the ceramic method and annealed at 100 Oe) of La0.7Ca0.3MnO3 prepared by the ceramic method and annealed at two diVerent temperatures various temperatures J.Mater. Chem., 1998, 8(5), 1219–1223 1221270 K with a broad transition at Tc#216 K. As the annealing by diVerent groups7,9,10,13 and identical Tc values for those samples processed at temperatures 1300 °C.6,7 temperature is increased, the contribution at ca. 270 K is decreased and the broad magnetic transition is shifted to lower The above arguments appear to be equally applicable to all the perovskite type compounds in the system La1-xAxMO3 temperatures and a well defined magnetic transition at 155 K is observed for the sample annealed at 1300 °C for 36 hours. (A=Ca, Sr, Ba, etc.; M=Mn, Co). For example, the compound La0.875Sr0.125MnO3 prepared at 1000 °C is reported19 to show Tc values and the Mn4+ content of this composition annealed at diVerent temperatures are compared, in Table 1, with that true paramagnetic behavior only above 360 K, the Tc of the composition with x=0.33 observed in the La1-xSrxMnO3 of the La0.7Ca0.3MnO3 ceramic sample synthesized and annealed under identical conditions.system.20 Our studies on the La1-xSrxCoO3 system also showed similar compositional inhomogeneity eVects for The broad nature of the magnetic transition as well as the apparent low-Tc of the combustion samples12 annealed at low samples processed at low temperatures.As the low-temperature-annealed samples show the presence temperatures [Fig. 1(a)] can be explained on the basis of compositional inhomogeneity.As both the ceramic and com- of phases with varying compositions between LaMnO3 and CaMnO3 which will have slightly varying lattice parameters, bustion samples processed below 1300 °C show an initial rise in the susceptibility at ca. 270 K, it can be attributed to the each peak in the powder X-ray diVraction pattern will be the sum of the individual peaks corresponding to the individual presence of the phase La0.67Ca0.33MnO3 in the samples.The sample will have diVerent compositions in La1-xCaxMnO3, phases. Therefore the overall XRD peak will be broader than that expected due to particle size broadening alone, for the whose Tc is a maximum for x=0.33 and continuously decreased as x is increased or decreased from this value. The sum of the low temperature synthesized compounds.One of the peaks in the powder XRD patterns of ceramic and combustion samples high field magnetization curves of all those phases with diVering transition temperatures will appear as a continuously increas- of La0.7Ca0.3MnO3 annealed at diVerent temperatures are shown in Fig. 6. The ceramic sample annealed at 1000 °C for ing curve as if the magnetic transition is very broad, with no well defined Curie temperature.The temperature at which 36 h shows a broad and asymmetric peak whereas after annealing at the same temperature for 72 h the shoulder at the maximum slope is obtained from the dM/dT curve (as reported in ref. 12) then corresponds to the Tc of the major phase higher angle side has disappeared. Similarly for the combustion sample annealed at 600 °C, a broad and asymmetric peak is present in the sample.For the 800 °C annealed sample, the dM/dT curve showed a broad maximum at ca. 235 K whereas observed, and the width and asymmetry of this peak are decreased after annealing at higher temperatures. Interestingly, the dxac/dT curve gave a maximum at 270 K. The dxac/dT curves of the combustion synthesized sample annealed at the grain size observed (from an SEM photograph) for the ceramic sample annealed at 1000 °C for 36 h is >1 mm which 1000 °C and 1300 °C shown in Fig. 2 (inset A) indicate that the Tc of the low temperature annealed sample is higher than is almost double that of the average grain size of the combustion sample (ca. 500 nm) annealed at 1000 °C for 12 h, though that of the high temperature annealed sample.The derivative curve of the 1000 °C annealed sample is very broad with a the width of the XRD peak is much higher for the ceramic sample. The asymmetry in the reflection of the low temperature shoulder at ca. 270 K and a broad feature below 250 K apart from the peak at 266 K, showing contribution from diVerent annealed samples indicates the presence of additional phases.Evaluation of the size of the particles from such broad and phases. These derivative curves are similar to the magnetic entropy change curves obtained by Guo et al.13 (magnetic asymmetric XRD peaks will be in error as it may not provide the actual size of the particles. The general conclusion that entropy, SM, is related to dM/dT ) and the higher entropy change for the high temperature annealed sample is then a ‘absence of any extra reflections in the powder XRD pattern is an indication of single phase nature of the compound’ is reflection of the increased homogeneity of that sample, rather than the change in the particle size as reported.Though a also not valid based on these arguments. From the present results it is concluded that for the calcium sharp increase in the magnetization at ca. 270 K for x=0.33 is reported,13 the magnetization increases continuously as the substituted lanthanum manganites, the anomalies reported in the magnetic behavior of low-temperature processed samples temperature is lowered to 78 K, and the total magnetization at this temperature is much less than that obtained for the x= 0.3 sample in the present study.The continuous increase in the magnetization below Tc as the temperature is decreased is due to the presence of ferromagnetic phase(s) which orders at a higher temperature, and paramagnetic contributions from those phases order at a relatively lower temperature (or from phases with very low Ca concentrations which are not ferromagnetic).The present results give evidence for the fact that compositional inhomogeneity is responsible for the dependence of Tc on the processing conditions. The diVerent values of Tc for the same composition, and almost identical values of Tc for diVerent compositions, in La1-xCaxMnO3, reported in the literature,6–11 thus arise due to non-homogeneity of the sample and also point to the fact that whatever the desired composition may be, the compound obtained upon processing below 1300 °C will be a mixture of diVerent phases.The eVect of compositional inhomogeneity may not be evident in the high field magnetization curves of the low temperature processed La0.67Ca0.33MnO3 (and those samples with substitution in the lanthanum site of this composition) because maximum Tc in the system La1-xCaxMnO3 is observed for x=0.33. However, if the Tc observed for this composition is less than that expected, then the observed Tc will be an indication of the major phase Fig. 6 Powder XRD pattern from the (200) plane of La0.7Ca0.3MnO3, present in the sample. This may be the reason for the large prepared by the ceramic and combustion methods, and annealed at diVerent temperatures diVerences in Tc of the composition La0.67Ca0.33MnO3 reported 1222 J.Mater. Chem., 1998, 8(5), 1219–122310 J. M. D. Coey, M. Viret, L. Ranno and K. Ounadjela, Phys. Rev. are due to compositional inhomogeneity which leads to mag- L ett., 1995, 75, 3910. netic as well as structural inhomogeneity. 11 R. Mahendiran, S. K. Tiwary, A. K. Raychaudhuri, T. V. Ramakrishnan, R.Mahesh, N. Rangavittal and C. N. R. Rao, Phys. P. S. A. K. is thankful to the University Grants Commission, Rev. B, 1996, 53, 3348. India, for financial support and Prof. D. S. Joag for his 12 R. D. Sanchez, J. Rivas, C. V. Vazquez, A. L. Quintela, continuous encouragement. M. T. Causa, M. Tovar and S. OseroV, Appl. Phys. L ett., 1996, 68, 134. 13 Z. B. Guo, Y. W.Du, J. S. Zhu, W. P. Ding and D. Feng, Phys. References Rev. L ett., 1997, 78, 1142; Z. B. Guo, J. R. Zhang, H. Huang, W. P. Ding and Y. W. Du, Appl. Phys. L ett., 1997, 70, 904. 1 A. P. Ramirez, J. Phys.: Condens. Matter, 1997, 9, 8171. 14 N. D. Mathur, G. Burnell, S. P. Isaac, T. J. Jackson, B.-S. Teo, 2 G. H. Jonker and J. H. Van Santen, Physica, 1950, 16, 337. J. L. MacManus-Driscoll, L.F. Cohen, J. E. Evetts and 3 E. O.Wollan and W. C. Koehler, Phys. Rev., 1955, 100, 545. M. G. Blamire, Nature, 1997, 387, 266. 4 C. Zener, Phys. Rev., 1951, 82, 403. 15 P. A. Joy, S. K. Date and P. S. Anil Kumar, Phys. Rev. B, 1997, 5 J. M. De Teresa, M. R. Ibarra, P. A. Algarabel, C. Ritter, 56, 2324. C. Marquina, J. Blasco, J. Garcia, A. del Moral and Z. Arnold, 16 P. SchiVer, A. P. Ramirez, W. Bao and S.-W. Cheong, Phys. Rev. Nature, 1997, 386, 256. L ett., 1995, 75, 3336. 6 S. J. L. Billinge, R. G. DiFrancesco, G. H. Kwei, J. J. Neumeier and 17 A. Tiwari and K. P. Rajeev, J.Mater. Sci. L ett., 1997, 16, 521. J. D. Thompson, Phys. Rev. L ett., 1996, 77, 715. 18 M. S. G. Baythoun and F. R. Sale, J.Mater. Sci., 1982, 17, 2757. 7 R. H. HeV ner, L. P. Lee, M. F. Hundley, J. J. Neumeier, 19 D. N. Argyriou, J. M. Mitchell, C. D. Potter, D. G. Hinks, G. M. Luke, K. Kojima, B. Nachumi, Y. J. Uemura, D. E. J. D. Jorgensen and S. D. Bader, Phys. Rev. L ett., 1996, 76, 3826. MacLaughlin and S.-W. Cheong, Phys. Rev. L ett., 1996, 77, 1869. 20 H. Y. Hwang, S.-W. Cheong, N. P. Ong and B. Batlogg, Phys. Rev. 8 J. W. Lynn, R. W. Erwin, J. A. Borchers, Q. Huang, A. Santoro, JL ett., 1996, 77, 2041. L. Peng and Z. Y. Li, Phys. Rev. L ett., 1996, 76, 4046. 9 G. H. Rao, J. R. Sun, J. K. Liang and W. Y. Zhou, Phys. Rev. B, 1997, 55, 3742. Paper 7/08235E; Received 17th November, 1997 J. Mater. Chem., 1998, 8(5), 1219–1223 1223