Valence band electronic structure of Nd1−xYxMnO3 using X-ray absorption, photoemission and GGA + U calculations

Valence band electronic structure of Nd1−xYxMnO3 using X-ray absorption, photoemission and GGA + U calculations

Journal of Electron Spectroscopy and Related Phenomena 189 (2013) 51–55 Contents lists available at ScienceDirect Journal of Electron Spectroscopy a...

1MB Sizes 0 Downloads 35 Views

Journal of Electron Spectroscopy and Related Phenomena 189 (2013) 51–55

Contents lists available at ScienceDirect

Journal of Electron Spectroscopy and Related Phenomena journal homepage: www.elsevier.com/locate/elspec

Valence band electronic structure of Nd1−x Yx MnO3 using X-ray absorption, photoemission and GGA + U calculations Padmanabhan Balasubramanian a,b,∗ , Harikrishnan. S. Nair c , H.M. Tsai a,g , S. Bhattacharjee d , M.T. Liu a , Ruchika Yadav e , J.W. Chiou f , H.J. Lin g , T.W. Pi g , M.H. Tsai h , Suja Elizabeth e , C.W. Pao a , B.Y. Wang a , C.H. Chuang a , W.F. Pong a a

Department of Physics, Tamkang University, Tamsui 251, Taiwan Institute of Physics, Bhubaneshwar 751005, India c Julich Center for Neutron Sciences, Forschungszentrum Julich, Outstation at FRM II, LichtenberGstr. 1, Garching 85747, Germany ˝ ˝ d Department of Physics and Astronomy, Uppsala University, Box 516, 75120 Uppsala, Sweden e Department of Physics, Indian Institute of Science, C.V. Raman Avenue, Bangalore 560012, India f Department of Applied Physics, National University of Kaohsiung, Kaohsiung 811, Taiwan g National Synchrotron Radiation Research Center, Hsinchu 300, Taiwan h Department of Physics, National Sun Yat-Sen University, Kaohsiung 804, Taiwan b

a r t i c l e

i n f o

Article history: Received 19 March 2013 Received in revised form 11 June 2013 Accepted 4 July 2013 Available online 12 July 2013 PACS: 78.70.Dm 74.72.−h 71.20.Eh Keywords: X-ray absorption Photoemission Density of states Strongly correlated systems

a b s t r a c t The electronic structures of Nd1−x Yx MnO3 (x = 0–0.5) were studied using X-ray absorption near-edge structure (XANES) at the Mn L3,2 - and O K-edge along with valence-band photoemission spectroscopy (VB-PES). The systematic increase in white-line intensity of the Mn L3,2 -edge with doping, suggests a decrease in the occupancy of Mn 3d orbitals. The O K-edge XANES shows a depletion of unoccupied states above the Fermi energy. The changes in the O K-edge spectra due to doping reflects an increase in the Jahn–Teller distortion. The VB-PES shows broadening of the features associated with Mn 3d and O 2p hybridized states and the shift of these features to a slightly higher binding energy in agreement with our GGA + U calculations. The system shows a net shift of the occupied and unoccupied states away from the Fermi energy with doping. The shift in theoretical site-projected density of states of x = 0.5 composition with respect to x = 0 suggest a subtle change from a charge transfer to Mott–Hubbard type insulator. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Perovskite rare-earth manganites (R1−x Ax MnO3 : R = La, Nd, Dy, etc.; A = Sr, Ca, etc.) are known for metal–insulator transition, colossal magnetoresistance, various structural transitions, magnetism, charge and orbital ordering [1]. The parent compound, RMnO3 , crystallizes in an orthorhombic Pbnm structure with highly distorted MnO6 octahedra [2]. For R = La, Nd and Pr, the magnetic ordering is simple A-type antiferromagnet (AFM). For higher R, the magnetic ordering is more complex with incommensurate spiral magnetic phases that give rise to spontaneous ferroelectric polarization [3]. Similarly in Nd1−x Yx MnO3 , the A-type AFM ordering of

∗ Corresponding author at: Institute of Physics, Bhubaneshwar 751005, India. Tel.: +91 674 2306444; fax: +91 674 2300142. E-mail addresses: [email protected], [email protected] (P. Balasubramanian). 0368-2048/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.elspec.2013.07.001

Mn sublattice gets suppressed with increasing x, resulting in spiral magnetic structure [4]. Between x = 0.3 and 0.5, A-type AFM coexists with an incommensurate spin density wave, while for x = 0.5 and above, the system exhibits only spiral phase along with appearance of spontaneous ferroelectric polarization [5]. Due to its trivalent nature, Y-doping does not introduce electrons or holes into the system which is essential for metallicity and ferromagnetism. Since Y3+ has a smaller ionic radii, there is a greater mismatch of the cations, followed by reduction in the tolerance factor and greater distortion of the MnO6 octahedra [1,4]. Even though there occurs no structural phase transitions in Nd1−x Yx MnO3 , the minor structural changes, can affect hybridization between Mn 3d–O 2p states and the charge transfer character of the system [6], thus modifying the local electronic structures around the Mn ions, which is the focus of this investigation. The electronic structures of pure and divalent doped manganites have been studied by X-ray absorption near-edge structure (XANES) and photoemission techniques [6–10]. XANES at the Mn

52

P. Balasubramanian et al. / Journal of Electron Spectroscopy and Related Phenomena 189 (2013) 51–55

incoming photon energy of 130 eV. The clean Au metal surface was used to calibrate the binding energy and to determine Ef . 3. Computational methods The spin polarized electronic structure and density of states calculations were performed using density functional theory within the generalized gradient approximation as implemented in the Vienna ab initio simulation package (VASP) [14]. We have performed GGA + U calculations to account for the correlation effects. The cut-off energy of the plane-wave basis set was 500 eV. The structural parameters that were obtained from X-ray diffraction spectra were adopted as inputs and the crystal structure was relaxed such that the forces on the ions are smaller than 0.02 eV/Å. A 5 × 5 × 5 k-mesh with 39 k-points was used for structural optimization and a 6 × 4 × 6 k-mesh with 72 k-points was used to calculate the DOS. The electronic states that were used in the calculations were 3d and 4s for Mn, 2s and 2p for O atoms, 6s, 6p and 6d for Nd, 5s 5p and 4d for Y, respectively. The Mn 3d on-site Coulomb and exchange energies were taken to be Udd = 5 eV and Jex = 1 eV, respectively for both compositions. A-type AFM order of the Mnmoments was considered and the Nd 4f was treated as core state and non-magnetic for both compositions. The Mn O Mn bond angles obtained from the structural relaxation for x = 0 and 0.5 are 147.34◦ and 142.64◦ , respectively, which are close to the experimental values. Fig. 1. Powder diffraction pattern (experimental and calculated) of Nd1−x Yx MnO3 for (a) x = 0.3 and (b) x = 0.5. Iobs and Icalc represent the observed and calculated intensities, respectively, and Idiff is the difference between them.

4. Results and discussion 4.1. Crystal structure

L3,2 -edge have been performed to determine valency and in addition, the effects of crystal field, hybridization and charge-transfer effects [6,7]. The O K-edge XANES and valence-band photoemission (VB-PES) spectroscopy have been used to probe the lowest unoccupied and highest occupied states near Fermi energy (EF ) and compared with the theoretical density of states (DOS) that are obtained from first-principles or local cluster calculations [10–12]. Thus to probe the local electronic structures of Nd1−x Yx MnO3 , the XANES at the O K-, Mn L3,2 -edge and VB-PES measurements were performed. Theoretical band structures were calculated using GGA + U method for x = 0 and 0.5 compositions to provide a complementary overview of the effects of Y doping on the DOS.

2. Experimental details Polycrystalline samples of Nd1−x Yx MnO3 were synthesized by the solid state reaction method for x = 0, 0.1, 0.2, 0.3, 0.4 and 0.5. The phase purity was verified by powder X-ray diffraction. Rietveld analysis was conducted using the Fullprof program to determine the lattice parameters along with Mn O and Mn Mn bond lengths [13]. The magnetic measurements were made using a physical property measurement system. XANES and VB-PES were performed at room temperature, which is well above the magnetic transition temperatures. All of the synchrotron-based experiments were carried out at the National Synchrotron Radiation Research Center in Hsinchu, Taiwan. The O K- and Mn L3,2 -edge XANES were obtained at the Dragon 11A-beamline using fluorescence and total electron yield modes, respectively. The energies in the spectra were calibrated using those of reference NiO. VB-PES measurements were made using a hemispherical energy analyzer at the lowenergy spherical monochromator 08A-beamline at a base pressure ∼5 × 10−10 Torr. The sample surface was scrapped with diamond file to get a clean surface. The spectra were obtained using fixed

The crystal structure of Nd1−x Yx MnO3 is orthorhombic (Pbnm) for x = 0–0.5. Fig. 1 confirms the phase formation and structural refinement for two selected compositions, x = 0.3 and 0.5. The atomic positions were refined by using the initial values obtained by Landsgesell et al. [4]. Fig. 2 presents the crystal structure of Nd1−x Yx MnO3 . Each Mn ion (blue) is surrounded by oxygen (red) octahedra that share corners and are arranged in a zig-zag fashion along the crystallographic a, b and c axes. Due to the highly distorted nature of the octahedral there are three distinct Mn O bond lengths denoted as Mn Ol , Mn Om and Mn Os , corresponding to long, medium and short bonds respectively, as presented in Table 1. The Mn O bond lengths undergo a slight but non-uniform variation upon doping. Additionally, two Mn O Mn bond angles and Mn Mn bond lengths, corresponding to Mn(1) and Mn(2), as shown in Fig. 2 are identified. Table 1 presents the values of the lattice parameters, the Mn O and Mn Mn bond lengths and the Mn O Mn bond angles for x = 0–0.5. The Mn O Mn bond angles show a large deviation from 180◦ , which is the ideal value for the undistorted cubic perovskite structure. With doping, the bond angle shows a systematic reduction. This also explains why the Néel temperature (TN ) (due to the ordering of the Mn ions) reduces from TN = 80 K for x = 0–51 K for x = 0.5. The magnetic behavior of all compositions are confirmed in good agreement with those by Landsgesell et al. [4]. 4.2. Mn L3,2 -edge Fig. 3 presents the Mn L3,2 -edge XANES spectra of Nd1−x Yx MnO3 . The area of the region above 660 eV was normalized to unity for all compositions considered to enable comparison of these spectra on the same footing. The spectrum includes two whitelines, L3 (∼642 eV) and L2 (∼653 eV), associated with Mn 2p3/2 and 2p1/2 → 3d transitions. The inset in Fig. 3 displays the spectra of x = 0 and 0.5 compositions along with that of Mn2 O3 (Mn3+ )

P. Balasubramanian et al. / Journal of Electron Spectroscopy and Related Phenomena 189 (2013) 51–55

53

Fig. 3. Normalized Mn L3,2 -edge spectra of Nd1−x Yx MnO3 . The inset shows L3 -edge near-edge spectra for x = 0 and 0.5 along with references Mn2 O3 and MnO2 . The dashed line in the inset shows that the spectrum resembles that of Mn2 O3 . Fig. 2. Crystal structure of Nd1−x Yx MnO3 in the orthorhombic Pbnm space group. The Nd atoms occupy the 4c Wycoff position and are randomly substituted by Y atom in the crystal.

and MnO2 (Mn4+ ) standard compounds. The spectral features of Nd1−x Yx MnO3 are very close to that of Mn2 O3 without any variation in energies/spectral widths upon doping, implying that the valency of Mn remains close to 3+. The increase in white-line intensity with x suggests decrease in occupancy of Mn 3d states with doping. 4.3. O K-edge Fig. 4(a) shows the normalized O K-edge XANES spectra of Nd1−x Yx MnO3 , due to the 1s → 2p dipole transition. The general line shapes are similar to those of RMnO3 (R = Y, Er) and YMn2/3 Me1/3 O3 (Me = Co, Ni and Cu) [15,16]. The spectra exhibit three distinct features. The pre-edge (528–532 eV) corresponds to the O 2p–Mn 3d hybridized states. The broad feature at 532–537 eV is associated with the hybridized O 2p-Nd 5d/Y 4d states. The feature in the region 538–542 eV corresponds to Mn 4s/4p hybrid bands (not fully shown). The edge-jump determined from the first derivative of the spectrum, is presented in the inset in Fig. 4(b). A systematic shift in the absorption edge to higher energy (arrow in inset of Fig. 4b) upon doping is observed with a maximum shift of 0.4 eV between x = 0 and 0.5. In Fig. 4(b) we plot the difference of the doped spectra obtained by subtracting undoped spectrum from each of

Nd1−x Yx MnO3 (x > 0). Even for lowest doping (x = 0.1) there occurs a negative peak whose intensity increases for higher doping. The negative peak implies a reduction in number of unoccupied states above EF . In the case of divalent doped manganites, the difference spectra shows a positive peak indicating formation of hole states in the band-gap region [17]. In Nd1−x Yx MnO3 , since the structural changes are negligible, the changes in the bare crystal field and hybridization strengths (Vpd and Vpd ) determined by Harrison’s formula do not show a systematic variation with doping [18]. However decrease in the number of unoccupied O 2p–Mn 3d hybridized states above Ef suggest changes in the charge-transfer character of the system. Thus depletion of states near Ef from O K-edge spectra along with the increase of the white-line intensity of Mn L3,2 -edge spectra with x, suggest a net decrease of the O 2p → Mn 3d charge transfer effect, implying an increase in the energy of charge transfer (). The local changes in Mn 3d states with doping can be understood from the pre-edge region. The pre-edge region contains two main features labeled A* and B* along with third feature at ∼532 eV labeled as C* as indicated by an arrow. The feature C* is more is more prominent for x = 0.4 and 0.5. For x = 0–0.3, feature C* appears to merge with the Nd 5d/Y 4d band. The strong J–T distortion causes splitting of eg state into eg (1) and eg (2) states, the effect on the t2g states being much weaker. Based on localized molecular orbital method, Toulemonde et al. [19], present a schematic diagram of the energy states in the Mn 3d ion. Based on their diagrams, the feature A* of our O K edge spectra would correspond to eg (2)↑ which is the lowest unoccupied state while features B* and C* would correspond

Table 1 Lattice parameters along with Mn O, Mn Mn bond lengths and Mn O Mn bond angles of Nd1−x Yx MnO3 . Composition

x = 0.0 x = 0.1 x = 0.2 x = 0.3 x = 0.4 x = 0.5

Lattice parameters (Å)

Mn O (Å)

Mn O Mn (degrees)

Mn Mn (Å)

a

b

c

Mn Om

Mn Os

Mn Ol

Mn O Mn (1)

Mn O Mn (2)

Mn Mn (1)

Mn Mn (2)

5.4155 5.3889 5.3704 5.3571 5.3361 5.3272

5.8477 5.7738 5.7602 5.7980 5.7905 5.8122

7.5426 7.5336 7.5171 7.5017 7.4761 7.4542

1.949 1.954 1.952 1.944 1.945 1.944

1.910 1.895 1.892 1.913 1.896 1.902

2.221 2.203 2.200 2.204 2.209 2.214

150.7 149.2 148.6 149.5 147.9 146.9

149.5 148.9 148.4 146.9 147.0 146.4

3.985 3.949 3.938 3.937 3.938 3.942

3.771 3.767 3.759 3.751 3.738 3.727

54

P. Balasubramanian et al. / Journal of Electron Spectroscopy and Related Phenomena 189 (2013) 51–55

(a) Valence band XPS

Au-expt fit-Au

h =130eV

Intensity (arb. units)

Ef

-1.2

-0.8

-0.4

0.0

0.4 x=0 x=0.1 x=0.2 x=0.3 x=0.4 x=0.5

C

(b)

B

A -12

Fig. 4. (a) Normalized absorption spectra at O K-edge of Nd1−x Yx MnO3 for x = 0–0.5. (b) The difference spectra of Nd1−x Yx MnO3 . The inset in (b) presents the first derivative spectra of the main spectra.

In Fig. 5, we shows the normalized VB-PES spectra formed by O 2p–Mn 3d hybrid bands. For incoming photon energy of 130 eV, the O 2p states have a larger photoionization cross-section than the Mn 3d states due to which the valence band reveal a strong O 2p character [20]. Fig. 5(a) presents the portion of VB-PES that corresponds to the rising edge of Nd1−x Yx MnO3 , along with the spectra of Au. The Fermi energy was determined by fitting the spectra of Au to the Fermi-Dirac function. The fitting gives a energy resolution of 100 meV. The spectrum of all the samples were then aligned with respect to EF . The variation of the rise of the valence-band spectra with doping is small and is not systematic unlike the absorption spectra. However we see a net shift of spectra toward higher binding energy. Fig. 5(b) displays the spectra of full valence band region of all compositions. For x = 0 composition the spectrum is similar to that for LaMnO3 [21]. The spectra consists of a main feature C and two shoulder features A and B. With doping, C shifts to higher binding energy. Previous electronic-structure calculations by Satpathy et al. [22] and configuration interaction cluster calculations by Kurata et al. [23] demonstrate that feature A corresponds to eg (1)↑, while feature B corresponds to t2g ↑. The main feature C is a combination of t2g ↑ and eg (1)↑, all of them strongly hybridized with the O 2p bands. To understand the net effect of doping, we show combined O K-edge XANES and VB-PES spectra of x = 0 and 0.5 in Fig. 6(a). From the combined XAS + XPS spectra we observe that between x = 0 and 0.5, there is a net shift of the both the spectra away from EF . This suggests that there is a net increase in the band

-6

-4

-2

0

Fig. 5. (a) Expansion of spectrum of Nd1−x Yx MnO3 in the region at/below Ef and spectrum of Au metal as reference. (b) VB-PES spectra of Nd1−x Yx MnO3 .

gap energy with doping, indicating an increase in the insulating character of Nd1−x Yx MnO3 . 4.5. Density of states In this section, we directly correlate the XPS and XAS spectra with the occupied and unoccupied density of states, respectively. Fig. 6(b) shows the total DOS of while in Fig. 6(c and d) we show 1.5 (a) 1.0

VB - PES C

XAS

* A* B

B

C*

A

0.5

x= 0 x= 0.5

0.0

x= 0 x= 0.5

12 (b) 6 0

0

t2g

x3

(c) x = 0

O 2p

-5

eg

0

x3

(d) x = 0.5 -6

-3

0

Ef

Partial DOS

4.4. Valence band photoemission

-8

Binding energy(eV)

Total DOS Intensity

to t2g ↓ and eg (1)↓ levels, respectively. The eg (2)↓ occurs at a much higher energy, overlapping the Nd 5d/Y 4d bands. With increase in doping, the shift in the rise of O K-edge spectrum to higher energy and the increase in energy separation between B* and C* suggest that there occurs an increase in the J–T splitting. This is also concurrent with the greater structural distortion arising from the substation of Nd with smaller cation Y.

-10

-3 3

6

Fig. 6. (a) Combined plots of O K-edge XANES and VB-PES of Nd1−x Yx MnO3 for x = 0 and 0.5. (b) Total DOS calculated for x = 0 and 0.5. (c and d) Partial DOS from the t2g and eg bands of Mn 3d and O 2p bands for x = 0 and 0.5, respectively with spin up and down. The O 2p DOS is multiplied by a factor of 3 for clearer viewing.

P. Balasubramanian et al. / Journal of Electron Spectroscopy and Related Phenomena 189 (2013) 51–55

spin polarized partial DOS containing the Mn t2g and eg bands along with O 2p band for x = 0 and 0.5, respectively. The t2g bands are sharp with higher intensity indicating their localized nature. The eg bands are more spread in energy because of greater hybridization with O 2p states. The eg band contains half occupied eg ↑ and unoccupied eg ↓ states. For x = 0.5, the t2g has smaller intensity, but broader in energy due to increased hybridization with the O 2p states. The occupied DOS for x = 0.5 moves slightly to higher binding energies relative to those of the x = 0 in agreement with the VB-PES (Fig. 6(a)). For x = 0, the highest occupied states are the O 2p states indicating the charge transfer insulating nature of NdMnO3 . For x = 0.5, the highest occupied states below Ef correspond to Mn 3d states thus suggesting a decrease in charge transfer nature. Similarly, above Ef we see a shift in the unoccupied states as well. For x = 0, the lowest unoccupied states corresponds to the eg band with no overlap with the O 2p bands. For x = 0.5 we observe strongly hybridized Mn eg -O 2p band at ∼2 eV. Our first-principles calculations thus almost reproduce the trends in experimental valence band. However, the theoretical shift in DOS (<0.5 eV) between x = 0 and 0.5 is smaller than the experimentally observed shift. Thus from our present calculations, the change in width of forbidden region is negligible. A possible reason could be due to the identical values of Udd and Jex used for calculations for both compositions. Since Y-doping causes decrease of lattice parameters, the band gap would tend to reduce rather than increase, due to larger overlap between the atomic orbitals. From constrained LDA calculations on RMnO3 , however it is found that the variation in Udd is very small and non-systematic as we vary the rare-earth ion [24]. Thus in Nd1−x Yx MnO3 , one can expect a small change in Udd and Jex values with x justifying the use of identical values for both compositions. However our first-principles calculations do not take into account the effects of charge transfer. Since manganites fall in the intermediate region between Mott–Hubbard and charge-transfer insulators in the Zannen–Allen–Sawatzky diagram [25], Udd is comparable with the charge transfer energy (). Thus in Nd1−x Yx MnO3 , we have the competing effects of Udd ,  along with increase in structural and J–T distortions with x. However the structural distortions can also lead to a decrease in band gap in certain rare-earth manganites [24]. In the well studied LaMnO3 the Mn 3d shell of has an effective configuration of 50% d4 + 40% d5 L + 10% d6 L2 (where L corresponds to the ligand hole) in the ground state which shows the large charge transfer nature of the system [25]. One can expect very similar configuration in case of NdMnO3 . Theoretically, we find that Nd1−x Yx MnO3 has 4.8 electrons in the charge sphere of the Mn 3d orbital for x = 0. This reduces to 4.7 electrons for x = 0.5. Though a small decrease, it indirectly hints toward a net decrease in the charge transfer character of Nd1−x Yx MnO3 with increasing x. Hence the increment in band gap observed from our experiments can be attributed to increase in charge transfer energy. Also our calculations suggest that in Nd1−x Yx MnO3 , the insulating nature changes from a charge transfer to Mott-like behavior. Similar behavior has been clearly observed in MF2 compounds (M: transition metal) by varying M from Ti to Cu, using X-ray emission spectroscopy [26]. However due to the greater covalency and hybridization in manganites such a direct transition is hard to observe. 5. Summary This study elucidates the effect of non-magnetic trivalent doping of Y into NdMnO3 on its local electronic structures. The changes

55

in the XANES features in terms of both energy and intensity are much smaller than those associated with the divalent doping of manganites. The Mn L3,2 -edge XANES spectra show a net reduction in 3d electron occupancy and do not exhibit any energy shift. The O K-edge XANES spectra reveal depletion of unoccupied states above Ef upon doping. The splitting of the pre-edge peaks suggests that the J–T distortion increases with doping. The main-feature in the VB-PES moves to higher binding energies upon doping. The band-gap of the system increases with doping. The calculated DOS demonstrate that for x = 0.5, doping increases the hybridization between the occupied Mn 3d and the O 2p orbitals. The combined XANES and VB-PES spectra along with our band structure calculations reflect that Y-doping effectively increases charge transfer energy which causes the increase in the band gap. Acknowledgement The author (WFP) is grateful to the National Science Council of Taiwan for financial support this research under Grant No. NSC 992119-M-032-004-MY3. References [1] Y. Tokura, Colossal Magnetoresistive Oxides, Gordon and Breach, New York, 2000. [2] J.A. Alonso, M.J. Martnez-Lopez, M.T. Casais, M.T. Fernandez-Daz, Inorg. Chem. 39 (2000) 917. [3] T. Kimura, G. Lawes, T. Goto, Y. Tokura, A.P. Raimrez, Phys. Rev. B 71 (2005) 224425. [4] S. Landsgesell, A. Maljuk, T.C. Hansen, O. Prokhnenko, N. Aliouane, D.N. Argyriou, Phys. Rev. B 80 (2009) 014412. [5] S. Landsgesell, K. Prokes, B. Ouladdiaf, B. Klemke, O. Prokhnenko, B. Hepp, K. Kiefer, D.N. Argyriou, Phys. Rev. B 86 (2012) 054429. [6] M. Abbate, F.M.F. deGroot, J.C. Fuggle, A. Fujimori, O. Strebel, F. Lopez, M. Domke, G. Kaindl, G.A. Sawatzky, M. Takano, Y. Takeda, Phys. Rev. 46B (1992) 4511. [7] T. Saitoh, A.E. Bocquet, T. Mizokawa, H. Namatame, A. Fujimori, M. Abbate, Y. Takeda, M. Takano, Phys. Rev. B 51 (1995) 13942. [8] O. Wessely, P. Roy, D. Aberg, C. Andersson, S. Edvardsson, O. Karis, B. Sanyal, P. Svedlindh, M.I. Katsnelson, R. Gunnarsson, D. Arvanitis, O. Bengone, O. Eriksson, Phys. Rev. B 68 (2003) 235109. [9] J.R. Hayes, A.P. Grosvenor, J. Phys.: Condens. Matter 23 (2011) 465502. [10] M.K. Dalai, P. Pal, B.R. Sekhar, M. Merz, P. Nagel, S. Schuppler, C. Martin, Phys. Rev. B 85 (2012) 155128. [11] J.-H Park, C.T. Chen, S.-W Cheong, W. Bao, G. Meigs, V. Chakarian, Y.U. Idzerda, Phys. Rev. Lett. 76 (1996) 4215. [12] J.-S Kang, S.W. Han, J.-G Park, S.C. Wi, S.S. Lee, G. Kim, H.J. Song, H.J. Shin, W. Jo, B.I. Min, Phys. Rev. B 71 (2005) 092405. [13] R. Carvajal, Physica B 192 (1993) 55. [14] G. Kresse, J. Furthmuller, Phys. Rev. B 54 (1996) 11169. [15] K. Asokan, J.C. Jan, K.V.R. Rao, J.W. Chiou, H.M. Tsai, S. Mookerjee, W.F. Pong, M.H. Tsai, R. Kumar, S. Husain, J.P. Srivastava, J. Phys.: Condens. Matter 16 (2004) 3791. [16] K. Asokan, Y.S. Chen, C.W. Pao, H.M. Tsai, C.W.O. Lee, C.H. Lin, H.C. Hsueh, D.C. Ling, W.F. Pong, J.W. Chiou, M.-H. Tsai, O. Pena, C. Moure, Appl. Phys. Lett. 95 (2009) 131901. [17] D.D. Sarma, O. Rader, T.T. Kachel, A.A. Chainani, M.M. Mathew, K.K. Holldack, W. Gudat, W. Eberhard, Phys. Rev. B 49 (1994) 14238. [18] N. Hollmann, Z. Hu, T. Willers, L. Bohat, P. Becker, A. Tanaka, H.H. Hsieh, H.-J Lin, C.T. Chen, L.H. Tjeng, Phys. Rev. B 82 (2010) 184429. [19] O. Toulemonde, F. Millange, F. Studer, B. Raveau, J.H. Park, C.T. Chen, J. Phys.: Condens. Matter 11 (1999) 109. [20] P. Pal, M.K. Dalai, B.R. Sekhar, S.N. Jha, S.V.N. Bhaskara Rao, N.C. Das, C. Martin, F. Studer, J. Phys.: Condens. Matter 17 (2005) 2293. [21] T. Sayito, A. Sekiyama, K. Kobayashi, T. Mizokawa, A. Fujimori, D.D. Sarma, Y. Takeda, M. Takano, Phys. Rev. B 56 (1997) 8836. [22] S. Satpathy, Zoran S. Popovic’, Filip R. Vukajlovic, Phys. Rev. Lett. 76 (1996) 960. [23] H. Kurata, C. Colliex, Phys. Rev. B 48 (1993) 2101. [24] I. Solovyev, J. Phys. Soc. Jpn. 78 (5) (2009) 054710. [25] J. Zannen, G.A. Sawatzky, J.W. Allen, Phys. Rev. Lett. 55 (1985) 418. [26] P. Olalde-Velasco, J. Jimenez-Mier, J.D. Denlinger, Z. Hussain, W.L. Yang, Phys. Rev. B 83 (2011), 241102(R).