High-Pressure Synthesis of a Novel Quadruple Perovskite Oxide AgCu₃Cr₄O₁₂

Abstract

A novel complex transition metal oxide AgCu₃Cr₄O₁₂ has been obtained using high-pressure and high-temperature conditions of 12 GPa and 1223 K. The crystal structure is refined to be a cubic AA′₃B₄O₁₂-type quadruple perovskite structure based on the Rietveld refinement of the synchrotron X-ray powder diffraction data. The density-functional theory calculation obtains a metallic band structure. The valence state is estimated to be Ag^∼1.3+Cu^∼2.2+₃Cr⁴⁺₄O₁₂ by bond valence sum and X-ray absorption spectroscopy analyses. The valence state on ACu₃Cr₄O₁₂ series (A = Ag, Ca, La, Ce) sequentially transforms from Ag^∼1.3+Cu^∼2.2+₃Cr⁴⁺₄O₁₂, Ca²⁺Cu²⁺₃Cr⁴⁺₄O₁₂, La³⁺Cu^(2+δ)+₃Cr^{(3.75−0.75δ)+}₄O₁₂, to Ce⁴⁺Cu²⁺₃Cr^3.5+₄O₁₂, where the electrons are doped into the A′-site (Cu^∼2.2+ → Cu²⁺), followed by predominant doping into the B-site (Cr⁴⁺ → Cr^{(3.75−0.75δ)+} → Cr^3.5+). Their electron doping sequence is distinguished from those reported in other quadruple perovskite oxides, proposing characteristic features of the ACu₃Cr₄O₁₂ family.

1. Introduction

Quadruple perovskite oxide AA′₃B₄O₁₂ has been investigated as one of the perovskite-related oxides. Its structure is derived from the 1:3-type A-site cationic ordering, in which the conventional A-site ions (alkali, alkaline earth, or rare earth metals) and Jahn-Teller active transition metals such as Cu²⁺ and Mn³⁺ ions are spatially ordered (Fig. 1). The electronic interactions between the transition metal ions at the A′- and B-sites realize various functional properties.^1–8) The aliovalent A-site substitutions induce drastic changes in electronic properties by electron/hole-doping into the A′- and/or B-site metals. CaCu₃Cr₄O₁₂ consisting of Cu²⁺(3d⁹) and Cr⁴⁺(3d²) ions, is an antiferromagnetic metal,^9,10) whereas LaCu₃Cr₄O₁₂ undergoes an antiferromagnetic transition simultaneously with charge transfer (3Cu^(2+δ)+ + 4Cr^{(3.75−0.75δ)+} $ \rightleftarrows $ 3Cu^(3−γ)+ + 4Cr^(3+0.75γ)+) at 220 K,¹¹⁾ followed by the ferrimagnetic metallic properties for Ce⁴⁺Cu²⁺₃Cr^3.5+₄O₁₂ (T_C = 333 K).¹²⁾ In their electron doping sequence, the electrons are primarily doped into the B-site Cr ions. However, the monovalent metal (A⁺) substitution for the A-site and its carrier-doping manner was not examined yet.

Fig. 1

Schematic of the crystal structure of AgCu₃Cr₄O₁₂.

This article reports the high-pressure synthesis of a novel quadruple perovskite AgCu₃Cr₄O₁₂, a missing piece in the ACu₃Cr₄O₁₂ family. The valence state of AgCu₃Cr₄O₁₂ was estimated to be Ag^∼1.3+Cu^∼2.2+₃Cr⁴⁺₄O₁₂ by using bond valence sum and X-ray absorption spectroscopy. The overall valence state transitions in the ACu₃Cr₄O₁₂ family were interpreted as the sequential electron doping into A′-site Cu (AgCu₃Cr₄O₁₂ → CaCu₃Cr₄O₁₂) and afterward mainly into B-site Cr (CaCu₃Cr₄O₁₂ → LaCu₃Cr₄O₁₂ → CeCu₃Cr₄O₁₂), which is distinguishable from those in other quadruple perovskite oxides reported so far.

2. Experimental

The raw regents of AgO (99%), CuO (99.9%), and CrO₂ (99%) were mixed using a mortar at a 1:3:4 molar ratio to prepare the starting mixture with stoichiometric composition of AgCu₃Cr₄O₁₂ based on the chemical reaction: AgO + 3CuO + 4CrO₂ → AgCu₃Cr₄O₁₂.

The mixture was filled into a Pt capsule, and the Pt capsule was placed in an octahedron-shaped pressure-transmitting medium made of (Mg, Co)O with an edge length of 14 mm. The high-pressure cell was surrounded by eight WC anvils with a truncation of 8 mm. The high-pressure cell was compressed up to 12 GPa using a Walker-type high-pressure apparatus. After the compression, the sample was heated to 1223 K for 15 min and held at this temperature for 30 min. The pressure was slowly released after the heat treatment, and a dense pellet of AgCu₃Cr₄O₁₂ was obtained. The reference sample of CaCu₃Cr₄O₁₂ was synthesized according to the reference.^9,10)

Synchrotron X-ray powder diffraction (SXRD) measurement was conducted at the BL02B2 beamline of SPring-8.¹³⁾ The powder sample obtained by crushing the pellet of AgCu₃Cr₄O₁₂ was filled into a Lindemann glass capillary with an inner diameter of 0.2 mm. The wavelength was determined to be 0.42034 Å using the standard CeO₂ reference. Structural parameters were refined using the RIETAN-FP program.¹⁴⁾ The crystal structure model was illustrated using the VESTA-3 program.¹⁵⁾ X-ray absorption spectra at Cu and Cr K-edges were collected in the transmission mode at the BL14B2 beamline of SPring-8. Electric resistivity and magnetization measurements were performed on the pellet sample using a Quantum Design Physical Properties Measurement System (PPMS, EverCool II, Quantum Design Inc.). Electric resistivity measurement was performed in a temperature range of 5–300 K using the DC four-probe method. Magnetic susceptibility data were obtained between 5 and 400 K in an external field of 1 kOe on field-cooling mode using a vibrating sample magnetometer (VSM) option. The isothermal magnetization curves were collected at 5 and 300 K in external magnetic fields between −50 and 50 kOe.

Density-functional theory (DFT) calculation was conducted using the WIEN2K package,¹⁶⁾ implementing the augmented plane wave and local orbital method. To include the electron-electron interaction effect on the Cu and Cr 3d shell, we employed a local-density approximation LDA+U scheme. The experimental structure of AgCu₃Cr₄O₁₂ was used, and the effective Hubbard parameters U_eff were set to 7.0 eV for Cu 3d and 3.0 eV for Cr 3d orbitals. The fully localized limit scheme was employed for the double counting.¹⁷⁾ We simulated an antiferromagnetic state, in which the Cu and Cr spins of the distinct magnetic sublattices were respectively coupled antiferromagnetically between the nearest-neighboring Cu and Cr sites. The spin polarization of Cr spins was retained in the converged state whereas that of Cu spins was lost.

3. Results and Discussion

AgCu₃Cr₄O₁₂ was successfully synthesized using high-pressure and high-temperature conditions of 12 GPa and 1223 K. Figure 2 shows the final Rietveld refinement result of the SXRD pattern at room temperature. The Bragg reflections of the primary phase were indexed by a cubic AA′₃B₄O₁₂-type quadruple perovskite structure with the $Im\bar{3}$ space group (No. 204), although CuO (21.2 mass%) and other unknown impurity phases were found. Although CuO was incorporated as the second phase in the refinement, the 2θ ranges, including unidentified Bragg reflections, were excluded from the Rietveld refinement. The structure parameters obtained from the final refinement are listed in Table 1. In addition to the small deviation between observed and calculated patterns in Fig. 2, the small reliability factors (R_wp = 4.948% and R_B = 6.041%) ensure that the refinement result is reasonable. The lattice constant a = 7.22361(6) Å of AgCu₃Cr₄O₁₂ was slightly smaller than that of CaCu₃Cr₄O₁₂ (7.23728 Å).¹⁰⁾ The Ag–O bond for AgCu₃Cr₄O₁₂ (2.608(5) Å) was longer than the Ca–O bond for CaCu₃Cr₄O₁₂ (2.556 Å). This is reasonable, considering the difference in Shannon’s effective ionic radius of Ag⁺ (1.28 Å, CN = 8) and Ca²⁺ (1.12 Å, CN = 8).¹⁸⁾ The Cu–O bond length for AgCu₃Cr₄O₁₂ (1.901(3) Å) was slightly smaller than that for CaCu₃Cr₄O₁₂ (1.926 Å), while the Cr–O bond lengths were almost identical (1.9205(11) and 1.9191 Å for AgCu₃Cr₄O₁₂ and CaCu₃Cr₄O₁₂, respectively). However, since the slight changes in the Cu–O bond length did not ensure the hole doping into the Cu site, the bond valence sums (BVSs) were calculated from the metal-oxygen bond lengths by using the following bond-valence parameters: b₀ = 0.37 Å for all atoms, r₀ = 1.805 Å for Ag⁺, r₀ = 1.679 Å for Cu²⁺, and r₀ = 1.81 Å for Cr⁴⁺.¹⁹⁾ The BVS of Ag for AgCu₃Cr₄O₁₂ was +1.37, almost the same as that (+1.33) for AgCu₃V₄O₁₂. This suggests that the Ag ions in AgCu₃Cr₄O₁₂ possess intermediate valences of Ag^(1+δ)+ as in AgCu₃V₄O₁₂.²⁰⁾ The BVS of Cr (+4.45) for AgCu₃Cr₄O₁₂ was also almost identical to that for CaCu₃Cr₄O₁₂ (+4.47), but these values are much larger than that expected for the Cr⁴⁺ ions. This is probably because of the inadequate bond valence parameters for Cr⁴⁺ ions. The fact that the BVS (+4.61) of Cr for CaCr⁴⁺O₃ calculated using the same bond-valence parameters has heavily deviated from the ideal value (+4)²¹⁾ supports the inadequacy in the bond-valence parameter for Cr⁴⁺. Anyway, the insignificant difference in the BVS value of Cr between AgCu₃Cr₄O₁₂ and CaCu₃Cr₄O₁₂ presumed that the Cr valences are almost the same for AgCu₃Cr₄O₁₂ and CaCu₃Cr₄O₁₂. The BVS of Cu in AgCu₃Cr₄O₁₂ was calculated to be +2.28, which is slightly larger than that (+2.12) in CaCu₃Cr₄O₁₂. This estimate that the holes generated by the Ag^(1+δ)+ substitution for Ca²⁺ were predominantly doped into the Cu ions. Accordingly, the BVS analysis implied the valence state transition from Ca²⁺Cu²⁺₃Cr⁴⁺₄O₁₂ to Ag^(1+x)+Cu^(2+y)+₃Cr⁴⁺₄O₁₂ (x + 3y = 1).

Fig. 2

Rietveld refinement result of the SXRD pattern for AgCu₃Cr₄O₁₂. The circles (black) and solid lines (red) represent the observed and calculated patterns, respectively. The difference between the observed and calculated patterns is shown at the bottom (blue). The vertical marks indicate the Bragg reflection positions of AgCu₃Cr₄O₁₂ (green) and CuO (orange).

Table 1 Structure parameters for AgCu₃Cr₄O₁₂ at room temperature obtained from the Rietveld refinement of SXRD data.

X-ray absorption spectroscopy measurements were conducted for further examination of the valence state. Figure 3(a) shows the X-ray absorption spectrum at Cr K-edge for AgCu₃Cr₄O₁₂. The onset of the absorption of AgCu₃Cr₄O₁₂ (∼6000 eV) was almost identical to that of CaCu²⁺₃Cr⁴⁺₄O₁₂, confirming that the Cr⁴⁺ valence state is predominant for both compounds. Figure 3(b) displays the X-ray absorption spectrum at Cu K-edge for AgCu₃Cr₄O₁₂. The overall spectrum of AgCu₃Cr₄O₁₂ was slightly shifted to the higher energy by ∼0.25 eV from that of CaCu₃Cr₄O₁₂. The edge shift indicates that the Cu valence in AgCu₃Cr₄O₁₂ is higher than divalent in CaCu₃Cr₄O₁₂, although the magnitude of the energy shift is underestimated because of the inclusion of a substantial amount of divalent copper oxide (CuO) in the present sample. If we assume the identical Ag valence for AgCu₃Cr₄O₁₂ and AgCu₃V₄O₁₂ (∼1.3), the electric neutrality conditions suggest the valence state of Ag^∼1.3+Cu^∼2.2+₃Cr⁴⁺₄O₁₂.

Fig. 3

X-ray absorption spectra at (a) Cr and (b) Cu K-edges for AgCu₃Cr₄O₁₂. The insets show the enlarged spectra near the absorption edge.

Figure 4 illustrates the temperature dependence of the electric resistivity for AgCu₃Cr₄O₁₂. The resistivity gradually increased on cooling, as typically observed in insulators. However, the change was tiny and included within one order of magnitude; thus, the insulating nature is not intrinsic and probably originates from the impurities of CuO and unidentified phases.

Fig. 4

Temperature dependence of the electric resistivity for AgCu₃Cr₄O₁₂.

The magnetic properties of AgCu₃Cr₄O₁₂ were also difficult to evaluate because of ferromagnetic impurities. Figure 5(a) shows the temperature dependence of the magnetic susceptibility for AgCu₃Cr₄O₁₂. Since the magnetic susceptibility was larger than that of paramagnet (∼10⁻³ emu mol⁻¹), a ferromagnetic component was predominant, as confirmed in the isothermal magnetization curves (Fig. 5(b)). The gradual decrease in the susceptibility likely dropped to near zero above 400 K, estimating the Curie temperature close to 400 K. Figure 5(b) displays the isothermal magnetization curves for AgCu₃Cr₄O₁₂. The ferromagnetic component was observed at all the temperatures measured between 5 and 400 K. The saturation magnetization at 5 K was relatively small (∼5 emu g⁻¹), which did not reach the value (∼40 emu g⁻¹) expected from the ferrimagnetism with the antiparallel alignment between Cu²⁺(S = 1/2)↓ and Cr⁴⁺ (S = 1)↑ for AgCu₃Cr₄O₁₂. Therefore, we expect that the intrinsic magnetic property of AgCu₃Cr₄O₁₂ is probably antiferromagnetic like CaCu₃Cr₄O₁₂ and LaCu₃Cr₄O₁₂ rather than ferrimagnetism like CeCu₃Cr₄O₁₂.^9,12) However, further study using microscopic probes such as µSR⁹⁾ is needed to elucidate the magnetic ground state for AgCu₃Cr₄O₁₂.

Fig. 5

(a) Temperature dependence of the magnetic susceptibility for AgCu₃Cr₄O₁₂ measured in an external field of 1000 Oe on field cooling. (b) Isothermal magnetization curves at 5, 100, 200, and 300 K for AgCu₃Cr₄O₁₂.

An LDA+U calculation was performed using the experimentally obtained crystal structure of AgCu₃Cr₄O₁₂, in which an antiferromagnetic structure was assumed. The electronic density of states (DOS) in Fig. 6 clearly shows a metallic feature. The metallic electronic structure was robust for the choice of the Hubbard parameters U_eff. The DOS near Fermi energy (E_F) predominated the Cr 3d character, while the Cu 3d and O 2p states were also found. Besides, the Ag 4d states partially contributed to the spectral weights just below and above E_F, which is consistent with the mixed valence state of Ag mentioned above.

Fig. 6

(a) Total density of states (Total DOS) and (b) partial DOS (PDOS) projected on Ag 4d, Cu 3d, Cr 3d, and O 2p states of AgCu₃Cr₄O₁₂ simulated by LDA+U calculation. The DOS of the down spin component is multiplied by −1. The energy origin is set to the Fermi level (dashed line).

In the present study, the novel Ag-containing quadruple perovskite oxide AgCu₃Cr₄O₁₂ was synthesized under high pressure, possessing a metallic electronic structure as well as CaCu₃Cr₄O₁₂. We discuss the electron doping sequence in the ACu₃Cr₄O₁₂ series (A = Ag^∼1.3+, Ca²⁺, La³⁺, Ce⁴⁺), such as ACu₃Mn₄O₁₂ (A = Ca²⁺, La³⁺, Ce⁴⁺; Ca²⁺Cu²⁺₃Mn⁴⁺₄O_12. → La³⁺Cu²⁺₃Mn^3.75+₄O₁₂ → Ce⁴⁺Cu²⁺₃Mn^3.5+₄O₁₂)^1,2,22) and ACu₃Co₄O₁₂ (A = Ca²⁺, Y³⁺, Ce⁴⁺; Ca²⁺Cu³⁺₃Co^3.25+₄O_12. → Y³⁺Cu³⁺₃Co³⁺₄O₁₂ → Ce⁴⁺Cu^2.67+₃Co³⁺₄O₁₂) families.^23,24) The ionic models of ACu₃Cr₄O₁₂ at room temperature are represented as Ag^∼1.3+Cu^∼2.2+₃Cr⁴⁺₄O₁₂, Ca²⁺Cu²⁺₃Cr⁴⁺₄O₁₂, La³⁺Cu^(2+δ)+₃Cr^{(3.75−0.75δ)+}₄O₁₂, and Ce⁴⁺Cu²⁺₃Cr^3.5+₄O₁₂.^9,12) The aliovalent substitutions for A-site ions from Ag^∼1.3+ to Ca²⁺ to La³⁺ to Ce⁴⁺ sequentially dope the electrons into A′-site Cu and B-site Cr, which can be featured by the clear difference from those reported in other ACu₃B₄O₁₂ oxides. On the other hand, the electronic properties are not substantially changed between AgCu₃Cr₄O₁₂ and CaCu₃Cr₄O₁₂, indicating the robustness of antiferromagnetic metallic nature against the electron doping into A′-site Cu ions, although further electron doping into B-site Cr induced the ferrimagnetic property in CeCu₃Cr₄O₁₂.

4. Conclusions

We synthesized the quadruple perovskite oxide AgCu₃Cr₄O₁₂ by using the high-pressure and high-temperature conditions of 12 GPa and 1223 K. The bond valence sums and X-ray absorption spectroscopy revealed the valence state of Ag^∼1.3+Cu^∼2.2+₃Cr⁴⁺₄O₁₂, which was derived from Ca²⁺Cu²⁺₃Cr⁴⁺₄O₁₂ by hole doping into A′-site Cu. The intrinsic metallic electronic structure was unveiled by the DFT calculation. The sequential electron doping for the ACu₃Cr₄O₁₂ system is distinguishable from the monotonical electron doping into only B-site metals for the other quadruple perovskite oxides.

Acknowledgments

The authors thank Fuminari Fujii for the support of sample preparation. The synchrotron X-ray experiment was performed at SPring-8 under the approval of the JASRI (proposal numbers 2022A1493 and 2022B1619). This work was supported by JSPS KAKENHI (grant number JP20H02825) and Tanikawa Fund Promotion of Thermal Technology.

REFERENCES

• 1)   Z.  Zeng,  M.  Greenblatt,  M.A.  Subramanian and  M.  Croft: Phys. Rev. Lett. 82 (1999) 3164–3167. doi:10.1103/PhysRevLett.82.3164
• 2)   J.A.  Alonso,  J.  Sánchez-Benítez,  A.  de Andrés,  M.J.  Martínez-Lope,  M.T.  Casais and  J.L.  Martínez: Appl. Phys. Lett. 83 (2003) 2623. doi:10.1063/1.1611647
• 3)   I.  Yamada,  K.  Takata,  N.  Hayashi,  S.  Shinohara,  M.  Azuma,  S.  Mori,  S.  Muranaka,  Y.  Shimakawa and  M.  Takano: Angew. Chem. Int. Ed. 47 (2008) 7032–7035. doi:10.1002/anie.200801482
• 4)   I.  Yamada,  K.  Tsuchida,  K.  Ohgushi,  N.  Hayashi,  J.  Kim,  N.  Tsuji,  R.  Takahashi,  M.  Matsushita,  N.  Nishiyama,  T.  Inoue,  T.  Irifune,  K.  Kato,  M.  Takata and  M.  Takano: Angew. Chem. Int. Ed. 50 (2011) 6579–6582. doi:10.1002/anie.201102228
• 5)   Y.W.  Long,  N.  Hayashi,  T.  Saito,  M.  Azuma,  S.  Muranaka and  Y.  Shimakawa: Nature 458 (2009) 60–63. doi:10.1038/nature07816
• 6)   I.  Yamada et al.: Inorg. Chem. 52 (2013) 13751–13761. doi:10.1021/ic402344m
• 7)   H.  Etani,  I.  Yamada,  K.  Ohgushi,  N.  Hayashi,  Y.  Kusano,  M.  Mizumaki,  J.  Kim,  N.  Tsuji,  R.  Takahashi,  N.  Nishiyama,  T.  Inoue,  T.  Irifune and  M.  Takano: J. Am. Chem. Soc. 135 (2013) 6100–6106. doi:10.1021/ja312015j
• 8)   I.  Yamada,  Y.  Kato,  H.  Nakajima,  H.  Ikeno,  S.  Mori and  S.  Kawaguchi: Mater. Trans. advance publication, doi:10.2320/matertrans.MT-MG2022005.
• 9)   J.  Sugiyama,  H.  Nozaki,  I.  Umegaki,  K.  Miwa,  W.  Higemoto,  E.J.  Ansaldo,  J.H.  Brewer,  H.  Sakurai,  M.  Isobe,  H.  Takagi and  M.  Månsson: Phys. Rev. B 97 (2018) 024416. doi:10.1103/PhysRevB.97.024416
• 10)   M.A.  Subramanian,  W.J.  Marshall,  T.G.  Calvarese and  A.W.  Sleight: J. Phys. Chem. Solids 64 (2003) 1569–1571. doi:10.1016/S0022-3697(03)00095-7
• 11)   S.  Zhang,  T.  Saito,  M.  Mizumaki and  Y.  Shimakawa: Chem. Eur. J. 20 (2014) 9510–9513. doi:10.1002/chem.201403692
• 12)   A.A.  Belik,  Y.  Katsuya,  M.  Tanaka and  K.  Yamaura: J. Alloy. Compd. 793 (2019) 42–48. doi:10.1016/j.jallcom.2019.04.128
• 13)   S.  Kawaguchi,  M.  Takemoto,  K.  Osaka,  E.  Nishibori,  C.  Moriyoshi,  Y.  Kubota,  Y.  Kuroiwa and  K.  Sugimoto: Rev. Sci. Instrum. 88 (2017) 085111. doi:10.1063/1.4999454
• 14)   F.  Izumi and  K.  Momma: Solid State Phenom. 130 (2007) 15–20. doi:10.4028/www.scientific.net/SSP.130.15
• 15)   K.  Momma and  F.  Izumi: J. Appl. Crystallogr. 44 (2011) 1272–1276. doi:10.1107/S0021889811038970
• 16)   P.  Blaha,  K.  Schwarz,  F.  Tran,  R.  Laskowski,  G.K.H.  Madsen and  L.D.  Marks: J. Chem. Phys. 152 (2020) 074101. doi:10.1063/1.5143061
• 17)   M.T.  Czyyk and  G.A.  Sawatzky: Phys. Rev. B 49 (1994) 14211–14228. doi:10.1103/PhysRevB.49.14211
• 18)   R.D.  Shannon: Acta Crystallogr. Sec. A 32 (1976) 751–767. doi:10.1107/S0567739476001551
• 19)   I.D.  Brown and  D.  Altermatt: Acta Crystallogr. Sec. B 41 (1985) 244–247. doi:10.1107/S0108768185002063
• 20)   Y.  Akizuki,  I.  Yamada,  K.  Fujita,  H.  Akamatsu,  T.  Irifune and  K.  Tanaka: Inorg. Chem. 52 (2013) 13824–13826. doi:10.1021/ic402579v
• 21)   C.A.  Juillerat,  Y.  Tsujimoto,  A.  Chikamatsu,  Y.  Masubuchi,  T.  Hasegawa and  K.  Yamaura: Dalton Trans. 49 (2020) 1997–2003. doi:10.1039/C9DT04321G
• 22)   J.  Sánchez-Benítez,  M.J.  Martínez-Lope and  J.A.  Alonso: J. Appl. Phys. 107 (2010) 103904. doi:10.1063/1.3369444
• 23)   I.  Yamada,  S.  Ishiwata,  I.  Terasaki,  M.  Azuma,  Y.  Shimakawa and  M.  Takano: Chem. Mater. 22 (2010) 5328–5332. doi:10.1021/cm101671j
• 24)   I.  Yamada,  T.  Odake,  A.  Tanaka,  Y.  Okazaki,  F.  Toda,  Y.  Ishii,  T.  Taniguchi,  S.  Kawaguchi and  A.  Hariki: Inorg. Chem. 59 (2020) 8699–8706. doi:10.1021/acs.inorgchem.0c00184