This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 159.226.35.216

This content was downloaded on 28/05/2015 at 04:22

Please note that terms and conditions apply.

Magnetism and magnetocaloric effect study of CaFe0.7Co0.3O3

View the table of contents for this issue, or go to the journal homepage for more

2015 Mater. Res. Express 2 046103

(http://iopscience.iop.org/2053-1591/2/4/046103)

Home Search Collections Journals About Contact us My IOPscience

Mater. Res. Express 2 (2015) 046103 doi:10.1088/2053-1591/2/4/046103

PAPER

Magnetism andmagnetocaloric effect study of CaFe0.7Co0.3O3HLXia1,2,3, Y YYin1, JHDai1, J YYang1, XMQin3, CQ Jin1,2 andYWLong1,2

1 BeijingNational Laboratory forCondensedMatter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190, People’sRepublic of China

2 Collaborative InnovationCenter ofQuantumMatter, Beijing 100190, People’s Republic of China3 Department of Physics, Shanghai NormalUniversity, Shanghai 200234, People’s Republic of China

E-mail: ywlong@iphy.ac.cn

Keywords: high pressure synthesis,magnetocaloric effect, ferromagnetic transition

AbstractTheCaFe0.7Co0.3O3 single crystal was grown for the first time by a two-stepmethod and itsmagnetismandmagnetocaloric effect were investigated. This compound experiences a second-order para-magnetism-to-ferromagnetism transition in awide temperature windowbetween 200 and 150 Kdueto the presence ofmultiple ferromagnetic interactions. Since the spin entropy is gradually releasedabove the ferromagnetic Curie temperature (∼177 K), no sharp λ-type anomaly is observed in specificheat. On the basis ofmagnetizationmeasurements, however, a considerable entropy change is foundin this perovskite oxide.More interesting, this compound exhibits a broadeningworking temperature,and a significant refrigerant capacity (∼355 J kg−1 at 6 T)which is comparable with those found insome giantmagnetocaloric alloys withfirst-ordermagnetic transitions. The present study thereforeprovides an example on how to enhance the refrigerant capacity by extending theworkingtemperature ofmagnetocaloricmaterial.

1. Introduction

Magnetic refrigeration is a promising technologywith the potential to replace conventional gas compressionrefrigeration duo to the advantages in energy efficiency, environment friendliness, and compact assembling etc[1–3].Magnetocaloric effect (MCE) is the basis formagnetic refrigeration technology. The applying of anexternalmagnetic field can decrease the spin entropy and therefore the heat will be delivered fromamagneticsystem to environment in an isothermal process. On the other hand, when themagnetic field is removedadiabatically,magnetic spins tend to randomize, giving rise to an increase in themagnetic entropy and a decreasein the lattice entropy and thus lowering the temperature of the system [4, 5]. The refrigerant capacity (RC) is animportant quality factor for a refrigerantmaterial, which determines the amount of heat transfer between the

cold and hot reservoirs in an ideal refrigeration cycle. The RC is defined as ∫ Δ= − S TRC d ,T

TM

1

2whereT1 andT2

are the temperatures corresponding to both sides of the half-maximumvalue of a temperature dependententropy change (ΔSM) peak. So it is desirable to combine a largeΔSMwith awideworking temperature for apromisingMCEmaterial in practical applications. At present, giantΔSMvalues have been found in severalmaterial systems such asGd5(Si2Ge2) [6], Ni2MnGa [7], andMnAs [8], etc.However, the giant entropy changesare often originated from afirst-ordermagnetic transition associatedwith a structural phase transition. Thenarrowworking temperature aswell as the thermal andmagnetic hysteresis severely limits the potentialapplications. Althoughmuch attention has been paid to searchMCEmaterials with largeΔSM, it is relativelylittle known on how to extend theworking temperature to enhance the refrigerant capacity.

ABO3 perovskite is a kind of very important functionalmaterials. Since the crystal structures and the A-Bionic combinations are quiteflexible, perovskite compounds show awide variety of intriguing physicalproperties [9, 10]. For example, interestingmagnetocaloric properties were found inMn-based perovskites [11–13]. CaFeO3 perovskite has attracted lots of interests due to the unique charge-disproportion transition of Fe

4+

ions (2Fe4+→ Fe3+ + Fe5+) aswell as the peculiar spiral spin structure [14, 15].WhenCo is used to partiallysubstitute Fe, themagnetismof CaFe1−xCoxO3 solid solutionwill change from antiferromagnetismwith x< 0.2

RECEIVED

9December 2014

REVISED

3 February 2015

ACCEPTED FOR PUBLICATION

26 February 2015

PUBLISHED

20March 2015

© 2015 IOPPublishing Ltd

to ferromagnetismwith x> 0.2. Because the ferromagnetic (FM) ordering in this family is related tomultipledouble exchange interactions between Fe andCo ions as well as between Fe/Co ions and ligand holes [16–19], abroader FM transition temperature region is possible to occur inCo-dopedCaFeO3magnets, which is favorableto enhance the RCby increasing theworking temperature. InCa/SrFe1−xCoxO3 family, a single-crystal sampleusually shows a larger saturated spinmoment than the related polycrystalline sample [18, 19].We thereforepreparedCaFe0.7Co0.3O3 (CFCO) single crystal for the first time and a considerableMCEwas found in thisperovskite compound.

2. Experimental details

Oxygen stoichiometric CaFe0.7Co0.3O3 single crystal was grownby a two-stepmethod [18, 19]. As thefirst step,oxygen-deficient polycrystalline sample CaFe0.7Co0.3O2.5 with brownmillerite-type structure was prepared by aconventional solid-state reactionmethod [16], and thenCaFe0.7Co0.3O2.5 single crystal was grownby using alamp-image-type floating-zone systemwithCaFeO2.5 single crystal as a seed. Secondly, the obtainedCaFe0.7Co0.3O2.5 single crystal was treated in a cubic-anvil-type high pressure apparatus at 6.0GPa and 1073 Kfor 30 minwith the presence of KClO4 oxidant. By combining these ambient and high pressure synthesistechniques, a CFCO single crystal with 3 mm in diameter and 4 mm in lengthwas obtained. Thermogravimetricanalysis confirmed the stoichiometry of oxygen content for the as-made single crystal.

The phase puritywas checked by powder x-ray diffraction using a Rigaku diffractometer withCu-Kαradiation (40 kV, 300 mA). The quality of single crystal was identified by Laue diffraction. Themagneticsusceptibility andmagnetization (M) weremeasured using aQuantumDesign vibrating samplemagnetometer.The resistivity and heat capacity (CP) data were collectedwith aQuantumDesign physical propertymeasurement system.

3. Results and discussion

The x-ray powder diffraction shows that CFCO single crystal crystallizes to an orthorhombic GdFeO3-typeperovskite structure with lattice parameters a= 5.323 Å, b= 5.312 Å, and c= 7.516 Å. The cell volume is a littlesmaller than that of the parent compoundCaFeO3 due to partial substitution of Fe

4+ byCo4+ ionswith a smallerionic radius [14, 20]. Unfortunately, however, for this orthorhombic compoundwith a≈ b, the obtained singlecrystal is highly twinned as found inCaFeO3 single crystal [21]. Therefore, amoderate crystal with a randomplanewas used for the presentMCE study.

Figure 1(a) shows the temperature dependence ofmagnetic susceptibilitymeasured atmagnetic fieldH= 0.1 T in zero-field-cooled (ZFC) andfield-cooled (FC)modes. BothZFC and FC curves show aparamagnetic (PM) to FM transition in awider temperature region between 200 and 150 K. Based on thetemperature derivative of the ZFCmagnetic susceptibility (figure 1(a)), the FMCurie temperature (TC)wasassigned to be about 177 K. As shown infigure 1(b), the inverse susceptibility above 250 K can bewellfitted bytheCurie–Weiss law, producing theCurie constantC= 3.38 emuKmolOe−1 and theWeiss constant θ= 206 K.The positiveWeiss constant is in agreement with the FM interactions. According to theCurie constant, the fittedeffectivemoment is μeff = 5.20 μB f.u.

−1. InCaFeO3 and SrFe1−xCoxO3 solid solution, high-spin andintermediate-spin configurationswere assigned to Fe4+ andCo4+ ions, respectively [14, 18, 19]. If identical spinconfigurations are considered in the present CFCO, the spin-only effectivemoment in theory should be4.62 μB f.u.

−1. This value is a little smaller than the fitted one probably due to the presence of some orbitalmoment of Co4+ ions as found in SrFe1−xCoxO3bymagnetic circular x-ray dichroismmeasurements [22].

The broader PM-FM transition is further confirmed by specific heat and electric transportmeasurements.As shown infigure 2(a), nearTC, a broadeningCP hump instead of a sharp λ-type anomaly, which is usuallypresented for a long-range second-ordermagnetic phase transition, is observed in the temperature dependenceof specific heat inCFCO. This observation can be attributed to that the spin entropy is partially releasedwellaboveTC due to the presence of some short-range FM interactions probably occurred between Fe andCo ionsand/or between Fe/Co ions and oxygen holes on account of the negative p-d charge transfer energy discovered inthe perovskite oxides with unusually high Fe4+ andCo4+ ions [23, 24]. Correspondingly, one cannot findcoherent anomaly in electric transport atTC. As shown infigure 2(b), the resistivity experiences a smoothchangewith temperature acrossTC. This behavior is different from that of SrCoO3, where a resistivity anomaly isfound near the FMCurie temperature [18].Whenmagnetic field is used, however, the resistivity of CFCO isconsiderably decreased in awider temperature region centered atTC (figures 2(b) and (c)). As a result, a negativemagnetoresistance effect (MR= [ρ(H)− ρ(0)]/ρ(0) × 100%) is obtained thanks to the reduction of spinfluctuation and spin scattering under field, as reported inmany other ferromagneticmetals [25, 26]. As shown infigure 2(c), themaximumMRvalue (∼−13%) is clearly observed aroundTC.

2

Mater. Res. Express 2 (2015) 046103 HLXia et al

Note that in addition to the presence ofmultiplemagnetic interactions, other effects such as imperfect andinhomogeneous crystalmay also lead to broad transitions inmagnetism, electricity and specific heat etc. In ordertofigure out this question, we prepared a high-quality SrFe0.8Co0.2O3 single crystal with a simple cubicperovskite structure [19].Ourmeasurement results show that SrFe0.8Co0.2O3 exhibits very similar phasetransitions inmagnetism and transport properties with those of CFCO. As an example, the specific heat ofSrFe0.8Co0.2O3 single crystal was shown in the inset offigure 2(a), and a broad humpwas also observed near theCurie temperature (∼250 K).We therefore conclude that the cooperated effects ofmultiple FM interactionsinstead of sample quality should be responsible for the broad FM transition in the present CFCO.

Figure 3 shows thefield dependent isothermalmagnetization curvesmeasured at several representativetemperatures. AboveTC, the observed linearmagnetization behavior is consistent with the PM feature. BelowTC, themagnetization increases sharply withfield and tends to saturate withfield up to about 2 T. The saturatedmoment obtained at 7 T and 5 K is 3.1 μB f.u.

−1, which is slightly smaller than the completely local spinmoment(3.7 μB f.u.

−1) for the high-spin Fe4+ and intermediate-spin Co4+ ions inCFCOdue to themetallic conductivityas presented infigure 2(b). The observed coercive field is about 0.1 T at 5 K.With increasing temperature to120 K, the coercive filed considerably decreases to <0.01 T (inset of figure 3). It is therefore expected that themagnetic hysteresis should be negligible aroundTC.

Figure 4(a) shows a series of isothermalmagnetization curves between 150 and 210 K (applied field 0–6T)spanning the ferromagnetic transition region to study theMCEofCFCO. The area between two isothermschanges slightly, indicating a gradual variation in themagnitude of entropy change. As is well known, theArrottplot (H/M versusM2) provides evidence to identify the nature ofmagnetic phase transition [27]. In general, anegative slope of this plot indicates the occurrence of afirst-ordermagnetic transition such as amagnetostrictivetransition as observed in La2/3Ca1/3MnO3 [28]. In contrast, a positive slope corresponds to a second-ordermagnetic transition [29, 30]. From the viewpoint of practical application ofMCE, thematerial with a second-ordermagnetic transition is advantaged over thatwith afirst-order one because of the lowhysteresis in boththermal andmagnetic cycles [5, 31]. As presented in figure 4(b), all the Arrott curves at the vicinity ofTC inCFCO showpositive slops, suggesting that the PM to FM transition in this compound is second-ordered. This isalso in accordancewith the temperature dependences ofmagnetic susceptibility and resistivity, where nothermal hysteresis behavior is found aroundTC (figures 1(a) and 2(b)).

Figure 1.Temperature dependence of (a)magnetic susceptibility and its temperature derivative, and (b) inverse susceptibility. Thered line in (b) shows theCurie-Weiss fitting between 250 and 300 K.

3

Mater. Res. Express 2 (2015) 046103 HLXia et al

The negative values of the entropy change -ΔSM(T) can be calculated by the following formula [32]:

Δ−′ + ′

=∑

′ + ′S

T T

T T21 2

2 1

⎛⎝⎜

⎞⎠⎟

HereΣ is the area between twoMH curvesmeasured at temperaturesT1′ andT2′, respectively. Figure 5 showsthe temperature dependence of entropy change in awide temperature region (15–295 K)measured underdifferentmagnetic fields. The peaks of these curves shift fromTC to∼190 Kwith increasingfield applied from0to 6 T.When the appliedfield is 2 T, which is available for permanentmagnet in industry, themaximumof−ΔSM is 2.5 J kgK−1. This is a largemagnitude for perovskiteMCEmaterials. Furthermore, when the appliedfield increases to 6 T, themaximal value of−ΔSM increases to 5.9 J kgK−1, which is similar with those of thewell-studiedMn-based perovskiteMCEmaterials such as La0.87Sr0.13MnO3 (5.8 J kgK

−1) [2, 33].Since the FMphase transition temperature window is rather wide in the present CFCO, a largeworking

temperature is expected to occur.We therefore calculated the full width at halfmaximum (TFWHM) for the

Figure 2.Temperature dependence of (a) heat capacitymeasured at zeromagnetic field, and (b) resistivity respectivelymeasured at 0and 7 Ton cooling and heating, and (c) themagnetoresistance (MR= [ρ(H)− ρ(0)]/ρ(0) × 100%) calculated using the cooling data.The inset in (a) shows the specific heat of high-quality SrFe0.8Co0.2O3 single crystal.

4

Mater. Res. Express 2 (2015) 046103 HLXia et al

temperature dependent−ΔSMpeak by a Lorenz fitting. Figure 6(a) shows themagnetic field dependentTFWHM

ofCFCO. For comparison, some reportedTFWHM values on other perovskite oxides are also plotted in this figure[28, 34, 35]. Apparently, theTFWHM values of SrFe0.5Co0.5O2.83 andCFCOare significantly larger than those ofBaFeO3 and the A-site doped LaMnO3, because several cooperated FM interactions coexist in the former two

Figure 3. Isothermalmagnetization curvesmeasured at three representative temperatures. Inset shows the low-fieldmagnetizationcurves to clarify the coercive fields.

Figure 4. (a) A series of isothermalmagnetization curves across the FM transition temperature in 0–6 T. (b) Arrott plot (H/M versusM2) constructed from themagnetization data.

5

Mater. Res. Express 2 (2015) 046103 HLXia et al

compounds. It indicates that the performance ofMCE can be enhanced by introducingmultiple FM interactionsto increase theworking temperature. According to the obtained−ΔSM,we calculated the refrigerant capacity by

the formula ∫ Δ= − S dTRCT

TM

1

2asmentioned before. The calculated RC values at different fields are shown in

figure 6(b). Since the present CFCOpossesses awider working temperature (as characterized by the large

Figure 5.Temperature dependence of entropy change−ΔSMat differentmagnetic fields.T1 andT2 represent the both sides of the half-maximumvalue of−ΔSM forH= 6 T.

Figure 6.Magnetic field dependent (a) full width at halfmaximum(TFWHM), and (b) refrigerant capacity (RC) of CFCOand otherperovskite oxides.

6

Mater. Res. Express 2 (2015) 046103 HLXia et al

TFWHM) and a considerable entropy change, it exhibits a promising refrigerant capacity. For example, the RCvalue of CFCOobtained at 5 T (6 T) is 290 J kg−1 (355 J kg−1), which is almost the largest one observed inperovskite oxides to the best of our knowledge and even is comparable with those exhibited by typical giantMCEalloys such asGd5Ge2Si2 (305 J kg

−1) with afirst-ordermagnetic transition [2, 6, 32, 36].

4. Conclusions

The oxygen-full CaFe0.7Co0.3O3 single crystal was obtained for the first time by integrating floating zonemethodand high-pressure treatment technique.Magnetization, specific heat, electric transport andmagnetocaloriceffect were studied in detail. A sluggish paramagnetic to ferromagnetic transitionwas observed in a broadertemperature region between 200 and 150 K. All the Arrott curves at the vicinity ofTC showed positive slops,suggesting that themagnetic transitionwas second-ordered in nature. Accompanyingwith the FM transition, nosharp anomaly was found in specific heat and resistivity at zeromagnetic field.However, at highmagnetic field,the resistivity decreased considerably near the FM transition temperature, and amaximumMRvaluewasobserved atTC.On the basis of detailed isothermalmagnetizationmeasurements, the entropy changewasobtained to be−ΔSM(T)≈ 5.9 J kgK−1 at 6 T, whichwas considerably large for perovskiteMCEmaterials.Because of the broadening PM-FM transition in the present CFCO single crystal, a wider working temperatureregionwhichwas characterized by the value ofTFWHMof aΔSM(T) peakwas found. As a result, a significantrefrigerant capacity (355 J kg−1 at 6 T)was obtained. This value is almost the largest one in perovskite oxides andeven comparable to conventional giantMEC alloys likeGd5Ge2Si2 with afirst-ordermagnetic transition. Inconsideration of thewideworking temperature, large−ΔSM(T) andRC, as well as excluding rare elements, thepresent CaFe0.7Co0.3O3 provides a promising candidate formagnetic refrigeration.

Acknowledgments

The authors thankWHe andY JKe for their useful discussion. This workwas partly supported by 973 Project oftheMinistry of Science andTechnology of China (GrantNo. 2014CB921500), and by the Strategic PriorityResearch Programof theChinese Academy of Sciences (GrantNo. XDB07030300).

References

[1] Ghodhbane S, Tka E,Dhahri J andHlil EK 2014 J. Alloys Compd. 600 172[2] PhanMHandYu SC2007 J.Magn.Magn.Mater. 308 325[3] Jerbi A, Thaljaoui R, KricheneA andBoujelbenW2014PhysicaB 442 21[4] Arayedh B, Kallel S, KallelN and PeñaO2014 J.Magn.Magn.Mater. 361 68[5] Gschneidner KA Jr, PecharskyVK andTsokol AO2005Rep. Prog. Phys. 68 1479[6] Pecharsky VK andGschneidner KA Jr 1997Phys. Rev. Lett. 78 4494[7] PasqualeM and SassoCP 2004 J. Appl. Phys. 95 6918[8] WadaH andTanabe Y 2001Appl. Phys. Lett. 79 3302[9] Rodgers J A,Williams A J andAttfield J P 2006Z.Naturforsch.B 61 1515[10] Goodenough J B 2004Rep. Prog. Phys. 67 1915[11] ZhongW,ChengW,DingWP, ZhangN,DuYWandYanQ J 1998 Solid State Commun. 106 55[12] PhanMH,Yu SC andHurNH2005Appl. Phys. Lett. 86 072504[13] Szewczyk A,GutowskaM,Dabrowski B, Plackowski T,DanilovaNP andGaidukov YP 2005Phys. Rev.B 71 224432[14] Woodward PM,CoxDE,Moshopoulou E, Sleight AWandMorimoto S 2000 Phys. Rev.B 62 844[15] Kawasaki S, TakanoM,KannoR, Takeda T and Fujimori A 1998 J. Phys. Soc. Japan 67 1529[16] Kawasaki S, TakanoMandTakeda Y 1998 Solid State Ion. 108 221[17] MostovoyM2005Phys. Rev. Lett. 94 137205[18] Long YW,KanekoY, Ishiwata S, Taguchi Y andTokura Y 2011 J. Phys.: Condens.Matter. 23 245601[19] Long YW,KanekoY, Ishiwata S, Tokunaga Y,Matsuda T,WadatiH, Tanaka Y, Shin S, Tokura Y andTaguchi Y 2012 Phys. Rev.B 86

064436[20] ShannonRD1976Acta Crystallogr.A32 751[21] Fujioka J, Ishiwata S, KanekoY, Taguchi Y andTokura Y 2012Phys. Rev.B 85 155141[22] Okamoto J et al 2005Phys. Rev.B 71 104401[23] Potze RH, SawatzkyGA andAbbateM1995Phys. Rev.B 51 11501[24] Lavigne B andCoxRT 1991Phys. Rev.B 43 12374[25] HwangHY,Cheong SW,OngNP andBatlogg B 1996Phys. Rev. Lett. 77 2041[26] Coey JMD,ViretM andMolnaír SV 1997Adv. Phys. 48 167[27] Banerjee S K 1964Phys. Lett. 12 16[28] Mira J andRivas J 2004Mod. Phys. Lett.B 18 725[29] MizumakiM, Yoshii K,HayashiN, Saito T, Shimakawa Y andTakanoM2013 J. Appl. Phys. 114 073901[30] Zhang YD, Lampen P J, PhanT L, Yu SC, SrikanthH andPhanMH2012 J. Appl. Phys. 111 063918[31] Shen BG, Sun J R,Hu FX, ZhangHWandCheng ZH2009Adv.Mater. 21 4545[32] Zou JD, Shen BG,GaoB, Shen J and Sun J R 2009Adv.Mater. 21 693

7

Mater. Res. Express 2 (2015) 046103 HLXia et al

[33] Szewczyk A, SzymczakH,Wisniewski A, Piotrowski K, Kartaszynski R,Dabrowski B, Kolesnik S andBukowski Z 2000Appl. Phys. Lett.77 1026

[34] ToulemondeO, Abel J, YinC,WattiauxA andGaudin E 2012Chem.Mater. 24 1128[35] IslamMS,HanhDT,Khan FA,HakimMA,MinhDL,HoangNN,HaiNHandChauN2009PhysicaB 404 2495[36] ProvenzanoV, Shapiro A J and Shull RD2004Nature 249 853

8

Mater. Res. Express 2 (2015) 046103 HLXia et al