High-pressure high-temperature preparation of CeGe₃

1 Introduction

The investigations into binary metal silicides and germanides, with metals being either alkali, alkaline-earth or rare-earth elements, have produced a variety of semiconducting Zintl phases [1, 2]. Interestingly, the use of high-pressure synthesis techniques allows amongst others for the preparation of compounds which violate the electron counting rules of the Zintl-Klemm concept. A large variety of compounds with the composition MT₃ (M = alkaline-earth or rare-earth metal, T = Si or Ge [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23]) have been prepared under elevated pressure and shown to exhibit a range of structural motifs which break the 8−N rule or contain diamagnetic metal cations and exhibit superconducting properties. One of the more recent examples, SmGe₃ [24], was initially thought to adopt a Cu₃Au−type structure, but upon further investigations was shown to in fact be a 2 × 2 × 2 superstructure of it. Here, we report the superstructure for CeGe₃, for which an earlier study reported a Cu₃Au-type subcell [25].

2 Experimental

CeGe₃ was synthesized under high-pressure high-temperature conditions. An intermediate was prepared by arc melting a mixture of Ce (ChemPur, 99.99%) and Ge (ChemPur, 99.99%) in the molar ratio of 1:4, and grinding the product to a powder before filling it into a BN crucible and transferring it into a MgO octahedron with an edge length of 18 mm. This octahedron was inserted into a Walker-type module [26] and subjected to 3 GPa of pressure and heated to 1873 K at a rate of 50 K/min. Under these conditions the sample was kept for 30 min. The reaction temperature was then lowered to 1173 K at a rate of 70 K/min, and the sample was annealed for 3 h and quenched under load. Optimization of the synthesis procedure was performed to minimize the presence of starting material and side phases. Although the samples were kept under argon except during transfer to the press and the actual high-pressure reaction, small amounts of β-Ce₂O₃ were always present in the reaction product. Calibration of pressure and temperature was conducted prior to the experiments via observations of resistivity changes in bismuth [27] and thermocouple calibrated runs. Sample handling, apart from the high pressure and high temperature synthesis, was conducted in argon-filled gloveboxes (MBraun, H₂O and O₂ < 0.1 ppm).

3 Sample characterization

Characterization of the samples was conducted by powder X-ray diffraction using a Huber Image Plate Guinier Camera G670 (Huber Diffraktionstechnik GmbH & Co. KG, Rimsting, Germany), using CuK_α1 radiation (λ = 1.54056 Å). Indexing of diffraction reflections, lattice parameter determination and structure refinements were performed with the WinCSD program package [28].

Microstructure analyses have been performed on grinded and polished cross-sections by visual (Zeiss optical microscope Axioplan2) and scanning electron microscopy (Jeol JSM 7800F). Back scattered electrons (BSE) as well as the intensities of the X-ray lines [GeK_α (9.86 keV), CeL_α (4.84 keV) and OK_α (0.5 keV)] were used to visualize the phase distribution and composition inhomogeneities by the material contrast in the element maps. X-ray intensities were recorded by the energy dispersive X-ray detector (silicon drift detector (SDD) of the Esprit Vers 2.3. system, Bruker) which is attached to the SEM. The SEM analyses have been realized with an acceleration voltage of 22 kV.

Differential scanning calorimetry experiments were carried out using a Netzsch DSC 404 C device (Netzsch-Gerätebau GmbH, Selb, Germany) with heating and cooling rates of 10 K/min, under Ar atmosphere, using corundum crucibles.

4 Crystal structure and properties

Differential scanning calorimetry investigations showed that CeGe₃ is metastable at ambient conditions (Figure 1). The exothermic peak at 520 K is assigned to the decomposition of the product into CeGe_2–x [29] and elemental Ge (Figure 2). Metallographic analysis of samples of reaction products indicated the presence of the majority target phase exhibiting an average chemical composition of CeGe_2.99(9), obtained by energy dispersive X-ray spectroscopy measurements. The finding agrees remarkably well to the composition CeGe₃. Besides the presence of unreacted Ge, a second side phase was detected, which consists of cerium and oxygen only according to the element mapping (Figure 3).

Figure 1:

Differential scanning calorimetry data of CeGe₃ at ambient pressure. In direction of increasing temperature, an exothermic effect is observed at 520(10) K, which is attributed to the decomposition of the phase hp-CeGe₃ to CeGe_2–x and Ge. The effect at 770(10) K upon cooling is assigned to the transition from β-CeGe_2–x to α-CeGe_2–x. It exhibits a pronounced hysteresis with respect to the reported transition temperature. Red curve – heating, blue curve – cooling.

Figure 2:

Powder X-ray diffraction pattern (CuK_α1 radiation) of a sample with composition Ce₂₅Ge₇₅ after DSC up to 1100 K. Black – experimental pattern, red – calculated diffraction pattern for CeGe_1.75, blue – calculated pattern of elemental germanium.

Figure 3:

Element mappings of a polished sample (grey) for cerium (red), germanium (yellow) and oxygen (blue). The results of the measurements confirm the existence of the desired product CeGe₃, excess Ge and the impurity β-Ce₂O₃.

Consistent with these observations, the powder X-ray diffraction pattern shows strong reflections related to the Cu₃Au-type structure [30]. This has been observed previously in other MGe_3–x phases (x ≈ 0.15: M = Tb, Dy, Yb [31]; x = 0: M = Sm [24], Ce [25], Np [32], Pu [33] and U [34]). In addition, several reflections not belonging to the Cu₃Au-type structure could be assigned to unreacted Ge and to the by-product β-Ce₂O₃ (Figure 4). The remaining unassigned reflections indicate the existence of a superstructure. A complete indexing of the diffraction lines assigned to CeGe₃ was successful on the basis of a doubled unit cell parameter, i.e., a_sc = 2 × a of the lt-Cu₃Au-type (aristotype). Systematic absences are compatible with space group F m 3 ‾ m .

Figure 4:

Crystal structure refinement using powder X-ray diffraction data (CuK_α1 radiation) for CeGe₃. Black – experimental pattern, red – calculated pattern, grey – difference. Vertical bars show the position of the Bragg lines.

These superstructure reflections were previously observed in the high-pressure phase of SmGe₃ [24]. This resulted in a doubling of the cubic unit cell, producing a 2 × 2 × 2 superstructure (Figure 5). The results of the crystal-structure refinement using powder X-ray diffraction data (Figure 4, Tables 1 and 2) confirm that the use of the 2 × 2 × 2 superstructure structural model is correct. The unit cell parameter of CeGe₃ (a = 8.6970(2) Å) is larger than that reported for isostructural SmGe₃ (a = 8.6719(2) Å [24]), which is to be expected since samarium is smaller due to the lanthanide contraction. The space groups of the refined F m 3 ‾ m superstructure and that of the undistorted aristotype (lt-Cu₃Au, P m 3 ‾ m ) are related by a group subgroup formalism (Figure 6) [35].

Figure 5:

Sections of the crystal structure of CeGe₃ showing the doubling of the unit cell with respect to lt-Cu₃Au (left) and the coordination sphere of the germanium atoms (right). Short distances are indicated in red, longer ones in yellow.

Figure 6:

Group-subgroup scheme [35] for the crystal structures of ht- and lt-Cu₃Au as well as CeGe₃. Indices for the klassengleiche (k) symmetry reductions, the unit cell transformations, and the evolution of the atomic positions are given. An expanded scheme additionally considering the structure types of CuPt₇ and SrPb₃ has been given earlier [24].

An overview of Ge–Ge distances in compounds REGe₃ (RE = rare earth metal [18–25, 31, 36, 37]) reveals that the 2 × 2 × 2 superstructure causes the Ge−Ge interatomic distances to diverge from the typical distances between 3.0 and 3.1 Å (rounded values) as observed for the Cu₃Au-type subcell structures (Figure 7, light blue region). While typical Ge–Ge bonds fall into the range between 2.45 and 2.65 Å (Figure 7, grey region), longer contacts range from approximately 2.72 to 2.94 Å (Figure 7, red region). After considering the splitting of distances Ge−Ge into longer and shorter ones because of the distortion in the F m 3 ‾ m superstructure (for CeGe₃ 2.895(1) and 3.254(1) Å, respectively), the shorter distances fall into the range observed for other rare-earth metal germanium compounds REGe₃. The same behavior is also observed for SmGe₃ (2.903(2) and 3.229(2) Å) [24]. The Ce−Ce interatomic distance, 4.3485(1) Å, fits in the range of what has been previously observed for Ce−Ge compounds, however, it is noticeably longer than that of the other reported high pressure phase CeGe₅, 4.000(5) Å [38].

Figure 7:

Average Ge–Ge distances, derived from crystal structure refinements, for selected compounds MGe₃ (M = La–Pr; Sm–Yb) [18–25, 36–38]. Structure types are indicated by different symbols; compounds synthesized at elevated pressures are represented by filled icons, those prepared at ambient pressure are indicated by open ones. The region shaded in grey indicates the region of covalent interactions, while red indicates longer distances and light blue shows the range of Ge–Ge distances in Cu₃Au-type phases.

5 Summary

Consistent with the presence of non-indexed reflections in the earlier study [25], X-ray powder diffraction data of the high-pressure phase CeGe₃ are evidence for a 2 × 2 × 2 superstructure (space group F m 3 ‾ m ) of the lt-Cu₃Au type (space group P m 3 ‾ m ). The Ge–Ge distances in the new structure model, which is revised with respect to the original report, fit the trend of Ge–Ge distances in the series of other rare-earth metal compounds of the type REGe₃.