The Mineral Stützite: a Zintl-Phase or Polar Intermetallic? A Case

Institute of Inorganic Chemistry, RWTH Aachen University, Landoltweg 1, 52074 Aachen, Germany. Inorg. Chem. , Article ASAP. DOI: 10.1021/acs.inorgchem...
4 downloads 36 Views 3MB Size
Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

pubs.acs.org/IC

The Mineral Stü tzite: a Zintl-Phase or Polar Intermetallic? A Case Study Using Experimental and Quantum-Chemical Techniques Kai C. Göbgen, Fabian C. Gladisch, and Simon Steinberg* Institute of Inorganic Chemistry, RWTH Aachen University, Landoltweg 1, 52074 Aachen, Germany S Supporting Information *

ABSTRACT: Differing reports regarding the structural features of the mineral stützite, Ag5−xTe3 (−0.25 ≤ x ≤ 1.44), and the quest for tellurides with low-dimensional fragments stimulated our impetus to review this system by employing experimental as well as quantum-chemical methods. Determination of the crystal structures for three samples with compositions Ag4.72(3)Te3 (I), Ag4.66(1)Te3 (II), and Ag4.96(2)Te3 (III) revealed considerable positional disorders for the Ag and Te sites and previously unknown structure models for I and II, which differ from that of III through the stacking sequences of honeycomb-fashioned Te layers. The crystal structures comprise [Te@Ag9]@Te14 units in the forms of bicapped hexagonal Te antiprisms that enclose Te-centered tricapped trigonal Ag prisms, while each Te atom is encapsulated by Ag atoms assembling diverse types of coordination polyhedra. The vibrational and electronic properties were determined for three models approximating the actual crystal structure of stützite by means of techniques based on first principles. From analyses of the electronic structures and projected crystal orbital Hamilton populations (pCOHP), it is clear that the amounts and distributions of the Ag atoms within the Te network should be influenced by the subtle interplay between the attempts to achieve an electronically favorable situation with a gap at EF and minimize the occupations of antibonding states.



INTRODUCTION Intermetallic compounds comprising low-dimensional polyanionic networks in their crystal structures are enduring subjects of research because they show versatile structural chemistry, diverse electronic structures and bonding motifs, and remarkable physical properties.1−6 The nature of bonding in many representatives of the aforementioned family of intermetallic compounds is frequently interpreted by applying the Zintl− Klemm concept,7−10 for some cases, with respect to hypervalent bonding;1 however, the applicableness of the Zintl−Klemm concept is restricted. For instance, applications of the Zintl− Klemm concept to certain rare-earth copper tellurides containing linear undistorted Te chains result in the assignments of nonelectron-precise charges to Te atoms in the chains,1 while quantum-chemical examinations on the nature of bonding for these rare-earth copper tellurides revealed extensive rare-earth− Te bonding besides minor, but evident, Te−Te and Cu−Te bonding.11 The presence of linear, undistorted Te chains in these rareearth copper tellurides is noteworthy because low-dimensional Te networks, mostly, in the form of layers of Te square nets, typically undergo structural distortions because of chargedensity-wave (CDW) formations.12−17 CDWs represent periodic modulations of the conducting electron density and occur in the presence of electronic instabilities, which are alleviated by approaching energetically more favorable states through the introductions of periodic lattice distortions below certain transition temperatures.18,19 As a result of the structural distortions within the low-dimensional fragments in a given © XXXX American Chemical Society

crystal structure, a gap opens at the Fermi level and the electronic energy is decreased relative to that of the metallic state.18,19 In that connection, it should be mentioned that more recent research demonstrated the emergence of a superconducting state after suppression of CDW transitions.20 Intermetallic compounds whose electronic structures can be inferred by applying the Zintl−Klemm concept are of great interest as candidates for materials for thermoelectric energy conversion.21,22 More recent predictions on the capabilities of undiscovered Te-rich (≥50 atom %) rare-earth silver tellurides for thermoelectric energy conversion23 based solely on quantumchemical techniques stimulated our impetus to explore the components for the ternary rare-earth−Ag−Te systems. In doing so, we also revisited the components of the binary Ag−Te system,24 which has attracted enormous attention since the discovery of large magnetoresistance effects in Ag2Te.25,26 Another intermetallic compound that has been identified for the binary Ag−Te system is the mineral stützite, Ag5−xTe3 (−0.25 ≤ x ≤ 1.44), which features a high Ag ionic conductivity;27 however, the actual crystal structure of Ag5−xTe3 (−0.25 ≤ x ≤ 1.44) has remained ambiguous particularly because different structure models have been proposed for stützite.28−30 In this contribution, we report on the synthesis, actual crystal structure, and electronic band structure of the mineral stützite. From examination of the electronic band structure and a subsequent (chemical) bonding Received: October 15, 2017

A

DOI: 10.1021/acs.inorgchem.7b02642 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 1. Details of the Crystal Structure Investigations and Refinements for I−III Ag4.72(3)Te3 (I) fw space group a, Å c, Å volume, Å3 Z density (calcd), g/cm3 μ, mm−1 F(000) θ range index ranges no. of reflns collected no. of indep reflns/Rint no. of reflns with I > 2σ (I)/Rσ refinement method data/restraints/parameters GOF on F2 final R indices [F2 > 2σ(F2)] R indices (all data) largest diff peak and hole, e/Å3

891.48 13.447(6) 8.464(4) 1325.5(13) 7 7.818 23.199 2643 1.75−24.14 −15 ≤ h ≤ 12, −15 ≤ k ≤ 15, −9 ≤ l ≤ 9 7118 825/0.139 721/0.083 825/0/76 1.10 R1 = 0.054, wR2 = 0.132 R1 = 0.065, wR2 = 0.139 2.59 and −1.54

Ag4.66(1)Te3 (II) 885.47 P6̅2m (No. 189) 13.476(1) 8.477(1) 1333.2(2) 7 7.720 22.925 2625 1.74−25.06 −14 ≤ h ≤ 16, −16 ≤ k ≤ 12, −8 ≤ l ≤ 10 7228 911/0.037 901/0.028 full matrix on least squares on F2 911/0/76 1.13 R1 = 0.044, wR2 = 0.128 R1 = 0.044, wR2 = 0.129 4.41 and −1.69

13.468(1) 16.942(1) 2661.4(2) 14 8.021 23.729 5451 1.20−24.16 −15 ≤ h ≤ 15, −14 ≤ k ≤ 15, −19 ≤ l ≤ 19 14321 1604/0.033 1423/0.020 1604/0/143 1.16 R1 = 0.062, wR2 = 0.20 R1 = 0.066, wR2 = 0.204 4.651 and −3.535

accomplished by utilizing the programs SAINT+ and SADABS, respectively.32 The crystal structures were initially solved in the space group P3 (No. 143) for Ag4.72(3)Te3 (I) and Ag4.66(1)Te3 (II) and P6 (No. 168) for Ag4.96(2)Te3 (III) using direct methods (SHELXS-97).33 Examinations of the Fourier maps for Ag4.72Te3 (I) and Ag4.66Te3 (II) revealed the presence of high residual electron densities in the vicinity of the Te5, Te7, Ag3, and Ag6/Ag7 sites. Additional inspections of the distances and geometries between the centers of the residual electron densities and the surrounding atoms revealed the presence of short interatomic contacts, which are too small to resemble Ag−Ag, Ag−Te, or Te−Te single bonds based on the atomic covalent radii of Ag (1.45 Å) and Te (1.38 Å).34 Because analyses of the coordination environments for the residual electron densities indicated the occurrence of positional disorders for the Te5, Te7, Ag3, and Ag6/ Ag7 sites, the residual electron densities were assigned to the Te6, Te8, Ag4, and Ag5 positions, respectively. To probe the occurrence of substitutional disorder for the aforementioned disordered positions, we also replaced each atom residing on the previously identified disordered sites by the respective opposite type of atom, yet examinations of the topologies and the respective anisotropic atomic displacement parameters pointed to the absence of substitutional disorder for these positions. A similar inspection of the Fourier maps and refined anisotropic atomic displacement parameters for III resulted in the identification of the Te10, Te12, Ag9, Ag11, and Ag12 sites. Additional cycles of refinements also included inspections of the Fourier maps for maxima and minima of the residual electron densities that could be hints to further positional or occupational disorders; however, further investigations considering the maxima and minima of the residual electron densities based on analyses of the anisotropic atomic displacement parameters and the coordination environments for the refined structure models and did not reveal any extra positional or occupational disorders. All least-squares refinements on F2 including the anisotropic atomic displacement parameters were carried out with the SHELXL code.33,35 Topological analyses of the refined structure models pointed to the space group P6̅2m (No. 189) as the highest possible space group for I−III. The transformations of the initial structure models for I−III into those corresponding to the space group P6̅2m (No. 189) were accomplished with the aid of the Platon code.36 Details of the crystal structure investigations and refinements may be extracted from Table 1, while atomic positions and equivalent isotropic displacement parameters of the inspected tellurides are listed in Table 2. Computational Details. The electronic and vibrational properties were examined for one “Ag36Te21” model and one “Ag34Te21” model, approximating the actual crystal structures of Ag5.00−xTe3 [x = 0.28(3)

analysis for this telluride, it becomes clear why Ag5−xTe3 exhibits substantial disorder in its crystal structure.



Ag4.96(2)Te3 (III) 918.30

EXPERIMENTAL TECHNIQUES

Syntheses. Ag5−xTe3 [x = 0.04(2), 0.28(3), 0.34(1)] were obtained from reactions of the pure elements Ag (Alfa, ≥99.99%), Te (Merck, >99%), and Y (smart elements, ≥99.95%) with the initial load of YAg5Te4 and, furthermore, from samples loaded with only Ag and Te in the target compositions. The mixtures of ∼230−280 mg total weight were first pestled, then loaded into fused-silica tubes, which were flamesealed under a vacuum of at least 2 × 10−3 mbar, and finally heated in computer-controlled tube furnaces. For the growth of single crystals prepared from samples with the initial YAg5Te4 loads, the following temperature program was used: heat to 1000 °C at a rate of 80 °C/h, keep this temperature for 5 days, slowly cool to 200 °C at a rate of 1 °C/ h, and rapidly cool to room temperature in 100 min. The following temperature program was utilized to obtain stützite from reactions of the composite elements: heat to 970 °C at a rate of 80 °C/h, keep this temperature for 96 h, and cool to room temperature at a rate of 20 °C/h. To prevent any contamination of the samples as a consequence of an exposure to air and moisture, all sample preparations were accomplished under a dry argon atmosphere in a glovebox (MBraun, Garching, Germany). The product appeared as a gray powder containing small crystals with metallic luster and, for instance, was accompanied by traces of hessite (see below; the powder X-ray diffraction patterns of two samples are provided in the Figures S1 and S2). X-ray Diffraction Studies and Structural Analyses. The purities of the obtained samples were examined based on phase analyses by comparing the measured powder X-ray diffraction (PXRD) patterns of the products to those simulated for the target compounds and feasible side products. The sets of PXRD data were collected on a STOE StadiP diffractometer at room temperature (Stoe & Cie GmbH, Darmstadt, Germany; area detector, Cu Kα radiation; λ = 1.54059 Å). For the measurements, the samples were first pestled and then loaded into capillaries, which were subsequently sealed. Data and phase analyses were accomplished with the aid of the WinXPow software package.31 Samples were selected from the bulk materials, fixed and sealed in glass capillaries, and subsequently transferred to a Bruker APEX CCD diffractometer (Bruker Inc., Madison, WI; Mo Kα radiation; λ = 0.71073 Å), which was used for the initial inspections of the quality of the selected samples and the collections of sets of single-crystal X-ray intensity data at room temperature. The integrations of the raw X-ray intensity data sets and multi-scan absorption corrections were B

DOI: 10.1021/acs.inorgchem.7b02642 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 2. Atomic Positions and Equivalent Isotropic Displacement Parameters for I−III atom

x

position

Te1 Te2 Te3 Te4 Ag1 Ag2 Te5 Te7 Ag3 Ag5 Te6 Te8 Ag4 Ag6 Ag7

3f 4h 6k 3g 6k 6i 3f 2e 12l 12l 3f 2e 12l 6j 6i

Te1 Te2 Te3 Te4 Ag1 Ag2 Te5 Te7 Ag3 Ag5 Te6 Te8 Ag4 Ag6 Ag7

3f 4h 6k 3g 6k 6i 3f 2e 12l 12l 3f 2e 12l 6j 6i

Te1 Te2 Te3 Te4 Te5 Te6 Te7 Te8 Ag1 Ag2 Ag3 Ag4 Ag5 Te9 Te11 Ag8 Ag10 Ag7 Ag6 Te10 Te12 Ag9 Ag11 Ag12

4h 3g 3g 4h 12l 3f 2e 6i 6k 6i 12l 6i 6i 2e 3f 12l 12l 6j 6i 2e 3f 12l 12l 12l

0 /3 0.1776(3) 0.5691(4) 0.3436(4) 0 −0.3151(4) 0 0.2113(5) 0.503(4) 0 0 0.1561(17) 0.1393(6) 0.4292(12)

1

0 /3 0.1777(1) 0.5696(2) 0.3442(2) 0 0.3154(2) 0 0.2109(2) 0.499(2) 0.1251(17) 0 0.1541(8) 0.1385(3) 0.4268(7)

1

2

/3 0.3122(3) 0.6722(3) 2 /3 0.7002(2) 0.6624(3) 0 0.4301(2) 0.5637(4) 0.8062(2) 0.6550(2) 0.4271(5) 0.8028(3) 0 0.6808(3) 0.7922(4) 0.5820(4) 0.5653(7) 0.5619(11) 0 −0.1263(12) 0.709(2) 0.5484(10) 0.552(2)

y Ag4.72(3)Te3 (I) 0.3322(4) 2 /3 0.2995(3) 0.5691(4) 0.5389(4) 0.1950(4) 0 0 0.4180(5) 0.451(3) −0.124(5) 0 0.4506(17) 0.5741(7) 0.4292(12) Ag4.66(1)Te3 (II) 0.3323(2) 2 /3 0.2997(1) 0.5696(2) 0.5394(2) 0.1952(2) 0.3154(2) 0 0.4183(2) 0.4521(11) 0.1251(17) 0 0.4510(8) 0.5735(4) 0.4268(7) Ag4.96(2)Te3 (III) 1 /3 0.3122(3) 0.6722(3) 1 /3 0.8780(1) 0.6624(3) 0 0 0.4283(4) 0.8062(2) 0.1956(3) 0.4271(5) 0.8028(3) 0 0 0.5811(3) 0.7948(5) 0.4235(7) 0 0 0 0.552(1) 0.7038(17) 0.052(3)

(I), 0.34(1) (II)], because the crystal structures of Ag5.00−xTe3 [x = 0.28(3) (I), 0.34(1) (II)] feature substantial disorders on diverse atomic

z

Uiso/Ueq, Å2

Occ. (