MERCOURI G. KANATZIDIS

Tel. (517)-355-9715 ext. 174 FAX (517)-353-1793

I. SOLID STATE INORGANIC CHEMISTRY

Exploratory Synthesis In Molten Salts. Thio- and Seleno-phosphate Compounds

The application of salt fluxes in the synthesis of new solid state compounds has witnessed significant development in the last few years. Particularly important has been the emergence of the molten polychalcogenide flux methodology in the exploratory synthesis of complex chalcogenides. This approach to new chalcogenides has simplified access to low and intermediate temperatures (160-600 ^oC) and contributed to the discovery of some very interesting materials. In many cases the compounds stabilized under polychalcogenide flux conditions are only kinetically stable and cannot be synthesized at higher temperatures. Thermodynamic influences, however, are not entirely avoided by this approach. Lower temperatures, coupled with the presence of a flux, also make possible the use of molecular assemblies as building blocks for incorporation into solid state structures, and consequently the construction of complex multinary solids. The latter two aspects are the subject of our current attention with the main emphasis being the generation, control, and understanding of the reactivity of flux-assembled [P_yQ_z]^n- anions (as well as [Ge_yQ_z]^n- and [Sn_yQ_z]^n- ) and their synthetic utility (Q=S, Se).

This project aims to produce novel materials, interesting from several points of view. Not only do we expect to learn new solid state chemistry and new synthetic methodology, we also anticipate exciting new materials with useful properties. Therefore, we routinely screen for various important chemical and physical properties. Promising materials are studied further and their potential for applications will be evaluated.

Unlike the huge class of (oxo)phosphate compounds, thio- and seleno-phosphates are still a relatively small group. There is a great divide between these two classes as the chemical, structural and physical properties differ substantially. Although more uses exist for the phosphates in the area of catalysis, ceramics, glasses and molecular sieves, the sulfur and seleno-counterparts include compounds which exhibit promising and unique properties. InPS4 is acentric and was found to have high non-linear optical susceptibility. Sn₂P₂S₆ is also acentric (space group Pc) and undergoes a second order, exothermic phase transition from ferroelectric (Pc) to paraelectric (P2₁/c) at 60 °C. This compound is a promising ferroelectric material for use in memory and storage devices. The M₂P₂Q₆ (M=first row transition metal) has been the most important family of compounds which contains the ethane-like [P₂Q₆]^4- fragment. This family has been extensively studied because various members exhibit interesting intercalation chemistry, ion-exchange properties, ferrimagnetism, sensor applications, and non-linear optical (NLO) properties.

Interesting new results have been obtained which has advanced the chemistry of the thiophosphate and selenophosphate class of compounds significantly. This work has already resulted in a number of unusual compounds; examples include

K₃Ag₃P₃Se₉, K₂Cu₂P₄Se₁₀ K₃AuP₂Se₈, Rb₂Au₂P₂Se₆, KAuP₂S₇

K₄In₂(PSe₅)₂(P₂Se₆), Rb₃Sn(PSe₅)(P₂Se₆), Rb₉Ce(PSe₄)₄,

Cs₄Pd(PSe₄)₂, Cs₂PdP₂Se₆, APbPSe₄,Rb₄Pb(PSe₄)₂, Rb₄Eu(PSe₄)₂,

Rb₄Ti₂(P₂Se₉)₂(P₂Se₇), and K₃RuP₅Se₁₀

The investigations with uranium gave some of the most unanticipated and interesting results. The large size of this atom, high anionic charge and ability to adopt different redox states combined to give two very unusual compounds K₂UP₃Se₉ and Rb₄U₄P₄Se₂₆.

Figure 1. Polyhedral representation of K₂UP₃Se₉ looking down the (U₂Se₁₄)x chains.

Figure 2. The (U₂Se₁₄)x chains. The complicated structure is formed by the side-by-side connection of the parallel (U₂Se₁₄)x chains, by [P₂Se₆]^4- ligands, to form what are essentially "pleated" layers.

K₂UP₃Se₉ is formally a U⁴⁺ compound and has a complicated layered structure, see Figures 1 and 2. The compound actually contains [P₂Se₆]^4- anions and so a more descriptive formula would be K⁺₄U⁴⁺₂[P₂Se₆]^4-₃.

Even more intriguing is Rb₄U₄P₄Se₂₆, a compound with a three-dimensional framework and pentavalent U centers. It contains [PSe₄]^3-, Se^2-, and (Se₂)^2- anions and so it is described as Rb⁺₄U⁵⁺₄(PSe₄)₄^3-(Se)₂^2-(Se₂)₄^2-. The U⁵⁺ state was confirmed with magnetic susceptibility measurements (m_eff ~1.85 mB). U⁵⁺ is relatively scarce because of its strong tendency to disproportionate to U⁴⁺ and U⁶⁺. Therefore, U⁵⁺ is of fundamental importance since it has the simplest 5f-electron configuration [Rn]5ƒ¹. This compound provides an unequivocal example of an air stable ƒ¹ actinide model compound for in-depth magnetic and spectroscopic studies of the ƒ¹ state of U.

Figure 3. The open three-dimensional structure of Rb₄U₄P₄Se₂₆ (two views). The mutually perpendicular tunnels ensure that the compound undergoes facile ion-exchange reactions with various alkali metal ions.

The structure possesses Rb⁺ filled interconnected channels which run both in the [100] and [010] directions. The largest size channels have a rectangular cross-section with dimensions 6.95Å x 5.34Å and run in the [010] direction, see Figure 3.

The Rb⁺ ions exchange readily with other smaller cations such as Li⁺, Na⁺ and K⁺ using a mild ion-exchange route (~110 ^oC) which we have developed using metal iodides, see eq. (1).

Undoubtedly, in Rb₄U₄P₄Se₂₆ we have a rare, three-dimensional selenophosphate host material which is robust enough to explore its soft chemistry.

The need to push for even more novel systems remains and so the complexity and diversity of the reactions will steadily increase. We anticipate that research activity in this area of solid state chemistry will grow substantially.

II. THERMOELECTRIC MATERIALS

Recently there has been renewed interest in thermoelectric materials. The efficiency of thermoelectric coolers operating near room temperature is only about 10% of Carnot efficiency; yet the thermodynamics of thermoelectric cooling suggest that achieving close to 100% of Carnot efficiency may be possible. An efficient thermoelectric device is fabricated from two materials, one n-type and the other a p-type conductor. Each material is separately chosen to optimize the figure of merit, zT, where zT = (S^2s /k )T; S is the thermopower, s the electrical conductivity, k the thermal conductivity and T is the temperature. Thus improving device performance means improving S^2s /k , or increasing S while keeping moderate to large carrier densities of one carrier type. All three of these materials properties are determined by the details of the electronic structure and scattering of charge carriers (electrons or holes) and thus are not independently controllable parameters. Thermopower is ability of a materials to fevelop voltage across it upon the application of a temperature differential across it, and vice versa.

Since the thermopower of optimally n- or p- doped Bi₂Te₃ is on the order of ± 200 m V/K, significant improvements in thermoelectric cooling efficiencies - from 100 to 400% - could occur if materials of reasonable carrier density with thermopowers of 300 to 450 m V/K can be found. This project seeks to discover new bulk materials with a vastly improved thermoelectric performance. The approach is to synthesize and study new ternary and quaternary chalcogenide materials with low symmetry crystal structures, narrow band-gaps and complex electronic structures near the band-gap.

This is a multidisciplinary research effort which will engage three teams: a synthetic solid state chemistry group, a materials charge transport and thermal transport characterization team (Prof. Hogan) and a theoretical physics group (Prof Mahanti).

We have initiated an exploratory synthesis program to identify new multinary phases with Bi and Sb with narrow band-gaps which may be suitable as thermoelectric materials.

BaBiTe₃. One such phase is BaBiTe₃, which was prepared at 750 ûC from a K₂Te₄ flux in which a mixture of Bi and Ba were dissolved. The structure of this material is two-dimensional with [BiTe₃]^2- layers alternating with Ba cations, see Figure 1.

Figure 1. The anisotropic structure of BaBiTe₃. The open white circles represent Te atoms.

Figure 2. Variable temperature single crystal thermopower data for p-type and n-type BaBiTe₃.

BaBiTe₃ is a semiconductor with a narrow band-gap of 0.35 eV and has promising electrical properties. The electrical conductivity is reasonably high at ~50 S/cm and its thermoelectric power reaches 200-220 µV/K at room temperature. The thermal conductivity, which is crucial in assessing the compound’s potential, is only 65-70% that of the rhombohedral Bi₂Te₃. The thermal conductivity is suppressed because of the low symmetry and because unit cell is much larger than that of Bi₂Te₃. At room temperature the Seebeck coefficients of both n-type and p-type samples are large, ~-200 µV/K and ~+200 µV/K respectively, see Figure 2. The p-type samples exhibit a kink of unknown origin at ~215 K and an unusual sign reversal below 185 K. This behavior is not observed in the n-type samples where the Seebeck coefficient decreases but remains negative.

The thermal conductivity of BaBiTe3 at room temperature is remarkably low with respect to that of optimized Bi2Te3 alloy (kL ~1.6 W/m-K).

K/Bi/Se phases. Our work also led to b-K2Bi8Se13 and K2.5Bi8.5Se14 which are interesting from the thermoelectric standpoint. Their structures are three-dimensional with K-filled tunnels running along one direction, see Figure 3. The K atoms are loosely held by the Bi/Se frameworks and this is reflected in their large thermal vibration amplitudes determined by the X-ray crystallographic analysis. We believe this is partly responsible for the low thermal conductivity of these materials. The structure of these materials have building blocks which can be thought of as fragments of basic structure types such as Bi2Te3, NaCl and CdI2 type, see Figure 3.

The electrical properties of b-K2Bi8Se13 and K2.5Bi8.5Se14 were measured on single crystal samples and polycrystalline ingot samples, see Figure 4. The highest room temperature conductivity value obtained for single crystals of b -K2Bi8Se13 was 240 S/cm with a weak negative temperature dependence consistent with a semi-metal or a narrow band-gap semiconducting material.

The thermopower data for b-K2Bi8Se13 and K2.5Bi8.5Se14 show large negative Seebeck coefficients (-200 and -100 µV/K at room temperature, respectively), which indicate the charge carriers are electrons (n-type), see Figures 4 and 5.

Figure 3. Left: The structure of K2Bi8S13 and b-K2Bi8Se13. Right: The structure of K2.5Bi8.5Se14.

The room temperature thermal conductivities of b-K2Bi8Se13 and K2.5Bi8.5Se14 are comparable (1.28 and 1.21 W/m-K, respectively) and lower than that of optimized Bi2Te3 alloy. In fact, some pressed pellets of b-K2Bi8Se13, however, showed room temperature k values of 0.83 W/m-K, almost one half that of Bi2Te3 alloy. This clearly suggests that it is possible to achieve lower thermal conductivity in ternary compounds with complex compositions and crystal structures compared to corresponding high symmetry binary compounds.

III. NOVEL NANOCOMPOSITES OF INORGANIC LAYERED MATERIALS WITH ORGANIC MACROMOLECULES

Inorganic layered materials exist in great variety. They possess well defined, ordered intralamellar space potentially accessible by foreign species. This ability enables them to act as matrices or hosts for polymers, yielding interesting hybrid nano-composite materials. During the decade we have focused our efforts on creating such materials with conjugated and saturated organic macromolecules. We have developed several general synthetic routes for inserting polymer chains into host structures and have designed many novel nanocomposites. We have now three synthetic strategies at our disposal: (a) In-situ intercalative polymerization (ISIP) of a monomer using the host itself as the oxidant. The rationale behind intercalative polymerization is that host matrices with high electron affinity can oxidatively polymerize appropriate monomers in their interior. (b) Monomer intercalation followed by topotactic intralamellar solid state polymerization. This route creates conjugated polymers inside non-oxidizing hosts. (c) Direct precipitative encapsulation of polymer chains by colloidally dispersed single layers of a host. This approach gives access to a large variety of nano-composites with many kinds of polymers and hosts.

In-situ intercalative polymerization (ISIP)

We have demonstrated and developed the ISIP concept using primarily V₂O₅.nH₂O xerogel and crystalline FeOCl as redox-intercalation hosts. Both materials possess highly oxidizing layered frameworks capable of delocalizing added electrons. FeOCl is available in single crystal form, whereas the V₂O₅.nH₂O xerogel is ordered only along one-dimension and can be fabricated into films. We have shown that polypyrrole, polythiophene and polyaniline (PANI) and their derivatives can be inserted in between the sheets of these lamellar materials. The resulting intercalation compounds are composed of alternating monolayers of positively charged polymer chains and negatively charged host layers. The in-situ intercalative polymerization of aniline in these two materials has been explored in detail.

Several years ago we suggested that intercalated polymers may order inside crystalline hosts. We have now demonstrated this in the polyaniline/FeOCl system where polyaniline (PANI) molecules order endotaxially due to the good lattice matching between the polymer NH-spacings and those of the Cl ions on the surface of the FeOCl layers. The PANI/FeOCl system is the first polymer intercalation compound in which substantial polymer ordering, endotaxy, is observed.

The PANI/FeOCl system is the first polymer intercalation compound in which substantial polymer ordering, "endotaxy", is observed inside a host material. This ordering is evident by X-ray diffraction. We were successful in inserting PANI into single crystals of FeOCl without significant loss of the single crystal character. The nano-composite crystals diffract X-rays relatively well. Oscillation photographs of such crystals showed the unit cell of the PANI/FeOCl compound to have a 2x2 superstructure in the ac-plane (the layer) due to substantial long-range order of PANI in FeOCl. This order is achieved by the orientation of the polymer chains along certain crystallographic directions, such that it causes a doubling of the periodicity along the a- and c- axes.

The Cl---Cl' distance on the FeOCl layer surface along the [101] direction (diagonal to the a-, c- axes) is 5.2 Å, almost one-half of the repeating unit of PANI. By orienting parallel to the [101] (diagonal) direction, see Figure 2, the polymer can spatially match each of its NH units with a Cl "partner". A diagonal orientation of PANI with respect to the a- and c- axes of the FeOCl (i.e. [101] direction) is most likely because it exploits the nearly exact but fortuitous lattice matching of a FeOCl layer and PANI. It also produces a new orthorhombic unit cell with crystallographic axes a'=2a and c'=2c.⁷ The results of this work set the foundation for the design of highly ordered lamellar organic/inorganic nanocomposites by proper selection (with respect to lattice matching) of the individual components.

The insertion of polyaniline in FeOCl represents a rare example of polymer "crystallization" inside a solid. However, crystal modeling studies show that maximum possible number of Cl---H interactions occur only if the diagonal chain orientation alternates in adjacent galleries. This arrangement is shown schematically in Figure 2.

Direct encapsulation of polymer chains using exfoliated suspensions

This approach is suitable when polymers which cannot be inserted by the above techniques are of interest. Examples include saturated long molecular weight (MW) polymers such as polyethylene oxide (PEO), nylons, cellulose etc. The host materials must be colloidally dispersed into single layers in various solvents. Examples include V₂O₅ layers prepared by a sol-gel process, MoS₂ and [MoO₃]^x-. Due to limited space, we will not elaborate on the preparation and properties of the (polymer)_xMoO₃ family of nano-composites.

We used the exfoliated suspension method to encapsulate several water-soluble polymers such as poly(vinylpyrrolidone) (PVP), poly(propylene-glycol) (PPG) and methyl-cellulose in V2O5 xerogel. A detailed account of this work has been published. We also prepared PEO and PANI intercalates of MoS₂. This methodology is general and suitable for intercalation of very large MW polymers provided these polymers are soluble.

Selected Publications

"Exploratory Synthesis with Molten Al as a Solvent and Routes to Multinary Aluminum Silicides. Sm2Ni(NixSi1-x)Al4Si6 (x=0.18 - 0.27): A New Silicide with a Ferromagnetic Transition at 17.5 K" X.Z. Chen, S. Sportouch, B. Sieve, P. Brazis, C.R. Kannewurf, J. A. Cowen, R. Patschke and M.G. Kanatzidis, Chem. Mater., 1998, 10, 3202-3211.

"Counterion Effects in Pd Polyselenides: Evolution from Molecular to Three-Dimensional Framework Structures" Kang-Woo Kim and Mercouri G. Kanatzidis, J. Am. Chem. Soc., 1998, 120 (32), 8124 -8135.

"Incorporation of A₂Q into HgQ and Dimensional Reduction to A₂Hg₃Q₄ and A₂Hg₆Q₇ (A=K, Rb, Cs; Q=S, Se). Access of Li Ions in A₂Hg₆Q₇ through Topotactic Ion-Exchange" Enos A. Axtell, III, Younbong Park, Konstantinos Chondroudis and Mercouri G. Kanatzidis, J. Am. Chem. Soc., 1998, 120, 124-136.

"High Thermopower and Low Thermal Conductivity in Semiconducting Ternary K-Bi-Se Compounds. Synthesis and Properties of b-K₂Bi₈Se₁₃ and K_2.5Bi_8.5Se₁₄ and their Sb Analogs" D.-Y. Chung, K.-S. Choi, L. Iordanidis, J.L. Schindler, P. W. Brazis, C. R. Kannewurf, B. Chen, S. Hu, C. Uher, and M. G. Kanatzidis, Chem. Mater., 1997, 9, 3060-3071.

"New Directions in Synthetic Solid State Chemistry. Chalcophosphate Salt Fluxes for Discovery of New Multinary Thiophosphate and Selenophosphate Solids." Mercouri G. Kanatzidis, Current Opinion in Solid State and Materials Science, 1997, 2, 139-149.

"Oligomerization Versus Polymerization of Te_x^2- in the Polytelluride Compound BaBiTe₃. Structural Characterization, Electronic Structure and Thermoelectric Properties." D.-Y. Chung, S. Jobic, T. Hogan, C. R. Kannewurf, R. Brec, J. Rouxel and M. G. Kanatzidis, J. Amer. Chem. Soc. 1997, 119, 2505-2515

"Plastic Superconducting Polymer- NbSe₂ Nanocomposites" Hui-Lien Tsai, Jon L. Schindler, Carl R. Kannewurf and Mercouri G. Kanatzidis, Chem. Mater. 1997, 9, 875-878.