Unconventional Clathrates with Transition Metal–Phosphorus

Dec 19, 2017 - In this Account, we focused on a unique class of inclusion compounds, intermetallic clathrates, which exist in a variety of structures ...
0 downloads 4 Views 5MB Size
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article Cite This: Acc. Chem. Res. 2018, 51, 31−39

pubs.acs.org/accounts

Unconventional Clathrates with Transition Metal−Phosphorus Frameworks Published as part of the Accounts of Chemical Research special issue “Advancing Chemistry through Intermetallic Compounds”. Jian Wang,†,‡,⊥ Juli-Anna Dolyniuk,§,⊥ and Kirill Kovnir*,†,‡ †

Department of Chemistry, Iowa State University, Ames, Iowa 50011, United States Ames Laboratory, U.S. Department of Energy, Ames, Iowa 50011, United States § Department of Chemistry, University of California, Davis, California 95616, United States ‡

CONSPECTUS: In this Account, we focused on a unique class of inclusion compounds, intermetallic clathrates, which exist in a variety of structures and exhibit diverse physical properties. These compounds combine covalent tetrahedral frameworks with rattling guest atoms situated inside their framework cages. Tetrels, the group 14 elements, are the basis for conventional clathrates because they fulfill the bonding requirement of four electrons per framework atom. In analogy to the replacement of Ge with GaAs in semiconductors, we focused on unconventional tetrel-free clathrates with frameworks composed of phosphorus and late transition metals. Compared to tetrels, these elements exhibit greater flexibility in their local coordinations and bonding. Tetrel elements cannot tolerate high deviations from regular tetrahedral coordinations. Thus, they exile a number of theoretically predicted framework topologies that are composed of a single type of polyhedral cage with square faces, such as the truncated octahedron. Unconventional clathrates are capable of stabilizing both envisaged and unique, unforeseen topologies. Clathrate structures with guest atoms held inside their cages by weak electrostatic interactions are predicted to be efficient thermoelectrics due to their low thermal conductivities. Unconventional clathrates may exhibit ultralow thermal conductivities, below 1 W m−1 K−1, without a need for heavy elements in their frameworks. The different chemical natures of transition metals and phosphorus led to their segregation over different framework positions, fulfilling the elements’ specific local coordination and bonding requirements. This resulted in the formation of short- and long-range ordered superstructures with complex phonon dispersions and ultralow thermal conductivities. Aliovalent substitutions are commonly used to tune charge carrier concentrations in materials science. They are often performed under the “doping” assumption that substitutions with neighboring elements in the periodic system should not affect the parent structure but only adjust the charge carrier concentrations. This is not the case for unconventional clathrates. We investigated the tunability of the prototype Ba8Cu16P30 clathrate by the aliovalent substitution of Cu with either Zn or Ge. These substitutions resulted in significant alterations of the local chemical bonding and led to the rearrangement of the whole crystal structure. Remarkable thermoelectric properties were achieved for the substituted unconventional clathrates, exhibiting an overall order of magnitude increase in the thermoelectric performance. Aliovalent substitution allowed us to vary the charge carrier concentration in one structure type within the limits of the structure’s stability. Exceeding these limits in the Ba−Cu−Zn−P system resulted in a transition from the covalent 2c−2e bonding found in conventional clathrates to the multicenter bonding common for metal-rich intermetallic compounds. This caused the complete rearrangement of the crystal structure into a new unique clathrate where a majority of the framework atoms are fiveand six-coordinated, and all metal atoms are joined in Cu−Zn dumbbells. Our work shows that unconventional clathrates exhibit diverse crystal structures and unique chemical bonding, which result in tunable transport properties. While they are similar to their tetrel counterparts in some ways, they are very different in others. Specifically, the high thermal and chemical stabilities and low thermal conductivities of unconventional clathrates make them promising bases for further development of thermoelectric materials. type of hydrate inclusion compound,2 citing a Latin root originating from the word clathri, meaning lattice. The structures of clathrates attracted significant attention from scientists around

1. CONVENTIONAL TETREL CLATHRATES The first clathrate to be documented in the scientific literature was synthesized more than 200 years ago by Sir Humphry Davy who passed chlorine gas into cold water and obtained a chlorine hydrate, [H2O]46(Cl2)6.4.1 The term “clathrate” was not introduced until 1948 when Powell used it to describe a specific © 2017 American Chemical Society

Received: September 25, 2017 Published: December 19, 2017 31

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39

Article

Accounts of Chemical Research the world, including Pauling,3 but the first complete crystal structure of a clathrate was not determined until 1959 by Allen, who studied a chlorine hydrate very similar to one Davy synthesized.4 Since then, numerous crystal structures have been determined for clathrate hydrates composed of tetrahedral frameworks of water molecules held together by hydrogen bonding. Molecular guest species, such as chlorine molecules, are trapped inside large cavities formed in the framework, and they form no covalent bonds to water. The clathrate hydrates are classified by the topology of their water frameworks and are labeled by Roman numbers I, II, ..., VII.5 The first intermetallic examples of clathrate structures were produced in 1965 upon the controlled decomposition of sodium silicides.6 In the crystal structures of Na8Si46 and Na24Si136, silicon atoms form tetrahedral frameworks analogous to those of the gas hydrate types I and II. Na atoms are situated inside the cages of the silicon framework, and they only interact with the Si framework by means of weak electrostatic forces. Since the first intermetallic clathrate was discovered, over 200 intermetallic clathrates have been synthesized, and the tetrahedral framework of the vast majority of them is based on one of the tetrels, the group 14 elements: Si, Ge, and Sn. Intermetallic clathrates crystallize in a variety of structure types that are similar to and different from the known clathrate hydrates. Tetrel-based clathrates exhibit diverse physical properties related to their unique crystal structures and a wide variety of framework compositions.7−10 Clathrates are currently researched for their applications in the fields of photovoltaics, thermoelectrics, superconductors, and Li-ion batteries.9,10 Unconventional clathrates are tetrel-free. Frameworks of such clathrates are based on the transition metals of groups 10−12 and pnictogens, P, As, or Sb. The rich chemistry of phosphorus-based unconventional clathrates is discussed below.

Figure 1. Crystal structures of the intermetallic clathrates based on tetrel elements. Clathrate-forming polyhedra are shown in the middle. Note that the pentagonal dodecahedron is overemphasized, it is the smallest polyhedron found in clathrates. Pentagonal dodecahedron, blue; tetrakaidecahedron, yellow; pentakaidecahedron, green; hexakaidecahedron, orange.

2. TOPOLOGY Intermetallic clathrates based on the tetrel elements form several different clathrate structure types, I, II, III, IV, VIII, and IX. The latter two types are unique to intermetallic clathrates and have yet to be found in the hydrates. The main building block of all but the type VIII structures is a pentagonal dodecahedron (Figure 1). The bond angle in this polyhedron is 108°, which is very close to the regular tetrahedral angle of 109.5°. This explains the popularity of this building block among the tetrel clathrates. From a space tiling point of view, the pentagonal dodecahedron alone cannot tile space in a periodic manner due to its 5-fold symmetry. To resolve this, in the structure of intermetallic clathrates, the pentagonal dodecahedron is combined with less symmetric polyhedra composed of pentagonal and hexagonal faces (Figure 1). Calculations of the hypothetical tetrel clathrate structures have shown that the structures formed by polyhedra with square faces are substantially strained, and they were predicted to be unstable when formed by Si, Ge, or Sn.11 The simplest example of a square-faced polyhedron is a truncated octahedron, present in the sodalite zeolite (Figure 2). The advantage of tiling space with square-faced polyhedra is that they would allow all space to be tiled by only one type of polyhedron. The crystallographic question of how to tile space with only one type of polyhedron with more than 8 faces was answered well before the first diffraction experiment was conducted. Lord Kelvin in 1887 suggested a truncated octahedron as a tile to achieve “the division of space with minimum partitional area”.12 While the original paper contains the simplest and most symmetrical answer to this question, a packing based on a

Figure 2. A comparison of the pentagonal dodecahedron to the truncated octahedron.

truncated octahedron (Figure 2), a more comprehensive answer, covering many other possibilities for tiles with less than 16 faces, was reported recently by O’Keeffe et al.13 In these proposed atomic arrangements, each framework atom is connected to another 4 framework atoms, and each polyhedron is composed of square and hexagonal faces, with or without additional pentagonal faces. The valence bond angles suggested in these polyhedra are quite different from the regular tetrahedral angle. Thus, all such polyhedra are considered to be unstable when formed by tetrel elements. In our work, we focused on unconventional clathrate compounds that are tetrel-free. The frameworks of these compounds are formed by group 15 elements, that is, phosphorus, and the late transition metals of groups 10−12, such as Ni, Cu, Zn, and their heavy 4d and 5d analogues. In such cases, the framework-building elements are known to adopt a range of local coordinations, including a square-planar coordination for Ni and Cu, and P3 and P4 rings for phosphorus.14,15 Two clathrate types predicted by O’Keeffe have been realized by unconventional clathrates with AM2P4 compositions (A = Sr, Ba; M = Ni, Pd, Cu).16−19 Both structure types are unknown for the tetrel-based intermetallic clathrates. In 32

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39

Article

Accounts of Chemical Research the crystal structures of BaNi2P4 and BaPd2P4, Ni(Pd) and P form truncated octahedra encapsulating Ba atoms (Figure 3).

Figure 4. Crystal structure of the novel clathrate Ba8M24P28+δ. (M = Cu, Zn; δ = 0.5−2.1). Top, two types of clathrate-forming polyhedra, the pentagonal dodecahedron (blue) and the 22-vertex polyhedron (cyan). Note that several rhomboid faces can be split into two triangular faces when M−M bonds are considered (not shown). Bottom, packing of the polyhedra in the crystal structure. M, black; P, yellow; Ba, cyan and blue.

Figure 3. Structures of two unique types of unconventional clathrates AM2P4: (left) clathrate-VII built from truncated octahedra; (right) twisted clathrate built from unique polyhedra called twisted Kelvin cells. Pictograms of the number and types of polyhedral faces, and Schlegel diagrams are also shown above. M, black; P, yellow; A, purple.

KCu4Te11,28,29 or fused together in 1D or 2D fragments, such as chains of Rb@Cd8Sb12 pentagonal dodecahedron in the structure of Rb16Cd25Sb36 or layers of Ba@Cu6Sb12 polyhedra in the structure of BaCu7.3Sb5.30,31 Truncated octahedral Cs@Cu12Te12 cages have also been found in the crystal structure of Cs3Cu20Te13.32 Such cages share hexagonal edges with the voids between the cages filled with additional Cu and Te atoms. Nevertheless, the complete space tiling typical of clathrates is not achieved in all of the aforementioned cases. To realize new tetrel-free clathrates, a combination of two factors is required: advanced synthetic techniques should be combined with a knowledge of the relationships between the composition, sizes of framework and guest atoms, electron count, chemical bonding, and transport properties. Some of those aspects are discussed below.

Alternatively, in the crystal structures of SrNi2P4 and BaCu2P4 a unique polyhedron called a twisted Kelvin cell is realized, which is composed of square, pentagonal, and hexagonal faces (Figure 3). The square M2P2 faces in the structures of AM2P4 clathrates have relatively short diagonal M-M distances of 2.8−2.9 Å. No localized electron density was revealed between transition metal atoms by Electron Localization Function (ELF) analysis, and this was confirmed by crystal orbital Hamilton population (COHP) analysis showing filled antibonding M−M states.19 Although only one type of polyhedral cage is possible in the structures of AM2P4, the allowance of two different cage types drastically increases the tiling possibilities.20−24 We have recently shown that a clathrate structure can be realized by the combination of pentagonal dodecahedra and another unique type of 22-vertex polyhedron composed of square, pentagonal, and hexagonal faces (Figure 4).25 The metal−phosphorus unconventional clathrates exhibit significant flexibility in their local coordinations, already realizing three new clathrate types not known for tetrel-based clathrates. We believe more compounds with new topologies are yet to be synthesized in this family. The family of tetrel-free clathrates is not limited to P-based compounds, and there are several hints that many members of this family with different pnictogens and chalcogens are yet to be discovered: 1. Inverse, tetrel-based clathrates with anionic guests where either pnictogens (P, As, and Sb) or chalcogens (S, Se, and Te) represent a significant fraction of the framework atoms have been reported.7−9 2. Clathrate-I structures with (Zn,Cd)-As and (Zn,Cd)-Sb frameworks have been reported.26,27 3. Compounds where the main building block is a pentagonal dodecahedron or similar high coordination number polyhedron composed of a transition metal and either pnictogens or chalcogens have been reported. Such building blocks may be isolated as A@Cu8Te12 polyhedra (A = K, Ba) in the crystal structures of BaCu6SeTe6 and

3. TUNING THE THERMAL CONDUCTIVITY Clathrates are promising thermoelectric materials, capable of converting heat into electrical energy or electrical energy into refrigeration.9,10,33 In thermoelectrics, a temperature gradient applied to a combination of p- and n-type semiconductors leads to the migration of main charge carriers from the hot side of the system to the cold side, creating an electrical current. Efficient thermoelectrics should exhibit low thermal conductivity (κ) with high electrical conductivity (σ) and high thermopower (S). Their efficiency can be characterized by the dimensionless figure-ofmerit, zT, zT = TS2σ/κ, where T is the absolute temperature. Realizing low thermal conductivities in crystalline, narrowbandgap semiconductors is not a trivial task. The total thermal conductivity is a sum of the electronic and lattice contributions: κE and κL, respectively. While the κE is dependent on the charge carrier concentration and mobility, κL is dictated by the crystal structure and phonon dispersion. The lowest thermal conductivities are achieved in amorphous materials, which lack longrange ordering (Figure 5). Due to poor electrical conductivities, such materials are not suitable for thermoelectric application. 33

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39

Article

Accounts of Chemical Research

atoms. A second important contributor to reduced thermal conductivity is the anharmonic vibrations of the heavy atoms present in the framework, such as Sn. Finally, Sn-based compounds are predominantly electron-balanced semiconductors with negligible contributions of the charge carriers to the total thermal conductivity. For unconventional clathrates, the lowest thermal conductivity was observed for the compound with the heavy element Au in the framework, Ba8Au16P30.35 The lattice thermal conductivity, κL, for this compound is as low as 0.2 W m−1 K−1 at room temperature, approaching the theoretical minimum for solids. Unlike in the tetrel-based clathrates, multiple unconventional clathrate compounds with frameworks composed of relatively light atoms, such as Cu, Zn, Ge, and P, exhibit thermal conductivities below 1.5 W m−1 K−1, comparable to the best values for tetrel clathrates.25,36−38 In addition to rattling guest atoms, this behavior can be attributed to the different chemical natures of the late transition metal and phosphorus atoms. In tetrel-based clathrates, joint occupancy of the same crystallographic position by two different elements is quite common, for example, Si/Al or Sn/In. In contrast, in the unconventional clathrates, phosphorus and transition metal atoms prefer to segregate over different framework positions with distinct coordinations. This often results in the formation of superstructural long-range ordering or short-range ordering and increase of the unit cell volume.10,25,38,39 The lattice thermal conductivity is inversely proportional to the structural complexity and number of atoms in the primitive unit cell.40 AM2P4 clathrates exhibit higher values of their total thermal conductivities due to charge carrier contributions, κE. Through a combination of heat capacity analysis, variable-temperature single crystal refinements, and phonon dispersion calculations, we have shown that the AM2P4 clathrates exhibit PGEC behavior and rattling of the guest cations. Despite only having a single guest atom situated in one unique crystallographic site, multiple Einstein vibrational modes are obtained from heat capacity fits, and multiple vibrational peaks are seen in their phonon density of states calculations.36 Thus, AM2P4 clathrates are similarly complex to their tetrel counterparts.

Figure 5. Examples of the room temperature thermal conductivities for different materials.

Traditional thermoelectric materials are narrow bandgap semiconductors composed of heavy elements, such as Bi2Te3 or PbTe, where low thermal conductivity is achieved by the combination of the anharmonic rattling of elements, engineered substitutional disorder, and defects. Two decades ago Slack introduced the phonon glass−electron crystal (PGEC) concept, pointing to semiconductors with cage-like structures and guest atoms trapped inside the cages as potentially efficient thermoelectrics.34 The rattling of guest atoms provides effective scattering of the heat carrying phonons, thus decreasing the lattice thermal conductivity, κL, while charge carrier transport occurs through the covalent framework. Clathrates are PGEC materials with thermal conductivities of below 3 W m−1 K−1 (Figure 6). The rattling of guest atoms was confirmed by X-ray and neutron scattering techniques, Raman spectroscopy, and computations.9,10 For the tetrel-based clathrates, the Sn-based compounds exhibit the lowest thermal conductivities (Figure 6) and the highest thermoelectric figures-of-merit, zT. This can be explained by several factors. First, increases in the average bond lengths of the framework going from Si−Si to Ge−Ge to Sn−Sn may result in larger cages with more space available for the rattling of guest

4. TUNING THE CHARGE CARRIER TRANSPORT Controlling and understanding the electronic structure of clathrates is crucial for the development of new technologies that take advantage of their unique combination of attributes. The Zintl concept is a tool often applied to clathrates as it can be used to predict the electronic properties. Zintl originally develop his concept for salt-like intermetallic compounds of the electropositive cations and anions of group 13−15.41,42 The premise of the Zintl concept rests on two key factors, first, a cationic guest is expected to donate its valence electrons to an anionic framework, and second, all atoms are expected to realize an electron octet configuration.42 The Zintl concept appeared to be very useful for the prediction of compositions and properties of conventional tetrel-based clathrates,9,10 although more comprehensive computational studies show that deviations from the predicted compositions are possible.43 Late transition metals of groups 10 and 11 can behave similarly to classical Zintl elements; thus within certain margins, the Zintl concept can be applied to predict properties and compositions of the unconventional clathrates. We investigated the tunability of the electronic properties of unconventional clathrates based on the prototype system Ba8Cu16P30, which crystallizes in the superstructure of the

Figure 6. Room temperature values of the total thermal conductivities for selected representatives of tetrel and unconventional clathrates. In the latter case, different clathrate types are denoted by polyhedra similar to the ones shown in Figures 3 and 4, while clathrate I systems are shown as pentagonal dodecahedra. 34

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39

Article

Accounts of Chemical Research clathrate I type, in the orthorhombic space group Pbcn. In the crystal structure of Ba8[Cu16P30], Cu and P atoms segregate over 23 framework sites in such a way that there are no Cu−Cu bonds and every Cu is surrounded by four P atoms.39 Ba atoms are encapsulated in the large cavities of the Cu−P framework (Figure 7). The arystotype of clathrate I (space group Pm3̅n) crystallizes

Tetrahedrally-coordinated Cu and P atoms require 4 electrons per atom to realize an electron octet. Per formula unit, 46 × 4 = 184 electrons are required. By assuming that all Cu atoms maintain a stable d10 electron configuration, each Cu atom can provide a single s-electron for bonding. With two valence electrons from Ba and five from P, this leads to a total valence electron count (VEC) of [1 × 16 (Cu)] + [5 × 30 (P)] + [2 × 8 (Ba)] = 182 electrons, two fewer than the required 184. From this count, metallic properties with holes as the main charge carriers are predicted, a result that has been verified experimentally and by state-of-the-art quantum-chemical computations (Figure 8, right).37,38,44 To achieve an electron balance and semiconducting properties in the Ba8Cu16P30 compound, a part of the framework atoms should be aliovalently substituted to add two electrons per formula unit. Our first choice was to replace Cu with Zn atoms.38 For the composition Ba8Cu14Zn2P30 a drastic increase in the thermopower was observed in accordance with our predictions (Figure 8). According to the Zintl concept, an increase in the Zn concentration over x = 2 in Ba8Cu16−xZnxP30 should result in metallic compounds with electrons as the main charge carriers. For example, for the composition Ba8Cu11Zn5P30, a VEC of 187 electrons is expected, [1 × 11 (Cu)] + [2 × 5 (Zn)] + [5 × 30 (P)] + [2 × 8 (Ba)]. Contrary to predictions, all compositions with x > 2 appeared to be p-type semiconductors. Moreover, an

Figure 7. Unit cells of orthorhombic (cyan) and cubic (black) clathrate I. Ba, yellow and blue; Cu and P, white.

in a simpler cubic structure with only three crystallographic sites in the framework and a volume four times smaller than the Pbcn superstructure.

Figure 8. Left, Measured thermopower (top) and electrical resistivities (bottom) are shown for Ba8M16+yP30−y (M = Cu/Zn). Note that the composition Ba8Cu14Zn2P30 corresponds to 12.5% Zn/Mtotal on the left figures. The different backgrounds separate regions of Ba8Cu16−xZnxP30 (gray) and Ba8Cu16−xZnxP30−yZny (white). Right, Density of states calculations for Ba8Cu16P30 (top) and Ba8Cu14Zn2P30 (bottom).36,38 35

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39

Article

Accounts of Chemical Research increase in the Zn concentration over 2 atoms per formula unit inadvertently resulted in complex rearrangements and a collapse of the superstructure into the cubic subcell (space group Pm3̅n), where every framework position is occupied by all three element types, Cu, Zn, and P. To understand these structural changes, we applied a set of advanced characterization techniques, including synchrotron X-ray and neutron powder diffractions, pair distribution function analysis, scanning-transmission electron microscopy, and solid state 31P NMR spectroscopy. We showed that for the compositions with x > 2, in addition to Cu/Zn substitution, the formation of Cu−Zn bonds in the framework is accompanied by P/Zn substitution. For the aforementioned compound with five Zn atoms this results in the following electron balanced composition: Ba8(Cu11.75Zn4.25)(P29.25Zn0.75) VEC = 184 = [1 × 11.75 (Cu)] + [2 × 5 (Zn)] + [5 × 29.25 (P)] + [2 × 8 (Ba)]. This result shows that, although there is segregation of M and P atoms over different framework positions in unconventional clathrates, those sublattices should not be considered fully independent, and the aliovalent modification of the clathrates’ compositions and structures will affect chemical bonding in both M and P sublattices. Although clathrates were discovered over 200 years ago, only three- and four-coordinated frameworks had been reported. In many ternary clathrates, all framework positions are fully occupied, leading to tetrahedral frameworks. In some cases, to accommodate excess electrons, framework vacancies are present, since five electrons are required to realize an electron octet by three-coordinated atoms due to the presence of an electron lone pair. To probe the tunability of the electronic structure of Ba− Cu−Zn−P clathrate I, we attempted to produce the electron-rich compositions of Ba8M16+yP30−y where nearly half of the Cu atoms were substituted with Zn. In contrast to the tetrel-based clathrates, no vacancies were formed in the clathrate framework. Instead, a complete structural rearrangement occurred, resulting in a new clathrate structure type mentioned earlier in the review (Figure 4). The new clathrate Ba8M24P28+δ (M = Cu, Zn; δ = 0.5−2.1) is electron-deficient with less than 4 electrons per framework atom. For example, for the composition Ba8Cu13Zn11P30: VEC/per framework atom = {[1 × 13 (Cu)] + [2 × 11 (Zn)] + [5 × 30 (P)] + [2 × 8 (Ba)]}/54 = 3.7. Electron-deficiency in Ba8M24P28+δ resulted in very different bonding environments than in the clathrate-I system Ba8M16+yP30−y (M = Cu, Zn). The new compound is the first example of a clathrate in which framework atoms have coordination numbers higher than four. The framework of the Ba8M24P28+δ clathrate is mainly composed of five- and sixcoordinated atoms (Figure 9). Through a combination of X-ray and neutron diffractions with solid-state 31P NMR, fine details of the structure were confirmed. In this structure, all metal atoms are joined in Cu−Zn dumbbells, in sharp contrast to Ba8Cu16P30 (no Cu−Cu bonds) and Ba8M16+yP30‑y (some Cu−Zn bonds). In addition, in the structure of Ba8M24P28+δ there is one partially occupied P site capping the 22-vertex polyhedra (P5 in Figure 9) which defines the charge transport properties of the system. With the Fermi level located very near the top of the valence band when δ = 0 (no P5 atoms), a slight increase in the P5 occupancy results in an order of magnitude decrease in the electrical resistivity from 23 mΩ·cm to 2.7 mΩ·cm at 300 K.25 This type of behavior was observed in other phosphorus frameworks with partially occupied P sites.15 Ba8M24P28+δ represents a transition from the covalent 2c−2e bonding found in Zintl clathrates,7−10 to the metal−metal multicenter bonding found in intermetallic compounds.45−47

Figure 9. Coordination environments of Cu (black), P (orange), and the partially occupied P5 site (yellow) in the crystal structure of Ba8Cu13Zn11P30.

5. BRIDGING THE GAP BETWEEN UNCONVENTIONAL AND TETREL-BASED CLATHRATES Combinations of tetrel elements with either P or As, result in positively charged tetrahedral frameworks unsuitable for hosting cationic guest atoms due to the presence of excess electrons. However, tetrel−pnictogen positively charged tetrahedral frameworks can be combined with anionic guests of halogens or chalcogens.7,8,48−52 Those examples of inverted clathrates inspired our research focused on the Ba−M−E−P frameworks were E is a tetrel element. We considered the tetrel elements to be suitable substitutes to regulate the electron count in the unconventional clathrates and focused on the Cu−Ge−P framework in analogy to the Ba−Cu−Zn−P systems. A new clathrate compound, Ba8Cu14Ge6P26, bridging the gap between tetrel-based and tetrel-free clathrates, was synthesized via solid state methods and grown as a large crystal by the vertical Bridgman method.37 Ba8Cu14Ge6P26 crystallizes in the clathrate I crystal structure in the cubic space group Pm3̅n (No. 223). The chemical composition of the new clathrate was determined by a combination of single crystal X-ray diffraction, elemental analysis, and Zintl counting combined with properties measurements. The uniform distribution of the three framework elements, Cu, Ge, and P, as well as the absence of any superstructural or local ordering in Ba8Cu14Ge6P26 were confirmed by synchrotron X-ray diffraction, electron diffraction, high angle annular dark field scanning-transmission electron microscopy, and neutron and X-ray pair distribution function analyses. The introduction of 13% of Ge atoms in the framework is sufficient to remove the preferential Cu−P bonding and suppress the formation of long- or short-range ordering, in sharp contrast to Cu−Zn−P frameworks. Ba8Cu14Ge6P26 is the first representative of an anionic clathrate with a framework composed of three atom types of very different chemical natures: a transition metal, a tetrel element, and a pnictogen. Ba8Cu14Ge6P26 significantly expands clathrate chemistry, not limiting it to only tetrel-based or tetrel-free families. Our recent investigations demonstrated that this chemistry is not unique to the Ge/P combination but can be extended to Si- or Ascontaining analogues. 36

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39

Article

Accounts of Chemical Research

Figure 10. Transport properties of Ba8Cu14Ge6P26 (blue circles) and Ba8Cu16P30 (orange triangles). Top left, thermal conductivity; top right, thermopower; bottom left, electrical resistivity; bottom right, thermoelectric figure of merit, zT. The zT figures are shown for two temperature ranges, 10−400 K and 300−850 K, measured with different instruments. Inset in the middle is the photograph of the cut Bridgman growth crystal of Ba8Cu14Ge6P26 on a background of millimeter-grid paper.37

properties. In a first approximation, the basic electronic structure and the type of charge carriers can be deduced from application of Zintl electron counting. With respect to thermoelectric applications, phosphorus-based clathrates have several advantages over traditional tetrel-based clathrates, such as significantly higher thermal and chemical stability compared to Sn-based clathrates and lower thermal conductivity and less expensive framework compositions compared to Ge-based clathrates. The thermoelectric performance of several Zn- and Ge-containing unconventional clathrates shows their promise as bases for further development of thermoelectric materials by means of aliovalent doping.

Ba8Cu14Ge6P26 holds promise for the development of thermoelectric materials due to its inexpensive framework elements compared to Ge-rich clathrates and its congruent melting. The prototype tetrel-free clathrate, Ba8Cu16P30, is predicted and was experimentally determined to exhibit metallic behavior: low thermopower and low electrical resistivity, which increase with increasing temperature (Figure 10). Ba8Cu14Ge6P26 is an electron-balanced phase with 184 electrons per formula unit: VEC = [1 × 14 (Cu)] + [4 × 6 (Ge)] + [5 × 26 (P)] + [2 × 8 (Ba)] = 184 electrons. As shown in Figure 10, the electrical resistivity of Ba8Cu14Ge6P26 exhibits an exponential decrease with temperature. This is accompanied by a 6-fold increase in the thermopower to 117 μV/K at 400 K compared to 18 μV/K at 400 K for Ba8Cu16P30. The combination of those properties resulted in a significantly enhanced thermoelectric performance for Ba8Cu14Ge6P26, zT = 0.63 at 812 K, almost an order of magnitude higher than the performance of the Ba8Cu16P30, zT = 0.07 at 812 K (Figure 10).



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Jian Wang: 0000-0003-1326-4470 Kirill Kovnir: 0000-0003-1152-1912

6. OUTLOOK In this Account, we have shown that unconventional transition metal−phosphorus clathrates represent a unique group of clathrates with diverse crystal structures and chemical bonding, which result in tunable transport properties. The rich structural chemistry of these compounds is underexplored and there are likely other compounds and new structure types yet to be discovered. We envision that earlier transition metals, chalcogens, and heavy pnictogens will contribute to the breadth of the chemistry of unconventional clathrates. The combination of several state-of-the-art characterization tools is necessary to establish the short- and long-range structure of such compounds. Further, chemical bonding is key to understanding the fine details of the crystal structures and

Author Contributions ⊥

J.W. and J.-A.D. contributed equally.

Funding

This research was supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Science and Engineering, under Award DE-SC0008931. Oak Ridge and Argonne National Laboratories are gratefully acknowledged for the beamtime allocation and general support. Use of the Advanced Photon Source at Argonne National Laboratory was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC0206CH11357. Research conducted at ORNL’s Spallation 37

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39

Article

Accounts of Chemical Research

(3) Pauling, L.; Marsh, R. E. The Structure of Chlorine Hydrate. Proc. Natl. Acad. Sci. U. S. A. 1952, 38, 112−118. (4) Allen, K. W. Chlorine hydrate. J. Chem. Soc. 1959, 4131−4132. (5) Jeffrey, G. A. Hydrate inclusion compounds. In Inclusion Compounds: Physical Properties and Applications; Atwood, J. L., Davies, J. E. D., MacNicol, D. D., Eds.; Academic Press: London, 1984; Vol. 1. (6) Kasper, J. S.; Hagenmuller, P.; Pouchard, M.; Cros, C. Clathrate Structure of Silicon Na8Si46 and NaxSi136 (x < 11). Science 1965, 150, 1713−1714. (7) Kovnir, K. A.; Shevelkov, A. V. Semiconducting clathrates: synthesis, structure and properties. Russ. Chem. Rev. 2004, 73, 923−938. (8) Shevelkov, A. V.; Kovnir, K. Zintl Clathrates. In Principles and Recent Developments in the field of Zintl Ions and Zintl Phases; Fässler, T. F., ed.; Structure & Bonding; Springer, 2011; Vol. 139, pp 97−142. (9) Nolas, G. S., Ed. The Physics and Chemistry of Inorganic Clathrates; Springer Science: Dordrecht, 2014. (10) Dolyniuk, J.; Owens-Baird, B.; Wang, J.; Zaikina, J. V.; Kovnir, K. Clathrate thermoelectrics. Mater. Sci. Eng., R 2016, 108, 1−46. (11) Karttunen, A. J.; Fässler, T. F.; Linnolahti, M.; Pakkanen, T. A. Structural Principles of Semiconducting Group 14 Clathrate Frameworks. Inorg. Chem. 2011, 50, 1733−1742. (12) Thomson, W. LXIII. On the division of space with minimum partitional area. Philos. Mag. 1887, 24, 503−516. (13) Delgado-Friedrichs, O.; O’Keeffe, M. Isohedral simple tilings: binodal and by tiles with ≤ 16 faces. Acta Crystallogr., Sect. A: Found. Crystallogr. 2005, 61, 358−362. (14) von Schnering, H. G.; Hönle, W. Chemistry and Structural Chemistry of Phosphides and Polyphosphides. 48. Bridging Chasms with Polyphosphides. Chem. Rev. 1988, 88, 243−273. (15) Wang, J.; Kaseman, D.; Lee, K.; Sen, S.; Kovnir, K. Enclathration of X@La4 tetrahedra in channels of Zn-P frameworks in La3Zn4P6X (X = Cl, Br). Chem. Mater. 2016, 28, 4741−4750. (16) Dünner, J.; Mewis, A. Barium Copper Phosphide (BaCu2P4) − a New Ternary Polyphosphide. J. Less-Common Met. 1990, 167, 127−134. (17) Johrendt, D.; Mewis, A. BaPdP, Ba2PdP3 and BaPd2P4 Zintl Compounds or Intermetallic Phases? J. Alloys Compd. 1994, 205, 183− 189. (18) Keimes, V.; Johrendt, D.; Mewis, A. BaNi2P4: Dimorphism by Peierls Instability? Z. Anorg. Allg. Chem. 1995, 621, 925−930. (19) Dolyniuk, J.; Wang, J.; Lee, K.; Kovnir, K. Twisted Kelvin cells and truncated octahedral cages in the crystal structures of unconventional clathrates, AM2P4 (A = Sr, Ba; M = Cu, Ni). Chem. Mater. 2015, 27, 4476−4484. (20) Delgado-Friedrichs, O.; O’Keeffe, M. Simple tilings by polyhedra with five- and six-sided faces. Acta Crystallogr., Sect. A: Found. Crystallogr. 2010, 66, 637−639. (21) Dutour Sikirić, M.; Delgado-Friedrichs, O.; Deza, M. Space fullerenes: A computer search for new Frank-Kasper structures. Acta Crystallogr., Sect. A: Found. Crystallogr. 2010, 66, 602−615. (22) Dutour Sikirić, M.; Deza, M. Space fullerenes: Computer search for new Frank-Kasper structures II. Struct. Chem. 2012, 23, 1103−1114. (23) O’Keeffe, M.; Peskov, M. A.; Ramsden, S. J.; Yaghi, O. M. The Reticular Chemistry Structure Resource (RCSR) Database of, and Symbols for, Crystal Nets. Acc. Chem. Res. 2008, 41, 1782−1789. (24) Karttunen, A. J.; Fässler, T. F. Structural Principles and Thermoelectric Properties of Polytypic Group 14 Clathrate-II Frameworks. ChemPhysChem 2013, 14, 1807−1817. (25) Dolyniuk, J.; Zaikina, J. V.; Kaseman, D. C.; Sen, S.; Kovnir, K. Breaking the Tetra-Coordinated Framework Rule: New Clathrate Ba8M24P28+δ (M = Cu/Zn). Angew. Chem., Int. Ed. 2017, 56, 2418−2422. (26) Liu, Y.; Wu, L.-M.; Li, L.-H.; Du, S.-W.; Corbett, J. D.; Chen, L. The antimony-based type I clathrate compounds Cs8Cd18Sb28 and Cs8Zn18Sb28. Angew. Chem., Int. Ed. 2009, 48, 5305−5308. (27) He, H.; Zevalkink, A.; Gibbs, Z. M.; Snyder, G. J.; Bobev, S. Synthesis, Structural Characterization, and Physical Properties of the Type-I Clathrates A8Zn18As28 (A = K, Rb, Cs) and Cs8Cd18As28. Chem. Mater. 2012, 24, 3596−3603. (28) Mayasree, O.; Sankar, C. R.; Cui, Y.; Assoud, A.; Kleinke, H. Synthesis, Structure, and Thermoelectric Properties of Barium Copper

Neutron Source was sponsored by the Scientific User Facilities Division, Office of Basic Energy Sciences, U.S. Department of Energy. Notes

The authors declare no competing financial interest. Biographies Jian Wang was born in LinYi, China. He received his B.A. in Inorganic Non-Metal Material Engineering from Changchun University of Science and Technology in 2008, and Ph.D. in Material Physics and Chemistry from Shandong University in 2013. From 2013 to now, he works as a postdoctoral researcher in Prof. Kovnir’s group. Jian develops various synthesis methods to discover novel inorganic compounds to reveal their structure−properties relationships. The characterization of thermoelectric, nonlinear optical, and magnetic properties of various types of inorganic compounds, such as but not limited to Zintl phases, clathrates, phosphides, and chalcogenides, are currently undertaken by Jian. Juli-Anna Dolyniuk (Johnson) grew up in Washington state, USA. She started analytical honors chemistry research in the lab as an undergraduate, working with Prof. J. Schreifels, and she received her B.S. in Chemistry with honors from George Mason University, Fairfax, Virginia, in 2011. She graduated from the University of California, Davis, in 2017 with a Ph.D. in Chemistry, inorganic focus. While at Davis, JuliAnna worked with Prof. K. Kovnir, pioneering research in the development and optimization of unconventional clathrates. Her work led to new discoveries of many different structure types, mainly relating to phosphides. Juli-Anna enjoys solving crystal structures, pushing the boundaries of traditional elemental substitutions, and evaluating complex structure−properties relationships. Kirill Kovnir grew up in Kirovograd (now Kropyvnytskyi), Ukraine. He studied chemistry and received Ph.D. at the Lomonosov Moscow State University with Prof. A. V. Shevelkov scrutinizing inverse clathrates. Afterwards he was tunneling between Max Planck Institute for Chemical Physics of Solids in Dresden and Fritz Haber Institute of the Max Planck Society in Berlin, exploring the potential of intermetallic compounds in heterogeneous catalysis supervised by Profs. Yu. Grin and R. Schlögl. In 2008, he moved to Florida State University where he acquired a comprehensive knowledge of magnetism of complex solids under guidance of Prof. M. Shatruk. Kirill started his independent career in 2011 at UC Davis where he was promoted to Associate Professor. In July 2017 Kirill relocated his group to Iowa State University where he accepted a John D. Corbett Associate Professor of Chemistry position. Kirill’s research interests are in the broad field of solid-state and materials chemistry. Research in his group is focused on synthesis of novel thermoelectric, superconducting, magnetic, catalytic, and lowdimensional materials and exploring their crystal structure, chemical bonding, and physical properties. Understanding the structure− property relationship is a key to the rational design of such materials.



ACKNOWLEDGMENTS We thank all current and former Kovnir group members and all collaborators, especially Profs. Susan M. Kauzlarich and Sabyasachi Sen (UC Davis) and Dr. Oleg I. Lebedev (CRYSMAT) for their contributions to the success of this research.



REFERENCES

(1) Davy, H. On a Combination of Oxymuriatic Gas and Oxygene Gas. Philos. Trans. R. Soc. London 1811, 101, 155−162. (2) Powell, H. M. The structure of molecular compounds. Part IV. Clathrate compounds. J. Chem. Soc. 1948, 61−73. 38

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39

Article

Accounts of Chemical Research Polychalcogenides with Chalcogen-Centered Cu Clusters and Te22‑ Dumbbells. Eur. J. Inorg. Chem. 2011, 2011, 4037−4042. (29) Park, Y.; Kanatzidis, M. G. Synthesis and characterization of potassium copper telluride K4Cu8Te11: a novel solid-state chalcogenide compound with a dodecahedral cluster as a building block. Chem. Mater. 1991, 3, 781−783. (30) Zheng, W.-Z.; Wang, P.; Wu, L.-M.; Chen, L.; Liu, Y. Rb16Cd25.39(3)Sb36: An Electron-Deficient Zintl Phase Containing Infinite Dodecahedron Chains. Inorg. Chem. 2010, 49, 5890−5896. (31) Zheng, W.-Z.; Wang, P.; Wu, L.-M.; Chen, L. Synthesis, Crystal and Electronic Structures, and Physical Properties of Caged Ternary CuRich Antimonide: BaCu7.31(3)Sb5. Inorg. Chem. 2010, 49, 7491−7496. (32) Huai, W. J.; Shen, J.-N.; Lin, H.; Chen, L.; Wu, L.-M. ElectronDeficient Telluride Cs3Cu20Te13 with Sodalite-Type Network: Syntheses, Structures, and Physical Properties. Inorg. Chem. 2014, 53, 5575−5580. (33) Owens-Baird, B.; Heinrich, S.; Kovnir, K. Thermoelectric Materials. Encyclop. Inorg. Bioinorg. Chem. 2017, 1−35. (34) Slack, G. A. CRC Handbook of Thermoelectrics; Rowe, D.M., Ed.; CRC: Boca Raton, FL, 1995; pp 407−440. (35) Fulmer, J.; Lebedev, O. I.; Roddatis, V. V.; Kaseman, D.; Sen, S.; Dolyniuk, J.; Lee, K.; Olenev, A. V.; Kovnir, K. Clathrate Ba8Au16P30: The ″Gold Standard″ for Lattice Thermal Conductivity. J. Am. Chem. Soc. 2013, 135, 12313−12323. (36) Dolyniuk, J. Clathrate Thermoelectrics Based on Late Transition Metals and Phosphorus: Synthesis, Properties, and Optimization. Ph.D. Dissertation, University of California, Davis, 2017. (37) Wang, J.; Lebedev, O. I.; Lee, K.; Dolyniuk, J.-A.; Klavins, P.; Bux, S.; Kovnir, K. High-efficiency Thermoelectric Ba8Cu14Ge6P26. Bridging the gap between tetrel-based and tetrel-free clathrates 2017, 8, 8030−8038. (38) Dolyniuk, J.; Whitfield, P. S.; Lee, K.; Lebedev, O. I.; Kovnir, K. Controlling Superstructural Ordering in the Clathrate-I Ba8M16P30 (M= Cu, Zn) through the Formation of Metal-Metal Bonds. Chem. Sci. 2017, 8, 3650−3659. (39) Kovnir, K.; Stockert, U.; Budnyk, S.; Prots, Yu.; Baitinger, M.; Paschen, S.; Shevelkov, A. V.; Grin, Yu. Introducing a Magnetic Guest to a Tetrel-Free Clathrate: Synthesis, Structure, and Properties of EuxBa8−xCu16P30 (0 ≤ x ≤ 1.5). Inorg. Chem. 2011, 50, 10387−10396. (40) Snyder, G. J.; Toberer, E. S. Complex thermoelectric materials. Nat. Mater. 2008, 7, 105−114. (41) Zintl, E. Intermetallische Verbindungen. Angew. Chem. 1939, 52, 1−6. (42) Kauzlarich, S. M., Ed. Chemistry, Structure and Bonding of Zintl Phases and Ions; VCH: Weinheim, Germany, 1996. (43) Bhattacharya, A.; Carbogno, C.; Böhme, B.; Baitinger, M.; Grin, Y.; Scheffler, M. Formation of Vacancies in Si- and Ge-based Clathrates: Role of Electron Localization and Symmetry Breaking. Phys. Rev. Lett. 2017, 118, 236401. (44) Huo, D.; Sasakawa, T.; Muro, Y.; Takabatake, T. Thermoelectric properties of a clathrate compound Ba8Cu16P30. Appl. Phys. Lett. 2003, 82, 2640−2642. (45) Wang, J.; Lee, K.; Kovnir, K. Distorted phosphorus and copper square-planar layers in LaCu1+xP2 and LaCu4P3: synthesis, crystal structure, and physical properties. Inorg. Chem. 2015, 54, 890−897. (46) Grin, Yu.; Savin, A.; Silvi, B. In The Chemical Bond: Fundamental Aspects of Chemical Bonding; Frenking, G.; Shaik, S., Eds.; Wiley-VCH: Weinheim, 2014; pp 345−382. (47) Hoffmann, R.; Zheng, C. Making and breaking bonds in the solid state: the thorium chromium silicide (ThCr2Si2) structure. J. Phys. Chem. 1985, 89, 4175−4181. (48) Von Schnering, H.-G.; Menke, H. Ge38P8I8 and Ge38As8I8, a New Class of Compounds with Clathrate Structure. Angew. Chem. 1972, 84, 30−31. (49) Zaikina, J. V.; Kovnir, K.; Burkhardt, U.; Schnelle, W.; Haarmann, F.; Schwarz, U.; Grin, Yu.; Shevelkov, A. V. Cationic Clathrate-I Si46‑xPxTey (6.6(1) ≤ y ≤ 7.5(1), x ≤ 2y): Crystal Structure, Homogeneity Range, and Physical Properties. Inorg. Chem. 2009, 48, 3720−3730.

(50) Kovnir, K.; Sobolev, A. V.; Presniakov, I. A.; Lebedev, O. I.; Van Tendeloo, G.; Schnelle, W.; Grin, Yu.; Shevelkov, A. V. Sn19.3Cu4.7As22I8: A New Clathrate-I Compound with TransitionMetal Atoms in the Cationic Framework. Inorg. Chem. 2005, 44, 8786− 8793. (51) Zaikina, J. V.; Kovnir, K.; Haarmann, F.; Schnelle, W.; Burkhardt, U.; Borrmann, H.; Schwarz, U.; Grin, Yu.; Shevelkov, A. V. Si172−xPxTey (x = 2y, y > 20) - The First Silicon-Based Cationic Clathrate III with High Thermal Stability. Chem. - Eur. J. 2008, 14, 5414−5422. (52) Zaikina, J. V.; Mori, T.; Kovnir, K.; Teschner, D.; Senyshyn, A.; Schwarz, U.; Grin, Yu.; Shevelkov, A. V. Bulk and surface structure and high temperature thermoelectric properties of inverse clathrate-III in the Si-P-Te system. Chem. - Eur. J. 2010, 16, 12582−12589.

39

DOI: 10.1021/acs.accounts.7b00469 Acc. Chem. Res. 2018, 51, 31−39