Crystalline and Electronic Structures and Magnetic andElectrical Properties of La-Doped Ca2Fe2O5 Compounds

T.L. PHAN,1 P.T. THO,1 N. TRAN,1 D.H. KIM,1 B.W. LEE,1,4 D.S. YANG,2

D.V. THIET,3 and S.L. CHO3

1.—Department of Physics and Oxide Research Center, Hankuk University of Foreign Studies,Yongin 449-791, South Korea. 2.—Department of Science Education, Chungbuk NationalUniversity, Cheongju 360-763, South Korea. 3.—Department of Physics, University of Ulsan,Ulsan 680-749, South Korea. 4.—e-mail: bwlee@hufs.ac.kr

Brownmillerite Ca2Fe2O5 has been observed to exhibit many outstandingproperties that are applicable to ecotechnology. However, very little work ondoped Ca2Fe2O5 compounds has been carried out to widen their applicationscope. We present herein a detailed study of the crystalline/geometric andelectronic structures and magnetic and electrical properties of Ca2�xLaxFe2O5

(x = 0 to 1) prepared by conventional solid-state reaction. X-ray diffractionpatterns indicated that the compounds with x = 0 to 0.05 exhibited brown-millerite-type single phase. La doping with higher content (x ‡ 0.1) stimulatedadditive formation of Grenier- (LaCa2Fe3O8) and perovskite-type (LaFeO3)phases. Extended x-ray absorption fine structure spectroscopy at the FeK-edge and electron spin resonance spectroscopy revealed presence of Fe3+ inthe parent Ca2Fe2O5 (x = 0) and both Fe3+ and Fe4+ in the doped compounds(x ‡ 0.05). The Fe4+ content tended to increase with increasing x. This stim-ulates ferromagnetic exchange interactions between Fe3+ and Fe4+ ions anddirectly influences the magnetic properties of Ca2�xLaxFe2O5. Electricalresistivity (q) measurements in the temperature range of T = 20 K to 400 Krevealed that all the compounds exhibit insulator behavior; the q(T) data forx ‡ 0.1 could be described based on the adiabatic small polaron hopping model.

Key words: Brownmillerite-type oxides, crystal structure, magneticproperties, electrical properties

INTRODUCTION

In recent years, Ca2Fe2O5+d with d = 0 hasattracted intensive interest from research groups,because of its novel properties applicable to ecotech-nology, such as its suitability for reliable resistive-switching memory,1 electrochemical energy conver-sion devices (i.e., solid-oxide fuel cells and metal airbatteries),2 gas sensors,3 chemical looping H2 pro-duction,4 and reversible CO2 capture for air purifi-cation.5 At room temperature, it has brownmillerite-type orthorhombic structure in space group Pcmnwith lattice constants a = 5.425 A, b = 14.769 A,

and c = 5.598 A.6,7 This structure can be describedas alternating layers of corner-sharing FeO6 octa-hedra and FeO4 tetrahedra. The Fe3+ spins arenearly directed along the c axis, leading to G-typeantiferromagnetic (AFM) order and weak ferromag-netism in Ca2Fe2O5.3,8

Detailed investigation of its thermal expansionhas revealed that this compound exhibits severaltransitions,8–10 including an antiferromagnetic–paramagnetic (AFM–PM) transition at TN � 725 K(the Neel temperature), transformation of the orig-inal Pcmn structure to a body-centered structureat � 970 K, an incommensurate modulated transi-tion caused by lattice expansion in the a and cdirections at � 1180 K, and significant lattice con-traction along the b direction at � 1308 K.10 Thesetransitions can change the oxygen content (d), and

(Received July 17, 2017; accepted September 27, 2017;published online October 16, 2017)

Journal of ELECTRONIC MATERIALS, Vol. 47, No. 1, 2018

DOI: 10.1007/s11664-017-5841-x� 2017 The Minerals, Metals & Materials Society

188

result in two symmetry types with left and rightFeO4 tetrahedral chains along the c axis in thesame layer.9,10

In terms of its electrical properties, Ca2Fe2O5 isan insulator and considered as a model system forunderstanding oxygen-ion-related conductivity. Alarge change in the oxygen content of Ca2Fe2O5+d

with d = 1 results in formation of the end systemCaFeO3 (with perovskite-type structure ofABO3),11,12 and a metal–insulator transitionat � 290 K.13 The electrical conductivity ofCa2Fe2O5 has been found to be enhanced by Ndoping2 or (Mg, Al) codoping.14 More recently,Dhankhar and coworkers found anomalous magne-toresistance below the TN of Ca2Fe2O5, and amemory effect when measuring the resistivityunder an external magnetic field.15

Though promising applications of this materialsystem have already been demonstrated, very littlework on rare-earth-doped Ca2Fe2O5 materials hasbeen carried out.2,14,16 Additionally, systematicstudies of the structural/geometric, electronic, mag-netic, and electrical properties have not yet beenconducted. Assessment of their applicability for usein electronic devices is also still lacking. In thiswork, we prepared La-doped Ca2Fe2O5 compounds,then studied their crystal/geometric and electronicstructures and magnetic and electrical properties.We suggest that La doping creates more holes (i.e.,more Fe4+ ions), because of La3+ substitution forCa2+ in Ca2�xLaxFe2O5. Our studies show that Ladoping causes structural phase separation and achange in the concentrations of Fe3+ and Fe4+ ions.This influences the strength of the ferromagnetic(FM) exchange interactions between Fe3+ and Fe4+

ions as well as the electrical conductivity of theCa2�xLaxFe2O5 compounds.

EXPERIMENTAL PROCEDURES

Polycrystalline Ca2�xLaxFe2O5 compounds (x = 0to 1) were fabricated by conventional solid-statereaction in air. Commercial powders of CaO, La2O3,and Fe2O3 with purity of 99.9% were used asprecursors. These powders were combined in stoi-chiometric amounts according to the nominal for-mula Ca2�xLaxFe2O5, then well mixed by using anagate pestle and mortar. The mixtures were thenpressed into pellets and preannealed at 1373 K for12 h. After several intermediate applications of thegrinding and preannealing processes at tempera-tures below 1523 K, the final pellets were annealedonce more at 1573 K for 24 h. The crystal structureof the resulting products was checked by x-raydiffractometer (Rigaku, MiniFlex) equipped with CuKa radiation source (wavelength k = 1.5406 A).Notably, before taking x-ray diffraction (XRD) pat-terns, a small amount (� 5 wt.%) of standard Sipowder was mixed with all compounds, to minimizeerrors caused by position calibration of the incidentx-ray beam. The electronic structure associated

with the Fe absorbing atom (E0 = 7112 eV)17 inCa2�xLaxFe2O5 was studied by using x-ray absorp-tion fine structure (XAFS) spectroscopy (PohangAccelerator Laboratory, South Korea). Data in thex-ray absorption near-edge structure (XANES)region were collected in steps of 0.2 eV to 0.5 eV.The XAFS data were analyzed using the IFEFFITsoftware package. For reference, we also recordedXAFS spectra of Fe3O4 (with a mixture of Fe2+ andFe3+) and Fe2O3 (Fe3+), corresponding to E0 valuesof about 7121 eV and 7123 eV, respectively. Elec-tron spin resonance (ESR) spectra were recordedfrom the powder compounds using a JEOL-TE300spectrometer operating in the X-band frequencyrange at f � 9.45 GHz. Magnetization (M) versusmagnetic field measurements were performed byvibrating-sample magnetometry (VSM), varying themagnetic field (H) from 0 kOe to 10 kOe. Theelectrical resistivity was measured in the tempera-ture range of T = 20 K to 400 K by four-probetechnique in van der Pauw configuration.

RESULTS AND DISCUSSION

Crystalline, Local Geometric, and ElectronicStructures

Figure 1 shows the powder XRD patterns of allthe Ca2�xLaxFe2O5 compounds, with fixed scan stepof 0.01�. The XRD features reveal structuralchanges of Ca2�xLaxFe2O5 as x was increased. Thiscan be seen from the enlarged view of the diffractionpeaks at around 31� to 35� in Fig. 2. Detailedanalysis revealed that the compounds with x = 0and 0.05 completely crystallized with orthorhombicbrownmillerite structure (in space group Pcmn),corresponding to the Miller-indexed peaks.

Fig. 1. Room-temperature XRD patterns of all Ca2�xLaxFe2O5+d

compounds prepared by solid-state reaction. The Miller-indexedpeaks belong to the Pnma brownmillerite phase (dotted lines).Additional peaks without Miller indices are assigned to Grenier- and/or perovskite-type phases.

Crystalline and Electronic Structures and Magnetic and Electrical Properties of La-DopedCa2Fe2O5 Compounds

189

However, for x ‡ 0.1, additional peaks started toappear and the brownmillerite peaks graduallydisappeared. By comparison with the referencecompounds (Fig. 2), we believe that the additivepeaks are related to development of Grenier-(LaCa2Fe3O8) and/or perovskite-type (LaFeO3)phases caused by the increase in oxygen content;Fig. 3 shows the 3D crystal geometry of three phasestructures for comparison. This means that thegeneral formula of our compounds must be rewrit-ten as Ca2�xLaxFe2O5+d. In fact, Ca2Fe2O5 andLaFeO3 have been found to be two end members ofLaCa2Fe3O8 (i.e., the LaCa2Fe3O8 structure isintermediate between the Ca2Fe2O5 and LaFeO3

structures), and the following process would happenduring sample fabrication7,18,19:

aLaCa2Fe3O8þz ! bCa2Fe2O5 þ bLaFeO3

þ cLaCa2Fe3O8þz0 ;ð1Þ

where a = b + c, z, and z¢ are variables. Coexistenceof brownmillerite, Grenier, and/or perovskitephases in the compounds with 0.1 £ x £ 1 leads tocomplex variation of the XRD patterns. Note that asmall amount of Ca could partially substitute for Lain perovskite LaFeO3, resulting in La1�yCayFeO3-type compounds, peak shifts, and secondary XRDpeaks.

We additionally analyzed the brownmilleritephase in the compounds with x = 0 to 0.2, anddetermined the lattice parameters (a, b, c) and unitcell volume (V). The values of a = 5.424 A,b = 14.761 A, and c = 5.589 A determined forCa2Fe2O5 (x = 0) in Table I are in good agreementwith those previously reported.6,7 With increasing xfrom 0 to 0.1, these parameters slightly increased,owing to replacement of La3+ with larger ionicradius (1.16 A) for Ca2+ with smaller radius (� 1.0A). However, increase of x above 0.1 reduced theunit cell parameters, owing to formation of Grenierand perovskite phases. These results indicate thatxc = 0.05 is the concentration threshold for forma-tion of brownmillerite Ca2�xLaxFe2O5 single-phasecompounds. For other systems reported previously,such as Ba2(In1�xGax)2O5,20 Ba2(In1�xAlx)2O5,21 andCa2FeAl1�xMgxO5,14 brownmillerite single phasewas also found for x< 0.2. However,Ca2Fe2�xAlxO5-based brownmillerite compoundscould be prepared with higher x values, withxc � 0.65.22 Synthesis of brownmillerite compoundswith all x values is very difficult and dependent onmany experimental factors.

Variation of x and d in Ca2�xLaxFe2O5+d alsochanged the valence state of Fe, as revealed byXAFS spectroscopy. Figure 4 shows normalized FeK-edge XANES spectra (upper panel) and their firstderivative (lower panel) for the compounds withx = 0, 0.1, 0.3, 0.5, 0.7, and 1 at room temperature.

Fig. 2. Enlarged view of XRD patterns of Ca2�xLaxFe2O5+d in thediffraction region of 31� to 35�, compared with those of referenceoxides LaCa2Fe3O8 and LaFeO3.

Fig. 3. Crystal geometry of (a) brownmillerite Ca2Fe2O5, (b) Grenier LaCa2Fe3O8, and (c) perovskite LaFeO3 structures.

Phan, Tho, Tran, Kim, Lee, Yang, Thiet, and Cho190

Careful analysis of the XANES spectra reveals threepoints of interest as follows:

(1) The intensity of the pre-edge peak (labeled P1)at around 7112 eV attributed to the 1s fi 3ddipole-forbidden transition of Fe ions23,24 is associ-ated with the symmetry of the crystal field aroundthe Fe absorbing atom. Tetrahedrally coordinatedFe ions give strong P1 intensity, due to the mixtureof p and d orbitals in the tetrahedral symmetry, forthe case of brownmillerite Ca2Fe2O5. Meanwhile,low P1 intensity will be observed for octahedrallycoordinated Fe ions, due to the quadrupole-allowedtransition in the octahedral symmetry, for the caseof perovskite LaFeO3.

(2) There is a shoulder (labeled P2 in Fig. 4) ataround 7119 eV, which is ascribed to the 1s fi 4p

transition with simultaneous ligand-to-metalcharge transfer.24 Similar to the case of P1, theintensity of P2 is also dependent on the crystal fieldsymmetry. It is strong for brownmillerite-typestructure with alternating sheets of FeO4 andFeO6, and weak for other (i.e., Grenier and per-ovskite) phases. The intensity decreased and thepeak shifted towards higher energy for P1 and P2

with increasing x (Fig. 4 and inset), demonstratingthe brownmillerite-to-perovskite transformationthrough the Grenier phase.

(3) The midpoint of the absorption edge (labeledEexp), corresponding to the maximum of the firstderivative curve (Fig. 4, lower panel), correlateswith the valence state of Fe atoms, being about7123 eV for the parent Ca2Fe2O5 (x = 0). This valueis the same as the E0 value of Fe3+ in Fe2O3.However, with increasing x, Eexp tended to shifttowards higher energy corresponding to Fe4+, beingabout 7125 eV for x = 1, as presented in Table I.

The local geometric structure around Fe ions canbe determined by plotting the Fourier-transformed(FT) spectra of the extended x-ray absorption finestructure (EXAFS) region. Specifically, the FT spec-tra of the Ca2�xLaxFe2O5+d compounds (Fig. 5)showed a strong peak centered at about 1.5 A,corresponding to the average bond distance of Fe–O(RFe–O). Due to the phase shift of photoelectronbackscattering, this peak was shifted by � 0.5 A onthe R axis from the standard value.25 Table I alsopresents the average values of RFe–O estimated forthe Ca2�xLaxFe2O5+d compounds. In previous work,Ryu et al.24 showed that the Fe–O distances in theparent Ca2Fe2O5 lie in the range of 1.963 A to1.977 A for Fe–O6 and 1.884 A to 1.858 A for FeO4.Their average distance of � 1.911 A is very close tothe value RFe–O = 1.910 A obtained for our Ca2Fe2O5

compound. In particular, with increasing x, therewas a gradual increase of RFe–O from 1.924 A forx = 0.1 to 2.001 A for x = 1, as shown in Table I. Thisenhancement of RFe–O is related to the structuraltransformation tendency from Ca2Fe2O5 throughLaCa2Fe3O8 to LaFeO3, because the perovskite

Fig. 4. Fe K-edge XANES spectra (upper panel) and their firstderivative (lower panel) for typical Ca2�xLaxFe2O5+d compounds.Inset shows enlarged view of pre-edge peak P1.

Table I. Experimental values of some parameters determined by studying the crystalline and electronicstructures and electrical and magnetic properties of La-doped Ca2Fe2O5 compounds. The structuralparameters (a, b, c, and V) were only calculated for compounds that crystallized mainly in brownmilleritephase

Compound a (A) b (A) c (A) V (A3) Eexp (eV) RFe–O (A) Ea (eV)

x = 0 5.424 14.761 5.589 447.48 7123.1 1.910 –x = 0.05 5.429 14.775 5.595 448.79 – – –x = 0.1 5.433 14.785 5.593 449.27 7123.3 1.924 123.8x = 0.15 5.427 14.780 5.592 448.54 – – –x = 0.2 5.421 14.786 5.594 448.39 – – 98.0x = 0.3 – – – – 7123.7 1.958 94.8x = 0.5 – – – – 7123.9 1.958 97.4x = 0.7 – – – – 7124.2 1.975 95.1x = 1 – – – – 7124.9 2.001 104.6

Crystalline and Electronic Structures and Magnetic and Electrical Properties of La-DopedCa2Fe2O5 Compounds

191

LaFeO3 contains only FeO6 octahedron with RFe–

O = 1.974 A to 2.006 A,24,26 fairly close to the RFe-O

values of our compounds with x = 0.7 and 1.Together with the peak shift of RFe–O, its intensity

also increased with increasing x in the Ca2�xLax-

Fe2O5+d compounds. This is thought to be due to anincrease in the coordination number of oxygenatoms surrounding the Fe atom, because theEXAFS spectrum is the sum of outgoing andincoming waves, which depend on the immediateenvironment of the absorbing atom.25 In otherwords, d in Ca2�xLaxFe2O5+d tends to increase withincreasing x, in good agreement with the XRDanalysis above. In summary, the studies of thegeometric and electronic structures demonstratethat: (1) only Fe3+ ions are present in Ca2Fe2O5

(x = 0), (2) both Fe3+ and Fe4+ ions coexist inCa2�xLaxFe2O5+d with x> 0, (3) the concentrationof Fe4+ ions tends to increase with increasing x, andthese ions are thought to be favored in the Grenierand perovskite structures, and (4) increase of xabove 0.05 leads to the structural phase transfor-mation Ca2Fe2O5 fi LaCa2Fe3O8 fi LaFeO3,due to the increase of the oxygen content. Thesechanges directly influence the magnetic and electri-cal properties of the Ca2�xLaxFe2O5+d compounds.

Magnetic Properties

The magnetic properties of the Ca2�xLaxFe2O5+d

compounds at room temperature were studied bymeans of ESR and magnetization measurements.Their ESR spectra indicated that the signal inten-sity (Int) was very weak for x = 0, but became strongon increasing x from 0.05 to 1 (Figs. 6 and 7a).Concurrently, the spectral line width (DH) andresonant field (Hr) also changed (Fig. 7b and c). Inour work, the Hr values (3040 Oe to 3200 Oe) are

smaller than the Hr value of free/unpaired electrons(� 3300 Oe),27 and depend on x. This proves thatthe resonant signals mainly arise from FM and/orAFM coupling between the magnetic moments ofFe3+ and Fe4+ ions. For the parent compoundCa2Fe2O5 (x = 0), its weak intensity indicatesFe3+–Fe3+ AFM pairs persisting up to

Fig. 6. Room-temperature ESR spectra of the Ca2�xLaxFe2O5+d

compounds.Fig. 5. k3-weighted FT EXAFS spectra collected at the Fe K-edgefor typical Ca2�xLaxFe2O5+d compounds.

Fig. 7. Variation of ESR spectral parameters for Ca2�xLaxFe2O5+d

compounds as function of x: (a) Int, (b) DH, and (c) Hr.

Phan, Tho, Tran, Kim, Lee, Yang, Thiet, and Cho192

TN � 700 K,8–10 because only Fe3+ ions are presentin this compound. With increasing x from 0.05 to0.2, in which range the brownmillerite and Grenierphases coexist, the appearance of Fe4+ ions resultsin Fe3+–Fe4+ FM and Fe4+–Fe4+ AFM exchangepairs, enhancing Int. For the compounds withx> 0.2, the complex variations in the spectralparameters are related to the coexistence of theGrenier and perovskite phases, and creation of moreFe4+–Fe4+ AFM pairs in addition to Fe3+–Fe4+ FMpairs. Here, the different variation tendencies ofInt, DH, and Hr are due to their different nature anddepend on the volume fractions of the crystal phasesin the compounds. While Int is mainly associatedwith the magnetic susceptibility, DH is associatedwith the total magnetic moments that contribute toresonance. Meanwhile, Hr is dependent on theLande g-factor characteristic of spin–spin andspin–orbit couplings27 of Fe3+ and Fe4+ ions. Obvi-ously, together with EXAFS studies, use of ESRspectroscopy as an auxiliary tool is very helpful todetect small changes in the concentration of Fe4+ inCa2�xLaxFe2O5+d, particularly for the compoundswith x = 0.05 to 0.2.

Considering the M(H) results (Fig. 8), it emergesthat all the Ca2�xLaxFe2O5+d compounds exhibitmagnetic hysteresis loops with weak FM/AFM orderat room temperature. These M(H) data can bedivided into two characteristic groups. The firstgroup includes the compounds with x £ 0.3, whoseloop areas at magnetic fields H< 2 kOe are narrow;at higher fields, no loop is observed and M increaseslinearly with increasing H. The other groupincludes the compounds with x> 0.3, whose loopareas are much larger than those of the first group;at magnetic fields H> 7 kOe, there is no hysteresisloop and the M(H) dependences are linear. Suchlinear M(H) dependence at high field could be

related to various effects such as (1) the magneticmoments in FM/AFM regions have not yet reachedsaturation state, (2) short-range FM order, and/or(3) paramagnetism of isolated Fe3+,4+ ions. For thehysteresis loop regions associated with FM/AFMinteractions, one can analyze the x dependence ofthe saturation magnetization (Mr) and coercivity(Hr), as shown in Fig. 9. For the first group, Mr

gradually decreased from 0.047 emu/g for x = 0to � 0.038 emu/g for x = 0.3, while Hc changed less(86 Oe to 129 Oe). However, for the second group,both Ms and Hc increased rapidly with increasing xfrom 0.4 to 1. These variations of Ms and Hc areassociated with the changes in the Fe3+/Fe4+ ratioand structural phases. As mentioned above, brown-millerite phase is dominant for x< 0.3, where theFe3+–Fe3+ AFM interaction plays a pivotal role. Theappearance of Fe4+ leads to addition of the Fe3+–Fe4+ FM interaction, which competes with the AFMone, slightly reducing Ms. For the Grenier phasewith dominantly AFM-ordered Fe3+ ions,18 webelieve that the presence of Fe4+ in this structurealso does not enhance Ms and Hc, because its volumefraction is relatively large, in the range of x = 0.1 to0.3. However, for the compounds with x> 0.3,higher x results in development of perovskite phase,creating more Fe3+–Fe4+ FM pairs. These factorsenhance Ms and Hc, similar to the case of La1�xCax-

FeO3 compounds.28,29 In other words, the enhancedFM behavior for x> 0.3 is related to Fe4+ ions in theperovskite rather than brownmillerite or Grenierstructure. Although Fe3+–Fe4+ FM pairs are presentin the Ca2�xLaxFe2O5+d compounds with x> 0.3,the AFM interaction due to Fe3+–Fe3+ and Fe4+–Fe4+ pairs remains dominant. The narrowing of theM(H) loop for fields H< 2.5 kOe for x = 1 is thusascribed to competition between the two FM andAFM phases in different crystal phases. The vari-ation tendency of Ms (Fig. 9) and Hr (Fig. 7c) versusx is more or less the same. This is due to their

Fig. 8. Room-temperature M(H) loops of typical Ca2�xLaxFe2O5+d

compounds. Inset shows enlarged view of M(H) curves for thecompounds with x = 0 to 0.3.

Fig. 9. Ms and Hc versus x in Ca2�xLaxFe2O5+d. Dotted lines areguides to the eye only; 5% error bars are shown.

Crystalline and Electronic Structures and Magnetic and Electrical Properties of La-DopedCa2Fe2O5 Compounds

193

correlation to FM resonance, as described by theKittel formulae.30

Electrical Properties

We measured the resistivity, q, of the Ca2�xLax

Fe2O5+d compounds in the temperature range ofT = 20 K to 400 K. However, the q for the com-pounds with x = 0 and 0.05 was above5 9 106 X cm, lying beyond the measurement rangeof our device. For the other compounds (x ‡ 0.1), werecorded their resistivity at temperatures ofT> 120 K, as shown in Fig. 10. Depending on xand T (from 120 K to 400 K), q varied over a widerange from 1 X cm to 106 X cm. For given x value, qdecreased exponentially with increasing T, being atypical characteristic of semiconductors. In particu-lar, the q–T dependences can be described based onthe adiabatic small polaron hopping (SPH) modelformulated as

qðTÞ ¼ q0 expEa

kBT

� �; ð2Þ

where Ea is the polaron formation energy and kB isthe Boltzmann constant. The resistivity coefficientq0 can be determined from the following equation:

q0 ¼ kB

nð1 � xÞe2d; ð3Þ

where n is the charge carrier density, x is the hole(Fe4+) concentration, and d is the polaron diffusionconstant, which depends on the lattice constantsand the characteristic frequency of the longitudinaloptical (LO) phonon carrying polarons through thelattice.31 The motion of charge carriers in theadiabatic region is considered to be faster thanlattice vibrations. Using Eq. 2 to fit the q(T) data inthe range of T = 130 K to 400 K (Fig. 11), wedetermined the values of Ea, as shown in Table Iand the inset of Fig. 11. The results reveal that Ea

varied as a function of x, being 123.8 meV for x = 0.1but decreasing rapidly to 94.8 meV for x = 0.2. Inthe range of x = 0.3 to 0.7, Ea (95 meV to 97 meV)remained relatively unchanged, but slightlyincreased at x = 1. These values indicate a closerelation between q, Ea, and Fe4+ ions. In otherwords, the decrease in q is due to an increase in theFe4+ concentration (with increased density of chargecarriers) that reduces Ea. Small polarons involved innearest-neighbor hopping become more mobile ver-sus thermal activation processes. Recently, Mal-veiro et al.14 studied brownmillerite Ca2FeAl1�x

MgxO5 (x = 0 and 0.05), finding large values of Ea

(260 meV to 620 meV), which tended to decrease ifmore Fe4+ ions were added to the brownmilleritelattice. Apart from the important role of Fe4+ ions,we believe that structural changes and grain-boundary-related effects also influence q and Ea,

because the LO phonon frequency and the hoppingof polarons would be modified by these factors.14,32

CONCLUSIONS

Detailed analysis of the XRD patterns, Fe K-edgeXANES spectra, and ESR spectra of Ca2�xLax

Fe2O5+d compounds (x = 0 to 1) at room temperaturerevealed the following: (1) only Fe3+ ions exist in theparent Ca2Fe2O5 (x = 0), (2) there is a mixture ofFe3+ and Fe4+ ions in Ca2�xLaxFe2O5+d with x> 0,(3) the Fe4+ content tends to increase with increas-ing x, and these Fe4+ ions preferably locate in theGrenier LaCa2Fe3O8 and perovskite LaFeO3 struc-tures, and (4) increase of x from 0.1 to 1 causes a

Fig. 10. Temperature-dependent resistivity, q(T), of Ca2�xLaxFe2O5+d compounds (x ‡ 0.1) as function of x.

Fig. 11. Log(q/T) versus 1000/T data (symbols) of Ca2�xLaxFe2O5+d

compounds with x> 0.1 fit to the SPH model (solid lines, Eq. 2).Inset shows x dependence of Ep varying on the meV scale; the solidline is a guide to the eye only.

Phan, Tho, Tran, Kim, Lee, Yang, Thiet, and Cho194

structural phase transformation due to the increas-ing oxygen and Fe4+ contents. M(H) measurementsat room temperature also revealed magnetic phaseseparation with two characteristic groups. For thefirst group, Mr gradually decreased from 0.047 emu/g for x = 0 to � 0.038 emu/g for x = 0.3, while Hc

changed less (86 Oe to 129 Oe). However, for thesecond group, both Ms and Hc increased rapidlywith increasing x from 0.4 to 1. These variations ofMs and Hc are tightly related to the changes in theFe3+/Fe4+ ratio and structural phases. Study of thetemperature-dependent resistivity in the range ofT = 20 K to 400 K revealed that all the compoundsare insulators. Although the q(T) values for thecompounds with x = 0 and 0.05 lay beyond ourmeasurement capability of 5 9 106 X cm, the com-pounds with x ‡ 0.1 showed lower q(T) values thatdecreased with increasing x from 0.1 to 1 and T from120 K to 400 K. We also found that the q(T) data forx ‡ 0.1 obeyed the adiabatic SPH model, withrapidly decreasing polaron formation energy Ea forx> 0.1. Apart from the important role of Fe4+ ions,structural changes and grain-boundary-relatedeffects are also thought to influence q and Ea.

ACKNOWLEDGEMENTS

This research was supported by the Basic ScienceResearch Program (NRF-2014R1A2A1A11051245,NRF-2017R1A2B4010490), and the InternationalCooperation Program (NRF-2015K2A1A2071001),through the National Research Foundation of Kor-ea, funded by the Ministry of Education, Science,and Technology.

REFERENCES

1. Q. Lu and B. Yildiz, Nano Lett. 16, 1186 (2016).2. C.P. Jijil, M. Lokanathan, S. Chithiravel, C. Nayak, D.

Bhattacharyya, S.N. Jha, P.D. Babu, B. Kakade, and R.N.Devi, ACS Appl. Mater. Interfaces 8, 34387 (2016).

3. M.D. Rossell, O.I. Lebedev, G. Van Tendeloo, N. Hayashi,T. Terashima, and M. Takano, J. Appl. Phys. 95, 5145(2004).

4. M. Ismail, W. Liu, M.S.C. Chan, M.T. Dunstan, and S.A.Scott, Energy Fuels 30, 6220 (2016).

5. K. Gupta, S. Singh, and M.S.R. Rao, Nano Energy 11, 146(2014).

6. T.C. Gibb, A.J. Herod, D.C. Munro, and N. Peng, J. Mater.Chem. 4, 1451 (1994).

7. J.M. Gonzalez-Calbet, M. Vallet-Regi, and M.A. Alario-Franco, J. Solid State Chem. 60, 320 (1985).

8. T. Takeda, Y. Yamaguchi, S. Tomiyoshi, M. Fukase, M.Sugimoto, and H. Watanabe, J. Phys. Soc. Jpn. 24, 446(1968).

9. K. Gupta, S. Singh, M. Ceretti, M.S.R. Rao, and W. Paulus,Phys. Status Solidi A 210, 1771 (2013).

10. T. Labii, M. Ceretti, A. Boubertakh, W. Paulus, and S.Hamamda, J. Therm. Anal. Calorim. 112, 865 (2013).

11. D. Petrov and B. Angelov, Acta Phys. Pol. A 122, 737(2012).

12. D. Petrov, B. Angelov, and V. Lovchinov, J. Alloys Compd.509, 5038 (2011).

13. P.M. Woodward, D.E. Cox, E. Moshopoulou, A.W. Sleight,and S. Morimoto, Phys. Rev. B 62, 844 (2000).

14. J. Malveiro, T. Ramos, L.P. Ferreira, J.C. Waerenborgh,M.R. Nunes, M. Godinho, and M.D. Carvalho, J. SolidState Chem. 180, 1863 (2007).

15. S. Dhankhar, K. Gupta, G. Bhalerao, N. Shukla, M.Chandran, B. Francis, B. Tiwari, K. Baskar, and S. Singh,RSC Adv. 5, 92549 (2015).

16. N. Abe, N.D. Khanh, T. Sasaki, and T. Arima, Phys. Rev. B89, 054437 (2014).

17. J.A. Bearden and A.F. Burr, Rev. Mod. Phys. 39, 125(1967).

18. J.M. Hudspeth, D.J. Goossens, A.J. Studer, R.L. Withers,and L. Noren, J. Phys. Condens. Matter 21, 124206(2009).

19. P.M. Price, N.D. Browning, and D.P. Butt, J. Am. Ceram.Soc. 98, 2248 (2015).

20. C. Didier, J. Claridge, and M. Rosseinsky, J. Solid StateChem. 218, 38 (2014).

21. I.V. Spesivtseva, N.A. Kochetova, E.M. Gorbunova, andI.E. Animitsa, Russ. J. Phys. Chem. A 85, 1816 (2011).

22. G.J. Redhammer, G. Tippelt, G. Roth, and G. Amthauer,Am. Miner. 89, 405 (2004).

23. A.E. True, R.C. Scarrow, C.R. Randall, R.C. Holz, and L.Que Jr, J. Am. Chem. Soc. 115, 4246 (1993).

24. K.H. Ryu, K. Sakaue, H. Terauchi, and C.H. Yo, Bull.Korean Chem. Soc. 18, 1047 (1997).

25. B.K. Teo, Basic Principles and Data Analysis (Berlin:Springer, 1986).

26. M. Marezio and P.D. Dernier, Mater. Res. Bull. 6, 23(1971).

27. M. Ikeya, New Applications of Electron Spin Resonance(Singapore: World Scientific, 1993).

28. Z.C. Xu, M.F. Liu, C.C. Chen, and X.N. Ying, J. Appl. Phys.115, 123516 (2014).

29. E.K. Abdel-Khalek and H.M. Mohame, Hyperfine Interact.222, S57 (2013).

30. C. Kittel, Introduction to Solid State Physics, 8th ed.(Hoboken: Wiley, 2004).

31. D. Varshney and N. Dodiya, J. Theor. Appl. Phys. 9, 45(2015).

32. W. Luo, H.M. Yao, P.H. Yang, and C.S. Chen, J. Am.Ceram. Soc. 92, 2682 (2009).

Crystalline and Electronic Structures and Magnetic and Electrical Properties of La-DopedCa2Fe2O5 Compounds

195