Magnetic properties and magnetocaloric effect in Fe

Magnetic properties and magnetocaloric effect in Fe-doped La0.6Ca0.4MnO3 with short-range ferromagnetic order T. A. Ho, T. D. Thanh, T. O. Ho, M. H. Phan, The-Long Phan, and S. C. Yu Citation: Journal of Applied Physics 117, 17A724 (2015); doi: 10.1063/1.4915103 View online: http://dx.doi.org/10.1063/1.4915103 View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/117/17?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Effect of hydrostatic pressure on magnetic and magnetocaloric properties of Mn-site doped perovskite manganites Pr0.6Ca0.4Mn0.96B0.04O3 (B=Co and Cr) J. Appl. Phys. 115, 043905 (2014); 10.1063/1.4862810 Impact of reduced dimensionality on the magnetic and magnetocaloric response of La0.7Ca0.3MnO3 Appl. Phys. Lett. 102, 062414 (2013); 10.1063/1.4792239 Tunable spin reorientation transition and magnetocaloric effect in Sm0.7−xLaxSr0.3MnO3 series J. Appl. Phys. 113, 013911 (2013); 10.1063/1.4773337 Critical behavior and magnetic-entropy change of orthorhombic La0.7Ca0.2Sr0.1MnO3 J. Appl. Phys. 112, 093906 (2012); 10.1063/1.4764097 Large reversible magnetocaloric effect in La0.7-xPrxCa0.3MnO3 J. Appl. Phys. 110, 013906 (2011); 10.1063/1.3603014

[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: 203.255.69.179 On: Mon, 23 Mar 2015 02:04:30

JOURNAL OF APPLIED PHYSICS 117, 17A724 (2015)

Magnetic properties and magnetocaloric effect in Fe-doped La0.6Ca0.4MnO3 with short-range ferromagnetic order T. A. Ho,1 T. D. Thanh,1,2 T. O. Ho,3 M. H. Phan,4 The-Long Phan,5 and S. C. Yu1,a) 1

Department of Physics, Chungbuk National University, Cheongju 361-763, South Korea Institute of Materials Science, Vietnam Academy of Science and Technology, 18-Hoang Quoc Viet, Hanoi, Vietnam 3 Institute of Chemistry, Vietnam Academy of Science and Technology, 18-Hoang Quoc Viet, Hanoi, Vietnam 4 Department of Physics, University of South Florida, Tampa, Florida 33620, USA 5 Department of Physics, Hankuk University of Foreign Studies, Yongin 449-791, South Korea 2

(Presented 6 November 2014; received 21 September 2014; accepted 6 November 2014; published online 19 March 2015) The magnetic properties and magnetocaloric effect of La0.6Ca0.4Mn1-xFexO3 (x ¼ 0–0.04) compounds fabricated by solid-state reaction have been studied. Magnetization measurements versus temperature revealed a decrease of the ferromagnetic-paramagnetic phase transition temperature (TC) with increasing Fe-doping content. The TC values determined for the samples with x ¼ 0, 0.02 and 0.04 are about 260, 254 and 236 K, respectively. Based on magnetic-field dependences of magnetization, M(H), the magnetic entropy change (DSm) of the samples were calculated. Under an applied field change DH ¼ 30 kOe, the maximum jDSmaxj value decreases from 5.74 Jkg1 K1 for x ¼ 0 to about 2.62 Jkg1 K1 for x ¼ 0.04. These values correspond to relative cooling powers 140–180 J/kg, which are comparable to those of other manganites. Analyzing magnetic-field dependences of jDSmj for the samples indicates their power-law relation. Based on Banerjee’s criteria and Franco’s universal curves related to the magnetic-entropy change, we assess magnetic order C 2015 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4915103] existing in the samples. V Currently, hole-doped perovskite-type manganites with a generally chemical formula Ln1-xAxMnO3 (Ln3þ and A2þ are rare-earth and alkaline-earth ions, respectively) are still of interest due to many intriguing features, typically the colossal magnetoresistance (CMR) and magnetocaloric effects (MCE).1–3 A lot of studies have revealed a strong dependence of magnetic phase transitions, CMR, and MCE on stoichiometry and A-dopant types.2,4 Among studied hole-doped lanthanum manganites, La1-xCaxMnO3 compounds are considered as one of the material systems showing very rich physical properties. For examples, they can exhibit a firstorder magnetic phase transition (FOMT) as x 0.3 and offer giant MCE tunable around room temperature. In these compounds, the substitution of La3þ by Ca2þ creates more Mn4þ ions besides pre-existing Mn3þ ions, leading to ferromagnetic (FM) double-exchange (DE) interactions via O2 ions between them,5 which are strongly affected by the changes related to the bond angle hMn-O-Mni and bond length hMn-Oi. Reviewing previous works, it can be seen much attention given to the influence of the Mn substitution by other transition metals (Me) on the crystal structure, magnetic and transport properties, and MCE of Me-doped La1-xCaxMnO3 compounds.6–9 For Fe-doped La1-xCaxMnO3 compounds,6,7,10,11 it has been pointed out that the Mn substitution by a small amount of Fe could change the magnetic and transport properties without changing significantly their crystal structure because of the same radii of Mn and Fe ions. Basically, increasing Fe-doping content (x) in Fe-doped La1-xCaxMnO3 compounds reduces a)

Author to whom correspondence should be addressed. Electronic mail: [email protected].

0021-8979/2015/117(17)/17A724/5/$30.00

their ferromagnetic-paramagnetic (FM-PM) phase transition temperature (TC) due to weakened DE interactions. Recently, Lampen et al. have studied the magnetic phase transition and MCE in La0.7Ca0.3Mn1-xFexO3 compounds, and concluded that when Fe content increased, the magnetic-entropy change (DSm) would be decreased.7 However, the peaks of the DSm(T) curves were broadened remarkably, and thus enhanced refrigerant capacity values up to 30% compared with the parent compound La0.7Ca0.3MnO3. Meanwhile, Othmani and co-worker6 studied La0.6Ca0.4Mn1-xFexO3 samples, and observed that the highest relative cooling power for x ¼ 0.05. To further understand this material system, we have prepared polycrystalline La0.6Ca0.4Mn1-xFexO3 samples. Based on a systematic investigation of the Arrott plot and universalcurve methods, the aim of this work is to show the effect of relevant disorder introduced by Fe doping in the universality class of La0.6Ca0.4MnO3 manganites, which the critical exponents are very close to those found out by the tricritical mean field model.12 Moreover, the field dependence of the magnetic entropy change is analyzed through a power law to get a deeper insight on the MCE of the samples. The results point out that the un-doped sample (La0.6Ca0.4MnO3) show the existence of the tricritical point sets a boundary between firstorder and second-order while Fe-doped samples undergo a SOMT. Also, Fe doping reduces DSm, but increases the relative cooling power (RCP) due to broadened DSm(T) peaks. Three polycrystalline samples La0.6Ca0.4Mn1-xFexO3 with x ¼ 0, 0.02, and 0.04 have been prepared by solid-state reaction from initially high-purity powders La2O3, CaCO3, Mn, and Fe (3 N). The powders combined with stoichiometric quantities were carefully mixed and ground, and

117, 17A724-1

C 2015 AIP Publishing LLC V


17A724-2

Ho et al.

pre-annealed at 1100 C for 24 h in air. After pre-annealing, the mixtures were re-ground and pressed into pellets. Finally, these pellets were sintered at 1300 C in air for 24 h. The crystal structure of the product after sintered was checked by an X-ray diffractometer (Brucker AXS, D8 Discover). Room-temperature X-ray diffraction patterns revealed the samples exhibiting a single phase in an orthorhombic structure. Magnetic measurements were performed on a superconducting quantum interference device (SQUID) worked in the temperature range of 5–350 K. Figure 1 shows temperature dependences of zero-fieldcooled magnetization, M(T), for the La0.6Ca0.4Mn1-xFexO3 samples under an applied field H ¼ 100 Oe. With increasing temperature above 230 K for x ¼ 0 and 0.02 and 190 K for x ¼ 0.04, there is a rapid decrease of magnetization due to the FM-PM phase, where magnetic moments become disordered under the impact of thermal energy. The Curie temperature (TC) determined from the minima of dM(T)/dT versus T curves (see the inset of Fig. 1) are about 260, 254, and 236 K for x ¼ 0, 0.02, and 0.04, respectively. As expected, the gradual decrease of TC with increasing Fe-doping content (x) in La0.6Ca0.4Mn1-xFexO3 indicates the weakening of FM DE interactions.13 We believe that the substitution of Fe3þ for Mn3þ ions changes not only hMn-Oi and hMn-O-Mni but also could be leads to the additional presence of anti-FM super-exchange (SE) interactions associated with pairs of Fe3þ-Fe3þ and/or Fe3þ-Mn3þ besides pre-existing anti-FM pairs of Mn3þ-Mn3þ and Mn4þ-Mn4þ. These anti-FM interactions compete with FM interactions due to DE pairs of Mn3þ-Mn4þ and thus reduce the TC of La0.6Ca0.4Mn1-xFexO3 compounds as increasing x. To learn about the nature of the PM-FM transition and MCE in La0.6Ca0.4Mn1-xFexO3, a series of isothermal magnetization curves, M(H), in the vicinity of TC, where a temperature increment is 2 K, were measured. Here, the MCE was assessed by the DSm versus temperature and magnetic field, which is calculated from the Maxwell relation14 ðH @M DSm ðT; H Þ ¼ dH; (1) @T H 0

FIG. 1. Temperature dependences of ZFC magnetization, M(T), for La0.6Ca0.4Mn1-xFexO3 samples with x ¼ 0, 0.02, and 0.04 under an applied field of 100 Oe. Inset: First derivative of magnetization.

J. Appl. Phys. 117, 17A724 (2015)

where H is the applied field. According to this equation, the DSm magnitude depends on the magnitudes of both M and (@M/@T)H. Because (@M/@T)H is related to magnetic-ordering transition type, a sharp change in the M(T) curve at the FMPM transition leads to a large DSm.15 In Fig. 2, it shows jDSm(T)j curves for the samples in applied-field intervals up to 30 kOe, with changing steps of 2.5 kOe. For all the samples, the maximum magnetic-entropy change (jDSmaxj) occurs near TC (where dM/dT is maximum), and increases with increasing H. Similar to the tendency of TC, the jDSmaxj position also shifts toward lower temperatures when Fe concentration increases. With H ¼ 30 kOe, jDSmaxj magnitudes are about 5.74, 5.57, and 2.62 J kg1 K1 for x ¼ 0, 0.02, and 0.04, respectively. Although the jDSmaxj magnitude is decreased with increasing Fe concentration, the DSm(T) peak is remarkably broadened. If taking into account for the fullwidth-at-half maximum of the jDSm(T)j curve (denoted as dTFWHM), the relative cooling power (RCP) is calculated from the relation RCP ¼ jDSmaxj dTFWHM.14 For H ¼ 30 kOe, dTFWHM values are about 24, 33, and 64 K, and thus RCP values are about 138, 184 and 168 J kg1 for x ¼ 0, 0.02, and 0.04, respectively. Clearly, with an appropriate Fe

FIG. 2. Temperature dependence of the magnetic-entropy change measured for applied fields ranging from 5 to 30 kOe (with increment steps of 2.5 kOe) for La0.6Ca0.4Mn1-xFexO3 samples with (a) x ¼ 0, (b) x ¼ 0.02, and (c) x ¼ 0.04.


17A724-3

Ho et al.

J. Appl. Phys. 117, 17A724 (2015)

concentration, it can be increase remarkably the RCP of La0.6Ca0.4Mn1-xFexO3 compounds compared to the parent compound La0.6Ca0.4MnO3. In reference to magnetic-field dependences of the DSm, it has been found that these dependences obey a power function DSm / Hn, where n is associated with the magnetic state. With the DSm(H) data obtained, n can be calculated from the following relation:16 n¼

d lnjDSm j : d ln H

(2)

For FM materials at temperatures well above and below its TC, n values are equal to 2 and 1, respectively. According to the mean-field theory, the DSm(H) data at the TC obey a power function with n ¼ 2/3. For materials that do not obey the mean-field theory, n at the TC can be determined by using the critical exponents.16 Figures 3(a)–3(c) show temperature dependences of n, n(T), for La0.6Ca0.4Mn1-xFexO3 samples at various magnetic fields change DH ¼ 10, 20 and 30 kOe, calculated from Eq. (2). For all the samples, n approaches 2 in the PM range at temperatures T TC, and is approximately equal to 1 in the FM region at T TC. Additionally, n at TC reaches to the minimum values. For DH ¼ 30 kOe, n(TC) values are about 0.53, 0.61, and 0.67 corresponding to x ¼ 0, 0.02, and 0.04, respectively. One can see that the n(TC) values of x ¼ 0 and 0.02 are quite smaller than the value n ¼ 2/3 expected for the mean-field theory, revealing FM short-range order existing

FIG. 3. n(T, H) curves for La0.6Ca0.4Mn1-xFexO3 samples with (a) x ¼ 0, (b) x ¼ 0.02, and (c) x ¼ 0.04.

in the samples x ¼ 0 and 0.02. At a higher Fe-doping concentration with x ¼ 0.04, however, its n(TC) value (¼0.67) is close to that of the mean-field theory for FM long-range order. The peak values in the DSm(T) curves of x ¼ 0.04 are thus attained in a wider temperature range. Such the results demonstrate that the x increase in La0.6Ca0.4Mn1-xFexO3 enhances the n value, and tendentiously leads to the establishment of FM long-range order as x > 0.02. This is ascribed to the changes in the magnetic interactions and structural parameters caused by additional Fe ions. M(H) curves and inverse Arrott plots (H/M versus M2)17 at different temperatures around TC of La0.6Ca0.4Mn1-xFexO3 are shown in Fig. 4. For the M(H) curves in the FM region, M increases quickly at low-magnetic fields, and then increases slowly at higher fields H > 10 kOe. A general tendency of a decrease in the magnetization with increasing Fe content is observed, which is related to the weakening of FM DE interactions. However, the initial increase in the moment that accompanies very small Fe doping (x ¼ 0.02) is an interesting exception. It may be due to some oxygen inhomogeneity or may be attributed by the La–Mn interaction.18 According to Banerjee’s criteria,19 the sign of the slope of the H/M versus M2 curve gives information related to the nature of the FM-PM transition; if all the H/M versus M2 curves have a positive slope the magnetic transition is of second order; if some of the H/M versus M2 curves show a negative slope at some point, the transition is of first order. Clearly, at some temperatures, the H/M versus M2 curves of La0.6Ca0.4MnO3, Fig. 4(b), show negative slopes at lowmagnetic fields (H < 10 kOe) but positive slopes at highmagnetic fields (H > 10 kOe). Particularly, around the TC, the slopes are positive over the entire field range. This indicates the existence of the tricritical point sets a boundary between FOMT and SOMT in un-doped sample. Meanwhile,

FIG. 4. M(H) curves and inverse Arrott plots for La0.6Ca0.4Mn1-xFexO3 samples with (a) and (b) x ¼ 0, (c) and (d) x ¼ 0.02, and (e) and (f) x ¼ 0.04.


17A724-4

Ho et al.

J. Appl. Phys. 117, 17A724 (2015)

those for the Fe-doped samples with x ¼ 0.02 and 0.04 show positive slopes over the entire field interval, Figs. 4(d)–4(f), demonstrating them undergoing the SOMT. Recently, a new criterion based on the entropy change curves to distinguish the magnetic order has been proposed by Franco et al.20–22 They have proven that for ferromagnets undergoing the SOMT DSm(T) curves measured with different maximum magnetic fields will collapse onto a universal curve after a scaling procedure. It is natural to expect a breakdown of the universal curve for compounds with the FOMT. As a consequence, whether or not breakdown of the universal behavior of DSm(T) is achieved can be applied as a method to distinguish first- and second-order transitions.22 The construction of the phenomenological universal curve requires to normalize all the DSm(T) curves to their respective maximum value DSmax, and then rescaling the temperature axis defines a new variable h as follows: ðT TC Þ=ðTr1 TC Þ; T TC h¼ (3) ðT TC Þ=ðTr2 TC Þ; T TC ; where Tr1 and Tr2 are two reference temperature points, which can be selected for each curve from temperature corresponding to DSm(Tr, Hmax)/DSmax(Hmax) ¼ a, where 0 < a < 1. This choice of a does not affect the actual construction of the universal curve. For this study, the reference temperatures Tr1 and Tr2 have been selected as those corresponding to 0.5DSmax for all the samples. Figure 5 shows the universal curve constructions for the samples by plotting DS0 (¼DSm/DSmax) versus h. It comes to our attention that the divergence of the curves is clearly observed in the parent compound La0.6Ca0.4MnO3, see Fig. 5(a). Combining with a check of the Arrott plot constructions, it can be conclude that the sample x ¼ 0 undergo the tricritical point sets a boundary between FOMT and SOMT. For two other samples x ¼ 0.02 and 0.04, the experimental point almost collapse into the master curves, as shown in Figs. 5(b) and 5(c), consistent with the SOMT. Notably, for the sample x ¼ 0.02, there is no perfect agreement of the data above TC due to magnetic inhomogeneity. However, a check of the Arrott plot constructions, Fig. 4(d), confirmed the secondorder nature of this sample. A similar behavior was also found in a previous report.23 Obviously, increasing Fe-doping content in La0.6Ca0.4Mn1-xFexO3 with x 0.02 leads to the SOMT. This conclusion is consistent with the trends observed in the Arrott plots, as shown in Fig. 4. In summary, the magnetic and MC properties of La0.6Ca0.4Mn1-xFexO3 were studied in detail. Experiment result show that around TC, the jDSmj curves show the maximum values of about 5.74, 5.57, and 2.62 J kg1 K1 for DH ¼ 30 kOe, corresponding to RCP values 138, 184, and 168 J kg1 for the samples with x ¼ 0, 0.02, and 0.04, respectively. Although the jDSmaxj magnitude was decreased, the peaks of the DSm(T) curves broaden remarkably, leading to an enhancement of the RCP as increasing Fe content. Large RCP values found in La0.6Ca0.4Mn1-xFexO3 compounds suggest them to be promising candidates for refrigeration applications. Interestingly, field dependences of jDSmj can be described by a power law jDSmj / Hn, where n ¼ 0.53, 0.61 and 0.67 x ¼ 0, 0.02 and 0.04, respectively. The value

FIG. 5. Scaled entropy change curves for La0.6Ca0.4Mn1-xFexO3 samples with (a) x ¼ 0, (b) x ¼ 0.02, and (c) x ¼ 0.04.

n ¼ 0.67 close to the value n ¼ 2/3 expected for the meanfield theory proves the establishment of FM long-range order in the sample x ¼ 0.04. Additionally, Arrott plot and universal curve methods indicated that the un-doped sample undergoes the tricritical point sets a boundary between FOMT and SOMT while Fe-doped samples undergo the SOMT. This research was supported by the Converging Research Center Program through the Ministry of Science, ICT and Future Planning, South Korea (2014048835). 1

R. V. Helmolt et al., Phys. Rev. Lett. 71, 2331 (1993). M. H. Phan et al., J. Magn. Magn. Mater. 308, 325 (2007). 3 C. R. H. Bahl et al., Appl. Phys. Lett. 100(12), 121905 (2012). 4 K. A. Gschneidner, Jr. et al., Rep. Prog. Phys. 68, 1479 (2005). 5 C. Zener et al., Phys. Rev. 81, 440 (1951). 6 S. Othmani et al., Solid State Commun. 192, 51–55 (2014). 7 P. J. Lampen et al., J. Appl. Phys. 112, 113901 (2012). 8 L. K. Leung et al., Phys. Rev. B 13, 4069 (1976). 9 N. Dhahri et al., J. Alloys. Compd. 507, 405–409 (2010). 10 A. Nasri et al., Eur. Phys. J. Plus 129, 14180 (2014). 11 K. H. Ahn et al., Phys. Rev. B 54, 15299 (1996). 12 P. Zhang et al., J. Magn. Magn. Mater. 348, 146 (2013). 13 S. K. Barik et al., J. Magn. Magn. Mater. 323, 1015 (2011). 14 M. H. Phan et al., Phys. Rev. B 84, 054436 (2011). 15 N. S. Bingham et al., J. Appl. Phys 106(2), 023909 (2009). 2


17A724-5

Ho et al.

J. Appl. Phys. 117, 17A724 (2015)

16

20

17

21

V. Franco et al., EPL 79, 47009 (2007). A. Arrott et al., Phys. Rev. Lett. 19, 786–789 (1967). 18 N. Kumar et al., J. Alloys Compd. 502, 283 (2010). 19 B. K. Banerjee et al., Phys. Lett. 12, 16–17 (1964).

V. Franco et al., Appl. Phys. Lett. 89, 222512 (2006). V. Franco et al., J. Phys.: Condens. Matter 20, 285207 (2008). 22 C. M. Bonilla et al., J. Appl. Phys. 107, 09E131 (2010). 23 P. Lampen et al., Appl. Phys. Lett. 102, 062414 (2013).
