Eu1 − xSrxAu4Cd2: a ferromagnetic solid solution with adjustable Curie temperature

Samples of the solid solution Eu1–xSrxAu4Cd2 (YbAl4Mo2 type, space group I4/mmm) with x = 0.2, 0.4, 0.6, and 0.8 were synthesized from the elements by annealing in sealed tantalum ampoules. The structures of Eu0.66Sr0.34Au4.12Cd1.88 and Eu0.22Sr0.78Au4.10Cd1.90 were refined from X-ray single crystal diffractometer data. Besides the expected Eu/Sr mixing on the 2a sites the diffraction data revealed also a small degree of Cd/Au mixing on the 4d sites of the cadmium chains. Temperature dependent magnetic susceptibility data show divalent europium and ferromagnetic ground states for all samples and a linear decrease of the Curie temperature from 16.3 K for EuAu4Cd2 to 2.9 K for Eu0.2Sr0.8Au4Cd2. Magnetization isotherms characterize the Eu1 − xSrxAu4Cd2 samples as soft ferromagnets. The divalent nature of europium is underpinned by 151Eu Mössbauer spectra. The decreasing Curie temperature goes along with a decreasing magnetic hyperfine field at 6 K.


Introduction
Strontium and europium have almost similar radii. To give an example, Shannon [1] lists values of 118 (Sr 2+ ) and 117 (Eu 2+ ) pm for coordination number 6. Thus, it is not surprising that one observes a large crystal chemical similarity between divalent europium and strontium compounds. The simplest example concerns the rocksalt-type phases SrO and EuO. Another prominent example concerns the structure of Eu 3 O 4 , where a complete substitution of the divalent europium by Sr 2+ leads to SrEu 2 O 4 [2].
This isovalent substitution is interesting with respect to the magnetic properties. Sr 2+ is a diamagnetic cation while Eu 2+ with its [Xe]4f 7 configuration is strongly paramagnetic. To keep the example of the rocksalt phases, SrO is a simple diamagnet, whereas EuO is a strong Curie-Weiss paramagnet which orders ferromagnetically below T C = 70 K [3,4]. These magnetic characteristics offer the possibility to study solid solutions. Due to the close size of the cations, these solid solutions are mostly strain free and one can investigate the force of the magnetic coupling as a function of the europium-strontium substitution.
Interesting results have been obtained for solid solutions of metal hydrides. Eu 2 IrH 5 orders ferromagnetically at T C = 20 K and the Curie temperature decreases with increasing strontium content: 11 K for Eu 1.5 Sr 0.5 IrH 5 , < 7 K for Eu 1.0 Sr 1.0 IrH 5 and < 5 K for Eu 0.5 Sr 1.5 IrH 5 [5]. Similar behavior was reported for the solid solutions Eu 2-x Sr x RuH 6 [6] and Eu 1 − x Sr x Mg 2 H 6 (T C = 21 K for x = 0.2 and 15 K for x = 0.4) [7]. In the three series, T C decreases with decreasing number of neighboring europium cations, i. e. a decrease of the strength of the Eu-Eu nearest neighbor exchange interactions. A different behavior was observed for the fullerides Eu 6 − x Sr x C 60 [8] where the Curie temperature shows little variation: 11.6 K for Eu 6 C 60 , 12.8 K for Eu 3 Sr 3 C 60 , and 10.4 K for EuSr 5 C 60 . It was concluded that the ferromagnetic interaction is mediated through the conduction carriers.
We have recently started a systematic investigation of europium-strontium substitution in the field of intermetallic compounds. The first complete solid solution was Eu 1-x Sr x PtIn 2 with MgCuAl 2 type structure [9]. The course of the cell parameters showed Vegard type behavior. EuPtIn 2 shows a comparatively high Curie temperature of 32.5 K and it is remarkable that the ferromagnetic ground state even remains in the Eu 0.1 Sr 0.9 PtIn 2 sample with T C = 3.1 K with a gradual decrease of the Curie temperature within the complete solid solution. This was monitored in parallel by 151 Eu Mössbauer spectroscopy.
In continuation of these studies we have now investigated the solid solution Eu 1-x Sr x Au 4 Cd 2 with YbAl 4 Mo 2 type structure which also shows adjustable Curie temperatures. The crystal chemistry and the magnetic behavior of this solid solution are reported herein.

Crystal chemistry
EuAu 4 Cd 2 and SrAu 4 Cd 2 [10] crystallize with the YbAl 4 Mo 2 type structure [11], space group I4/mmm. The slightly larger radius of strontium (see "Introduction") leads to a small increase of the a and c lattice parameters (Table 1) and consequently a larger cell volume for SrAu 4 Cd 2 . We have studied the complete solid solution Eu 1 − x Sr x Au 4 Cd 2 for samples with x = 0.2, 0.4, 0.6, and 0.8. Within the standard deviations, we do not strictly observe Vegard-type behavior. The small differences observed for the two single crystals and the Eu 0.2 Sr 0.8 Au 4 Cd 2 sample can be ascribed to the Cd/ Au mixing (formation of small homogeneity ranges) on the 4d sites (vide infra).
The prototype YbAl 4 Mo 2 [11] has more than 40 isotypic representatives [12]. The crystal chemistry and chemical bonding of YbAl 4 Mo 2 type phases has repeatedly been discussed in the original contributions [13][14][15][16][17][18][19][20]. Herein we therefore concentrate only on the structural peculiarities of the solid solution Eu 1 − x Sr x Au 4 Cd 2 . As an example, we present the unit cell of Eu 0.66 Sr 0.34 Au 4.12 Cd 1.88 in Fig. 1. The projection onto the ab plane ( Fig. 1, top) shows the Au 4 squares with 279 pm Au-Au. These squares condense with adjacent squares that are shifted by half the translation period c and form the Au 12 clusters that surround the europium/strontium atoms. The Au-Au distances between the clusters of 299 pm are slightly longer. In average, these Au-Au distances compare well with fcc gold (12 × 288 pm) [21].
The second striking motif concerns the linear cadmium chains that extend in c direction. The Cd-Cd distances of 276 pm are indicative of substantial Cd-Cd bonding given that they are shorter than in hcp cadmium (6 × 298 and 6 × 329 pm) [21]. Each cadmium atom is surrounded by eight gold atoms in strongly distorted square prismatic coordination. These polyhedra are capped by cadmium atoms on the rectangular faces in c direction.
The refinements of the occupancy parameters of the 2a sites (vide infra) clearly revealed the Eu/Sr mixing, in approximate agreement with the nominal starting compositions of the samples. These deviations indicate that all samples from the solid solution definitely contain quantities of domains with compositions that slightly deviate from the starting ones. The mixed Eu/Sr occupancy for Eu 0.66 Sr 0.34 Au 4.12 Cd 1.88 is presented by light and dark green segments.

Magnetic properties
Results of the magnetic property measurements of the solid solution Eu 1-x Sr x Au 4 Cd 2 are summarized in Table 2. Measurements of M(T,H) of Eu 0.4 Sr 0.6 Au 4 Cd 2 are shown as an example in Fig. 2.
Measurements of M(T) performed on the samples of the solid solution in zero-field-cooled/field-cooled (ZFC/FC) mode from 2.5 to 100 K, with an applied field of 100 Oe, show ferromagnetic ordering for all members of the solid solution at low temperatures, which is in accordance with the properties of EuAu 4 Cd 2 [10]. As expected, the Curietemperature T C decreases ( Fig. 3) with decreasing europium content (increasing Eu/Sr disorder). Similar behaviour has been observed in other solid solutions of europium and strontium such as Eu 1 − x Sr x PtIn 2 [9].
The effective magnetic moments determined from zerofield-cooled (ZFC) measurements with an applied field of 10 kOe in a temperature range of 3-300 K are in good accordance with the theoretical value of 7.94 μ B per Eu Magnetization isotherms were measured for every member of the solid solution at 3, 10, and 50 K with applied fields of up to 80 kOe. All magnetization isotherms at 50 K show a linear course as expected for paramagnetic compounds. The isotherms at 3 K, which are below or close to the ordering temperatures of the compounds, show typical behaviour for soft ferromagnetic compounds as the magnetization rises rapidly at low fields. The saturation magnetization of Eu 0.8 Sr 0.2 Au 4 Cd 2 and Eu 0.4 Sr 0.6 Au 4 Cd 2 at 3 K and 80 kOe is in good accordance with the theoretical

Eu Mössbauer spectroscopy
The europium valence and the magnetic ordering within the solid solution Eu 1 − x Sr x Au 4 Cd 2 were further studied through temperature dependent 151 Eu Mössbauer spectra. Figure 4 presents the 151 Eu Mössbauer spectra of these samples at 78 and 6 K, respectively. The corresponding fitting parameters are listed in Table 3. The extremely high absorption caused by the high content of highly absorbing heavy elements leads to a low resolution of the experimental spectra. The 78 K 151 Eu Mössbauer spectroscopic investigations confirm divalent europium in the Eu 1 − x Sr x Au 4 Cd 2 samples. Only the Eu 0.6 Sr 0.4 Au 4 Cd 2 sample revealed a small degree of an additional Eu(III) component, most likely resulting from surface oxidation/surface hydrolyses of the polycrystalline material, frequently observed for intermetallic europium compounds (even if samples are handled under inert conditions, small amounts of Eu(III) deterioration products occur; this is decomposition, not intrinsic trivalent europium of the main sample) [25,26]. The isomer shifts values are in the usual range observed for structurally related intermetallics [25,26]. The small quadrupole splitting parameters account for the non-cubic site symmetry of the europium atoms.
The magnetic susceptibility data (vide ultra) showed a decreasing T C with increasing strontium content. The Mössbauer spectrometer allowed data collections only above a temperature of 6 K. Thus, in Fig. 4 we only    [27] in small tantalum tubes under an argon atmosphere of ca. 800 mbar. The argon (Westfalen, 99.998%) was purified over titanium sponge (870 K), silica gel, and molecular sieves. A slight excess of europium and strontium was necessary to minimize the amount of by-products. The tantalum ampoules were then sealed in quartz tubes for protection against oxidation and were subsequently heated within one hour to 1273 K. This temperature was held for five minutes before the samples were cooled to 1073 K within 20 min, kept at that temperature for another three hours and finally cooled to room temperature at a rate of 10 K h -1 . The polycrystalline samples were powdered in an agate mortar under dry cyclohexane, cold pressed into pellets (∅ 6 mm) and thermally post-treated in a muffle furnace (680 K for four days, followed by slow cooling to room temperature at a rate of 5 K h -1 ) to increase the phase purity and crystallinity. The samples were mechanically separated from the ampoules. No reactions with the ampoule material were evident. All samples are stable in air over several weeks and have metallic luster, ground powders are dark grey.

X-ray diffraction on powders and single crystals
The polycrystalline Eu 1 − x Sr x Au 4 Cd 2 samples were ground to fine powders and studied by X-ray powder diffraction using the Guinier technique (Enraf-Nonius FR 552 camera) with CuK α1 radiation and α-quartz (a = 491.30 and c = 540.46 pm) as an internal standard. The diffraction patterns were collected through imaging plate technique (Fuji film, BAS-READER 1800). The lattice parameters (Table 1) were obtained from least-squares refinements. Intensity calculations (Lazy-Pulverix routine [28]) facilitated the correct indexing of the diffraction lines.
Conchoidally fractured crystal splinters were selected from the crushed samples with the nominal compositions Eu 0.6 Sr 0.4 Au 4 Cd 2 and Eu 0.2 Sr 0.8 Au 4 Cd 2 and glued to glass fibers with bees wax. The crystals were first investigated on a Buerger precession camera (white Mo radiation, Fuji-film imaging plate) in order to check their quality for intensity data collection. Complete data sets were collected at room temperature by using a Stoe IPDS-II image plate system (graphite monochromatized Mo radiation; λ = 71.073 pm) in oscillation mode. Numerical absorption corrections were applied to the data sets. Details on the crystallographic data are given in Table 4.

Structure refinements
Both data sets showed tetragonal body-centered lattices with high Laue symmetry and no further extinctions. Similar to EuAu 4 Cd 2 and SrAu 4 Cd 2 [10], the centrosymmetric space group I4/mmm was found to be correct. The positional parameters of SrAu 4 Cd 2 [10] were taken as starting values and the structures were refined on F 2 with the Jana2006 [29] software package using anisotropic displacement parameters for all atoms. Both data sets showed mixed occupancies. Besides the Eu/Sr mixing on the 2a sites we observed small degrees of Cd/Au mixing on the 4d sites, similar to EuAu 4.8 Cd 1.2 reported earlier [10]. These site occupancies were refined as least-squares variables. The final difference Fourier syntheses were flat. The final atomic coordinates, displacement parameters and interatomic distances of both structures are listed in Tables 5, 6. CCDC-1995352 (Eu 0.66 Sr 0.34 Au 4.12 Cd 1.88 ) and CCDC-1995314 (Eu 0.22 Sr 0.78 Au 4.10 Cd 1.90 ) contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_reque st/cif.

Magnetic characterization
The samples of the solid solution Eu 1 − x Sr x Au 4 Cd 2 (x = 0.2, 0.4, 0.6, and 0.8) were investigated using the vibrating sample magnetometer (VSM) option of a physical property measurement system (PPMS) by Quantum Design. The samples were ground to a fine powder and each was filled into a polypropylene capsule, which was then inserted into a brass sample holder, which was attached to the sample holder rod of the VSM. The samples of the solid solution with  x ranging between 0.2 and 0.8 were investigated through measurements of M(T,H) in a temperature range from 2.5 to 300 K and fields up to 80 kOe. Measurements of M(T) were performed for the end member of the solid solution, SrAu 4 Cd 2 , in the temperature range from 3 to 300 K and a field of 10 kOe.

Mössbauer spectroscopy
The samples of the solid solution Eu 1 − x Sr x Au 4 Cd 2 (x = 0.2, 0.4, 0.6, and 0.8) were investigated via 151 Eu Mössbauer spectroscopic investigations. For the measurements a 151 Sm:EuF 3 source (the 21.53 keV transition was utilized, 55 MBq, 1% of the total activity) was used in either a continuous flow helium cryostat system (Janis Research Co LLC) equipped with a resistance thermometer (accuracy ± 0.5 K) or a commercial liquid nitrogen bath cryostat in the usual transmission geometry. The source was kept at room temperature while the samples were cooled to the respective temperature ( Table 3). The sample amounts were adjusted to the ideal absorber thickness [30] and mixing with α-glucose assured homogeneous coverage. The samples were then placed in thin-walled PMMA containers. The high gold and europium content (see also the high absorption coefficients in Table 4) caused long measuring times between one and two weeks with a barely satisfactory resolution of the spectra especially for the spectra that were recorded below the magnetic ordering temperature. Fitting of the spectra was done with the WinNormos for Igor6 program package [31].