# Fileset

[Gd2CuZnMn4O12-25-12-2023-R.doc](https://mdr.nims.go.jp/filesets/c47e7684-efd6-4ccf-87a2-33191800cdb3/download)

## Creator

[Alexei A. Belik](https://orcid.org/0000-0001-9031-2355)

## Rights

[In Copyright](http://rightsstatements.org/vocab/InC/1.0/)

## Other metadata

[Gd2CuZnMn4O12: A-site columnar-ordered perovskite with anisotropic thermal expansion and a gradual charge-order transition](https://mdr.nims.go.jp/datasets/9256295c-c237-4cb7-b633-5abc8ea9844c)

## Fulltext

Crystal structures of quadruple perovskites RMn3O6 (R = Gd, Er, and Tm)Gd2CuZnMn4O12: A-site columnar-ordered perovskite with anisotropic thermal expansion and a gradual charge-order transitionAlexei A. Belik*Research Center for Materials Nanoarchitectonics (MANA), National Institute for Materials Science (NIMS), Namiki 1-1, Tsukuba, Ibaraki 305-0044, JapanAbstractA new A-site columnar-ordered quadruple perovskite, Gd2CuZnMn4O12, was synthesized by a high-pressure, high-temperature method at 6 GPa and 1500 K. Its structural properties between 100 K and 800 K were investigated by synchrotron powder X-ray diffraction. It has space group Pmmn (No. 59) between 100 K and about 450 K with a layered charge order (CO) of Mn3+ and Mn4+ cations at the perovskite B sites and space group P42/nmc (No. 137) above 475 K with one perovskite B site (having an average manganese oxidation state of +3.5). Anisotropic thermal expansion was observed as the c lattice parameter decreases with increasing temperature above 325 K. Temperature dependence of the unit-cell volume and the lattice parameters and differential scanning calorimetry measurements gave evidence that the CO transition (with TCO ( 475 K) is of the second order. Gd2CuZnMn4O12 shows a ferrimagnetic transition below TC = 105 K with an additional decrease of magnetic susceptibilities below about 45 K, probably caused by a gradual increase of the ordered moments on the Gd sublattices coupled antiferromagnetically with the existent ferrimagnetic structure.1. IntroductionHole-doped perovskite manganites, R1(xAxMnO3 (where R3+ is a trivalent rare-earth cation and A2+ is a divalent alkali-earth cation or Pb2+), have been a playground for materials and solid-state physicists and chemists for decades [1–6]. Because of a mixture of R3+/A2+ they have a mixture of Mn3+/Mn4+ cations and, therefore, a charge degree of freedom. This degree of freedom gives rise to a large variety of physical phenomena. The most famous one is the colossal magnetoresistance effect [5] closely related with a phase separation [6]. The R1(xAxMnO3 systems have rich phase diagrams [4] with a variety of different electronic and magnetic ground states, such as, charge-ordered and orbital-ordered insulating antiferromagnets (with a plethora of antiferromagnetic orders: A-, C-, G-, and CE-types), ferromagnetic metals, charge-ordered and orbital-ordered stripe structures, and spin-glass states [1–4]. The important physical concepts of Jahn–Teller distortions, small polarons, double exchange, and electron-phonon couplings were intensively tested and developed using R1(xAxMnO3 perovskite manganites as model systems.A special attention has always been paid in the literature to half-doped samples with x = 0.5 or near-half-doped samples as some properties could be maximized at these doping levels [1–6], such as, the colossal magnetoresistance effect. In addition, other properties could emerge, for example, a first-order phase transition induced by a magnetic field was found in Nd0.5Sr0.5MnO3 [7]. Charge order (CO) phenomena could appear at other values of doping levels, such as, x = 2/3 and N/8 (where N is a positive integer) [1].There is a large family of perovskite manganites belonging to the A-site-ordered quadruple perovskite family, AMn7O12 or AMn3+3Mn4O12 [8], with intrinsic A-site ordering. They show 1:1 Mn3+:Mn4+ charge ordering for A = Na [9], 1:3 Mn4+:Mn3+ charge ordering for A = Cd, Ca, Sr, and Pb [8] with complex orbital ordering as they contain 75 % of Jahn–Teller-active Mn3+ cations at the B sites, and complex magnetic and crystal structures for solid solutions with intermediate doping levels [10,11].There is a special family of perovskite oxides with the general composition of A2A(A((B4O12, where there is intrinsic triple A-site cation ordering [12], and which are formed through a large a+a+c( type octahedral tilt in the Glazer’s notations [13]. The highest symmetry of such perovskites is P42/nmc. By the appropriate selection of cations for the A, A(, and A(( sites, one can retain the triple A-site cation ordering and, at the same time, introduce the required hole doping of cations at the B sites. In comparison with R1(xAxMnO3 perovskites, where the sizes of R3+ and A2+ cations are comparable, the sizes of cations for the A, A(, and A(( sites should be quite different. In addition, the coordination of the A( and A(( sites in A2A(A((B4O12 is quite different: A( has a square-planar coordination and A(( has a tetrahedral coordination [12]. Therefore, very specific cations can only be located at the A( and A(( sites. For example, Cu2+, Mn2+, and Mn3+ cations are typically located at the A( site, while Zn2+, Mn2+, and Ga3+ cations are typically located at the A(( site [12]. Some (pure B site) manganites, belonging to the A2A(A((B4O12 family, have already been studied in the literature, RMn3O6 [14–17], R2CuMnMn4O12 [18], and Y2CuGaMn4O12 [19]. While a charge-ordered structure with space group Pmmn was found in RMn3O6 [14–17] and R2CuMnMn4O12 [18] at room temperature, no temperature-driven structural transitions to the parent structure with space group P42/nmc were found/studied. On the other hand, Y2CuGaMn4O12 [19] remains in a charge-mixed state (that is, it crystallizes in space group P42/nmc) down to 1.5 K despite the presence of a 1:3 Mn4+:Mn3+ mixture.In our current work, we investigated an A2A(A((B4O12-type pure manganite (B = Mn) with A = Gd3+, A( = Cu2+, and A(( = Zn2+, in other words, the Gd2CuZnMn4O12 composition. This combination of the A/A(/A(( cations corresponds to the half-doped case of R1(xAxMnO3 and gives an average manganese oxidation state as +3.5 – an ideal value for the realization of possible charge ordering. We found that Gd3+2Cu2+Zn2+Mn3+2Mn4+2O12 has a 1:1 Mn3+:Mn4+ charge-ordered structure between 100 K and 450 K. Moreover, we could observe a structural transition above about 475 K to a charge-mixed state, Gd3+2Cu2+Zn2+Mn3.5+4O12, using direct high-temperature structural studies. In addition, Gd2CuZnMn4O12 exhibits a ferrimagnetic transition below 105 K with a strong decrease of magnetic susceptibilities at lower temperatures almost reaching a magnetization reversal phenomenon. 2. Experimental SectionGd2CuZnMn4O12 was prepared from a stoichiometric mixture of o-GdMnO3, CuO (99.9 %), ZnO (99.9 %), and MnO2 (Alfa Aesar, 99.9 %). The oxygen content of a commercial well-crystallized single-phase MnO2 chemical was confirmed by the thermal analysis. Single-phase o-GdMnO3 was prepared from a stoichiometric mixture of Gd2O3 (99.9 %) and Mn2O3 (99.99 %) by annealing in air at 1430 K for 60 h with several intermediate grindings. The synthesis of Gd2CuZnMn4O12 was performed at high pressure of about 6 GPa and high temperature of about 1500 K for 2 h in an Au capsule. The sample was cooled down to room temperature by turning off a heating current after annealing at 1500 K, and then the pressure was slowly released. One high-pressure high-temperature synthesis run gave about 0.4 g of one sample (and one high-pressure cell could hold 4 samples).X-ray powder diffraction (XRPD) data were collected on a RIGAKU MiniFlex600 diffractometer using CuK( radiation (in a 2( range from 8( to 100(, a step width of 0.02(, and scan speed of 2 (/min) at room temperature. Synchrotron XRPD data were collected on the beamline BL02B2 of SPring-8 [20] from 100 K to 800 K on heating and then from 800 K to 295 K on cooling. The data were taken between 2.08 and 78.22 at 0.006 intervals in 2 using a wavelength of  = 0.420186 Å. The data with good statistics for structural refinements were collected at 100 K, 295 K, and 600 K with the measurement time of 200 s (at 100 and 295 K) and 400 s (at 600 K); the measurement time was 10 s at all other temperatures. The sample was filled into an open Lindemann glass capillary tube (inner diameter: 0.2 mm), which was rotated during measurements. The Rietveld analysis of all XRPD data was performed using the RIETAN-2000 program [21]. No signs of decomposition were detected on synchrotron XRPD patterns after heating to 800 K, and widths of reflections remained the same (Figure S1 of Supporting Information).Magnetic measurements were performed on a SQUID magnetometer (Quantum Design, MPMS3) between 2 and 400 K in different applied magnetic fields under both zero-field-cooled (ZFC) and field-cooled on cooling (FCC) conditions. Isothermal magnetization measurements were performed between (70 and +70 kOe at different temperatures. Frequency dependent alternating current (ac) susceptibility measurements were performed on cooling with a Quantum Design MPMS3 instrument at zero static dc field and at different frequencies (f) and different applied oscillating magnetic fields (Hac).Specific heat, Cp, at different magnetic fields between 0 Oe and 90 kOe was recorded from 270 K to 2 K by a pulse relaxation method using a commercial calorimeter (Quantum Design PPMS). Magnetic and specific heat measurements were performed using pieces of pellets: 13.22 mg for dc magnetic, 38.74 mg for ac magnetic, and 9.57 mg for specific heat.Differential scanning calorimetry (DSC) curves of a powder sample (68.10 mg) were recorded on a Mettler Toledo DSC1 STARe system between 300 K and 700 K in an open Al capsule with a heating/cooling rate of 10 K/min under N2 flow. No DSC anomalies were found in the measurement temperature range. Three DSC runs were performed to check the reproducibility. Laboratory XRPD data collected after the DSC experiments showed no changes in comparison with the as-synthesized sample.3. Results and DiscussionGd2CuZnMn4O12 was found to contain an impurity with the cubic A-site-ordered quadruple perovskite structure AA(3B4O12 and the lattice parameter of a = 7.30981(7) Å (at room temperature). Judging from the impurity magnetic transition temperature of about 330 K (see below), its chemical composition should be close to GdCu3Mn4O12, whose magnetic transition temperature is 394 K and the lattice parameter is 7.2744 Å [22], and it is known that magnetic transition temperatures decrease in RCu3(xMnxMn4O12 with increasing x. The weight fraction of the impurity was estimated to be about 4 % from refined scale factors during the Rietveld analysis (see below). Reflections belonging to Gd2CuZnMn4O12 could be indexed in orthorhombic symmetry with a = 7.2714 Å, b = 7.2870 Å, and c = 7.8388 Å. Therefore, Gd2CuZnMn4O12 should crystallize in the orthorhombic space group Pmmn of the A-site columnar-ordered quadruple perovskites [12], and we used the structural parameters of DyMn3O6 [14] as the initial ones in the Rietveld analysis of Gd2CuZnMn4O12 at 100 K and 295 K.We note that we tried to synthesize Gd2CuZnMn4O12 at a higher temperature of 1730 K (at 6 GPa in a Pt capsule for 2 hours); however, the amount of the impurity remained nearly the same (about 5 weight %). The lattice parameters of the sample prepared at 1730 K were a = 7.27738 Å, b = 7.30207 Å, and c = 7.84091 Å (based on synchrotron XRPD at room temperature), and the orthorhombic distortion was visibly larger in agreement with the larger difference between the a and b lattice parameters; the magnetic transition temperature was at 108 K. The impurity had a = 7.32160(8) Å and the magnetic transition at 305 K.Synchrotron XRPD measurements (non-resonant) cannot distinguish Cu2+ and Zn2+ cations. Therefore, distribution of Cu2+ and Zn2+ was assigned based on their coordination environment preferences as Zn2+ cations have a strong preference for a tetrahedral coordination; on the other hand, Cu2+ cations have a strong preference for a square-planar coordination [23] (between tetrahedral and square-planar coordination). The structural analysis of Gd2CuZnMn4O12 also showed that all cation occupation factors converged to unity within experimental errors. We will give the refined occupation factors at 600 K because the symmetry at 600K is higher, and the number of refined structural parameters is smaller (see below). We obtained the following values: g(Gd) = 1.000(2), g(Cu) = 0.490(3) (for a disordered model with the ideal g = 0.5), and g(Zn) = 1.019(5) with fixed g(Mn) = 1; and g(Cu) = 0.490(3), g(Zn) = 1.019(5), and g(Mn) = 1.001(2) with fixed g(Gd) = 1. Note that all other structural and non-structural parameters were refined simultaneously including atomic displacement parameters, and one cation occupation factor should be fixed to avoid strong correlations with atomic displacement parameters. Therefore, within the sensitivity of synchrotron XRPD data, an ideal cation distribution is realized in Gd2CuZnMn4O12 without detectable anti-site disorder.Structural parameters, primary bond lengths, and bond-valence sums (BVS) [24] of Gd2CuZnMn4O12 at 100 K and 295 K after the Rietveld refinements are summarized in Tables 1 and 2. Experimental, calculated, and difference synchrotron XRPD patterns of Gd2CuZnMn4O12 at 295 K are shown on Figure 1a, and the crystal structure is shown on Figure 1b.The orthorhombic Pmmn structural model was applied to the synchrotron XRPD data from 100 K to 670 K, and it was observed that the orthorhombic (O) aO and bO lattice parameters nearly merged at 525 K (Figure S2 of Supporting Information). This fact suggests the existence of a phase transition to a structure with tetragonal (T) symmetry. However, the structural parameters refined in the Pmmn model between 475 K and 670 K gave strange values of some bond lengths deviating from the observed monotonic tendencies (Figure S2 of Supporting Information) suggesting strong correlations among refined structural parameters. Synchrotron XRPD data could be well fitted with the tetragonal P42/nmc model at 475 K and above [12, 19]. Figures 2, 3, and 4 show temperature dependence of the lattice parameters, unit cell volume, and bond lengths, respectively. No visible orthorhombic splitting of reflections was observed at 475 K and above (Figure 5). Considering that no DSC anomalies were detected between 300 K and 700 K (Figure S3 of Supporting Information), it was difficult to determine the exact phase transition temperature. Therefore, we assigned the phase transition temperature to a point where visible orthorhombic splitting of reflections disappears and the tetragonal P42/nmc model can be applied (that is, at TCO = 475 K). We note that space group Pmmn has additional reflections in comparison with space group P42/nmc, for example, 00l reflections with l = 2n + 1. However, the intensity of all additional reflections is very weak, and they often overlap with other stronger reflections. Therefore, they can be readily detected only on the high-statistics data (Figure 5d) measured with 200 s. They are hardly detectable on fast data measured with 10 s. Therefore, temperature evolution of such additional reflections could not be used to determine the value of TCO.Temperature dependence of the lattice parameters and unit-cell volume on heating and cooling showed a small hysteresis between 625 K and 800 K of an unknown origin. Temperature dependence of the unit-cell volume shows no anomalies at TCO. This fact, coupled with the absence of DSC anomalies, gives evidence that the structural phase transition at TCO = 475 K could be of the second order. Moreover, space group Pmmn is a direct subgroup of P42/nmc; therefore, second-order transitions are allowed by symmetry. Structural parameters, bond lengths, and BVS values obtained at T = 600 K in the P42/nmc model are summarized in Tables 3 and 4, and the fitting patterns are shown in the inset of Figure 1a, the crystal structure is shown on Figure 1c. We note that the Cu site could be split from the ideal 2a site (0.75, 0.25, 0.75) with some reduction in the refined value of the atomic displacement parameter.The aO and bO lattice parameters (from 100 K to 450 K) and aT lattice parameter (from 475 K to 800 K) increase monotonically with temperature (Figure 2). On the other hand, the cO lattice parameter first increases from 100 K to 325 K and then decreases from 350 K to 450 K, and it continues to monotonically decrease as cT from 475 K to 800 K. Therefore, Gd2CuZnMn4O12 shows anisotropic thermal expansion above 325 K; but the origin of its anisotropic thermal expansion is not clear and will require additional experiments to understand, for example, using neutron diffraction data, which can located oxygen atoms more precisely.The BVS parameters for the Mn1 and Mn2 sites at room temperature were +3.29 and +3.64. We note that the BVS approach just shows tendencies of oxidation states, and BVS values rarely match with formal oxidation states. For example, in R3+MnO3 perovskites, where the formal oxidation state of Mn is definitely +3, the BVS values vary from +3.15 to +3.25 [25, 26]. More importantly, the octahedral distortion parameters of the Mn1 and Mn2 sites differed by about one order of magnitude, 30.0(10–4 and 2.6(10–4, respectively. These values, coupled with the BVS parameters, suggest that the Mn1 site should be occupied by Mn3+ cations with a strong Jahn–Teller distortion of the Mn1O6 octahedra. On the other hand, the Mn2 site should be occupied by Mn4+ cations without strong Jahn–Teller distortions of the Mn2O6 octahedra. There is one crystallographic site (Mn) for the B cations in the P42/nmc model, and the BVS value of +3.44 was close to the expected average manganese oxidation state of +3.5. The octahedral distortion parameter of 10.2(10–4 at 600 K had an intermediate value, reflecting the statistical presence of a 1:1 mixture of Mn3+ and Mn4+ cations. Therefore, from these structural features we can suggest that a CO structure is realized below TCO = 475 K in Gd2CuZnMn4O12 with a transition from a formal charge-mixed Mn3.5+ state to a mixture of the formal Mn3+ and Mn4+ states.FCC magnetic susceptibilities showed sharp rises below TC = 105 K in all measurement fields from 5 Oe to 10 kOe suggesting the development of a strong ferromagnetic (FM) component (Figure 6). This transition could be seen as sharp, negative peaks on the differential d(T/dT versus T curves (insets of Figure 6). Almost no difference was observed between the ZFC and FCC ( versus T curves at H = 10 kOe, while there was difference below TC at H = 100 Oe. A noticeable extrapolated (between 40 kOe and 70 kOe) magnetization was observed on the M versus H curves (Figure 7), for example, about 3.48(B/f.u. at T = 100 K and 4.75(B/f.u. at T = 60 K, where f.u. stands for formula unit. Such large values cannot be caused by canted antiferromagnetic structures. On the other hand, they are too small for full FM alignments (29(B/f.u. for all magnetic cations and 15(B/f.u. for 3d magnetic cations). Therefore, a ferrimagnetic structure should be realized. The sharp S-type shape of the M versus H curves at T = 60 K and 100 K was typical for soft ferrimagnets with nearly zero coercive fields. At high fields, the M versus H curves did not saturate but continue to increase linearly, probably due to paramagnetic contributions from Gd3+ cations.Below about 40–48 K there was a gradual decrease of magnetic susceptibilities in all measurement fields from 5 Oe to 10 kOe, with the ( values reaching low values at the lowest temperature of 2 K in the small measurement fields of 5–100 Oe (Figure 6c). The decrease was sharper in larger measurement fields (e.g., at H = 10 kOe) and could be seen as a sharp, positive peak on the differential d(T/dT versus T curve (note that a very broad, positive anomaly was observed at H = 100 Oe). We note that an additional (FM-like) anomaly was observed at 330 K. Taking into account the presence of an impurity with the cubic A-site-ordered quadruple perovskite structure, the 330 K anomaly should originate from this impurity [22]. The M versus H curves (Figure 7) qualitatively changed their behavior below 40–48 K (at T = 5 K and 25 K), they now had the smeared S-type shape (with narrow hysteresis near the origin) instead of the sharp S-type shape with negligible hysteresis. The remnant magnetization at 5 K (0.47(B/f.u.) was smaller than that at 25 K (1.39(B/f.u.).Specific heat measurements showed (1) a small anomaly at TC = 105 K, (2) an additional, small anomaly near 45 K, and (3) a strong rise of the Cp/T values below 20 K (Figure 8). A magnetic field of 90 kOe had (1) a strong effect on the anomaly near TC = 105 K and moved the magnetic entropy to much higher temperatures (the effect of the magnetic field was observed up to 200 K), (2) had a weak effect on the anomaly near 45 K, and (3) moved the magnetic entropy to slightly higher temperatures with the suppression of the peak in the Cp/T values below 20 K. Considering that both specific heat and magnetic susceptibilities showed anomalies near 40–48 K we can suggest that they are intrinsic.We can suggest that below TC = 105 K a ferrimagnetic structure is established including only the 3d transition metals of Gd2CuZnMn4O12 in analogy with Y2CuGaMn4O12 [19] (see also Figure S4 of Supporting Information). Then, below about 45 K the Gd3+ sublattices start to order but with small ordered moments, and the ordered moments of the Gd3+ sublattices align opposite to the established ferrimagnetic structure resulting in a decrease of magnetic susceptibilities. Below about 20 K, the ordered moments of the Gd3+ sublattices increase more rapidly giving a strong, broad peak on specific heat. The similar situation was realized, for example, in HoMnO3 [27], where the Ho3+ moments appeared below about 24 K and rapidly increased below about 10 K. Temperature dependence of magnetic susceptibilities of ferrimagnetic Gd2CuZnMn4O12 was close to that of materials with the magnetization reversal phenomenon, e.g., in Ho1(xMnxMnO3 [28]. But negative magnetization was not reached at lower temperatures because the Gd3+ sublattices could not fully overtake the Mn/Cu sublattices. It is interesting that the Dy3+ sublattices give ferromagnetic contributions in Dy2CuZnMn4O12 [29] in comparison with Gd2CuZnMn4O12.To get more information about the magnetic properties we measured ac susceptibilities (Figure 9). The ac susceptibility measurements showed just one transition with peaks on both (( versus T and ((( versus T curves near TC = 105 K and additional shoulders below TC on both curves. The absence of any sharp anomalies near 40–48 K supports our conclusion that the Gd3+ moments are developed gradually with decreasing temperature. Sharp peaks are usually observed (on both (( versus T and ((( versus T curves) when an uncompensated FM moment appears in canted antiferromagnets, ferrimagnets, and ferromagnets. However, the behavior of ac susceptibilities below TC depends on further evolution of an established magnetic structure and coercive fields. If coercive fields are small even small ac fields can interact with and alter established domain structures resulting in additional contributions to (( versus T and ((( versus T curves. For example, large contributions were observed in a soft ferromagnet BiMnO3 below TC = 100 K down to 2 K [30]. Negligible coercive fields were observed in Gd2CuZnMn4O12 between 60 K and 100 K (Figure 10 and Figure S5 of Supporting Information) while the coercive field was about 500 Oe at 25 K; this fact could explain the appearance of shoulders on both (( versus T and ((( versus T curves below TC.4. ConclusionsIn conclusion, we prepared a new member of the A-site columnar-ordered quadruple perovskites, Gd2CuZnMn4O12. This compound crystallizes in space group Pmmn between 100 K and 450 K with the 1:1 Mn3+:Mn4+ charge-order structure. We could find a structural phase transition to a charge-mixed state above about 475 K using direct high-temperature structural studies while thermal methods could not detect such a transition suggesting its second-order nature. A new combination of the A/A(/A(( cations (Gd3+/Cu2+/Zn2+) was used to realize triple A-site cation ordering. Gd2CuZnMn4O12 exhibits a ferrimagnetic transition below 105 K with a strong decrease of magnetic susceptibilities at lower temperatures due to antiferromagnetic ordering of the Gd3+ sublattices.Author InformationCorresponding AuthorAlexei.Belik@nims.go.jpNotesThe authors declare no competing financial interest. Data will be available on reasonable request.AcknowledgementsThe synchrotron radiation experiments were performed at SPring-8 with the approval of Japan Synchrotron Radiation Research Institute (proposal number: 2023A1496). We thank Dr. S. Kobayashi for his help at BL02B2 of SPring-8.Supporting InformationThe Supporting information is available at www…. and includes bond distances of Gd2CuZnMn4O12 at 100 K, DSC curves, details of M versus H curves, and details of synchrotron X-ray diffraction patterns.References[1] M. B. Salamon, M. Jaime, The physics of magnanites: structure and transport. Rev. Mod. Phys. 73 (2001) 583–628.[2] J. M. D. Coey, M. Viret, S. von Molnar, Mixed-valence manganites. Adv. Phys. 58 (2009) 571–697.[3] D. M. Edwards, Ferromagnetism and electron-phonon coupling in the manganites. Adv. Phys. 51 (2002) 1259–1318.[4] C. Martin, A. Maignan, M. Hervieu, B. Raveau, Magnetic phase diagrams of L1-xAxMnO3 manganites (L = Pr, Sm; A = Ca, Sr). Phys. Rev. B 60 (1999) 12191–12199.[5] C. N. R. Rao, A. K. Cheetham, R. Mahesh, Giant magetoresistance and related properties of rare-earth manganates and other oxide systems. Chem. Mater. 8 (1996) 2421–2432.[6] A. Moreo, S. Yunoki, E. Dagotto, Phase separation scenario for manganese oxides and related materials. Science 283 (1999) 2034–2040.[7] H. Kuwahara, Y. Tomioka, A. Asamitsu, Y. Moritomo, Y. Tokura, A first-order phase transition induced by a magnetic field. Science 270 (1997) 961–963.[8] A. A. Belik, R. D. Johnson, D. D. Khalyavin, The rich physics of A-site-ordered quadruple perovskite manganites AMn7O12. Dalton Trans. 50 (2021) 15458–15472.[9] A. Prodi, A. Daoud-Aladine, F. Gozzo, B. Schmitt, O. Lebedev, G. van Tendeloo, E. Gilioli, F. Bolzoni, H. Aruga-Katori, H. Takagi, M. Marezio, A. Gauzzi, Commensurate structural modulation in the charge- and orbitally ordered phase of the quadruple perovskite (NaMn3)Mn4O12. Phys. Rev. B 90 (2014) 180101.[10] R. D. Johnson, F. Mezzadri, P. Manuel, D. D. Khalyavin, E. Gilioli, P. G. Radaelli, Evolution of magneto-orbital order upon B-site electron doping in Na1−xCaxMn7O12 quadruple perovskite manganites. Phys. Rev. Lett. 120 (2018) 257202.[11] W.-T. Chen, C.-W. Wang, C.-C. Cheng, Y.-C. Chuang, A. Simonov, N. C. Brislowe, M. S. Senn, Striping of orbital-order with charge-disorder in optimally doped magnanites. Nat. Comm. 12 (2021) 6319.[12] A. A. Belik, Rise of A-site columnar-ordered A2A(A((B4O12 quadruple perovskites with intrinsic triple order. Dalton Trans. 47 (2018) 3209–3217.[13] A. M. Glazer, The classification of tilted octahedra in perovskites. Acta Cryst. B28 (1972) 3384–3392.[14] L. Zhang, Y. Matsushita, K. Yamaura, A. A. Belik, Five-fold ordering in high-pressure perovskites RMn3O6 (R = Gd-Tm and Y). Inorg. Chem. 56 (2017) 5210–5218.[15] A. A. Belik, L. Zhang, Y. Matsushita, Y. Katsuya, M. Tanaka, K. Yamaura, Crystal structures of cation non-stoichiometric RMn3O6 (R = Gd, Er, and Tm) manganites belonging to A-site columnar-ordered quadruple perovskite family. J. Solid State Chem. 275 (2019) 43–48.[16] A. M. Vibhakar, D. D. Khalyavin, P. Manuel, L. Zhang, K. Yamaura, P. G. Radaelli, A. A. Belik, R. D. Johnson, Magnetic structure and spin-flop transition in the A-site columnar-ordered quadruple perovskite TmMn3O6. Phys. Rev. B 99 (2019) 104424.[17] A. M. Vibhakar, D. D. Khalyavin, P. Manuel, N. J. Steinke, L. Zhang, K. Yamaura, A. A. Belik, R. D. Johnson, Magnetic inhomogeneities in the quadruple perovskite manganite [Y2−xMnx]Mn6O12. Phys. Rev. B 108 (2023) 054403.[18] A. M. Vibhakar, D. D. Khalyavin, P. Manuel, J. Liu, A. A. Belik R. D. Johnson, Spontaneous rotation of ferrimagnetism driven by antiferromagnetic spin canting. Phys. Rev. Lett. 124 (2020) 127201.[19] A. A. Belik, D. D. Khalyavin, Y. Matsushita, K. Yamaura, Triple A-site cation ordering in the ferrimagnetic Y2CuGaMn4O12 perovskite. Inorg. Chem. 61 (2022) 14428–14435.[20] S. Kawaguchi, M. Takemoto, K. Osaka, E. Nishibori, C. Moriyoshi, Y. Kubota, Y. Kuroiwa, K. Sugimoto, High-throughput powder diffraction measurement system consisting of multiple MYTHEN detectors at beamline BL02B2 of SPring-8. Rev. Sci. Instrum. 88 (2017) 085111.[21] F. Izumi, T. Ikeda, A Rietveld-analysis program RIETAN-98 and its applications to zeolites. Mater. Sci. Forum 321–324 (2000) 198–205.[22] J. Sanchez-Benitez, J. A. Alonso, M. J. Martinez-Lope, A. de Andres, M. T. Fernandez-Diaz, Enhancement of the Curie temperature along the perovskite series RCu3Mn4O12 driven by chemical pressure of R3+ cations (R = rare earths). Inorg. Chem. 49 (2010) 5679–5685.[23] D. Waroquiers, X. Gonze, G.-M. Rignanese, C. Welker-Nieuwoudt, F. Rosowski, M. Göbel, S. Schenk, P. Degelmann, R. André, R. Glaum, G. Hautier, Statistical analysis of coordination environments in oxides. Chem. Mater. 29 (2017) 8346–8360.[24] N. E. Brese, M. O’Keeffe, Bond-valence parameters for solids. Acta Crystallogr. B47 (1991) 192–197.[25] J. A. Alonso, M. J. Martıinez-Lope, M. T. Casais, M. T. Fernandez-Dıaz, Evolution of the Jahn–Teller distortion of MnO6 octahedra in RMnO3 perovskites (R = Pr, Nd, Dy, Tb, Ho, Er, Y): a neutron diffraction study. Inorg. Chem. 39 (2000) 917–923.[26] M. Tachibana, T. Shimoyama, H. Kawaji, T. Atake, E. Takayama-Muromachi, Jahn–Teller distortion and magnetic transitions in perovskite RMnO3 (R = Ho, Er, Tm, Yb, and Lu). Phys. Rev. B 75 (2007) 144425.[27] A. Munoz, M. T. Casais, J. A. Alonso, M. J. Martinez-Lope, J. L. Martinez, M. T. Fernandez-Diaz, Complex magnetism and magnetic structures of the metastable HoMnO3 perovskite. Inorg. Chem. 40 (2001) 1020–1028.[28] A. Dönni, V. Y. Pomjakushin, L. Zhang, K. Yamaura, A. A. Belik, Magnetic properties and ferrimagnetic structures of Mn self-doped perovskite solid solutions (Ho1(xMnx)MnO3. J. Alloys Compd. 857 (2021) 158230.[29] A. A. Belik, Anisotropic thermal expansion and a second-order charge order transition in the ferrimagnetic Dy2CuZnMn4O12 perovskite with triple A-site cation ordering. Inorg. Chem. 62 (2023) 20042–20049.[30] A. A. Belik, E. Takayama-Muromachi, Magnetic properties of BiMnO3 studied with dc and ac magnetization and specific heat. Inorg. Chem. 45 (2006) 10224–10229.Table 1. Structure Parameters of Gd2CuZnMn4O12 at 100 K (the first line for each site) and 295 K (the second line for each site) from synchrotron XRPD data. Site WP x y z Biso (Å2) Gd1 2a 0.250.25 0.250.25 0.7768(2)0.7766(2) 0.21(2)0.40(3) Gd2 2a 0.250.25 0.250.25 0.2783(2)0.2782(2) 0.28(3)0.51(3) Cu 2b 0.750.75 0.250.25 0.7317(4)0.7327(4) 0.62(4)0.92(5) Zn 2b 0.750.75 0.250.25 0.2466(4)0.2474(4) 0.01(2)0.21(3) Mn1 4c 0.50.5 00 00 0.04(4)0.13(4) Mn2 4d 00 0.50.5 0.50.5 0.35(5)0.54(5) O1 8g 0.4417(7)0.4407(7) 0.9391(7)0.9402(7) 0.2643(9)0.2627(10) 0.94(9)1.32(10) O2 4f 0.0611(10)0.0613(10) 0.250.25 0.0331(13)0.0331(12) 0.27(11)0.39(12) O3 4e 0.250.25 0.5415(11)0.5415(11) 0.9202(12)0.9187(12) 0.62(14)0.89(16) O4 4f 0.5426(11)0.5429(11) 0.250.25 0.4141(12)0.4151(12) 0.57(15)0.67(16) O5 4e 0.250.25 0.4355(10)0.4367(10) 0.5354(12)0.5351(12) 0.38(12)0.56(13)Source: synchrotron powder X-ray diffraction (( = 0.420186 Å); used d-space range: from 0.3663 Å to 12.04 Å. Crystal system: orthorhombic; space group Pmmn (No. 59, origin choice 2); Z = 2.The occupation factors of all the sites are unity. WP: Wyckoff position. Impurity: GdCu3Mn4O12-type with a weight fraction of 4.3 %.100 K: a = 7.26357(2) Å, b = 7.27678(2) Å, c = 7.83344(2) Å, and V = 414.0398(18) Å3; Rwp = 6.84 %, Rp = 4.94 %, RB = 2.70 %, and RF = 1.25 %; (cal = 6.860 g/cm3. 295 K: a = 7.27141(3) Å, b = 7.28702(3) Å, c = 7.83880(2) Å, and V = 415.354(3) Å3; Rwp = 6.80 %, Rp = 4.91 %, RB = 3.07 %, and RF = 1.73 %; (cal = 6.838 g/cm3. Table 2. Selected bond lengths (l (Å) < 2.8 Å), bond angles (deg), bond valence sum, BVS, and distortion parameters of MnO6, (, in Gd2CuZnMn4O12 at 295 K a Gd1–O5 (2Gd1–O3 (2Gd1–O2 (2Gd1–O1 (4BVS(Gd13+) 2.331(9)2.398(8)2.434(9)2.660(6)+3.33 Gd2–O2 (2Gd2–O4 (2Gd2–O5 (2Gd2–O1 (4BVS(Gd23+) 2.361(9)2.385(8)2.431(9)2.653(6)+3.30 Cu–O1 (4BVS(Cu2+) 1.961(4)+1.87 Zn–O4 (2Zn–O3 (2BVS(Zn2+) 1.999(9)2.001(9)+1.80 Mn1–O2 (2Mn1–O3 (2Mn1–O1 (2BVS(Mn13+)((Mn1–O)Mn1–O1–Mn2 (2Mn1–O2–Mn1Mn1–O3–Mn1 1.893(2)1.950(3)2.149(7)+3.2930.0(10–4145.17(9)148.38(9)137.59(9) Mn2–O5 (2Mn2–O1 (2Mn2–O4 (2BVS(Mn24+)((Mn2–O)Mn2–O4–Mn2Mn2–O5–Mn2 1.896(2)1.959(7)1.964(3)+3.642.6(10–4136.07(9)147.09(9)a BVS =    010203040050100150200250300ZFC, 100 OeFCC, 100 Oe0.00.51.01.52.02.50100200300400ZFC, 10 kOeFCC, 10 kOe-3-2-1012340100200300400ZFC, 10 kOe0306090120050100150200FCC, 5 OeFCC, 20 OeFCC, 100 Oec (emu´mol-1´Oe-1) Temperature (K) c (emu´mol-1´Oe-1) (b) (c) -350-250-150-5050050100150(a) ~ 48 K ~ 43 K ~ 40 K dcT/dT (emu´mol-1´Oe-1) dcT/dT (emu´mol-1´Oe-1) Temperature (K) Temperature (K) TC TC Timp Timp 5 Oe 20 Oe 100 Oe ZFC FCC ZFC , (i = exp[(R0 ( li)/B], N is the coordination number, B = 0.37, R0(Gd3+) = 2.065, R0(Cu2+) = 1.679, R0(Zn2+) = 1.704, R0(Mn4+) = 1.753, and R0(Mn3+) = 1.76 [24].Table 3. Structure parameters of Gd2CuZnMn4O12 at 600 K from synchrotron powder diffraction data. Atom WP g x y z Biso (Å2) Gd 4d 1 0.25 0.25 0.22259(5) 0.791(7) Cu 4c 0.5 0.75 0.25 0.7697(6) 1.26(7) Zn 2b 1 0.75 0.25 0.25 0.65(4) Mn 8e 1 0 0 0 0.591(12) O1 8g 1 0.25 0.0606(5) 0.9654(5) 0.90(7) O2 8g 1 0.25 0.5409(5) 0.5814(5) 1.23(9) O3 8f 1 0.4394(4) (x 0.25 1.74(10)Source: synchrotron powder X-ray diffraction (( = 0.420186 Å); used d-space range: from 0.3663 Å to 12.04 Å.Crystal system: tetragonal; space group: P42/nmc (No. 137, origin choice 2); Z = 2. Molecular weight: 855.1708 g/mol.a = 7.31430(2) Å, c = 7.82275(2) Å, and V = 418.510(2) Å3; Rwp = 6.79 %, Rp = 4.93 %, RB = 3.78 %, and RF = 3.29 %; (cal = 6.786 g/cm3.Table 4. Bond lengths (in Å; below 2.8 Å), bond angles (in deg), bond valence sum (BVS), and distortion parameters of MnO6 (() in Gd2CuZnMn4O12 at 600 K. Gd–O1 (2 2.351(4) Mn–O1 (2 1.901(1) Gd–O2 (2 2.397(4) Mn–O2 (2 1.959(2) Gd–O1 (2 2.443(4) Mn–O3 (2 2.054(1) Gd–O3 (4 2.669(1) ((MnO6) 10.2(10–4 BVS(Gd3+) +3.24 BVS(Mn3+) +3.44 Cu–O3 (4 1.966(4) Mn–O1–Mn (2 148.30(6) BVS(Cu2+) +1.84 Mn–O2–Mn (2 137.91(6) Zn–O2 (4 2.020(4) Mn–O3–Mn (2 144.48(6) BVS(Zn2+) +1.70  å=Nii1nFigure 1. (a) Experimental (black crosses) between 2( and 42(, calculated (red line), and difference (blue line at the bottom) synchrotron powder X-ray diffraction patterns of Gd2CuZnMn4O12 at T = 295 K. The tick marks show possible Bragg reflection positions for Gd2CuZnMn4O12 (the first row) and the impurity (the second row). The inset shows similar curves at T = 600 K between 2( and 17(. (b) The crystal structure at T = 295 K. (c) The crystal structure at T = 600 K.   7.817.827.837.8401002003004005006007008007.267.287.307.327.340100200300400500600700800Lattice parameters (Å) Lattice parameters (Å) Temperature (K) (b) (a) aO, Pmmn bO, Pmmn aT, P42/nmc aO bO aT cO cT cO, Pmmn cT, P42/nmc cT, P42/nmc cO, Pmmn aT, P42/nmc aO, Pmmn bO, Pmmn heating cooling TCO = 475 K heating cooling Figure 2. (a) Temperature dependence of the aO, bO, and aT lattice parameters and (b) temperature dependence of the cO and cT lattice parameters of Gd2CuZnMn4O12 between 100 K and 800 K on heating and cooling. O: orthorhombic, T: tetragonal.     13.013.213.420.620.821.021.214.514.714.92q  (deg): l = 0.420186 Å 500 K 450 K 100 K 200 K 295 K 350 K 400 K 400 402 412 420 602 620 400 040 062 602 260 620 042 402 240 420 142 412 010002000300040002.02.53.03.54.0Intensity (counts) 001 100 K, 200 s 295 K, 200 s 600 K, 400 s 200 K, 10 s 400 K, 10 s ´8 ´13 O T O T (a) (b) (c) (d) O O O O ~O O O O Figure 3. Temperature dependence of the unit-cell volume of Gd2CuZnMn4O12 between 100 K and 800 K on heating and cooling. (V is the volumetric coefficient of thermal expansion calculated in the P42/nmc phase between 500 K and 800 K using the cooling data.   0.91.11.31.51.70501001502000 Oe90 kOe0.00.51.01.52.02.53.03.502040600 Oe10 kOe30 kOe50 kOe70 kOe90 kOeCp / T (J´K-2´mol-1) Cp / T (J´K-2´mol-1) Temperature (K) (a) (b) TC = 105 K 45 K 20 K Figure 4. (a) Temperature dependence of the Mn–O bond lengths in Gd2CuZnMn4O12 from 100 K to 800 K. The bond lengths were obtained in the Pmmn model between 100 K and 450 K and in the P42/nmc model between 475 K and 800 K. The heating SXRD data were used. (b) Temperature dependence of the Cu–O and Zn–O bond lengths.  1.901.952.002.052.100100200300400500600700800Cu-O1Zn-O3Zn-O4Cu-O3Zn-O21.851.901.952.002.052.102.150100200300400500600700800Mn1-O2Mn1-O3Mn1-O1Mn2-O5Mn2-O4Mn2-O1Mn-O1Mn-O2Mn-O3P42/nmc Pmmn Temperature (K) (a) (b) Bond Lengths (Å) Bond Lengths (Å) Gd2CuZnMn4O12 Mn3+ Mn4+ Mn3.5+ TCO = 475 K Mn–O3 Mn–O2 Mn–O1 Zn–O2 Cu–O3 Mn1–O1 Figure 5. (a, b, c) Temperature dependence of (normalized) experimental synchrotron X-ray powder diffraction patterns of Gd2CuZnMn4O12 from 100 K to 500 K. Zoomed details in selected regions are shown to emphasize the disappearance of the orthorhombic distortion. Indexes of some reflections are given. (d) Zoomed details of synchrotron X-ray powder diffraction patterns at different temperatures between 2( and 4(. O: orthorhombic. T: tetragonal. Measurement times are given.  2q  (deg): l = 0.420186 Å Intensity (counts/106) -0.10.00.10.20.30.40.5212223242-0.20.00.20.40.60.81.01.2261014T = 295 K Pmmn T = 600 K P42/nmc (b) 295 K Gd1 Gd2 Mn1 Mn2 (c) 600 K b c Mn Zn Cu (a) Figure 6. ZFC (filled symbols) and FCC (empty symbols) dc magnetic susceptibility (( = M/H) curves of Gd2CuZnMn4O12 measured at (a) 100 Oe and (b) 10 kOe. (c) FCC ( versus T curves at 5 Oe, 20 Oe, and 100 Oe. Insets on panels (a) and (b) show differential d(T/dT versus T curves to emphasize the phase transition temperatures. Timp: a magnetic transition temperature of the impurity.   0.00.40.81.21.60204060801000500100015002000Remnant Magnetization  (mB / f.u.) Temperature (K) Coercive Field (Oe) Figure 7. (a) M versus H curves of Gd2CuZnMn4O12 at 5 K, 25 K, 60 K, and 100 K (f.u.: formula unit). (b) Zoomed parts of the same M versus H curves near the origin.    051015200204060801001201400.5 Hz2 Hz7 Hz25 Hz110 Hz500 Hz-0.10.10.30.50.70.90204060801001201400.5 Hz2 Hz7 Hz25 Hz110 Hz500 Hz051015200204060801001201400.1 Oe0.5 Oe5 Oe-0.20.20.61.01.41.82.22.60204060801001201400.1 Oe0.5 Oe5 Oec ¢ (emu´mol-1´Oe-1) c ¢¢ (emu´mol-1´Oe-1) Temperature (K) Hac = 0.5 Oe Hdc = 0 Oe Hac = 0.5 Oe Hdc = 0 Oe (a) (b) (c) (d) f = 300 Hz Hdc = 0 Oe f = 300 Hz Hdc = 0 Oe Hac =  Hac =  f =  f =  Figure 8. (a) Zoomed parts of the Cp/T versus T curves of Gd2CuZnMn4O12 at H = 0 Oe and 90 kOe to emphasize weak anomalies near 45 K and 105 K. (b) The Cp/T versus T curves at different magnetic fields below 70 K.  -4-2024-8-4048-14-12-10-8-6-4-202468101214-80-60-40-200204060805 K25 K60 K100 KMagnetization  (mB / f.u.) Magnetization  (mB / f.u.) Magnetic Field (kOe) (a) (b) 100 K 100 K 5 K 5 K 25 K 25 K 60 K 60 K Figure 9. (a) The (( versus T and (b) ((( versus T curves of Gd2CuZnMn4O12 at different frequencies measured with Hac = 0.5 Oe and Hdc = 0 Oe. (c) The (( versus T and (d) ((( versus T curves at different Hac = 0.1, 0.5, and 5 Oe and one frequency of 300 Hz (Hdc = 0 Oe).  4144154164174184194204210100200300400500600700800VO, Pmmn VT, P42/nmc VT, P42/nmc VO, Pmmn heating cooling Temperature (K) Unit cell volume (Å3) TCO = 475 K aV = 1.8(2)´10-5 K-1 800–500 K cooling Figure 10. Temperature dependence of the remnant magnetization (black circles, the left-hand axis) and absolute values of the coercive field (red squares, the right-hand axis) in Gd2CuZnMn4O12.PAGE  1_1209387283.unknown