betaFeedback

WorksPhysics and Astronomy

The 3d and 5d electronic structures and orbital hybridization in Ba- and Ca-doped La2CoIrO6 double perovskite

Abstract

Here we present a detailed investigation of the Co and Ir local electronic structures in La1.5A0.5CoIrO6 (A = Ba, Ca) compounds in order to unravel the orbital hybridization mechanism in these CoIr-based double perovskites. Our results of x-ray powder diffraction, ac and dc magnetization, Co and Ir L2,3-edges and Co K-edge x-ray absorption spectroscopy and x-ray magnetic circular dichroism suggest a competition between magnetic interactions. A dominant antiferromagnetic coupling is found to be responsible for the ferrimagnetic behavior observed for A = Ca below approximately 96 K, the competing magnetic phases and the cationic disorder in this compound giving rise to a spin-glass state at low temperatures. For the A = Ba, on the other hand, there is no evidence of long range order down to its spin-glass transition temperature. The remarkably different magnetic properties observed between these two compounds is discussed in terms of the structural distortion that alters the strength of the Co - Ir couplings, with a relevant role played by the Co 3d eg - Ir 5d j = 1/2 hybridization.

Introduction

Since the discovery of ferrimagnetism (FIM) and half-metallic behavior above room temperature (TT) in Sr2_2FeMoO6_6 [1], double perovskites (DP) with various compositions have been produced and characterized in view of tuning the electronic and magnetic properties for spintronic applications. The most interesting properties are observed when the systems contain a combination of 3d3d and 4d/5d4d/5d transition-metal (TM) ions [2, 3]. Ir-based DPs have recently received considerable attention due to the peculiar interplay between the strong spin-orbit coupling (SOC), the Coulomb repulsion, and the crystal field splitting which, in octahedral coordination, lifts the Ir t2gt_{2g} orbital degeneracy into a fourfold jeff=3/2j_{\mathrm{eff}} = 3/2 and a jeff=1/2j_{\mathrm{eff}} = 1/2 doublet, giving rise to unusual magnetic properties [4, 5].

The use of Co as 3d3d TM in addition to 5d5d TM adds further complexity to the stabilization of magnetism in these Ir-based DPs. A delicate balance between crystal field and interatomic exchange interactions can lead to different valences and spin states of the Co ions in the octahedral symmetry [6]. In particular, La2_2CoIrO6_6 shows exciting properties such as magnetodielectric effect [7] and cooperative octahedral breathing distortion [8]. There is, however, some controversy concerning the microscopic mechanisms governing its magnetic properties.

N. Narayanan et al., from the analysis of neutron powder diffraction (NPD) data, ascribed the small ferromagnetic (FM)-like behavior in La2_2CoIrO6_6 to spin canting in the two interpenetrating antiferromagnetic (AFM) Co and Ir sublattices [9]. However, the strong neutron absorption of Ir ions prevents a reliable determination of its magnetic structure. On the other hand, recent studies using x-ray absorption spectroscopy (XAS), x-ray magnetic circular dichroism (XMCD), and muon spin spectroscopy suggest an AFM coupling between Co and Ir FM sublattices, resulting in a ferrimagnetic (FIM) behavior [10, 11]. The small magnetization observed in the system can be ascribed to the antisite disorder (ASD) at Co/Ir sites and the competing magnetic interactions that lead to frustration. There are also other interpretations. Min-Cheol Lee et al., using Co and O K-edge XAS, claim that the spin frustration and extended paramagnetic (PM) phase are due to the non spin-selective nature of the Co2+^{2+} (3d3d eg⊕4pe_g \oplus4p) – Ir4+^{4+} 5d5d ege_g coupling [12]. At the same time, recent Ir L3L_3-edge resonant inelastic x-ray scattering show the importance of the 3d3d-5d5d hybridization in the magnetism of Ir-based DPs [13]. Additionally, x-ray photoelectron spectroscopy measurements point to a large covalence in several DP iridates, questionning the description of these materials by a simple ionic picture [14].

The observations reported above indicate how far we are from reaching a consensus on the nature of the magnetic ordering in Ir-based DPs. This is also valid when La is partially replaced by alkaline-earth in La2_2CoIrO6_6. For La2−x_{2-x}Srx_xCoIrO6_6, a structural transition accompanied by changes in the Co/Ir valences and magnetic structures is observed even for small Sr-doping [9, 15]. For La2−x_{2-x}Cax_xCoIrO6_6, on the other hand, there is no structural transition, and up to 25% of Ca to La, the substitution acts mainly at changing the Co valence, with Ir remaining nearly tetravalent [11]. The x=0.5x = 0.5 concentration is of particular interest due to its compensation temperatures and spontaneous exchange bias effect [16]. which seems to be directly related to a re-entrant spin glass (RSG) phase [17, 18]. Likewise, the Sr to La partial substitution also induces competition between magnetic phases that leads to RSG [19], whereas the La2_2CoIrO6_6 parent compound does not show glassy magnetic behavior.

In order to shed light on the role of the Co 3d3d – Ir 5d5d orbital hybridization in stabilizing the magnetic coupling in CoIr-based DPs, we have performed a detailed investigation of how the crystal structure affects the electronic and magnetic structures of Co and Ir sublattices in La1.5_{1.5}(Ba,Ca)0.5_{0.5}CoIrO6_6 compounds using x-ray powder diffraction (XRD), ac and dc magnetization, XAS and XMCD. Our results show remarkably different magnetic properties for these compounds, that are discussed in terms of the lattice distortions which ones alter the orbital hybridization between Co and Ir.

Experimental details

The polycrystalline samples here investigated were synthesized by conventional solid-state reaction. To obtain La1.5_{1.5}Ba0.5_{0.5}CoIrO6_6 (hereafter called Ba0.5), stoichiometric amounts of La2_2O3_3, BaCO3_3, Co3_3O4_4 and metallic Ir in powder form were mixed and heated at 650°C for 24 hours in air atmosphere. Later, the sample was regrinded before a second step at 800°C for 48 hours. Finally, the material was grinded, pressed into a pellet, and heated at 1200°C for additional 48 hours. The synthesis route used to produce La1.5_{1.5}Ca0.5_{0.5}CoIrO6_6 (herein called Ca0.5) is described elsewhere [11].

High-resolution XRD data were collected at room temperature using a Cu Kα_\alpha radiation operating at 40 kV and 40 mA at Centro Brasileiro de Pesquisas Físicas (CBPF), Brazil. The XRD data was carried over the angular range of 10∘≤2θ≤110∘10^\circ\le2\theta\le110^\circ, with a 2θ2\theta step size of 0.01∘0.01^\circ. Rietveld refinement was performed using GSAS software and its graphical interface program [20]. The ac and dc magnetic measurements were carried out using a Quantum Design PPMS magnetometer, at CBPF.

Low-temperature XAS measurements were carried out at the Co K- and Ir L2,3_{2,3}-absorption edges on beamline P09 at PETRA III at the DESY [21] and at Co L2,3_{2,3}-absorption edges on beamline I06 at Diamond Light Source (DLS). For the hard x-ray measurements, fine powder samples were mixed and pressed into a low-Z BN material to produce pellets for transmission measurements. XMCD measurements were performed at 5 K by fast-switching the beam helicity between left and right circular polarization [22]. In order to align the magnetic domains and correct for nonmagnetic artifacts, an external magnetic field of ±5\pm5 T was applied parallel and antiparallel to the incident beam wave vector kk using a 6T/2T/2T vector magnet. For the soft x-ray measurements, powder samples from the same batches were spaded on conductive carbon tapes and measured in total electron yield (TEY). The samples were cooled at 2 K, and an external magnetic field of ±5\pm5 T was applied along the beam axis to reach saturation.

Results and discussion

x-ray diffraction

Figure 1 shows the XRD pattern of Ba0.5, which confirms the formation of a perovskite structure with no trace of impurity phase. Previous studies on La2_2CoIrO6_6 and LaBaCoIrO6_6 show that the former compound grows in the monoclinic P21/nP2_1/n space group whereas the latter was reported as belonging to the triclinic Iˉ1\bar{I}1 non-standard space group [9, 11, 23]. Such non-standard setting is commonly adopted to facilitate the comparison of this structure with that of other DPs [3, 24]. In our case, a detailed investigation of the XRD data of the Ba0.5 sample revealed that some of the experimentally observed reflections are not predicted for the triclinic symmetry. In contrast, all peaks match with that of the monoclinic structure. Therefore, the 25% of Ba substitution at the La site is not sufficient to induce a structural transition, with Ba0.5 belonging to the P21/nP2_1/n space group, the same as that of Ca0.5 [11]. The main results obtained from the Rietveld refinement are displayed in Table I, where the structural parameters of Ca0.5 (some of them already reported in Ref. 11) are also shown for comparison.

Rietveld refinement fitting of Ba0.5 XRD

Figure 1. Rietveld refinement fitting of Ba0.5 XRD. The vertical lines represent the Bragg reflections for the P21/nP2_1/n space group. Inset shows the crystal structure, in which IrO6_6 and CoO6_6 are drawn as blue and green octahedra, respectively.

From Table I we observe an increase in the unit cell volume (VV) of Ba0.5 as compared to Ca0.5, in agreement with the Ba2+^{2+} ionic radius being larger than Ca2+^{2+} (respectively 1.61 and 1.34 Å in XII coordination [25]). Interestingly, although both compounds belong to the P21/nP2_1/n monoclinic space group, there is a tendency towards a more tetragonal symmetry for Ba0.5, which is manifested by the closeness of the aa and bb lattice paparameters and a significant increase in the cc parameter. The increased crystallographic symmetry is expected for larger rare-earth/alkaline-earth ionic radii in DPs [2, 3], manifested in the increased Co–O–Ir bond angles and the more similar Co–O and Ir–O bond lengths for Ba0.5. It must also be noticed in Table I that the Co/Ir ASD, i.e. the amount of Co(Ir) ions lying at Ir(Co) site, is rather similar for both compounds. Though this may look unexpected at a first glance, since an increased crystallographic symmetry usually leads to larger ASD, we recall that Ca0.5 was synthesized at 1000°C [11] while Ba0.5 was produced at 1200°C. The synthesis condition was already shown to play a role on the B/B’ cationic ordering of DPs, which commonly increases with increasing the synthesis temperature [26].

ac and dc magnetization

Figure 2(a) shows the zero-field cooled (ZFC) and field cooled (FC) dc magnetic susceptibility (χdc\chi_{\mathrm{dc}}) data for Ba0.5, measured with a magnetic field (HH) of 0.05 T. For comparison, Figure 2(b) displays χdc\chi_{\mathrm{dc}} curves for Ca0.5 also measured with H=0.05H = 0.05 T, as reported in Ref. 11. For Ca0.5, the higher-TT FM-like magnetic transition at T∼96T \sim96 K is ascribed to Co2+^{2+}–O–Ir4+^{4+} AFM coupling that results in FIM. At the same time, the peak at T∼27T \sim27 K is related to the emergence of a spin glass (SG) phase caused by competing magnetic interactions and ASD [11].

ZFC-FC susceptibility, ac susceptibility, and magnetization curves for Ba0.5 and Ca0.5

Figure 2. (a) ZFC-FC χdc\chi_{dc} curves for Ba0.5, measured with H=0.05H = 0.05 T. The arrow indicates the temperature where the curves bifurcate, and the inset shows the best fit of PM region with the CW law. (b) ZFC-FC χdc\chi_{dc} curves measured with H=0.05H = 0.05 T for Ca0.5, adapted from Ref. [11]. Inset shows the χac\chi_{ac} curve measured with Hac=10H_{ac} = 10 Oe, f=100f = 100 Hz. (c) χac\chi_{ac} as a function of TT for Ba0.5 at different ff, measured with Hac=10H_{ac} = 10 Oe. Inset shows TfT_f as a function of ff, where the solid line represents the best fit to the data using Eq. (1). (d) M(H)M(H) curves for Ba0.5 and Ca0.5 samples, measured at 5 K.

For Ba0.5, although there is only one clear peak at T∼24T \sim24 K in the χdc\chi_{\mathrm{dc}} data, a bifurcation of the ZFC and FC curves at T∼85T \sim85 K is observed, at which temperature a deviation of the Curie-Weiss law is also observed in the inverse susceptibility [inset of Figure 2(a)]. This could indicate that a non-percolating short-range magnetic coupling starts to develop around this temperature, which could be in turn related to the appearance of a spin-glass phase at much lower temperatures. The emergence of short-range correlations and fluctuations far above the freezing temperature is commonly observed in spin-glasses [27].

The real part of the ac magnetic susceptibility (χac\chi_{\mathrm{ac}}) curves measured for Ba0.5 in an oscillating field Hac=10H_{\mathrm{ac}} = 10 Oe with six ac frequencies (ff) ranging from 30 to 10000 Hz is shown on Figure 2(c). In contrast to the ac magnetic susceptibility data for Ca0.5, where a clear cusp is observed at TCT_C [inset of Figure 2(d)], the χac\chi_{\mathrm{ac}} curves of Ba0.5 gives no evidence of any anomaly close to 85 K, further indicating that the weak magnetic coupling between Co and Ir does not percolate on this compound. At lower temperatures, it is observed a systematic decrease in the amplitude of the T∼24T \sim24 K peak and its shift to higher temperatures as ff increases. These are characteristic features of glassy magnetism, signaling that this peak corresponds to the freezing temperature (TfT_f) of a SG-like phase [28]. The evolution of TfT_f with ff turns out to be well fitted by the power-law equation of the dynamic scaling theory, commonly used to investigate SG-like systems

ττ0=[Tf−TgTg]−zν,(1)\frac{\tau}{\tau_0} = \left[\frac{T_f - T_g}{T_g}\right]^{-z\nu}, \tag*{(1)}

where τ\tau is the relaxation time corresponding to the measured frequency, τ0\tau_0 is the characteristic relaxation time of spin flip, TgT_g is the SG transition temperature (TfT_f as ff tends to zero), zz is the dynamical critical exponent and ν\nu is the critical exponent of the correlation length [28, 29]. The solid line on inset of Figure 2(c) represents the fit to (1), yielding Tg≃24.4T_g \simeq24.4 K, τ0≃6.3×10−12\tau_0 \simeq6.3 \times10^{-12} s and zν≃5.9z\nu\simeq5.9, being these values typical of conventional SG systems [16, 29, 30].

Usually, the relative shift δTf=ΔTf/Tf(Δln⁡f)\delta T_f = \Delta T_f/T_f(\Delta\ln f) is calculated to classify the material as canonical SG, cluster SG or superparamagnet. For Ba0.5 we find δTf≃0.006\delta T_f \simeq0.006, which is within the range usually found for canonical SG (δTf<0.01\delta T_f < 0.01 [30, 31, 32]). This confirms that, as for Ca0.5, the ASD and competing magnetic interactions also lead to the emergence of a SG phase in Ba0.5.

The fit of the PM region of 1/χdc1/\chi_{\mathrm{dc}} raw data with the Curie-Weiss (CW) law, inset of Figure 2(a), yields a CW temperature θCW=−58\theta_{\mathrm{CW}} = -58 K for Ba0.5. The negative sign indicates that AFM couplings dominate the magnetic interactions in this compound, as observed for Ca0.5 [11]. The μeff=5.3 μB\mu_{\mathrm{eff}} = 5.3\ \mu_B/f.u. obtained from the fit is fairly close to the value found for Ca0.5 (5.4 μB\mu_B/f.u.) [11]. It is important to mention that the situation here is quite different from that of some Ir5+^{5+}-based compounds where a χ0\chi_0 term is usually included in the CW equation to account for a diamagnetic contribution to the susceptibility [5, 33, 34, 35]. In our Ir4+^{4+}/Co2+,3+^{2+,3+} system, the van-Vleck constant term and the core diamagnetic contribution are expected to be negligible regarding the paramagnetic moments of the TM ions.

Figure 2(d) displays the magnetization as a function of HH [M(H)] curves measured at 5 K for Ba0.5 and Ca0.5 samples. Both curves present typical of FIM systems, with a FM-like hysteresis but a linear increase of magnetization at high HH, further indicating the presence of AFM interactions. It is interesting to notice the remarkable decrease in the coercive field (HCH_C) of Ba0.5 with respect to Ca0.5. This resembles other 3d-5d DPs for which HCH_C dramatically decreases with the expansion of the unit cell [36, 37, 38, 39].

XAS and XMCD measurements

XAS spectra at the TM L2,3L_{2,3}-edges are very sensitive to the valence state, corresponding to electron transitions from the 2p2p to 3d3d (5d5d) orbitals for ions of the first (third) row. Figure 3(a) shows the Co L2,3L_{2,3}-edge spectra for the samples of interest together with those of CoO and LaCoO3_3, used here as standards for Co2+^{2+} and Co3+^{3+}, respectively. The presence of Co2+^{2+} in Ba0.5 and Ca0.5 samples is confirmed by the similarity of the spectra with that of CoO. However, for Ba0.5 it is observed the presence of pronounced peaks at ∼\sim785 and ∼\sim800 eV, associated with Ba M4,5M_{4,5} absorption edges. A closer inspection of the higher energy side of the L3L_3-edge (∼\sim780 eV) of Ca0.5 and Ba0.5 reveals a relative increase in the intensity of the spectra with respect to that of CoO near the energy position of Co3+^{3+} L3L_3-edge absorption, suggesting that some fraction of Co is in trivalent state. This is in agreement with the effective magnetic moments obtained for both compounds [11].

Co $L_{2,3}$-edge XAS and XMCD spectra of Ca0.5 and Ba0.5 samples

Figure 3. (a) Co L2,3L_{2,3}-edge XAS spectra of Ca0.5 and Ba0.5 samples, together with those of CoO (reference for Co2+^{2+}) and LaCoO3_3 (reference for Co3+^{3+}). The presence of Ba M4,5M_{4,5}-edges in Ba0.5 sample is highlighted by green arrows. (b) and (c) show respectively the XMCD spectra of Ca0.5 and Ba0.5 samples, taken at 2 K with H=5H = 5 T. The photon spin was aligned parallel (μ+\mu^{+}, black) and antiparallel (μ−\mu^{-}, red) to HH, and the difference spectra are shown in blue.

Sampleaa (Å)bb (Å)cc (Å)β\beta (°)VV (Å3^3)RpR_p (%)RwpR_{wp} (%)
Ca0.55.5566(1)5.6427(2)7.8808(2)89.98(1)247.09(2)8.511.2
Ba0.55.6326(2)5.6152(1)7.9402(3)90.02(1)251.13(1)4.05.6
SampleCo-O1 (Å)Co-O2 (Å)Co-O3 (Å)⟨Co-O⟩ (Å)Ir-O1 (Å)Ir-O2 (Å)Ir-O3 (Å)⟨Ir-O⟩ (Å)Co-O1-Ir (°)Co-O2-Ir (°)Co-O3-Ir (°)⟨Co-O-Ir⟩ (°)
Ca0.52.062(13)2.127(13)2.072(15)2.087(8)1.980(13)2.033(13)2.076(15)2.030(8)156.7(8)144.4(7)143.6(5)148.2(4)
Ba0.52.023(1)2.060(1)1.996(1)2.023(1)2.012(1)2.043(1)2.071(1)2.042(1)162.2(1)151.4(1)154.8(1)156.1(1)
SampleatomWyckoff positionoccupancyxxyyzzUisoU_{\mathrm{iso}} (Å2^2)
Ca0.5La/Ca4e0.75/0.250.5103(3)0.5437(5)0.2503(3)0.0105(5)
Ca0.5Co1/Ir12c0.927/0.073(3)00.500.0121(3)
Ca0.5Ir2/Co22d0.927/0.073(3)0.5000.0121(3)
Ca0.5O14e10.2188(21)0.2093(23)0.9629(27)0.0045(24)
Ca0.5O24e10.3064(21)0.7076(24)0.9363(23)0.0045(24)
Ca0.5O34e10.3842(22)0.9867(18)0.2503(20)0.0045(24)
Ba0.5La/Ba4e0.75/0.250.5140(2)0.5271(1)0.2474(5)0.0104(2)
Ba0.5Co1/Ir12c0.936/0.064(2)00.500.0129(2)
Ba0.5Ir2/Co22d0.936/0.064(2)0.5000.0129(2)
Ba0.5O14e10.2475(26)0.2474(25)0.9608(29)0.0072(12)
Ba0.5O24e10.3039(30)0.6980(31)0.9645(23)0.0072(12)
Ba0.5O34e10.4214(28)0.9996(11)0.2548(29)0.0072(12)

Table 1. Main results obtained from the Rietveld refinements of the room temperature XRD data. Some of the parameters displayed for Ca0.5 were already reported in Ref. [11].

A quantitative estimate of the Co oxidation state can be made for our Ca0.5 sample by comparing the center of mass of its L3L_3-edge spectrum with those of the reference samples. Assuming a linear variation in energy between the center of mass of Co2+^{2+} and Co3+^{3+}, the Co formal valence was estimated to be ∼\sim2.4 for Ca0.5, which would mean the presence of ∼\sim10% of Ir5+^{5+} or oxygen vacancy in this sample. Surely this is just a rough estimate, since other parameters such as the crystal field and structure may play a role on the integrated intensity of XAS curves. Nevertheless, a +2.4 formal valence is reasonably close to the +2.5 value expected for a 25% Ca2+^{2+} to La3+^{3+} substitution, which one would act mainly at changing the Co valence, a scenario corroborated by the magnetometry results and Ir L3L_{3}-edge XAS. For Ba0.5, the presence of the Ba-M5M_5 absorption peak nearby the Co L3L_3-edge curve affects the calculation of the center of mass, preventing a reliable estimate of its valence state. Notwithstanding, the similarity of its L2,3L_{2,3}-edge curves with that of Ca0.5, as well as the closeness of the μeff\mu_{\mathrm{eff}} observed for both samples, suggest that the Co valence for Ba0.5 is similar to that of Ca0.5.

Figure 3(b) and (Figure 3c) show the Co L2,3L_{2,3} XMCD of Ca0.5 and Ba0.5 samples, respectively. The same overall behavior suggests nearly the same magnetic structure for both compounds. The black and red curves stand respectively for μ+\mu^+ and μ−\mu^-, i.e., for parallel and antiparallel alignments between the photon spin and HH, and the difference spectra Δμ=μ+−μ−\Delta\mu= \mu^+ - \mu^- corresponds to the blue line. As can be noticed, the XMCD signal is largely negative at L3L_3-edge but with a very small positive value at L2L_2-edge, indicating non-negligible orbital contribution to the magnetic moment, as commonly observed for Co in octahedral coordination [6]. For a quantitative analysis of the Co-L2,3L_{2,3} XMCD spectra, we have used the sum rules to derive the orbital and spin contributions to the magnetization [41, 42]

ml=−43∫L3+L2(μ+−μ−)dω∫L3+L2(μ++μ−)dωNh,(2)m_l = -\frac{4}{3}\frac{\int_{L_3+L_2}(\mu^{+}-\mu^{-})d\omega}{\int_{L_3+L_2}(\mu^{+}+\mu^{-})d\omega}N_h, \tag*{(2)}
ms=−6∫L3(μ+−μ−)dω−4∫L3+L2(μ+−μ−)dω∫L3+L2(μ++μ−)dωNh(1+7⟨Tz⟩2⟨Sz⟩)−1,(3)m_s = -\frac{6\int_{L_3}(\mu^{+}-\mu^{-})d\omega-4\int_{L_3+L_2}(\mu^{+}-\mu^{-})d\omega}{\int_{L_3+L_2}(\mu^{+}+\mu^{-})d\omega}N_h\left(1+\frac{7\langle T_z\rangle}{2\langle S_z\rangle}\right)^{-1}, \tag*{(3)}

where mlm_l and msm_s are the angular and spin magnetic moments in units of μB\mu_B/atom, SzS_z denotes the projection along zz of the spin magnetic momentum, NhN_h represent the number of empty 3d3d states, TzT_z denotes the magnetic dipole moment, and L2L_2 and L3L_3 represent the integration range. Here, we used Nh=3.5±0.5N_h = 3.5\pm0.5, which is an approximated atomistic value corresponding to a mixture of Co2+^{2+} (Nh=3N_h = 3) and Co3+^{3+} (Nh=4N_h = 4). Moreover, TzT_z is estimated to be negligible against SzS_z for TM ions in octahedral symmetry [43, 44]. With these considerations, we obtained ml=0.15 μBm_l = 0.15\ \mu_B and ms=0.29 μBm_s = 0.29\ \mu_B. Considering sources of deviations such as electronic interactions, imprecision in the integral calculations, and the NhN_h value assumed, we estimate an uncertainty of ∼20%\sim20\% on these values [43, 44].

The calculated moments are similar to those reported for La2_2CoIrO6_6 (ml=0.18 μBm_l = 0.18\ \mu_B, ms=0.31 μBm_s = 0.31\ \mu_B) [10], which is somewhat unexpected considering that from our XAS spectra we can expect about 40% of the Co ions in a trivalent state, as opposed to a presumably 100% of Co2+^{2+} in La2_2CoIrO6_6. However, in Ref. 10 a quantitative investigation of the Co formal valence was not performed for the parent compound, whilst another study performed on a La2_2CoIrO6_6 polycrystalline sample produced by a similar synthesis route reported the presence of some amount of Co3+^{3+} [11]. Furthermore, the previous XMCD study reported for La2_2CoIrO6_6 in Ref. 10 was carried out at 43 K using H=0.8H = 0.8 T, while element-specific hysteresis curves measured for CoIr-DPs have already shown that Co is far from saturation at these conditions [15].

For the Ba0.5 sample, one must recall the pronounced Ba M4,5M_{4,5}-edge peaks observed near the Co L2,3L_{2,3} absorption edge. Although one may not expect a XMCD signal at the Ba M4,5M_{4,5}-edge, the large XAS white line intensity may influence the background of the Co L2,3L_{2,3} pre- and post-edge regions, and affect the adjustment of the XAS curve for the calculation of the L2+L3L_2+L_3 integral. Therefore, mlm_l and msm_s were not estimated for this compound. The ml/msm_l/m_s ratio, however, is independent of the XAS integrated intensity, resulting in a trustworthy value that is fairly close to that found for Ca0.5.

In order to avoid the problem of Ba M4,5M_{4,5}-edges, we investigated the Co KK-edge XAS and XMCD, as shown in Fig. 4. The Co KK-edge XAS reflects mainly the electric dipole transitions from 1s to 4p level, where the edge position in energy may probe the Co valence. At the same time, local arrangements of neighboring ions determine the region after the absorption threshold. The similarity between the curves displayed in Fig. 4(a) is in accordance with the same monoclinic structure and similar Co valence states for both compounds.

The inset of Fig. 4(a) shows a magnified view of the pre-edge region. Although the KK-edge XAS reflects mostly the transitions to unoccupied Co 4p states, M.-C. Lee et al. showed that the pre-edge anomalies α\alpha and β1\beta_1 are related to Co 4p partial electron filling via intra-site mixing with Co 3d orbitals [12]. Thus, the orbital hybridization with Ir 5d orbitals may largely contribute to these pre-edge features (see Figure 6). Interestingly, a third anomaly, β2∼77.0\beta_2 \sim77.0 eV, is observed for Ca0.5 but not Ba0.5, which may be related to Co3+^{3+} (3d eg⊕4p3d\ e_g\oplus4p) – Ir 5d ege_g hybridization via O 2p.

Schematic diagrams of orbital hybridization and spin-selective exchange pathways

Figure 6. (a) Schematic diagram of the mechanisms of hybridization between Co2+/3+^{2+/3+} (3d7/3d63d^7/3d^6) and Ir4+^{4+} (5d55d^5), where the red arrow represents an additional spin down electron for Co2+^{2+} t2gt_{2g}, which is absent for Co3+^{3+}. The green arrows represent virtual electron hopping. The blue double-arrow in Ir jeff=1/2j_{eff} = 1/2 is spin down for the case of AFM interaction and spin up for the FM case. (b) (Co/Ir)O6_6 octahedra geometries for Co–O–Ir bond angle θ=180∘\theta= 180^\circ and θ<180∘\theta< 180^\circ. (c) The Co t2gt_{2g} – Ir jeff=1/2j_{eff} = 1/2 (left) and Co (3d3d eg⊕4pe_g \oplus4p) – Ir jeff=1/2j_{eff} = 1/2 (right) orbital hybridizations.

The Co KK-edge XMCD signals measured at T=5T = 5 K and H=5H = 5 T are very similar for both Ca0.5 and Ba0.5 Fig. 4(b), suggesting that although the Co 3d3d bands present a relative extended character, the changes in the crystallographic environment does not drastically affect the Co magnetism.

SampleCo mlm_l (μB\mu_B)Co msm_s (μB\mu_B)Co mtm_t (μB\mu_B)Co ml/msm_l/m_sIr mlm_l (μB\mu_B)Ir msm_s (μB\mu_B)Ir mtm_t (μB\mu_B)Ir ml/msm_l/m_s
Ca0.50.15±0.030.15 \pm0.030.29±0.060.29 \pm0.060.44±0.070.44 \pm0.070.52±0.110.52 \pm0.11−0.11±0.02-0.11 \pm0.02−0.11±0.02-0.11 \pm0.02−0.22±0.03-0.22 \pm0.031.00±0.211.00 \pm0.21
Ba0.5---0.54±0.150.54 \pm0.15−0.07±0.01-0.07 \pm0.01−0.06±0.01-0.06 \pm0.01−0.13±0.01-0.13 \pm0.011.17±0.271.17 \pm0.27

Table 2. Co and Ir orbital (mlm_l), spin (msm_s) and total (mtm_t) magnetic moments obtained from the sum rules calculations of Co and Ir L3,2L_{3,2} XMCD spectra.

Co $K$-edge XAS spectra

Figure 4a. Co KK-edge XAS spectra of Ba0.5 and Ca0.5 samples, measured at 5 K. Inset shows a magnified view of the pre-edge region, highlighting the α\alpha, β1\beta_1 and β2\beta_2 peaks.

Co $K$-edge XMCD spectra

Figure 4b. The XMCD spectra of these samples.

Ir $L_{2,3}$-edges XAS spectra

Figure 5a. Ir L2,3L_{2,3}-edges: (a) XAS and (b) XMCD spectra for Ba0.5 and Ca0.5 measured at 5 K and under 5 T.

Ir $L_{2,3}$-edges XMCD spectra

Figure 5b. Ir L2,3L_{2,3}-edges: (a) XAS and (b) XMCD spectra for Ba0.5 and Ca0.5 measured at 5 K and under 5 T.

Magnified view of the Ir $L_3$-edge

Figure 5c. Magnified view of the L3L_3-edge, where the IrO2_2 standard is added for comparison.

Element-specific XMCD hysteresis loops

Figure 5e. Element-specific XMCD hysteresis loops of Ir L3L_3-edge.

Similar scenario cannot be extrapolated to the Ir ions, since the 5d5d orbitals are even more extended than the Co 3d3d, which makes Ir ions more sensitive to structural and electronic changes. Fig. 5(a) shows the Ir L2,3L_{2,3}-edge XAS spectra of Ba0.5 and Ca0.5 samples together with that of IrO2_2, reference for Ir4+^{4+} state. The spectra of Ba0.5 and Ca0.5 are reasonably similar to that of IrO2_2, indicating that the majority of Ir ions are in a tetravalent state for both compounds. Nonetheless, the similarity between the curves is characteristic of the L2,3L_{2,3}-edge spectra of

5d TM ions, being related to the more diffuse valence orbitals compared to the 3d ones [15, 45]. A magnified view of the L3L_3-edge region depicted in Fig. 5(c) indicates some subtle differences between the spectra. The Ca0.5 and Ba0.5 spectra are shifted of respectively ∼0.2\sim0.2 eV and ∼0.3\sim0.3 eV with respect to the IrO2_2 L3L_3 XAS, suggesting the presence of a small amount of Ir5+^{5+}. However, these shifts lie within the instrumental resolution (∼1.5\sim1.5 eV), while the difference in energy between the white line positions of Ir4+^{4+} and Ir5+^{5+} is of about 1.2 eV [34]. With these considerations, we can only conclude that Ba0.5_{0.5} and Ca0.5_{0.5} presents nearly the same Ir valence state, with the majority of ions in a tetravalent state.

Fig. 5(b) shows the Ir L2,3L_{2,3}-edge XMCD spectra measured at T=5T = 5 K and H=5H = 5 T. The positive signs of the Ir XMCD signals, combined with the negative sign for the Co signal, indicate that those two ions develop an AFM coupling in both compounds. Table II shows the Ir orbital and spin moments obtained from the sum rules, for which it was assumed Nh=5±0.5N_h = 5 \pm0.5 (Ir4+^{4+}). The mlm_l and msm_s values found are similar to those previously reported for La1.5_{1.5}Sr0.5_{0.5}CoIrO6_6 [15], and the large ml/msm_l/m_s ratio observed is typical of iridates and other 5d5d-based materials presenting strong SOC [46, 47]. The somewhat smaller moments observed for Ba0.5_{0.5} with respect to Ca0.5_{0.5} helps to explain its slightly smaller μeff\mu_{eff} obtained from the χdc\chi_{dc} curves, also suggesting that the Ir magnetization is closely related to its orbital hybridization with neighboring Co ions.

Fig. 5(c) shows element-specific hysteresis loops obtained by monitoring the Ir L3L_3-edge XMCD signal as a function of HH. As can be noticed, the Ir moment is nearly saturated for Ca0.5_{0.5} under H=3H = 3 T, whereas for Ba0.5_{0.5} the saturation might be reach only for much higher magnetic fields. A remarkable point in these curves is the pronounced decrease in the HCH_C of Ba0.5_{0.5} with respect to Ca0.5_{0.5}, as also observed in the macroscopic M(H)M(H) curves measured in PPMS [Fig. 2(d)].

Discussion

In general, the magnetic ordering of TM ions in DPs can be described by the orbital hybridization mechanism, according to which the dd orbital states reside close to the Fermi surface, being respectively slightly below the chemical potential for one TM ion and slightly above for another. Then, a hybridization between these orbitals, mediated by O 2p2p orbitals, enables the virtual spin-dependent electron hopping from one TM ion to another and can stabilize either FM or AFM coupling [2, 3]. For HS Co2+^{2+} (Co3+^{3+}), there are two (one) completely filled and one (two) half filled t2gt_{2g} orbital(s), while both ege_g orbitals are partially filled. The 5d5d Ir4+^{4+} ion, on the other hand, is at LS configuration, with the SOC lifting the t2gt_{2g} degeneracy to a fourfold jeff=3/2j_{eff} = 3/2 and a twofold jeff=1/2j_{eff} = 1/2 state. Thus, Ir4+^{4+} has one jeff=1/2j_{eff} = 1/2 electron while its jeff=3/2j_{eff} = 3/2 orbitals are fully occupied [4, 48].

Recent first-principles calculations for La2_2CoIrO6_6 have shown that the energy of the Ir-jeff=1/2j_{eff} = 1/2 state resides in between those of Co t2gt_{2g} and ege_g, whereas the strong crystal field splitting on Ir puts its ege_g states ∼3\sim3 eV higher in energy, making these later orbitals irrelevant to our discussion [49]. As a consequence, only the t2gt_{2g} – O 2p2p – t2gt_{2g} π\pi bonding would be a relevant exchange path between Co and Ir. However, although the t2gt_{2g} – O 2p2p – ege_g coupling is in principle forbidden by symmetry in a cubic perovskite (i.e. for a 180∘180^\circ B–O–B′' bond angle), it can become possible with lattice distortion, as is the case for the samples here investigated.

Figure 6(a) depicts the spin-selective exchange pathways between Co2+^{2+}/Co3+^{3+} and Ir4+^{4+} in our system. There is one possible FM virtual hopping from the Co t2gt_{2g} minority spin state to Ir t2gt_{2g} (jeff=1/2j_{eff} = 1/2), and a number of possible AFM couplings. Here we must recall the highly covalent character of the Ir–O bond, with an O-electron partially occupying the Ir-t2gt_{2g} orbital [14], which in principle inhibits the virtual electron transfer from Co to Ir. This partial population of Ir-jeff=1/2j_{eff} = 1/2 orbital with O-electrons, together with the SOC, may explain the dramatic decrease of the Ir magnetic moment observed here and for other Ir4+^{4+}-based perovskites [8, 9, 11, 15]. Therefore, AFM charge transfer from Ir to Co seems to dominate the orbital hybridization in these samples, which is supported by the fact that all experimental and theoretical works reported so far for CoIr-based DPs points towards AFM coupling between these ions. As Figure 6(a) shows, the intermediate charge transfer from Ir t2gt_{2g} can occur to a Co t2gt_{2g} or ege_g orbital. A transfer to Co t2gt_{2g}, through the Ir t2gt_{2g} – O 2p2p – Co t2gt_{2g} pathway, is expected to be stronger for Ba0.5 than for Ca0.5, since the former compound presents increased Co–O–Ir bond angles [see Figure 6(b) and (Figure 6c)]. The fact that Ca0.5 shows a clear magnetic ordering at a relatively high temperature, contrasting the results found for Ba0.5, suggests that the Ir t2gt_{2g} – O 2p2p – Co t2gt_{2g} is not the most relevant hybridization channel in these compounds. On the other hand, the Ir t2gt_{2g} – O 2p2p – Co ege_g channel is expected to strengthen as the Co–O–Ir angle decreases, subsequently becoming the most relevant hybridization pathway between Co and Ir. This is reminiscent of other 3d-5d DPs, for which a systematic increase of the magnetic ordering temperature is observed as the B–O–B’ angle decreases [36, 38, 39, 50]. The absence of a clear magnetic ordering for Ba0.5 is thus likely related to the weakened Ir t2gt_{2g} – O 2p2p – Co ege_g and strengthened Ir t2gt_{2g} – O 2p2p – Co t2gt_{2g} hybridization, with the competing magnetic interactions leading to spin frustration.

Another striking difference between Ba0.5 and Ca0.5 resides in their HCH_C. In other 3d-5d DPs, as well, changes in the orbital hybridization caused by structural distortions remarkably alter their magnetic properties. For AA′AA'FeReO6_6 (AA′AA' = Ba, Ca, Sr), the increase of HCH_C and TCT_C was observed with decreasing the average AA radius, which was ascribed to the increase of Re orbital moment due to lattice distortion [36]. Contrastingly, high-pressure studies in A2A_2FeReO6_6 (AA = Ba, Ca) showed a remarkable increase of HCH_C with increasing pressure that was not related to the increase of Re orbital contribution. Instead, such changes in HCH_C were explained in terms of reducing the strength of SOC caused by the structural distortion [37].

For our CoIr-based samples, the dramatic increase in the HCH_C of Ca0.5 with respect to Ba0.5 is probably not related to increase of the orbital moments since the Ir mlm_l value obtained for Ca0.5 is only slightly larger than that of Ba0.5, while the Co ml/msm_l/m_s ratio is higher for the latter compound. Here we must recall that FeRe-based DPs are somewhat different from our CoIr-system. Re5+^{5+} has two unpaired electrons at jeff=3/2j_{\mathrm{eff}} = 3/2 orbitals whereas for Ir4+^{4+} these orbitals are complete and there is one unpaired spin at jeff=1/2j_{\mathrm{eff}} = 1/2. Moreover, a quenched orbital moment is expected for Fe3+^{3+}, whereas Co2+^{2+} is known to present a non-negligible orbital contribution, as our XMCD results clearly indicate. Our results may find more resemblance with those observed for NiIr-based DPs, since Ni2+^{2+} may also present orbital contribution to the magnetic moment [51, 52]. For A2A_2NiIrO6_6 (AA = La-Lu) the systematic increase of TCT_C and HCH_C with decreasing AA ionic radius is attributed to the increased strength of the AFM Ni2+^{2+} ege_g – Ir4+^{4+} t2gt_{2g} orbital hybridization [38, 39]. The enhancement of HCH_C with bending of B–O–B’ angle in the La2−x_{2-x}Ax_xCoIrO6_6 samples here investigated gives further evidence that the t2gt_{2g} – O 2p2p – ege_g hybridization plays a major role on the CoIr-based DPs.

Summary

In summary, we thoroughly investigated the structural, electronic and magnetic properties of La1.5_{1.5}Ba0.5_{0.5}CoIrO6_6 (Ba0.5) and La1.5_{1.5}Ca0.5_{0.5}CoIrO6_6 (Ca0.5) samples by means of XRD, ac and dc magnetization, Co and Ir L2,3L_{2,3}-edges and Co KK-edge XAS and XMCD. Our results show that the presence of 25% of Ba2+^{2+}/Ca2+^{2+} at La3+^{3+} site leads to Co2+^{2+}/Co3+^{3+} mixed valence, while Ir remains mostly in a tetravalent state. For Ca0.5 it is observed a magnetic ordering at ∼\sim96 K and a large HCH_C, while for Ba0.5 a magnetic ordering is not clear and the HCH_C is much smaller. Since both compounds presents nearly the same Co and Ir valence states and effective magnetic moments, the remarkable differences between the magnetic properties of these samples can be understood in terms of changes in the exchange paths between Co ege_g/t2gt_{2g} and Ir t2gt_{2g} (jeff=1/2j_{\mathrm{eff}} = 1/2) orbitals. For Ca0.5, the AFM Co ege_g – Ir t2gt_{2g} coupling seems to dominate due to its highly distorted structure. For Ba0.5, the increased structural symmetry weakens such hybridization but strengthen the Co t2gt_{2g} – Ir t2gt_{2g} pathway, with the competing interactions leading to a magnetic frustration on this sample. Together with the ASD, the competing magnetic couplings explain the emergence of a SG phase at low temperatures in both compounds.

Acknowledgements

This work was supported by the Brazilian funding agencies: Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) [Nos. E-26/202.798/2019 and E-26/211.291/2021], Fundação de Amparo à Pesquisa do Estado de Goiás (FAPEG) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) [No. 400633/2016-7]. We thank Diamond Light Source for time on beamline I06 under proposal MM29620-1. We acknowledge DESY (Hamburg, Germany), a member of the Helmholtz Association HGF, for the provision of experimental facilities. Parts of this research were carried out at beamline P09 under proposal I-20200348 and we would like to thank J. Bergtholdt and O. Leupold for assistance in setting up the experiment and 6T/2T/2T vector magnet, respectively.

References

  1. [1]K.-I. Kobayashi, T. Kimura, H. Sawada, K. Terakura, and Y. Tokura. Room-temperature magnetoresistance in an oxide material with an ordered double-perovskite structure, Nature 395, 677-680 (1998).DOI
  2. [2]D. Serrate, J. M. De Teresa, and M. R. Ibarra. Double perovskites with ferromagnetism above room temperature, J. Phys.: Condens. Matter 19, 023201 (2007).
  3. [3]S. Vasala and M. Karppinen. A2B'B''O6 perovskites: A review, Prog. Solid State Chem. 43, 1 (2015).DOI
  4. [4]B. J. Kim, H. Ohsumi, T. Komesu, S. Sakai, T. Morita, H. Takagi, and T. Arima. Phase-Sensitive Observation of a Spin-Orbital Mott State in Sr2IrO4, Science 323, 1329 (2009).DOI
  5. [5]G. Cao, T. F. Qi, L. Li, J. Terzic, S. J. Yuan, L. E. DeLong, G. Murthy, and R. K. Kaul. Novel Magnetism of Ir5+ (5d4) Ions in the Double Perovskite Sr2YIrO6, Phys. Rev. Lett. 112, 056402 (2014).pubmed.ncbi.nlm.nih.gov/24580616
  6. [6]B. Raveau and M. Motin Seikh, Cobalt Oxides: From Crystal Chemistry to Physics (Wiley-VCH, Weinheim, 2012).
  7. [7]J. Song, B. Zhao, L. Yin, Y. Qin, J. Zhou, D. Wang, W. Song, and Y. Sun. Reentrant spin glass behavior and magnetodielectric coupling of an Ir-based double perovskite compound, La2CoIrO6, Dalton Trans. 46, 11691 (2017).DOI
  8. [8]S. Lee, M. C. Lee, Y. Ishikawa, P. Miao, S. Torii, C. J. Won, K. D. Lee, N. Hur, D. Y. Cho, and T. Kamiyama. Magnetoelastic octahedral breathing mode in the ferrimagnetic La2CoIrO6 double perovskite, Phys. Rev. B 98, 104409 (2018).DOI
  9. [9]N. Narayanan, D. Mikhailova, A. Senyshyn, D. M. Trots, R. Laskowski, P. Blaha, K. Schwarz, H. Fuess, and H. Ehrenberg. Temperature and composition dependence of crystal structures and magnetic and electronic properties of the double perovskites La2−xSrxCoIrO6 (0≤x≤2), Phys. Rev. B 82, 024403 (2010).DOI
  10. [10]M.-C. Lee, C. H. Sohn, S. Y. Kim, K. D. Lee, C. J. Won, N. Hur, J.-Y. Kim, D.-Y. Cho, and T. W. Noh. Stabilization of ferromagnetic ordering in cobaltite double perovskites of La2CoIrO6 and La2CoPtO6, J. Phys.: Condens. Matter 27, 336002 (2015).DOI
  11. [11]L. Bufaical, E. Sadrollahi, F. J. Litterst, D. Rigitano, E. Granado, L. T. Coutrim, E. B. Araújo, M. B. Fontes, E. Baggio-Saitovitch, and E. M. Bittar. Effect of structural and magnetic disorder on the 3d-5d exchange interactions in La2−xCaxCoIrO6, Phys. Rev. B 102 (2020) 024436.
  12. [12]M.-C. Lee, S. Lee, C. J. Won, K. D. Lee, N. Hur, J. L. Chen, D. Y. Cho, and T. W. Noh. Hybridized orbital states in spin-orbit coupled 3d-5d double perovskites studied by x-ray absorption spectroscopy, Phys. Rev. B 97, 125123 (2018).DOI
  13. [13]W. Jin, S. H. Chun, J. Kim, D. Casa, J. P. C. Ruff, C. J. Won, K. D. Lee, N. Hur, and Y.-J. Kim. Magnetic excitations in the double-perovskite iridates La2MIrO6 (M = Co, Ni, and Zn) mediated by 3d-5d hybridization, Phys. Rev. B 105, 054419 (2022).DOI
  14. [14]D. Takegami, D. Kasinathan, K. K. Wolff, S. G. Altendorf, C. F. Chang, K. Hoefer, A. Melendez-Sans, Y. Utsumi, F. Meneghin, T. D. Ha, C. H. Yen, K. Chen, C. Y. Kuo, Y. F. Liao, K. D. Tsuei, R. Morrow, S. Wurmehl, B. Büchner, B. E. Prasad, M. Jansen, A. C. Komarek, P. Hansmann, and L. H. Tjeng. Charge-transfer energy in iridates: A hard x-ray photoelectron spectroscopy study, Phys. Rev. B 102, 045119 (2020).DOI
  15. [15]A. Kolchinskaya, P. Komissinskiy, M. B. Yazdi, M. Vafaeae, D. Mikhailova, N. Narayanan, H. Ehrenberg, F. Wilhelm, A. Rogalev, and L. Alff. Magnetism and spin-orbit coupling in Ir-based double perovskites La2−xSrxCoIrO6, Phys. Rev. B 85, 224422 (2012).
  16. [16]L. T. Coutrim, E. M. Bittar, F. Stavale, F. Garcia, E. Baggio-Saitovitch, M. Abbate, R. J. O. Mossanek, H. P. Martins, D. Tobia, P. G. Pagluso, and L. Bufaiçal. Compensation temperatures and exchange bias in La1.5Ca0.5CoIrO6, Phys. Rev. B 93, 174406 (2016).DOI
  17. [17]L. T. Coutrim, E. M. Bittar, F. Garcia, and L. Bufaiçal. Influence of spin glass-like magnetic relaxation on the zero-field-cooled exchange bias effect, Phys. Rev. B 98 (2018) 064426.DOI
  18. [18]L. Bufaiçal, L. T. Coutrim, E. M. Bittar, and F. Garcia. A phenomenological model for the spontaneous exchange bias effect, J. Magn. Magn. Mater. 512 (2020) 167048.DOI
  19. [19]L. T. Coutrim, E. M. Bittar, E. Baggio-Saitovitch, and L. Bufaiçal. Spin glass-like properties and exchange bias in La1.5Sr0.5CoIrO6, J. Magn. Magn. Mater. 441, 243 (2017).
  20. [20]A. C. Larson and R. B. Von Dreele. General structure analysis system - GSAS/EXPGUI, Los Alamos National Laboratory Report No. LAUR 86-748, 2000; B. H. Toby. EXPGUI, a graphical user interface for GSAS, J. Appl. Crystallogr. 34, 210 (2001).
  21. [21]J. Strempfer, S. Francoual, D. Reuther, D. K. Shukla, A. Skaugen, H. Schulte-Schrepping, T. kracht and H. Franz. Resonant scattering and diffraction beamline P09 at PETRA III, J. Synchrotron Rad. 20, 541 (2013).arxiv.org/abs/1304.1333
  22. [22]J. Strempfer, J. R. L. Mardegan, S. Francoual, L. Veiga, L. Bouchenoire, T Spitzbart and H. Zink. Fast helicity switching of x-ray circular polarization at beamline P09 at PETRA III, AIP conference proceedings 1741, 030017 (2016).DOI
  23. [23]P. D. Battle, J. G. Gore, R. C. Hollyman, A. V. Powell. The magnetic properties of Ba2CoIrO6 and Ba3CoIr2−xRu2−xO9, J. Alloy. Compd. 218, 110-116 (1995).
  24. [24]C. J. Howard, B. J. Kennedy and P. M. Woodward. Ordered double perovskites - a group-theoretical analysis, Acta Cryst. B59, 463-471 (2003).pubmed.ncbi.nlm.nih.gov/12947230
  25. [25]R. D. Shannon. Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides, Acta Cryst. A32, 751 (1976).DOI
  26. [26]P. Woodward, R-D. Hoffmann, and A. W. Sleight. Order-disorder in A2M3+M5+O6 perovskites, J. Mater. Res. 9, 8 (1994).
  27. [27]A. F. J. Morgownik and J. A. Mydosh. High-temperature susceptibility of the CuMn spin-glass, Phys. Rev. B 24, 5277 (1981).DOI
  28. [28]J. A. Mydosh. Spin glasses: redux: an updated experimental/materials survey, Rep. Prog. Phys. 78, 052501 (2015).DOI
  29. [29]J. Souletie and J. L. Tholence. Critical slowing down in spin glasses and other glasses: Fulcher versus power law, Phys. Rev. B 32, 516 (1985).pubmed.ncbi.nlm.nih.gov/9936703
  30. [30]J. A. Mydosh. Spin Glasses: An Experimental Introduction (Taylor & Francis, London, 1993).
  31. [31]A. Malinowski, V. L. Bezusyy, R. Minikayev, P. Dziawa, Y. Syryanyy, and M. Sawicki. Spin-glass behavior in Ni-doped La₁.₈₅Sr₀.₁₅CuO₄, Phys. Rev. B 84, 024409 (2011).DOI
  32. [32]V. K. Anand, D. T. Adroja, and A. D. Hillier. Ferromagnetic cluster spin-glass behavior in PrRhSn₃, Phys. Rev. B 85, 014418 (2012).DOI
  33. [33]T. Dey, A. Maljuk, D. V. Efremov, O. Kataeva, S. Gass, C. G. F. Blum, F. Steckel, D. Gruner, T. Ritschl, A. U. B. Wolter, J. Geck, C. Hess, K. Koep­ernik, J. van den Brink, S. Wurmehl, and B. Büchner. Ba₂YIrO₆: A cubic double perovskite material with Ir⁵⁺ ions, Phys. Rev. B 93, 014434 (2016).DOI
  34. [34]S. Agrestini, K. Chen, C.-Y. Kuo, L. Zhao, H.-J. Lin, C.-T. Chen, A. Rogalev, P. Ohresser, T.-S. Chan, S.-C. Weng, G. Auffermann, A. Völkze, A. C. Komarek, K. Yamaura, M. W. Haverkort, Z. Hu, and L. H. Tjeng. Nature of the magnetism of iridium in the double perovskite Sr₂CoIrO₆, Phys. Rev. B 100, 014443 (2019).DOI
  35. [35]K. K. Wolff, L. H. Tjeng, and M. Jansen. The new ordered double perovskite SrLaCuIrO₆, Solid State Commun. 289, 43-46 (2018).
  36. [36]M. Sikora, Cz. Kapusta, M. Borowiec, C. J. Oates, V. Prochazka, D. Ry­bicki, D. Zajac, J. M. De Teresa, C. Marquina, and M. R. Ibarra. Evidence of unquenched Re orbital magnetic moment in AA′FeReO₆ double perovskites, Appl. Phys. Lett. 89, 062509 (2006).DOI
  37. [37]C. A. Eschenkoha, G. Fabbris, F. Sun, C. Park, J. Gopalakrishnan, K. Ramesha, E. Granado, N. M. Souza-Neto, M. van Veenendaal, and D. Haskel. Tuning magnetoelectricity with external pressure in iron-rhenium based ferrimagnetic double perovskites, Phys. Rev. B 98, 054402 (2018).
  38. [38]Timothy Ferreira, Gregory Morrison, Jeongho Yeon, and Hans-Conrad zur Loye. Design and Crystal Growth of Magnetic Double Perovskite Iridates: Ln₂MIrO₆ (Ln = La, Pr, Nd, Sm-Gd; M = Mg, Ni), Cryst. Growth Des. 16, 5 (2016).DOI
  39. [39]Hai L. Feng, Zheng Deng, Meixia Wu, Mark Croft, Saul H. Lapidus, Sizhan Liu, Trevor A. Tyson, Bruce D. Ravel, Nicholas F. Quackenbush, Corey E. Frank, Changqing Jin, Man-Rong Li, David Walker, and Martha Greenblatt. High-Pressure Synthesis of Lu₂NiIrO₆ with Ferrimagnetism and Large Coercivity, Inorg. Chem. 58, 1 (2019).DOI
  40. [41]B. T. Thole, P. Carra, F. Sette, and G. van der Laan. X-Ray Circular Dichroism as a Probe of Orbital Magnetization, Phys. Rev. Lett. 68, 1943 (1992).pubmed.ncbi.nlm.nih.gov/10045260
  41. [42]P. Carra, B. T. Thole, M. Altarelli, and X. Wang. X-Ray Circular Dichroism and Local Magnetic Fields, Phys. Rev. Lett. 70, 694 (1993).pubmed.ncbi.nlm.nih.gov/10054179
  42. [43]Y. Teramura, A. Tanaka, and T. Jo. Effect of Coulomb Interaction on the X-Ray Magnetic Circular Dichroism in the X-Ray Magnetic Circular Dichroism, J. Phys. Soc. Jpn. 65, 1053 (1996).DOI
  43. [44]C. Piamonteze, P. Miedema, and F. M. F. de Groot. Accuracy of the spin sum rule in XMCD for the transition-metal L edges from manganese to copper, Phys. Rev. B 80, 184410 (2009).DOI
  44. [45]X. Liu, M. Kotiuga, H. Kim, A. T. N’Diaye, Y. Choi, Q. Zhang, Y. Cao, M. Kareev, F. Wen, B. Pal, J. W. Freeland, L. Gu, D. Haskel, P. Shafer, E. Arenholz, K. Haule, D. Vanderbilt, K. M. Rabe, and J. Chakhalian. Interfacial charge-transfer Mott state in iridate−nickelate superlattices, P. Natl. Acad. Sci. USA 116, 19863-19868 (2019).DOI
  45. [46]D. Yi, J. Liu, S.-L. Hsu, L. Zhang, Y. Choi, J.-W. Kim, Z. Chen, J. D. Clarkson, C. R. Serrao, E. Arenholz, P. J. Ryan, H. Xu, R. J. Birgeneau, and A. Ramesh. Atomic-scale control of magnetic anisotropy via novel spin–orbit coupling effect in La₂⁄₃Sr₁⁄₃MnO₃/SrIrO₃ superlattices, P. Natl. Acad. Sci. USA 113, 23 (2016).
  46. [47]S. Fujiyama, H. Ohsumi, K. Ohashi, D. Hirai, B. J. Kim, T. Arima, M. Takata, and H. Takagi. Spin and Orbital Contributions to Magnetically Ordered Moments in 5d Layered Perovskite Sr₂IrO₄, Phys. Rev. Lett. 112, 016405 (2014).pubmed.ncbi.nlm.nih.gov/24483915
  47. [48]G. Jackeli and G. Khaliullin. Mott Insulators in the Strong Spin-Orbit Coupling Limit: From Heisenberg to a Quantum Compass and Kitaev Models, Phys. Rev. Lett. 102, 017205 (2009).DOI
  48. [49]S. Ganguly, S. Bhowal. Study of nontrivial magnetism in 3d-5d transition metal based double perovskites, Phys. Rev. B. 101, 085104 (2020).DOI
  49. [50]R. Morrow, K. Samanta, T. S. Dasgupta, J. Xiong, J. W. Freeland, D. Haskel, and P. M. Woodward. Magnetism in Ca₂CoOsO₆ and Ca₂NiOsO₆: Unraveling the Mystery of Superexchange Interactions between 3d and 5d Ions, Chem. Mater. 28, 3666-3675 (2016).DOI
  50. [51]V. Fernandez, C. Vettier, F. de Bergevin, C. Giles, and W. Neubeck. Observation of orbital moment in NiO, Phys. Rev. B 57, 7870 (1998).DOI
  51. [52]S. K. Kwon and B. I. Min. Unquenched large orbital magnetic moment in NiO, Phys. Rev. B 62, 73 (2000).DOI

Paper details

vlnc:W_tghhsqvt
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 8

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents