Oliver
Janka
*ab,
Rolf-Dieter
Hoffmann
b,
Matthias
Eilers-Rethwisch
b,
Ute Ch.
Rodewald
b,
Oliver
Niehaus
bc and
Rainer
Pöttgen
b
aInstitut für Chemie, Carl von Ossietzky Universität, Carl-von-Ossietzky Strasse 9-11, D-26129 Oldenburg, Germany. E-mail: ocjanka@uni-muenster.de
bInstitut für Anorganische und Analytische Chemie, Universität Münster, Corrensstraße 30, D-48149 Münster, Germany
cNRW Graduate School of Chemistry, Universität Münster, Corrensstraße 30, D-48149 Münster, Germany
First published on 19th August 2016
Structural reinvestigations of Ce3Pt4Ge6 on high quality single crystals revealed additional reflections contradicting the previously reported structure featuring half occupied crystallographic sites for this compound. The structure could be solved and refined in the orthorhombic (3 + 1)D superspace group Cmcm(α,0,0)0s0 with α = 0.5a* and lattice parameters of a = 441.17(2), b = 2618.26(14) and c = 441.33(2) pm using the super space approach. The description of the modulated structure as commensurate case allows for an ordering of the Ce2/Ge3 atoms and the respective voids caused by the 50% occupation. The corresponding approximant can be described in the orthorhombic crystal system with space group Pnma and lattice parameters of a = 2618.26(14), b = 441.33(2) and c = 882.34(2) pm. Investigations of the magnetic properties revealed a magnetic moment of μeff = 2.48(1)μB/Ce atom and a Weiss constant of θp = −39(5) K, indicating stable trivalent cerium. No clear magnetic ordering was evident from the susceptibility measurements. Heat capacity investigations showed a λ-shaped anomaly at T = 2.1(1) K. Resistivity measurements show values corresponding to Ce3Pt4Ge6 being a metal, however nearly temperature independent behavior is observed down to low temperatures. A shallow minimum and an abrupt drop suggest the title compound to be a Kondo material.
In the original work, the structures of CePtGe3 and Ce3Pt4Ge611 were reported with mixed and partially occupied sites, respectively. Reinvestigation of the CePtGe structure18,19 revealed weak superstructure reflections, originating from platinum–germanium ordering. The single crystal data clearly stated the orthorhombic YPdSi structure type. The Ce3Pt4Ge6 structure11 was refined in space group Bmmb (non-standard setting of Cmcm) with half-occupied Ce2 (4c) and Ge3 (8g) sites. Similar composition, but different space group symmetry was observed for the RE3Pt4Ge6 representatives with the smaller rare earth (RE) elements.24–26 The Y3Pt4Ge6 structure24 as well as the silicide Yb∼3Pt∼4Si6−x (x = 0.3)26 were described in the monoclinic space group P21/m and the germanides RE3Pt4Ge6 (RE = Pr, Nd, Sm, Gd, Tb, Dy)25 were reported with space group Pnma which allows for an ordering of all atomic sites. Reinvestigation of the Ce3Pt4Ge6 structure by these authors gave hints for cerium–germanium ordering. Single crystals taken from a long-term annealed sample showed diffuse intensities.
We became interested in these order–disorder problems when solving the structure of a new monoclinic cerium–gold–germanide with an approximate composition Ce3Au∼5.5Ge∼4.527 and presumably Y3Pt4Ge6 structure. In the course of these systematic investigations we also synthesized new Ce3Pt4Ge6 samples and obtained well shaped single crystals upon annealing arc-melted precursors in a water-cooled sample chamber of an induction furnace. Herein we report on crystal chemical data on the ordered phase and the magnetic properties of Ce3Pt4Ge6. The order–disorder difficulty is discussed on the basis of a group–subgroup scheme.
Correct indexing of the diffraction lines was ensured through an intensity calculation. The lattice parameters were obtained through least-squares fits30 with standard deviations smaller than ±0.1 pm for all axes.
Irregularly shaped crystals of Ce3Pt4Ge6 were obtained by mechanical fragmentation of the annealed arc-melted button. These fragments were glued to thin quartz fibers using beeswax. The crystal quality was tested by Laue photographs on a Buerger camera (white molybdenum radiation, image plate technique, Fuji film, BAS-1800). Intensity data sets of a suitable crystal were collected at room temperature by use of a Stoe IPDS II diffractometer (graphite-monochromatized MoKα radiation; oscillation mode). Numerical absorption correction was applied to the data set. All relevant crystallographic data and details of the data collections and evaluations are listed in Table 1.
Empirical formula | Ce3Pt4Ge6 |
Superspace group | Cmcm(α,0,0)0s0 |
α = 1/2 a* | |
Molar mass | 1636.2 g mol−1 |
Unit cell dimensions | a = 441.17(2) pm |
b = 2618.26(14) pm | |
c = 441.33(2) pm | |
V = 0.5098 nm3 | |
Calculated density | 10.66 g cm−3 |
Linear absorption coeff. | 85.0 mm−1 |
Transm. ratio (min/max) | 0.096, 0.190 |
F(000) | 1356 |
Crystal size/μm | 30 × 30 × 35 |
Diffractometer | IPDS-II |
Wavelength | MoKα (71.073 pm) |
Detector distance | 90 mm |
Exposure time | 8 min |
Integr. param./A, B, EMS | 9.9/1.5/0.012 |
θ range | 2.4–30.7 ° |
Range in hklm | ±6; ±37; ±6; ±1 |
Total no. reflections | 23224 |
Independent reflections | 1360 (Rint = 0.0519) |
Reflections with I ≥ 3σ(I) | 930 (Rσ = 0.0112) |
Main + satellite | 930 + 430 |
Data/parameters | 1360/63 |
Goodness-of-fit | 2.07 |
Final R indices [I ≥ 3σ(I)] | R M = 0.0186/wRM = 0.0451 |
R S = 0.0751/wRS = 0.1480 | |
R indices for all data | R M = 0.0227/wRM = 0.0459 |
R S = 0.1243/wRS = 0.1557 | |
Extinction scheme | Lorentzian isotropic31 |
Extinction coefficient | 222(12) |
Largest diff. peak and hole/e Å−3 | 8.71, −3.64 |
Further details on the structure refinement are available from Fachinformationszentrum Karlsruhe, D-76344 Eggenstein-Leopoldshafen (Germany), E-mail: crysdata@fiz-karlsruhe.de, by quoting the Registry no. CSD 431566.
Finally the description in the P-centered monoclinic cell lacks the interpretation of the additional reflections with 1/4 a* (green circles) and exhibits systematically absent reflection violations (red squares) for the interpretation as a primitive lattice (Fig. 1, top). A closer look at the reconstructed diffraction patterns reveals two possibilities to index all observed reflections. One possibility is to use the depicted Bbmm cell shown in Fig. 1 (right) along with a q-vector q = (0, 0, 1/2 c*), the other one is to use the depicted monoclinic cell with q = (1/4 a*, 0, 1/2 c*). Both possible cell settings have been integrated including all satellite reflections. The structure was subsequently solved by the program package Superflip,33 included in the Jana2006 package, which was used for the refinement.34,35 The corresponding superspace groups Cmcm(α,0,0)0s0 (SSG 63.1.13.2) and P21/m(α,0,γ)00 (SSG 11.1.2.1) have been identified. The notation of Stokes, Campbell and van Smaalen for the superspace groups was used.36,37 Since both refinements gave identical residuals, the higher symmetry was chosen and will be described in the following. The symmetry operations for the orthorhombic superspace group are listed in Table 2.
x1 | x2 | x3 | x4 |
−x1 | −x2 | x3 + 1/2 | −x4 |
−x1 | x2 | −x3 + 1/2 | −x4 + 1/2 |
x1 | −x2 | −x3 | x4 + 1/2 |
−x1 | −x2 | −x3 | −x4 |
x1 | x2 | −x3 + 1/2 | x4 |
x1 | −x2 | x3 + 1/2 | x4 + 1/2 |
−x1 | x2 | x3 | −x4 + 1/2 |
x1 + 1/2 | x2 + 1/2 | x3 | x4 |
−x1 + 1/2 | −x2 + 1/2 | x3 + 1/2 | −x4 |
−x1 + 1/2 | x2 + 1/2 | −x3 + 1/2 | −x4 + 1/2 |
x1 + 1/2 | −x2 + 1/2 | −x3 | x4 + 1/2 |
−x1 + 1/2 | −x2 + 1/2 | −x3 | −x4 |
x1 + 1/2 | x2 + 1/2 | −x3 + 1/2 | x4 |
x1 + 1/2 | −x2 + 1/2 | x3 + 1/2 | x4 + 1/2 |
−x1 + 1/2 | x2 + 1/2 | x3 | −x4 + 1/2 |
After refinement of the basic structure the satellites were added to the refinement and positional or occupational modulations for all atoms were applied. In addition, it was possible to refine modulated anisotropic atomic displacement parameters (ADP) for all atoms. The Fourier maps for all atoms showing significant modulations are depicted in Fig. 2. The space group allows for additional degrees of freedom, however some modulation components are very small (Table 3). While Ce1, Pt1, Pt2, Ge1 and Ge2 show positional modulations along x1/x4 and x3/x4, Ce2 and Ge3 exhibit occupational modulations along x3/x4. The occupational modulations were fitted using a crenel function; for the positional modulations simple harmonic functions were used. As a check for the correct composition, the occupancy parameters were refined in a separate series of least-squares cycles. All sites were fully occupied within three standard deviations. There was no indication of mixing on any site. The final difference electron-density synthesis was flat.
Atom | Occ. | Wave | x | y | z | U eq |
---|---|---|---|---|---|---|
Ce1 | 1 | 0 | 0.34219(3) | 1/4 | 96(2) | |
sin | 0.0174(3) | 0 | 0 | |||
cos | 0 | 0.00002(5) | 0 | |||
Ce2 | 0.5 | 1/2 | 0.47112(4) | 1/4 | 94(3) | |
x40 = 0.25 | crenel | — | — | — | ||
Pt1 | 1 | 1/2 | 0.06271(1) | 1/4 | 77(1) | |
sin | 0.00094(17) | 0 | 0 | |||
cos | 0 | −0.00009(3) | 0 | |||
Pt2 | 1 | 0 | 0.20433(1) | 1/4 | 97(1) | |
sin | −0.00671(19) | 0 | 0 | |||
cos | 0 | −0.00259(3) | 0 | |||
Ge1 | 1 | 1/2 | 0.24984(5) | 1/4 | 94(3) | |
sin | 0.0061(5) | 0 | 0 | |||
cos | 0 | −0.00197(9) | 0 | |||
Ge2 | 0.5 | 0 | 0.11020(6) | 1/4 | 146(4) | |
sin | 0.0003(5) | 0 | 0 | |||
cos | 0 | −0.00203(10) | 0 | |||
Ge3 | 1 | 0.2039(3) | 0.46437(6) | 1/4 | 120(4) | |
x40 = 0.6 | crenel | — | — | — |
Due to the q-vector component α = 1/2 a*, the crystal structure can be refined as commensurate case in the corresponding orthorhombic superstructure with space group Pbnm and a doubled a axis. Standardization finally leads to space group Pnma and lattice parameters of a = 2618.26(14), b = 441.33(2) and c = 882.34(2) pm. This unit cell corresponds to the solution of the compounds in the RE3Pt4Ge6 (RE = Pr–Dy) series,25 however in our description all intensities were handled avoiding non-space group specific extinctions.
Using the motifs highlighted by Venturini et al. for Y3Pt4Ge6, the structure can be described as an intergrowth between CaBe2Ge2- and YIrGe2-type slabs. The latter forms pentagonal prisms built by Ir and Ge (here Pt and Ge), which are linked alternatively (up/down) to form layers that are connected further over the pentagonal faces to form a three-dimensional framework. The pentagonal prisms are occupied by the Y3+ cations, here Ce2. The CaBe2Ge2 arrangements consist of tetrahedral [GeBe4/4] layers (here Pt and Ge), which are stacked along [100] and exhibit only Ge–Be bonds. The six-membered rings formed by the stacking of the layers are occupied by the Ca2+ cations, here Ce1. Due to the occupational modulation of the Ce2 and Ge3 atoms, the pentagonal prisms are allowed to order, facing either to the right or to the left along the a axis. In contrast to the description of Gribanov et al., no half occupied sites are needed in this description. The positional modulation can be easily understood since the orientation of the five membered rings influences the surrounding atoms of the remaining framework. The Pt1 atoms are the nearest neighbors of the five-membered rings, Ge2 is a part of the ring itself; therefore these atoms exhibit the strongest positional modulations along [100]. The remaining atoms follow the deformation of the framework, resulting in slightly lower amplitudes of the sin and cos functions (Table 3). In order to describe the crystal structure of Ce3Pt4Ge6 a commensurate description can be used, which is obtained via a doubling of the a axis followed by a standardization. The detailed group–subgroup relations will be discussed in the following paragraph (vide infra). The resulting orthorhombic 3D-approximant (Pnma, a = 2618.26(14), b = 441.33(2) and c = 882.34(2) pm) describes the crystal structure without the need of half-occupied sites.
The crystallographic sites present in the modulated description split, resulting in three Ce, four Pt and six Ge sites. The structural elements described before are found as well in this model and are combined as shown in Fig. 3. In Fig. 4 the coordination polyhedra surrounding the Ce1/2 (left) and Ce3 (right) atoms are depicted. The refined anisotropic displacement parameters are listed in Table 4, interatomic distances in Table 5.
Atom | U 11 | U 22 | U 33 | U 12 |
---|---|---|---|---|
Ce1 | 101(3) | 93(3) | 94(3) | 0 |
Ce2 | 89(4) | 98(4) | 96(5) | 0 |
Pt1 | 66(2) | 93(2) | 73(2) | 0 |
Pt2 | 87(2) | 109(2) | 94(2) | 0 |
Ge1 | 79(5) | 122(6) | 81(6) | 0 |
Ge2 | 80(5) | 96(6) | 262(8) | 0 |
Ge3 | 116(6) | 147(6) | 97(7) | −28(5) |
Ce1 | 2 | Ge4 | 251.66(1) | Ce2 | 2 | Pt3 | 249.77(1) | Ce3 | 2 | Pt2 | 237.84(1) | Pt1 | 1 | Ce3 | 239.61(1) | Pt2 | 2 | Ce3 | 237.84(1) |
2 | Pt4 | 254.45(1) | 2 | Ge3 | 255.40(1) | 1 | Pt1 | 239.61(1) | 1 | Ge3 | 251.14(1) | 2 | Ge6 | 248.77(1) | |||||
1 | Ge1 | 325.20(1) | 1 | Ge1 | 327.63(1) | 2 | Ge5 | 307.05(1) | 1 | Ge4 | 255.34(1) | 1 | Ge3 | 251.13(1) | |||||
1 | Ge2 | 327.05(1) | 1 | Ge2 | 329.60(1) | 1 | Ge5 | 311.04(1) | 2 | Ge5 | 265.86(1) | 1 | Ge4 | 255.33(1) | |||||
1 | Ge6 | 343.50(2) | 1 | Ge5 | 333.86(2) | 2 | Ge3 | 375.68(1) | 1 | Ge6 | 272.66(1) | 1 | Ge5 | 288.73(1) | |||||
1 | Pt3 | 365.87(2) | 1 | Pt4 | 356.18(2) | 2 | Ge4 | 379.92(1) | 2 | Pt2 | 395.62(2) | 2 | Ge5 | 387.29(1) | |||||
2 | Ge2 | 394.54(1) | 2 | Ge2 | 392.41(1) | 2 | Pt1 | 395.62(2) | |||||||||||
2 | Ge1 | 396.19(1) | 2 | Ge1 | 393.96(1) | 2 | Ce2 | 395.99(1) | |||||||||||
2 | Pt2 | 396.25(1) | 2 | Pt2 | 395.99(1) | 2 | Ce1 | 396.25(1) | |||||||||||
Pt3 | 1 | Ge3 | 245.44(1) | Pt4 | 1 | Ge4 | 247.52(1) | Ge1 | 2 | Ge2 | 220.70(1) | Ge2 | 2 | Ge1 | 220.70(1) | Ge3 | 1 | Pt3 | 245.44(1) |
2 | Ce2 | 249.77(1) | 1 | Ge2 | 249.99(1) | 1 | Pt4 | 250.38(1) | 1 | Pt4 | 249.99(1) | 1 | Pt2 | 251.13(1) | |||||
1 | Ge2 | 251.38(1) | 1 | Ge1 | 250.38(1) | 1 | Pt3 | 251.43(1) | 1 | Pt3 | 251.38(1) | 1 | Pt1 | 251.14(1) | |||||
1 | Ge1 | 251.43(1) | 2 | Ce1 | 254.45(1) | 1 | Ce1 | 325.20(1) | 1 | Ce1 | 327.05(1) | 2 | Ce2 | 255.40(1) | |||||
2 | Pt4 | 325.42(1) | 2 | Pt3 | 325.42(1) | 1 | Ce2 | 327.63(1) | 1 | Ce2 | 329.60(1) | 2 | Ge5 | 305.67(1) | |||||
2 | Ge1 | 337.29(1) | 2 | Ge1 | 331.25(1) | 2 | Pt4 | 331.25(1) | 2 | Pt4 | 331.48(1) | 2 | Ce3 | 375.68(1) | |||||
2 | Ge2 | 337.35(1) | 2 | Ge2 | 331.48(1) | 2 | Pt3 | 337.29(1) | 2 | Pt3 | 337.35(1) | 1 | Ge5 | 388.62(2) | |||||
1 | Ce1 | 365.87(2) | 1 | Ce2 | 356.18(2) | 2 | Ce2 | 393.96(1) | 2 | Ce2 | 392.41(1) | ||||||||
2 | Ce1 | 396.19(1) | 2 | Ce1 | 394.54(1) | ||||||||||||||
Ge4 | 1 | Pt4 | 247.52(1) | Ge5 | 2 | Pt1 | 265.86(1) | Ge6 | 2 | Pt2 | 248.77(1) | ||||||||
2 | Ce1 | 251.66(1) | 1 | Pt2 | 288.73(1) | 1 | Pt1 | 272.66(1) | |||||||||||
1 | Pt2 | 255.33(1) | 2 | Ge3 | 305.67(1) | 2 | Ge4 | 324.63(1) | |||||||||||
1 | Pt1 | 255.34(1) | 2 | Ce3 | 307.05(1) | 1 | Ce1 | 343.50(2) | |||||||||||
2 | Ge6 | 324.63(1) | 1 | Ce3 | 311.04(1) | 2 | Ge5 | 363.54(1) | |||||||||||
2 | Ce3 | 379.92(1) | 1 | Ce2 | 333.86(2) | 2 | Ge6 | 389.56(1) | |||||||||||
2 | Ge5 | 340.39(1) | |||||||||||||||||
2 | Ge6 | 363.54(1) | |||||||||||||||||
2 | Pt2 | 387.29(1) | |||||||||||||||||
1 | Ge3 | 388.62(2) |
The annealing temperature and the cooling rate are important parameters for the growth of high quality crystals. The structural details of Ce3Pt4Ge6 are comparable to the recently reported stannide Ce3Pt4Sn6.38 The stannide crystals showed smeared out reflections and disorder.
Fig. 5 Relation of the (3 + 1)D and 3D space groups of the modulated structure of Ce3Pt4Ge6. Group–supgroup scheme in the Bärnighausen formalism40–43 for the subcell of Ce3Pt4Ge6 (*occupancy parameter listed as 50% in the literature24) and the supercell. The indices for the klassengleiche (k) symmetry reduction as well as the evolution of the atomic parameters are given. The index a is introduced for the formation of the approximant of the modulated structure.39 The last line gives the atomic coordinates for Gd3Pt4Ge6, published by Mewis in space group Pnma.25 |
Here two transitions are involved, a doubling of the a axis along with a loss of the centering and an origin shift of (1/4, 1/4, 0). The a4 transition is paralleled by two k2 transitions in 3D space. First a symmetry reduction from Cmcm to Pbcm by a klassengleiche transition of index 2 takes place, which is accompanied by an origin shift of (1/4, 1/4, 0). In the second step again a klassengleiche transition of index 2 takes place, allowing the doubling of the a axis. The final step is used to reach the standard setting of Pbnm, the orthorhombic space group Pnma. The group–supgroup scheme in the Bärnighausen formalism40–43 is shown in Fig. 5. The observed disorder and diffuse superstructure reflections of38 have their origin most likely in the occurrence of anti phase boundaries triggered by the klassengleiche transitions.
To obtain more information about the magnetic ground state, a low-field measurement was performed in a zero-field- and field-cooled mode (ZFC/FC) which is shown in the middle graph of Fig. 6. No anomaly indicating clear magnetic ordering could be detected in this measurement, however in the low temperature region saturation effects can be observed, while the derivative dχ/dT shows a double feature at 2.5(1) K and 3.0(1) K (Fig. 6, inset). No bifurcation between the ZFC and FC curve is visible.
The bottom graph in Fig. 6 displays the magnetization isotherms of Ce3Pt4Ge6 measured at 3, 10, and 50 K. The isotherms at 10 and 50 K both display a linear field dependency of the magnetization as expected for a paramagnetic material. With respect to the 10 K isotherm less saturation effects are visible at 50 K. At 3 K a curvature of the magnetization isotherm can be observed with weak saturation effects towards higher fields. The magnetic moment at 3 K and 80 kOe is 0.61(5)μB/Ce atom which is significantly below the expected saturation magnetization of 2.14μB/Ce according to gJ × J. Magnetic susceptibility measurements reported in the literature1 found no ordering phenomena and a reduced magnetic moment of μeff = 2.19μB between 40 and 140 K and a moment of μeff = 1.86μB above 150 K. Between 140 and 150 K an anomaly is observed which is in agreement with their resistivity measurements (vide supra). A magnetization isotherm at 5.5 K shows a small hysteresis. Both, the reduced moment and the anomaly between 140–150 K, are in contrast to our investigations. However the authors report their sample to be “predominantly single phase”, which might be an explanation for the missing features in our investigations.
Footnote |
† Dedicated to Professor Mercouri G. Kanatzidis on the occasion of his 60th birthday. |
This journal is © the Partner Organisations 2016 |