English

• Unlike 3d or 4d electrons, 5d electrons are significantly different due to the radial distribution function and strong spin-orbit coupling (SOC) effect [1-9]. The strong SOC of 5d electrons drives energy level splitting to generate new electronic states, which interact with spatial and temporal symmetry breaking, delivering new quantum effects and functionalities in the large class of perovskite materials, such as spin Hall effect [10-12], Weyl semimetal [13-15], Rashba effect [16-18], topological surface states [19], and magnetic skyrmion [20,21]. The representative 5d compounds, perovskite iridates, have garnered significant interest in SOC system. Particularly, Ruddlesden-Popper iridates Sr_n+1Ir_nO_3n+1 display metal-insulator transition (MIT) driven by the modulation of bandwidth, from strongly insulating Sr₂IrO₄ to metallic non-Fermi liquid SrIrO₃ [22-24]. Of special interest is the strong SOC combined with the modest electronic correlations, leading to a novel spin-orbital Mott insulator with a J_1/2 ground state [25,26]. Thus, a slightly perturbation can affect the stability of insulating phase, such as defect [27], carrier doping [28], magnetic field [29], or external pressure [30-33]. For example, Sr₂Ir_1−xO₄ (T = Ru, Rh) provides the first demonstration of a spin-orbit-controlled MIT phenomenon [34], which is consistent with theoretical calculations in the finite SOC systems Sr₂IrO₄ (n = 1) [35-37]. In addition, the spin-orbit Mott insulator Sr₃Ir₂O₇ (n = 2), a double-layered perovskite, provides a valuable platform to explore the collapse of Mott gap under high pressure (HP) [38]. At 59.5 GPa, Sr₃Ir₂O₇ transforms into a confined metal, exhibiting metallicity in the ab plane but insulating behavior along the c axis, which is similar to the peculiar metal phase observed in cuprate superconductors. A small amount of La substitution for Sr in (Sr_1-xLa_x)₃Ir₂O₇ can also melt away its insulating gap, and thus lead to a correlated-metallic state [39]. These findings emphasize the intricate correlation between SOC and MIT.

  Metastable orthorhombic SrIrO₃ (denoted as O-SrIrO₃, space group of Pnma) is the end member of the Ruddlesden-Popper Sr_n+1Ir_nO_3n+1 series with n being infinite, which is prepared under HP (5.0 GPa) [40,41]. Nevertheless, the thermodynamically stable phase of SrIrO₃ is in monoclinic structure (named as 6H-SrIrO₃, space group of C2/c) prepared at atmospheric pressure (AP) [42]. The SOC and electronic correlation in SrIrO₃ polymorphs manifest interesting magnetic and electrical properties, providing an ideal platform for chemical/electronic structure-dependent phase engineering. Previous studies have demonstrated that the 6H-SrIrO₃ exhibits metallic conductivity with non-Fermi-liquid behavior, being the first known paramagnet in ternary iridate oxides [43]. The O-SrIrO₃ also exhibits paramagnetic (PM) behavior and metallic response with a MIT near 44 K with a positive magnetoresistance (MR) effect at low temperature, where the metallic property is likely due to the large electron hopping [44]. The 6H-SrIrO₃ can be destabilized by partial Ir substitution with larger size and less electronegative cations (such as Li, Mg, Fe, Co, Ni, Zn) by the alteration of electron interactions, resulting in the O-SrIrO₃ polymorph under AP [45]. It is noteworthy that, the uneven distribution of electrons in the degenerate 5d orbital (t_2g or e_g) of the central ion would lead to the distortion of IrO₆ octahedra, which can be altered by crystal field splitting regulation. Thus, O-SrIrO₃ provides a fascinating playground to explore MIT driven by intertwined charge, spin, and lattice degrees of freedom. For example, the Dirac semi-metal state (x = 0) of SrIr_1-xSn_xO₃ can be suppressed by the locally interposed Sn⁴⁺ (4d¹⁰), leading to an antiferromagnetic (AFM) insulator ground state [46]. The orbital inactive Sn⁴⁺ (4d¹⁰) predominantly promotes the Mott localization of the J_1/2 state by reducing the effective one-electron bandwidth, while that of J_3/2 state is merely moderately changed. Nevertheless, the former attempts at revealing that the role of SOC in O-SrIrO₃ have been hindered by concurrently occurring changes to the filling, thus the direct evidence for the role of SOC in stabilizing the insulating state still remains unclear, where the novel MIT is potentially driven by alteration of the bandwidth for J_1/2 and J_3/2 states.

  To address these concerns, we realize the preservation of metastable O-SrIrO₃ polymorph by employing physical and chemical pressures, where the chemical pressure is driven by Ru and/or Mg doping into Ir and Sr site, respectively. Comprehensive characterizations on the crystal structure, magnetic and electrical properties and DFT calculations reveal that Ru doping weakens the SOC to support metallic state, and Mg doping enhances the SOC with modest electron-electron correlations to stabilize insulating state, manifesting SOC controlled MIT.

  The phase transition from 6H- to O-SrIrO₃ can be initiated by elevated physical pressure (prepared at 5.0 GPa and 1273 K). The purity of two different polymorphs is firstly investigated by powder X-ray diffraction (PXD) measurements (Fig. S1a in Supporting information). Rietveld refinements of PXD data of 6H-SrIrO₃ in monoclinic C2/c symmetry is displayed in Fig. S1b (Supporting information). The target HP sample is further investigated using synchrotron PXD (SPXD) measurements, which intensify the identification of crystalline structure (Fig. S2a in Supporting information). Tables S1 and S2 (Supporting information) show that crystallographic information including selected interatomic distances, bond angles and bond valence sums (BVS). As a rule of thumb, the structure of ABO₃ compounds can be predicted from the ionic radii by the Goldschmidt tolerance factor, [47]. When t is close to 1, cubic symmetry is expected, and distortions from the ideal cubic structure to tetragonal, orthorhombic and monoclinic symmetries are anticipated for lowering t values. As for SrIrO₃, t = 0.992 indicate a cubic or pseudo-cubic symmetry. Nevertheless, the thermodynamically stable phase of SrIrO₃ adopts distorted monoclinic 6H polymorph, which is very unusual for ABO₃ with t around 1. Fig. 1a schematically describes the crystal structure of 6H-SrIrO₃, in which corner-shared Ir(1)O₆ octahedra and Ir(2)₂O₉ dimers formed by two face-sharing Ir(2)O₆ octahedra are alternatively arranged along the c direction, giving a cchcch stacking sequence. The Ir(2)-Ir(2) distances across face sharing octahedra are relatively short (~2.75 Å), implying strong metal-metal bonding and stretched Ir(2)-O bond within the highly distorted octahedral dimers in 6H-SrIrO₃. Owing to the subtle balance between strong Ir-Ir bonding and Coulombic repulsion across the face-sharing octahedra, a reconstructive phase transition would be ignited in 6H-SrIrO₃ when dwelled under HP (5.0 GPa) and high temperature, giving the metastable orthorhombic structure contained only corner-sharing IrO₆ octahedra (Fig. 1a). Despite the slight difference in ionic radii of Ir⁴⁺ (r_Ir ~ 0.625 Å) and Ru⁴⁺ (r_Ru ~ 0.62 Å), SrRuO₃ (t = 0.994) crystallizes in orthorhombic symmetry isostructural to the HP O-SrIrO₃ (Fig. 1a). Obviously, the ionic size dependent geometric factor is insufficient to judge the crystal structure of SrIrO₃. This abnormal phenomenon emphasizes the intricate correlation between SOC (~0.4 eV for iridates and ~0.15 eV for ruthenates) and lattice.

  Figure 1

  

  Figure 1.  (a) Comparison of the crystal structures in monoclinic C2/c, orthorhombic Pnma of SrIrO₃ and Pnma-SrRuO₃. (b) PXD patterns of SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0) prepared by solid-state reaction combined with high-temperature and high-pressure synthesis. (c) x dependence of lattice parameter a (Å), b (Å) and c (Å) in orthorhombic SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0) refined from SPXD data. (d) PXD patterns of Sr_1-yMg_yIrO₃ (y = 0–0.4) prepared by solid-state reaction at AP. (e) E-V curve of HP O-SrIrO₃ to show the equivalent chemical pressure by Ru or Mg doping.

  DownLoad: Full-Size Img PowerPoint

  In contrast to physical pressure, "chemical compression" (also known as chemical pressure, can be induced by substitution with smaller ion to simulate the lattice contraction under external HP) offers a feasible route to mimic the effect of external physical pressure [28,48-50]. There are two main functions of chemical pressure: structural and electronic effects. The former one alters the crystal structure, such as bond length and bond angle, while the latter route can adjust the electronic configuration of B-site ions by isovalent/aliovalent substitution. Both effects can modulate the structure and properties of materials, particularly in the strong correlated 5d systems, where additional SOC effect renders enriched behaviors. Accordingly, replacing Ir⁴⁺ by Ru⁴⁺ in 6H-SrIrO₃ is expected to intercept the metastable O-SrIrO₃ phase at AP. As shown in Fig. S1c (Supporting information), single-phase solid solutions in the metastable orthorhombic structure of SrIr_1-xRu_xO₃ (0 ≤ x ≤ 0.5) can be prepared by substituting Ru⁴⁺ (4d⁴) for Ir⁴⁺ (5d⁵) with solid-solution limit of 50%, which is largely ascribed to the key role of SOC. At high-temperature and high-pressure reaction, the solid-solution limit can be further improved, and finally single-phase solid solution with a full proportion can be prepared below 5.0 GPa (Fig. 1b). The refined crystallographic information from SPXD data (Fig. S2 in Supporting information) are listed in Tables S1 and S2. As shown in Fig. 1c, the substituting Ru⁴⁺ for Ir⁴⁺ results in a monotonically decreasing trend in the unit cell parameters (a, b, and c) with increasing Ru content x, which together with the linear decrease of the B-O bond lengths and increase of the B-O-B bond angles, underlying the alteration of physical properties.

  Apart from smaller B-site doping, substituting smaller size Mg²⁺ (^VIIIr_Mg = 0.89 Å) for Sr²⁺ (^VIIIr_Sr = 1.26 Å) on A-sites also evokes lattice contraction (chemical pressure), which is expected to destabilize the monoclinic C2/c symmetry and induce a phase transition to the orthorhombic Pnma symmetry in Sr_1-yMg_yIrO₃ as displayed in Fig. 1d. Sr_1-yMg_yIrO₃ crystallizes in the individually orthorhombic structure with Mg at 20% (Sr_0.8Mg_0.2IrO₃) according to refined crystallographic information from PXD data in Fig. S1d (Supporting information), which corroborates an orthorhombic Pnma structure. Further increasing of Mg²⁺ content (y ≥ 25%) results in the appearance of MgO impurity. In order to evaluate the chemical pressure stemming from volumetric difference between HP O-SrIrO₃ and the Sr_0.8Mg_0.2IrO₃ solid solution, the energy-volume (E-V) curve is calculated by the Murnaghan equation of state equations (Eqs. S1-S3 in Supporting information). The detailed results are shown in Fig. 1e, which indicates equivalent pressure up to 6.58 GPa for Sr_0.8Mg_0.2IrO₃ and 3.51 GPa for SrIr_0.6Ru_0.4IrO₃, further confirming the feasible trapping of metastable phase with chemical pressure at AP. Assisted by chemical pressure, simultaneous co-doping smaller ionic radii Mg²⁺ and Ru⁴⁺ in the Sr²⁺ and Ir⁴⁺ sites can further increase the solid solution limit at AP, to easier synthesize the O-SrIrO₃ than the solely electron-doped approach in SrIr_1-xRu_xO₃, seeing PXD diagram with 10% and 20% Mg content in Figs. S3 and S4 (Supporting information). While 20% Mg²⁺ doping at A-site, the full solid solution by Ru substitution can be obtained at AP. So, chemical pressure over the Ru/Ir sublattice can partially stabilize the O-SrIrO₃ type solid solution in a limited chemical space when the A-site compression is simultaneously applied, yielding the HP polymorph in Sr_1-yMg_yIr_1-xRu_xO₃ (0 ≤ y ≤ 0.2; 0 ≤ x ≤ 1.0) models.

  Ru and/or Mg doping is expected to induce pronounced changes in a wide range of magnetic properties of single-phase Sr_1-yMg_yIr_1-xRu_xO₃ (0 ≤ y ≤ 0.2; 0 ≤ x ≤ 1.0). The Ru doping induces obvious changes in a wide range of M (T) of SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0) with zero-field-cooling (ZFC) and field-cooling (FC) curves (Fig. 2a). The hysteresis loops M (H) of all series at different temperatures under magnetic field between −5 T and 5 T are shown in Figs. S5, S6a and b (Supporting information). Initial negative magnetization on some ZFC curves could be caused by negative trapped fields inside a magnetometer or the sample insertion procedure [51,52]. Previous studies have demonstrated that SrRuO₃ shows ferromagnetic (FM) ordering with Curie temperature (T_C) ~160 K [53], and O-SrIrO₃ exhibits metallic conductivity and Pauli PM [4]. In orthorhombic samples with x = 0, 0.2, 0.4 (synthesized at HP), there are no obvious deviation between ZFC and FC curves, indicating the basic PM. With an increase of the Ru dopant, the x = 0.6, 0.8, 1.0 (synthesized at AP) cases exhibit FM ordering with rising T_C at 61.2, 112.9, and 160.2 K and effective magnetic moment μ_eff ~ 2.37, 2.41 and 2.59 μ_B, respectively (Table S3 in Supporting information), which attributed to the Ru⁴⁺ at Ir⁴⁺ site involved in the long-range magnetic order with strengthened B-O-B exchange interaction. Fig. S7a (Supporting information) vividly shows the trend of T_C and μ_eff evolution. With the increase of Ir content, the μ_eff and T_C of SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0) series gradually decreased, indicating that the introduction of Ir can weaken the exchange of adjacent ions. As for the solely Mg dopant in Sr²⁺ site case, Sr_0.8Mg_0.2IrO₃ (y = 0.2) captures the orthorhombic single phase and keeps PM similar to O-SrIrO₃ (Fig. 2b, Figs. S6c and d in Supporting information). It is worth noting that there is obvious deviation between ZFC and FC data, which can be attributed to spin glass-like or short-range ordering transitions. In the case of Ru and Mg co-doped Sr_1-yMg_yIr_1-xRu_xO₃ series, the FM ordering can be further enhanced with the increase of Ru content (Figs. S7bc and S8 in Supporting information).

  Figure 2

  

  Figure 2.  Temperature-dependent magnetization of (a) SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0) and (b) Sr_0.8Mg_0.2IrO₃ in ZFC/FC at 0.1 T between 10 K and 300 K. (c) The evolution of B-O bond lengths and B-O-B bond angles in SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0). (d) BO₆ octahedron distortion (Σ and Θ) of orthorhombic SrRuO₃, SrIr_0.6Ru_0.4O₃, SrIrO₃ and Sr_0.8Mg_0.2IrO₃.

  DownLoad: Full-Size Img PowerPoint

  In light of the orthorhombic structure, the distance between the B-site ions within the corner-shared BO₆ octahedra ranges from 3.92 Å (SrRuO₃) to 3.94 Å (SrIrO₃). Thus, the strength of the magnetic exchange interactions is expected to be governed by the B-O-B exchange interactions through the bridging O sites, which can be reflected by the evolution of B-O-B angle (Fig. 2c). In SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0), the B-O-B angle increases linearly with increasing x and closer to nearly 180° for x = 1.0, leading to a less distorted lattice with enhanced magnetic super-exchange coupling (Fig. 2d), where the symbols Σ and Θ represent the summation of deviations from of 6 and 24 distinct θ angles in the octahedra, respectively [54,55]. In Sr_0.8Mg_0.2IrO₃, the substitution of Mg²⁺ for Sr²⁺ leads to the change of ionic potential at A site with the altered Sr/Mg-O and Ir-O bond lengths and Ir-O-Ir bond angles, giving rise to larger distortion of the IrO₆ octahedra, and weakening the super-exchange interactions between adjacent electronic spins of Ir-Ir cations (AFM interaction). The measured PM properties agree well with the expected electronic exchange interactions in these orthorhombic iridates.

  The measurements of electrical properties are further executed to uncover the effect of local crystal structure on the electronic transport behavior. The temperature-dependent resistivity ρ (T) curves measured at zero-field of representative cases are shown in Fig. 3a. Clearly, the metallic state of O-SrIrO₃, SrIr_0.6Ru_0.4O₃ and SrRuO₃ are maintained, where the resistivity values decrease with the increase of doping amount of Ru⁴⁺. In contrast, the isostructural Sr_0.8Mg_0.2IrO₃ exhibits semiconducting behavior, which is radically different from metallic O-SrIrO₃, indicating the appropriate level of Mg²⁺ dopant at Sr²⁺ site can induce an exotic MIT in O-SrIrO₃. In addition, compared with the metallic Mg-free case SrIr_1-xRu_xO₃, the series of Mg doped Sr_0.9Mg_0.1Ir_1-xRu_xO₃ (0.4 ≤ x ≤ 1.0) and Sr_0.8Mg_0.2Ir_1-xRu_xO₃ (0 ≤ x ≤ 1.0) all transform into semiconducting response except for samples containing only Ru at the B site as shown in Figs. S9a and b (Supporting information). The corresponding phase diagram of magnetic and electrical properties related to Mg²⁺ and Ru⁴⁺ doping levels is depicted in Fig. 4, exhibiting interesting magnetic and electrical variations. In particular, the MIT phenomenon due to the doping of Mg at Sr site is highlighted.

  Figure 3

  

  Figure 3.  (a) Temperature dependence of the resistivity of SrRuO₃, SrIr_0.6Ru_0.4O₃, O-SrIrO₃ and Sr_0.8Mg_0.2IrO₃ between 10 K and 300 K. (b) Local crystal structure of SrRuO₃, SrIr_0.6Ru_0.4O₃, O-SrIrO₃ and Sr_0.8Mg_0.2IrO₃. (c) Schematic of Ir⁴⁺and Ru⁴⁺ t_2g orbital distribution.

  DownLoad: Full-Size Img PowerPoint

  Figure 4

  

  Figure 4.  Dopant-dependent structure-property phase diagram of Sr_1-yMg_yIr_1-xRu_xO₃(0 ≤ y ≤ 0.2; 0 ≤ x ≤ 1.0). M and O stand for the monoclinic and orthorhombic phase, respectively.

  DownLoad: Full-Size Img PowerPoint

  To assess the origin of the MIT in the orthorhombic iridates, the local crystal structure evolution of corner-shared BO₆ octahedra of representative samples are schematically described in Fig. 3b. Clearly, the angles of B-O-B decrease from 162.9° (SrRuO₃), 157.4° (SrIr_0.6Ru_0.4O₃), 154.5° (O-SrIrO₃) to 153.5° (Sr_0.8Mg_0.2IrO₃), indicating a stronger GdFeO₃-type distortion of the IrO₆ octahedra by Mg substitution for Sr. The crystal field and SOC are deterministic of the electronic structure of the 5d⁵ (Ir⁴⁺) and 4d⁴ (Ru⁴⁺) distribution (Fig. 3c). In 5d⁵ (Ir⁴⁺) orbital, four out of five d electrons occupy the J_eff = 3/2 state, leaving one electron in the J_eff = 1/2 state. In 4d⁴ (Ru⁴⁺) orbital, three out of four d electrons occupy the J_eff = 3/2 state, leaving one electron in the J_eff = 1/2 state. Thus, Ru doping, which has one more hole than Ir, would reduce the magnitude of the SOC and structural distortion, and add holes to the t_2g orbit.

  To gain more insight on the correlations between SOC and ground state in the iridates, DFT calculations are further performed to elucidate the effect of SOC on electrical transport property when Ru and Mg are doped in O-SrIrO₃. The band structure of O-SrIrO₃ subjected to SOC is shown in Fig. 5a. For iridate Ir⁴⁺ (5d⁵), the stronger SOC leads to the J_1/2 and J_3/2 with a large energy difference λ_5d ~ 0.47 eV, which is in line with λ_5d ~ 0.43 eV in O-SrIrO₃ from DFT+U method with U_{eff (Ir)} = 0 eV [56]. Conversely, Ru⁴⁺ (4d⁴) with weaker SOC manifests small energy difference (λ_4d ~ 0.16 eV) between J_1/2 and J_3/2, as the band structure of SrRuO₃ subjected to SOC shown in Fig. 5b. Here, SrIr_0.6Ru_0.4O₃ is opted as a reasonable structure model, which can intensify the understanding of the influence on SOC effects with Ru doping. The band structure of SrIr_0.6Ru_0.4O₃ subjected to SOC is shown in Fig. 5c. The competition between SOC interaction and Hund's effect leads to diluted SOC in Ir site in SrIr_0.6Ru_0.4O₃ compared to that in O-SrIrO₃. Thus, the weaker SOC combined with more effectively screened coulomb interactions between O 2p and Ru/Ir 4d/5d electrons can cause reduction of t_2g orbital difference and lower E_F, driving to the system toward a more robust metallic state. This is consistent with the above electronic transport measurements. As for the computing model of Sr_0.8Mg_0.2IrO₃, a supposed O-SrIrO₃ with 20% Mg dopant concentration in Sr sites is constructed. The band structure of Sr_0.8Mg_0.2IrO₃ subjected to SOC is shown in Fig. 5d. The stronger GdFeO₃-type distortion induces J_1/2 and J_3/2 states further splitting due to enhanced SOC. Compared with O-SrIrO₃, Sr_0.8Mg_0.2IrO₃ exhibits further energy level splitting induced by the finite coulomb repulsion energy U (~0.1 eV). Moving down the periodic table from 3d to 4d and then to 5d, the orbitals in the solids that contain the corresponding d orbitals become increasingly extended and so does the bandwidth (W_3d < W_4d < W_5d). As the bandwidth increases, the corresponding on-site Coulomb repulsion decreases in a sequential manner (U_3d > U_4d > U_5d) [57]. So, the 5d orbits have smaller on-site Coulomb interaction U than the 3d and 4d orbits. Thus, the electron correlation should contribute less to the energy band structure, giving rise to the metallic ground states in many 5d transition metal oxides (TMOs) described by the band theory of solid. However, some 5d TMOs, such as Sr₂IrO₄, Sr₃Ir₂O₇, and Ba₂NaOsO₆, have insulating ground states, which can be attribute to the key role of electron correlation effects U [58-60]. To achieve the insulating bands, the on-site Coulomb interactions U is considered in the Hamiltonian to drive the localization of charge carriers [61]. Thus, the band structure is further calculated combined SOC and on-site Coulomb interactions, where electron correlation U_{eff (Ir)} values are tested from 0.5 eV to 3 eV. When the U_{eff (Ir)} value is 0.5 and 1.0 eV, the electronic structures are metallic from GGA+SOC+U method (Fig. 5e and Fig. S10a in Supporting information). As the U_{eff (Ir)} value is adjusted to 2.0 and 3.0 eV, finite indirect band gaps of 0.18 and 0.50 eV can be observed in Fig. 5f and Fig. S10 (Supporting information), respectively. It is seen that the occupied electronic states of Ir 5d electrons become isolated and localized. Thus, the band structure derived from GGA+SOC+U calculations is consistent with our experimental results, demonstrating that the insulator behavior can be ignited by combining SOC and the on-site Coulomb interactions. That is, electron correlation U and SOC effect are non-negligible factors in MIT phenomenon in DFT calculations. Electron-electron Coulomb repulsion interactions (U) for Ru 4d orbitals are considered in GGA+SOC+U method with U_{eff (Ru)} = 2.9 eV as used to predict metallic behavior of SrRuO₃ [62]. When the U_{eff (Ir)} value is 2.0 eV and U_{eff (Ru)} value is 2.9 eV, the electronic structure is metallic from GGA+SOC+U method (Fig. S11 in Supporting information) in SrIr_0.6Ru_0.4O₃ case, which aligns with our experimental observations, confirming the existence of metallic behavior in the context of Ru doping. All in all, the ground state in strong correlated 5d elements systems is governed by the cooperative effect involving SOC, electron-electron correlations, and crystal field effect, which are expected to raise exotic physical properties in strong correlated iridate oxides.

  Figure 5

  

  Figure 5.  The magnified view of band structure near E_F of (a) O-SrIrO₃, (b) SrRuO₃, (c) SrIr_0.6Ru_0.4O₃ and (d) Sr_0.8Mg_0.2IrO₃ from GGA+SOC method. And band structure near E_F of Sr_0.8Mg_0.2IrO₃ from GGA+SOC+U method as U_{eff (Ir)} = (e) 0.5, (f) 2.0 eV.

  DownLoad: Full-Size Img PowerPoint

  Bond-angle distortion in transition-metal ABO₃ perovskites would reduce the effective d-electron hopping energy and/or the electron bandwidth, via the reduced hybridization between transition-metal d and oxygen p states. In fact, the variation of bond angle with the change of A-site ionic size occasionally causes drastic electronic changes, such as the bandwidth-controlled Mott transition and the colossal MR. The metastable HP O-SrIrO₃ perovskite is an exceptional compound among known iridates in terms of its unusual positive MR (PMR) effect (12% at 20 K and 7 T) [44]. To assess the influence of resistivity in Sr_0.8Mg_0.2IrO₃ under magnetic field, the temperature dependence of resistivity is measured at 7.0 T. As shown in Fig. 6a, the resistivity decreases with applied magnetic field, indicating negative MR (NMR) effect. According to the relationships of MR [MR = (ρ_H - ρ₀)/ρ₀] and temperatures, we consequently investigate the NMR in the orthorhombic Sr_0.8Mg_0.2IrO₃ perovskite (Fig. 6b), giving maximum NMR (−10.1%) at 10 K under 7 T. In general, a variety of MR behavior can be observed by changing the temperature, magnetic field, doping level, and so on. Therein, the NMR is induced by the reduction of scattering and field suppression of localization, while the PMR can be arising from Zeeman spin splitting and spin-orbit scattering [63,64]. In the orthorhombic Sr_0.8Mg_0.2IrO₃, NMR at low temperatures can be ascribed to the reduced scattering ignited by disorder-induced localization under magnetic field. Due to the p-d exchange in Mg-doped O-SrIrO₃ system, charge carrier is surrounded by the polarized magnetic ion electron cloud, thus the spin magnetic moment will be arranged in parallel with the direction of external magnetic field, leading to the enhancement of charge carrier mobility. Therefore, the resistivity will decrease with the increase of magnetic field. As temperature increases, the magnetic moment alignment is gradually disordered due to intensified thermal motion of the molecules, leading to the decrease of NMR. Similar phenomenon has been reported in other oxides [65-67]. The NMR of Sr_0.8Mg_0.2IrO₃ tends to be stable above 150 K, which can be attributed to reduced effect of magnetic field on electron scattering, resulting in a weakened temperature dependence of MR.

  Figure 6

  

  Figure 6.  Temperature dependence of the (a) resistivity and (b) MR of Sr_0.8Mg_0.2IrO₃ between 10 K and 300 K measured at 0 and 7 T.

  DownLoad: Full-Size Img PowerPoint

  In conclusion, we have captured metastable orthorhombic O-SrIrO₃ by both physical and chemical pressures, and investigated the correlation between lattice and SOC in determining the ground state in SOC system. In chemical pressure strategy, Sr_1-yMg_yIr_1-xRu_xO₃ with Ru⁴⁺ doped in Ir⁴⁺ site and/or Mg²⁺ doped in Sr²⁺ sites can stabilize the orthorhombic metastable O-SrIrO₃ phase. The Ru-doping keeps ferromagnetic metallic state, while the Mg-dopant successfully traps paramagnetic semiconducting state with NMR compared to the pristine paramagnetic and metallic O-SrIrO₃ with PMR. Co-doping smaller ionic radii Mg²⁺ and Ru⁴⁺ in the Sr²⁺ and Ir⁴⁺ sites can further increase the solid solution concentration and regulate the magnetic and electrical properties. The exotic electronic phase transition is further uncovered by the DFT calculations, emphasizing the key role of cooperative effect involving SOC, electron-electron correlations, and crystal field effect in determining the ground state of these iridate oxides. The present findings provide a plausible strategy for stabilizing metastable phases by regulating electronic structure through chemical pressure.

  Declaration of competing interest

  The authors report no competing financial interest.

  CRediT authorship contribution statement

  Jinjin Yang: Writing – original draft, Writing – review & editing. Chuanhui Zhu: Formal analysis, Supervision, Writing – original draft, Writing – review & editing. Shuang Zhao: Data curation, Investigation. Tao Xia: Investigation. Pengfei Tan: Investigation. Yutian Zhang: Investigation. Mei-Huan Zhao: Investigation. Yijie Zeng: Methodology, Software. Man-Rong Li: Funding acquisition, Validation, Writing – original draft, Writing – review & editing, Supervision.

  Acknowledgments

  This work was financially supported by the National Natural Science Foundation of China (NSFC, No. 22090041), the Guangdong Basic and Applied Basic Research Foundation (No. 2022B1515120014).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109891.

  
  
• 1. [1]

     Y.N. Liu, S.F. Wang, Y.T. Tao, et al., Chin. Chem. Lett. 27 (2016) 1250–1258. doi: 10.1016/j.cclet.2016.07.018

  2. [2]

     Q. Cui, J.G. Cheng, W. Fan, et al., Phys. Rev. Lett. 117 (2016) 176603. doi: 10.1103/PhysRevLett.117.176603

  3. [3]

     S. Bhowal, I. Dasgupta, J. Phys. Condens Matter. 33 (2021) 1–42.

  4. [4]

     P.E.R. Blanchard, E. Reynolds, B.J. Kennedy, et al., Phys. Rev. B 89 (2014) 214106. doi: 10.1103/PhysRevB.89.214106

  5. [5]

     J.P. Clancy, N. Chen, C.Y. Kim, et al., Phys. Rev. B 86 (2012) 1–9.

  6. [6]

     Y. Qi, S. Matsuishi, J. Guo, et al., Phys. Rev. Lett. 109 (2012) 217002. doi: 10.1103/PhysRevLett.109.217002

  7. [7]

     J. Guo, Y. Qi, S. Matsuishi, et al., J. Am. Chem. Soc. 134 (2012) 20001–20004. doi: 10.1021/ja309724w

  8. [8]

     C. Lu, J.M. Liu, Adv. Mater. 32 (2020) 1904508. doi: 10.1002/adma.201904508

  9. [9]

     H.W. Wang, L.Y. Zhang, N. Hu, et al., Phys. Rev. Mater. 5 (2021) 104412. doi: 10.1103/PhysRevMaterials.5.104412

  10. [10]

      Y. Ohuchi, J. Matsuno, N. Ogawa, et al., Nat. Commun. 9 (2018) 213. doi: 10.1038/s41467-017-02629-3

  11. [11]

      A.S. Patri, K. Hwang, H.W. Lee, et al., Sci. Rep. 8 (2018) 8052. doi: 10.1038/s41598-018-26355-y

  12. [12]

      H.R. Fuh, B. Yan, S.C. Wu, et al., New J. Phys. 18 (2016) 113038. doi: 10.1088/1367-2630/18/11/113038

  13. [13]

      J.M. Kim, M.F. Haque, E.Y. Hsieh, et al., Adv. Mater, 35 (2023) e2107362. doi: 10.1002/adma.202107362

  14. [14]

      X. Ou, H. Wu, et al., Sci. Rep. 4 (2014) 4609. doi: 10.1038/srep04609

  15. [15]

      Y. Yang, F. Yu, X. Wen, et al., Nat. Commun. 14 (2023) 2260. doi: 10.1038/s41467-023-37971-2

  16. [16]

      D. Bhowmik, L. You, S. Salahuddin, et al., Nat. Nanotechnol. 9 (2014) 59–63. doi: 10.1038/nnano.2013.241

  17. [17]

      D. Maryenko, M. Kawamura, A. Ernst, et al., Nat. Commun. 12 (2021) 3180. doi: 10.1038/s41467-021-23483-4

  18. [18]

      L. Zhang, Y.B. Chen, B. Zhang, et al., J. Phys. Soc. Jpn. 83 (2014) 054707. doi: 10.7566/JPSJ.83.054707

  19. [19]

      M.M. Otrokov, Klimovskikh, D. Estyunin, et al., Nature 576 (2019) 416–422. doi: 10.1038/s41586-019-1840-9

  20. [20]

      K.Y. Meng, A.S. Ahmed, M. Bacani, et al., Nano. Lett. 19 (2019) 3169–3175. doi: 10.1021/acs.nanolett.9b00596

  21. [21]

      X. Yao, J. Chen, S. Dong, et al., New J. Phys. 22 (2020) 083032. doi: 10.1088/1367-2630/aba1b3

  22. [22]

      B.J. Kim, H. Jin, S.J. Moon, et al., Phys. Rev. Lett. 101 (2008) 076402. doi: 10.1103/PhysRevLett.101.076402

  23. [23]

      B.J. Kim, H. Ohsumi, T. Komesu, et al., Science 323 (2009) 1329–1332. doi: 10.1126/science.1167106

  24. [24]

      S.J. Moon, H. Jin, K.W. Kim, et al., Phys. Rev. Lett. 101 (2008) 226402. doi: 10.1103/PhysRevLett.101.226402

  25. [25]

      L. Hao, D. Meyers, M.P.M. Dean, et al., J. Phys. Chem. Solids. 128 (2019) 39–53. doi: 10.1016/j.jpcs.2017.11.018

  26. [26]

      I. Qasim, B.J. Kennedy, M. Avdeev, et al., J. Mater. Chem. A 1 (2013) 13357–13362. doi: 10.1039/c3ta12326j

  27. [27]

      P. Tan, C. Zhu, J. Yang, et al., Chin. Chem. Lett. 35 (2023) 108485.

  28. [28]

      J.L. García-Muñoz, M. Suaaidi, M.J. Martínez-Lope, et al., Phys. Rev. B 52 (1995) 13563–13569. doi: 10.1103/PhysRevB.52.13563

  29. [29]

      Z.L. Sun, A.F. Wang, H.M. Mu, et al., npj Quantum Mater. 6 (2021) 94. doi: 10.1038/s41535-021-00397-4

  30. [30]

      T. Hussain, M.J. Oh, M. Nauman, et al., Phys. B: Condens. Matter. 536 (2018) 235–238. doi: 10.1016/j.physb.2017.11.032

  31. [31]

      G. Duvjir, B.K. Choi, I. Jang, et al., Nano Lett. 18 (2018) 5432–5438. doi: 10.1021/acs.nanolett.8b01764

  32. [32]

      S. Zhao, J. Yang, Y. Han, et al., Chin. Chem. Lett. 34 (2023) 107355. doi: 10.1016/j.cclet.2022.03.078

  33. [33]

      M. Xi, C. He, H. Yang, et al., Chin. Chem. Lett. 33 (2022) 2595–2599. doi: 10.1016/j.cclet.2021.12.041

  34. [34]

      H. Wang, M. Marshall, Z. Wang, et al., Inorg. Chem. 62 (2023) 2161–2168. doi: 10.1021/acs.inorgchem.2c03835

  35. [35]

      V. Brouet, P. Foulquier, A. Louat, et al., Phys. Rev. B 104 (2021) 121104. doi: 10.1103/PhysRevB.104.L121104

  36. [36]

      H. Jin, H. Jeong, T. Ozaki, et al., Phys. Rev. B 80 (2009) 075112. doi: 10.1103/PhysRevB.80.075112

  37. [37]

      T.F. Qi, O.B. Korneta, L. Li, et al., Phys. Rev. B 86 (2012) 125105. doi: 10.1103/PhysRevB.86.125105

  38. [38]

      P. Schutz, D. Di Sante, L. Dudy, et al., Phys. Rev. Lett. 119 (2017) 256404. doi: 10.1103/PhysRevLett.119.256404

  39. [39]

      V. Singh, J.J. Pulikkotil, Conference: 59th DAE-BRNS 1665 (2015) 090034. doi: 10.1063/1.4918014

  40. [40]

      J.M. Carter, V.V. Shankar, M.A. Zeb, et al., Phys. Rev. B 85 (2012) 115105. doi: 10.1103/PhysRevB.85.115105

  41. [41]

      V. Singh, J.J. Pulikkotil, J. Phys-condens. Mat. 27 (2015) 335502. doi: 10.1088/0953-8984/27/33/335502

  42. [42]

      J. Yu, X. Wu, D. Guan, et al., Chem. Mater. 32 (2020) 4509–4517. doi: 10.1021/acs.chemmater.0c00149

  43. [43]

      L. Yang, H. Chen, L. Shi, et al., ACS Appl. Mater. Interfaces 11 (2019) 42006–42013. doi: 10.1021/acsami.9b11287

  44. [44]

      J.G. Zhao, L.X. Yang, Y. Yu, et al., J. Appl. Phys. 103 (2008) 103706. doi: 10.1063/1.2908879

  45. [45]

      I. Qasim, B.J. Kennedy, M. Avdeev, et al., J. Mater. Chem. A 1 (2013) 3127–3132. doi: 10.1039/c3ta00540b

  46. [46]

      J. Fujioka, T. Okawa, M. Masuko, et al., J. Phys. Soc. Jpn. 87 (2018) 123706. doi: 10.7566/jpsj.87.123706

  47. [47]

      R.A. Jishi, M.A. Lucas, Int. J. Photoenergy 2016 (2016) 1–9. doi: 10.1155/2016/6193502

  48. [48]

      R. Sun, S. Jin, J. Deng, et al., Adv. Funct. Mater. 31 (2021) 2102917. doi: 10.1002/adfm.202102917

  49. [49]

      M.H. Zhao, X. Zhou, Y. Han, et al., Chem. Mater. 34 (2022) 10153–10161. doi: 10.1021/acs.chemmater.2c02957

  50. [50]

      X. Zhou, M.H. Zhao, J. Yang, Y. et al., Mater. Today Chem. 25 (2022) 100902. doi: 10.1016/j.mtchem.2022.100902

  51. [51]

      L. Zhang, N. Terada, R.D. Johnson, et al., Inorg. Chem. 57 (2018) 5987–5998. doi: 10.1021/acs.inorgchem.8b00479

  52. [52]

      R. Liu, M. Tanaka, K. Yamaura, et al., J. Alloys Compd. 825 (2020) 154019. doi: 10.1016/j.jallcom.2020.154019

  53. [53]

      C. Sow, D. Samal, A.K. Bera, et al., AIP Conf. Proc. 1512 (2013) 90–91. doi: 10.1063/1.4790925

  54. [54]

      J.A. Alonso, M.J. Martı´nez-Lope, M.T. Casais, Inorg. Chem. 39 (2000) 917–923. doi: 10.1021/ic990921e

  55. [55]

      M. Buron-Le Cointe, J. Hébert, C. Baldé, et al., Phys. Rev. B 85 (2012) 064114. doi: 10.1103/PhysRevB.85.064114

  56. [56]

      A. Chauhan, B.R.K. Nanda, Phys. Rev. B 105 (2022) 045127. doi: 10.1103/PhysRevB.105.045127

  57. [57]

      K. Samanta, J. Noky, I. Robredo, et al., npj Comput. Mater. 9 (2023) 167. doi: 10.1038/s41524-023-01106-4

  58. [58]

      J.M. Rondinelli, N.M. Caffrey, S. Sanvito, et al., Phys. Rev. B 78 (2008) 155107. doi: 10.1103/PhysRevB.78.155107

  59. [59]

      S. Gangopadhyay, W.E. Pickett, Phys. Rev. B 91 (2015) 045133. doi: 10.1103/PhysRevB.91.045133

  60. [60]

      C. Martins, M. Aichhorn, S. Biermann, J. Phys. Condens Matter. 29 (2017) 263001. doi: 10.1088/1361-648X/aa648f

  61. [61]

      W. Ju, G.Q. Liu, Z. Yang, Phys. Rev. B 87 (2013) 075112. doi: 10.1103/PhysRevB.87.075112

  62. [62]

      R.D. Wulandari, S. Muhammady, Y. Darma, et al., J. Phys. Chem. Solids. 137 (2020) 109225. doi: 10.1016/j.jpcs.2019.109225

  63. [63]

      K. Liu, X. Ma, S. Xu, et al., npj Comput. Mater. 9 (2023) 16. doi: 10.1109/icpads60453.2023.00012

  64. [64]

      G. Popov, M. Greenblatt, Phys. Rev. B 67 (2003) 024406. doi: 10.1103/PhysRevB.67.024406

  65. [65]

      J. Zhao, B. Jiang, J. Yang, et al., Phys. Rev. B 107 (2023) L060408. doi: 10.1103/PhysRevB.107.L060408

  66. [66]

      P. Salev, L. Fratino, D. Sasaki, Phys. Rev. B 108 (2023) 174434. doi: 10.1103/PhysRevB.108.174434

  67. [67]

      X. Wang, K. Huang, X. Wu, et al., Chin. Chem. Lett. 34 (2023) 108267. doi: 10.1016/j.cclet.2023.108267

• 

  Figure 1  (a) Comparison of the crystal structures in monoclinic C2/c, orthorhombic Pnma of SrIrO₃ and Pnma-SrRuO₃. (b) PXD patterns of SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0) prepared by solid-state reaction combined with high-temperature and high-pressure synthesis. (c) x dependence of lattice parameter a (Å), b (Å) and c (Å) in orthorhombic SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0) refined from SPXD data. (d) PXD patterns of Sr_1-yMg_yIrO₃ (y = 0–0.4) prepared by solid-state reaction at AP. (e) E-V curve of HP O-SrIrO₃ to show the equivalent chemical pressure by Ru or Mg doping.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Temperature-dependent magnetization of (a) SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0) and (b) Sr_0.8Mg_0.2IrO₃ in ZFC/FC at 0.1 T between 10 K and 300 K. (c) The evolution of B-O bond lengths and B-O-B bond angles in SrIr_1-xRu_xO₃ (0 ≤ x ≤ 1.0). (d) BO₆ octahedron distortion (Σ and Θ) of orthorhombic SrRuO₃, SrIr_0.6Ru_0.4O₃, SrIrO₃ and Sr_0.8Mg_0.2IrO₃.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Temperature dependence of the resistivity of SrRuO₃, SrIr_0.6Ru_0.4O₃, O-SrIrO₃ and Sr_0.8Mg_0.2IrO₃ between 10 K and 300 K. (b) Local crystal structure of SrRuO₃, SrIr_0.6Ru_0.4O₃, O-SrIrO₃ and Sr_0.8Mg_0.2IrO₃. (c) Schematic of Ir⁴⁺and Ru⁴⁺ t_2g orbital distribution.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Dopant-dependent structure-property phase diagram of Sr_1-yMg_yIr_1-xRu_xO₃(0 ≤ y ≤ 0.2; 0 ≤ x ≤ 1.0). M and O stand for the monoclinic and orthorhombic phase, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  The magnified view of band structure near E_F of (a) O-SrIrO₃, (b) SrRuO₃, (c) SrIr_0.6Ru_0.4O₃ and (d) Sr_0.8Mg_0.2IrO₃ from GGA+SOC method. And band structure near E_F of Sr_0.8Mg_0.2IrO₃ from GGA+SOC+U method as U_{eff (Ir)} = (e) 0.5, (f) 2.0 eV.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Temperature dependence of the (a) resistivity and (b) MR of Sr_0.8Mg_0.2IrO₃ between 10 K and 300 K measured at 0 and 7 T.

  下载: 全尺寸图片 幻灯片
•