LIBRARY MIchIgan State Unlverslty PLACE ll RETURN BOXtonmovothb mum yum 1'0 AVOID FINES Man on or him duo duo. DATE DUE DATE DUE DATE DUE MSUIsMWWVEmeWIm W v __._—__..—w _ __ _ MOLTEN ALKALI METAL POLYCHALCOGENIDE AND POLYCHALCOPHOSPHATE SALT SYNTHESIS OF TERNARY AND QUATERNARY CHALCOGENIDES By Timothy J. McCarthy A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1994 ABSTRACT MOLTEN ALKALI METAL POLYCHALCOGENIDE AND POLYCHALCOPHOSPHATE SALT SYNTHESIS OF TERNARY AND QUATERNARY CHALCOGENIDES By Timothy J. McCarthy Using the now proven alkali metal polychalcogenide flux technique, we decided to explore the chemistry of two Group 15 main group metals, antimony and bismuth. We have isolated a number of interesting compounds with novel structural architectures such as fi-CsBiSz, y-CsBiSZ, KzBiSe13, KBi3Ss, C528b483, and CsSbS6. y-CsBiSZ and KzBiSe13 are potential candidates for thermoelectric cooling applications. Thermal analysis studies were conducted to probe the mechanism of compound formation and the stability of these phases. In addition, high temperature reactions with K23 and Bi283 have resulted in the synthesis of KBi6,33810 and K2Bi3$13 with interesting semi-metallic electrical conductivity and thermoelectric power. We utilized our experience with alkali polychalcogenide fluxes to develop a general methodology by which new thio(seleno)phosphate (or chalcophosphate) compounds can be consistently obtained. These polychalcophosphate fluxes, Ax[PyQ], provide highly charged [PnyP' anions that coordinate to the W“ center. Using [P287]4', [PS4]3', and [P286]4' anions as building blocks, we synthesized several quaternary compounds with bismuth and antimony metals, ABiP287 (A=K,Rb), A3M(PS4)2 (A=K,Rb,Cs; M=Sb,Bi), CS3Bi2(PS4)3, and Na0,163i1,23PZS6. Further exploration using alkali metal polyselenophosphate fluxes has allowed us to investigate the coordination chemistry of the [PZSe6]4' ligand with metals such as Sb, Bi, Mn, Fe, Cu, Ag, and Au. These reactions have resulted in the formation of a variety of new quaternary compounds including KMPZSe6 (M=Sb,Bi), C83M4(PzSe6)5 (M=Sb,Bi), AzM'PzSe6 (A=K,Rb,Cs; M'=Mn,Fe), A2M'2P28e6 (A=K,Rb,Cs; M'=Cu,Ag), and K4Au2(P28e6)2 with novel structure types. Several of these [P28e6]4‘ containing compounds possess band gaps that are potentially attractive for solar cell applications. In this dissertation, the synthesis, characterization and properties of the above compounds will be discussed. Dedicated to Helen McCarthy 1 906-1 994 ACKNOWLEDGMENTS None of the work described in this dissertation would have been possible without the enthusiasm, patience, guidance and support of my research advisor, Professor Mercouri G. Kanatzidis. I would also like to thank Professor Younbong Park for assisting me in the development of my synthetic methodogy, Dr. Sandeep Dhingra for his helpful suggestions and support, and Dr. Ju-Hsiou Liao for his helpful discussions concerning X-ray crystallography. All three set a superior example of what is expected of a Kanatzidis group member. I would also like to thank the rest of the Kanatzidis group for providing an enjoyable and stimulating research environment. I also thank Professor Kannewurf and his research group at Northwestern University for charge transport measurements, Professor Stanley-Pierre Ngeyi (Madonna University) for help with calorimetric measurements, Professor Jerry Cowen for his helpful comments concering magnetic susceptibility measurements, and Dr. Don Ward for helpful discussions on single crystal X-ray diffraction studies. I would also like to thank Troy Tanzer (REU Summer Program in Solid State Chemistry) for his help with the ion-exchange reactions and high temperature syntheses. Most importantly I extend by deepest gratitude to my parents, Joseph and Kathleen McCarthy, for their love and support over the years. Their sacrifices for my benefit have been limitless. The only way I can repay them is by completion of this dissertation. Financial support given by the Herbert T. Graham Fellowship (1989-1992) at the Department of Chemistry at Michigan State University, the National Science Foundation, and the Center for Fundamental Materials Research is greatly appreciated. TABLE OF CONTENTS List of Tables ....................................................................................................... xiv List of Figures..................................................................................................... xxi I . Introduction ........................................................................................... l 1. A Review of Antimony and Bismuth Chalcogenides ................... 2 2. A Review of Metal Thio(seleno)phosphates .................................. 25 3. Alkali Metal Polychalcogenide Fluxes ........................................... 48 II. Molten Salt Synthesis and Properties of Three New Solid State Ternary Bismuth Chalcogenides, fiCsBiSz, y-CsBiSZ and KzBi38e13 (Part 1) ........................................ 67 1. Introduction ...................................................................................... 67 2. Experimental Section ....................................................................... 69 2.1. Reagents.................................................................................... 69 2.2. Synthesis ................................................................................... 69 C528 ............................................................................................ 69 K2$e ........................................................................................... 7O B-CsBiSz (I) ............................................................................... 7O y-CsBiSZ (l I) ............................................................................. 71 KzBi38e13 (l I I) ....................................................................... 72 2.3. Physical Measurements .......................................................... 73 2.4. X—ray Crystallography ............................................................ 76 3. Results and Discussion ..................................................................... 83 3.1. Synthesis and Spectroscopy ................................................ 83 vii 3.2. Description of Structures ..................................................... 85 Structure of B—CsBiSz (I) ....................................................... 85 Structure of KzBi38e13 (I I I) ................................................ 87 3.3. Thermal Analysis.................................................................... 97 3.4. Electrical Conductivity and Thermoelectric Power Measurements ................................ 103 3.5. UV-Visible-Near IR Spectroscopy .......................................... 109 3.6. Conclusion................................................................................. 112 Synthesis, Structural Characterization, Ion-Exchange Properties and Reactivity of KBi3Ss: A New Three-dimensional Sulfide With Large Tunnels (Part 2) ..................................................................................................... 116 1. Introduction....................................................................................... 116 2. Experimental Section ....................................................................... 118 2.1. Reagents .................................................................................... 118 2.2. Synthesis ................................................................................... 119 K28 ............................................................................................. 119 KBi3Ss ....................................................................................... 119 (H20)Bi3S4,5-(83)0,06 .............................................................. 119 fi-RbBi3Ss ..................................... 120 2.3. Physical Measurements .......................................................... 121 2.4. X-ray Crystallography ........................................................... 122 3. Results and Discussion................................................................... 129 3.1. Synthesis and Spectroscopy ................................................ 129 3.2. Description of Structure ....................................................... 132 Structure of KBi3Ss ................................................................. 132 3.3. Thermal Analysis.................................................................... 138 viii 3.4. Ion-exchange Studies ............................................................. 140 3.5. Solid State Ion-exchange Reactions with Halide Salts .............................................................................. 143 3.6. Conclusion ................................................................................. 150 High Temperature Synthesis and Properties of KBi6,33810 and KzBi3813 (Part 3) ........................................................ 153 1. Introduction....................................................................................... 153 2. Experimental Section ...................................................................... 155 2.1. Reagents .................................................................................... 155 2.2. Synthesis ................................................................................... 156 Bi283 .......................................................................................... 156 KBi6,33810 (I) .......................................................................... 156 KzBi3813 (I I) ........................................................................... 157 2.3. Physical Measurements ......................................................... 157 2.4. X-ray Crystallography ........................................................... 158 3. Results and Discussion ................................................................... 169 3.1. Synthesis and Spectroscopy ................................................. 169 3.2. Description of Structures ..................................................... 173 Structure of KBi6,33810 (I) ................................................... 173 Structure of K2Bi3$ 13 (I I) .................................................... 175 3.3. Electrical Conductivity and Thermoelectric Power Measurements ............................................................ 185 3.4. Conclusion ................................................................................. 189 III. Polysulfide ligands in Solid State Antimony Compounds. Isolation and Structural Characterization of CszS4Sg and CsSbS6. ........................................... 192 1. Introduction ....................................................................................... 192 ix 2. Experimental Section ...................................................................... 193 2.1. Reagents.................................................................................... 193 2.2. Synthesis ................................................................................... 193 Cssz4Sg (I) ............................................................................. 194 CsSbe, (I I) ............................................................................... 194 2.3. Physical Measurements ......................................................... 195 2.4. X-ray Crystallography........................................................... 195 3. Results and Discussion ................................................................... 201 3.1. Synthesis and Spectroscopy ................................................ 201 3.2. Description of Structures ..................................................... 204 Structure of Cssz4Sg (I) ...................................................... 204 Structure of CsSbS6 (I I) ........................................................ 205 3.3. Thermal Analysis ................................................................... 212 3.4. Conclusion ................................................................................ 218 IV. New Quaternary Bismuth and Antimony Thiophosphates ABiPZS7 (A=K,Rb), A3M(PS4)2, (A=K,Rb,Cs; M=Sb,Bi), CS3Bi2(PS4) 3, and Na0,16Bi1_23P286 in Molten Alkali Polythiophosphate Fluxes .................................................................... 221 1. Introduction....................................................................................... 221 2. Experimental Section ...................................................................... 224 2.1. Reagents .................................................................................... 224 2.2. Synthesis ................................................................................... 224 NaZS ........................................................................................... 225 KBiPZS7 (I) ............................................................................... 225 RbBiPZS7 (I I) ........................................................................... 225 K3Sb(PS4)2 (I I I) ..................................................................... 226 Rb3Sb(PS4)2 (IV) .................................................................... 226 X CS3Sb(PS4)2 (V) ....................................................................... 227 K3Bi(PS4)2 (V I) ....................................................................... 227 Rb3Bi(PS4)2 (V I I) .................................................................. 227 CS3Bi(PS4)2 (V I I I) ................................................................. 228 CS3Bi2(PS4)3 (IX) .................................................................... 228 Na0.16Bi1.28P286 (X) .............................................................. 229 2.3. Physical Measurements ........................................................ 229 2.4. X—ray Crystallography ........................................................... 229 3. Results and Discussion ................................................................... 242 3.1. Synthesis, Spectroscopy, and Thermal Analysis .............. 242 3.2. Description of Structures ..................................................... 252 Structure of KBiPZS7 (I) ....................................................... 252 Structure of K3Bi(PS4)2 (V l) ............................................... 254 Structure of CS3Bi2(PS4)3 (IX) ............................................. 255 Structure of N30.16Bi1.23P2$6 (X) ....................................... 257 3.3. Conclusions ............................................................................... 274 Coordination Chemistry of [P28e6]4' in Molten Alkali Metal Polyselenophosphate Fluxes. Isolation of KMPZSe6 and CSgM4( P28e6) 5 (M=Sb,Bi) ...................................................................... 278 1. Introduction....................................................................................... 278 2. Experimental Section ...................................................................... 280 2.1. Reagents.................................................................................... 280 2.2. Synthesis ................................................................................... 280 CszSe.......................................................................................... 280 P2885 .......................................................................................... 281 KBiPZSe6 (I) ............................................................................. 281 KSbPzSe6 (I I) .......................................................................... 281 xi CSng4( P2$e6)5 (I I I) .............................................................. 282 C338i4( PZSe6)5 (IV) ................................................................ 282 2.3. Physical Measurements ........................................................ 283 2.4. X—ray Crystallography ........................................................... 283 3. Results and Discussion ................................................................... 294 3.1. Synthesis, Spectroscopy, and Thermal Analysis .............. 294 3.2. Description of Structures ..................................................... 298 Structure of KMPZSe6 ............................................................ 298 Structure of ngSb4( P28e6)5 ................................................. 299 3.3. Electrical Conductivity Measurements ................................. 317 3.4. Conclusions ............................................................................... 319 VI. Synthesis in Molten Alkali Metal Polyselenophosphate Fluxes. A New Family of Transition Metal Selenophosphate Compounds. AzMPZSe6 (A=K,Rb,Cs; M=Mn,Fe), A2M'2P2866 (A=K,Cs; M'=Cu,Ag) and K4Auz( P2886)2 ....................... 322 1. Introduction ...................................................................................... 322 2. Experimental Section ...................................................................... 324 2.1. Reagents.................................................................................... 324 2.2. Synthesis ................................................................................... 325 CszSe .......................................................................................... 325 RbZSe......................................................................................... 325 PZSes ......................................................................................... 325 Kzanzsef, (I) ........................................................................ 325 szManSe6 (I I) .................................................................... 326 CszMnPZSe6 (I I I) ................................................................... 326 K2F€P28e6 (IV) ....................................................................... 327 CstePZSe6 (V) ........................................................................ 327 xii CszCuszSe6 (V I) .................................................................... 328 KzAgzPZSe6 (V I I) ................................................................... 328 CszAgszSe6 (V I I I) ................................................................ 328 K4Auz( P2866) 2 (IX) ................................................................ 329 KzMn0.5F80.5PzSe(, (X) .......................................................... 329 2.3. Physical Measurements ......................................................... 331 2.4. X-ray Crystallography ........................................................... 332 3. Results and Discussion ................................................................... 347 3.1. Synthesis, Spectroscopy, and Thermal Analysis .............. 347 3.2. Description of Structures ..................................................... 359 Structure of AZMPZSe6 ........................................................... 359 Structure of Cs2Cu2PZSe6 ...................................................... 365 Structure of CszAg2P28e6 ...................................................... 366 Structure of K2Ag2P28e6 ....................................................... 368 Structure of K4Au2( PZSe6)2 .................................................. 370 3.3. Magnetic Susceptibility Measurements ............................... 381 3.4. Conclusions ............................................................................... 384 LIST OF TABLES 1-1. Properties of Binary Group 15 Chalcogenides .............................. 4 Properties of ASbTez Compounds ..................................................... 19 1-3. Ternary Antimony Chalcogenide Compounds ............................... 24 1-4. Magnetic Data for MszQ, Compounds ............................................ 33 1-5. Optical Properties of SnszCb and szPzQ,................................... 37 1-6. Types of Thiophosphate ligands ..................................................... 42 1-7. Melting Points (°C) of Some Known Alkali Metal Polychalcogenides .............................................................................. 49 2-1. Calculated and Observed X-ray Powder Diffraction Patterns for B—CsBiSz.......................................................................... 79 2-2. Calculated and Observed X-ray Powder Diffraction Patterns for KzBi38e13 ...................................................................... 80 2-3. Summary of Crystallographic Data and Structure Analysis for B—CsBiSz and K28i38e13 ............................................... 81 Fractional Atomic Coordinates and Beq Values for fi-CsBiSz with Estimated Standard Deviations in Parentheses ......................................................................................... 82 2-5. Fractional Atomic Coordinates and Beq Values for KzBi38e13 with Estimated Standard Deviations in Parentheses.......................................................................................... 82 2-6. Selected Distances (A) and Angles (deg) for B-CsBiSz with Standard Deviations in Parentheses ....................... 95 2-7. Comparison of Bond Distances (A) and Angles (deg) in Selected Group 15 Sulfides (M=As,Sb,Bi) ......................................... 95 2-8. Selected Distances (A) and Angles (deg) in the xiv [Bi38e13]2' Framework with Standard Deviations in Parentheses .................................................................................... 96 Solid State Reactions of KBi3Ss with Alkali Metal Halides ................................................................................................... 125 2-10. Calculated and Observed X—ray Powder Diffraction Patterns for KBi3Ss ............................................................................. 126 2-11. Summary of Crystallographic Data and Structure Analysis for KBi3Ss ............................................................................ 127 2-12. Fractional Atomic Coordinates and Egg Values for KBi3Ss with Estimated Standard Deviations in Parentheses......................................................................................... 128 2-13. Selected Distances (A) and Angles (deg) for KBi3Ss with Standard Deviations in Parentheses ......................... 137 2-14. Calculated and Observed X-ray Powder Diffraction Patterns for KBi6,3381() ..................................................................... 162 2-15. Calculated and Observed X-ray Powder Diffraction Patterns for K2Bi3813 ........................................................................ 164 2—16. Summary of Crystallographic Data and Structure Analysis for KBi6,33810 and [(23133 13 ............................................ 166 2-17. Fractional Atomic Coordinates and Beq Values for KBi6,33810with Estimated Standard Deviations in Parentheses.......................................................................................... 167 2-18. Fractional Atomic Coordinates and Beq Values for KzBi3813 with Estimated Standard Deviations in Parentheses.......................................................................................... 168 2-19. High Temperature K28/Bi283 Reactions ......................................... 171 2-20. Selected Distances (A) in KBi6,3381() with XV Standard Deviations in Parentheses ................................................ 181 2-21. Selected Angles (deg) in KBi6,3381() with Standard Deviations in Parentheses ................................................ 182 2-22. Selected Distances (A) in KzBi3813 with Standard Deviations in Parentheses ................................................ 183 2-23. Selected Angles (deg) in KzBi3$13 with Standard Deviations in Parentheses ................................................ 184 3-1. Calculated and Observed X-ray Powder Diffraction Patterns for Cssz4Sg......................................................................... 197 Calculated and Observed X-ray Powder Diffraction Patterns for CsSbS6. ............................................................................ 198 3-3. Summary of Crystallographic Data and Structure Analysis for Cssz4Sg and CsSbS6. ................................................... 199 3-4. Fractional Atomic Coordinates and Beq Values for Cs28b483 with Estimated Standard Deviations in Parentheses......................................................................................... 200 3-5. Fractional Atomic Coordinates and Beq Values for CsSbe, with Estimated Standard Deviations in Parentheses......................................................................................... 200 3-6. Selected Distances (A) and Angles (deg) for Cszsb488 with Standard Deviations in Parentheses ..................... 210 Selected Distances (A) and Angles (deg) for CsSbe, with Standard Deviations in Parentheses ......................... 211 4-1. Calculated and Observed X-ray Powder Diffraction Patterns for KBiPZS7.......................................................................... 232 4-2. Calculated and Observed X—ray Powder Diffraction Patterns for K3Bi(PS4)2 ..................................................................... 234 xvi 4—3. Calculated and Observed X-ray Powder Diffraction Patterns for Na0,16Bi1,23PzS6 .......................................................... 235 4—4. Crystallographic Data for 1, VI, IX, and X ....................................... 237 4—5. Fractional Atomic Coordinates and Beq Values for KBiPzS7 with Estimated Standard Deviations in Parentheses......................................................................................... 239 4—6. Fractional Atomic Coordinates and Beq Values for K3Bi(PS4)2 with Estimated Standard Deviations in Parentheses ......................................................................................... 239 4—7. Fractional Atomic Coordinates and Beq Values for CS3Bi2(PS4) 3 with Estimated Standard Deviations in Parentheses ......................................................................................... 240 4—8. Fractional Atomic Coordinates and Beq Values for Nao,16Bi1,28P286 with Estimated Standard Deviations in Parentheses ......................................................................................... 240 4—9. Far-IR Spectra for (I-X) ..................................................................... 245 4—10. Optical Band Gaps and Melting Point Data for (I-X) ...................... 246 4—11. Selected Distances (A) and Angles (deg) for KBiPZS7 with Standard Deviations in Parentheses ...................... 269 4—12. Selected Distances (A) and Angles (deg) for K3Bi(PS4)2 with Standard Deviations in Parentheses ................. 270 4-13. Selected Distances (A) for CS3Bi2(PS4) 3 with Standard Deviations in Parentheses ................................................ 271 4—14. Selected Angles (deg) for K3Bi(PS4)2 with Standard Deviations in Parentheses ................................................ 272 4—15. Selected Distances (A) and Angles (deg) for Na0,16Bi1,23PZS6 with Standard Deviations in Parentheses ...................................... 273 xvii 5-1. Far-IR Data for KMPZSe6 and Cs8M4( P28e6)5 ............................... 285 5-2. Calculated and Observed X-ray Powder Diffraction Patterns for KBiPZSe6 ........................................................................ 286 Calculated and Observed X-ray Powder Diffraction Patterns for CSng4( P28e6)5 ............................................................. 288 Calculated and Observed X-ray Powder Diffraction Patterns for CSgBi4( P28e6)5 ............................................................. 289 5-5. Crystallographic Data for KBiPZSe6, CSng4(PZSe6)5 and CSgBi4( P2366) 5 ............................................................................ 290 5—6. Fractional Atomic Coordinates and Beq Values for KBiPZSe6 with Estimated Standard Deviations in Parentheses......................................................................................... 291 5-7. Fractional Atomic Coordinates and Beq Values for CSng4( P28e6)5 with Estimated Standard Deviations in Parentheses......................................................................................... 292 5—8. Fractional Atomic Coordinates and Beq Values for C338i4( P28e6)5 with Estimated Standard Deviations in Parentheses......................................................................................... 293 5-9. Selected Distances (A) and Angles (deg) for KBiPZSe6 with Standard Deviations in Parentheses ...................................... 312 5-10. Selected Distances (A) for C88M4( P28e6)5 with Standard Deviations in Parentheses ...................................... 313 5-11. Selected Angles (deg) for CSgM4( PZSe6)5 with Standard Deviations in Parentheses ...................................... 316 6-1. Calculated and Observed X-ray Powder Diffraction Patterns for KzMnPZSe6 .................................................................... 334 Calculated and Observed X-ray Powder Diffraction Patterns for CszMnPZSe6 .................................................................. 335 Calculated and Observed X-ray Powder Diffraction Patterns for KzAgzPZSe6 .................................................................. 336 6-4. Calculated and Observed X-ray Powder Diffraction Patterns for CszAg2P28e6 ................................................................. 338 Calculated and Observed X-ray Powder Diffraction Patterns for CSZCuzPZSe6 .................................................................. 339 Calculated and Observed X-ray Powder Diffraction Patterns for K4Au2(PzSe6)2 ............................................................. 340 Crystallographic Data for (I), (III), (IV), (VI), (VII), (VIII), and (IX) ................................................................................................. 341 Fractional Atomic Coordinates and Beq Values for KzMnPZSe6 with Estimated Standard Deviations in Parentheses......................................................................................... 343 Fractional Atomic Coordinates and Beq Values for CszMnPZSe6 with Estimated Standard Deviations in Parentheses......................................................................................... 343 6-10. Fractional Atomic Coordinates and Beq Values for KzFePZSe6 with Estimated Standard Deviations in Parentheses. ........................................................................................ 343 6-11. Fractional Atomic Coordinates and Beq Values for Cs2Cu2P28e6 with Estimated Standard Deviations in Parentheses......................................................................................... 344 6-12. Fractional Atomic Coordinates and Beq Values for K2Ag2P28e6 with Estimated Standard Deviations in Parentheses......................................................................................... 345 6—13. Fractional Atomic Coordinates and Beq Values for xix CszAgzPZSe6 with Estimated Standard Deviations in Parentheses......................................................................................... 346 6-14. Fractional Atomic Coordinates and Beq Values for K4Au2( P28e6)2 with Estimated Standard Deviations in Parentheses......................................................................................... 346 6—15. Far-IR Data (cm' 1) for AZMPZSeg, AZM'2P28e6, and K4Au2( P28e6) 2 Compounds ....................................................... 351 6—16. Optical Band Gaps and Melting Point Data for AzMPzSeg, A2M'2P28e6, and K4Auz( P28e6)2 ................................. 353 6-17. Selected Distances (A) and Angles (deg) for (I), (III), and (IV) with Standard Deviations in Parentheses ...................... 364 6—18. Selected Distances (A) and Angles (deg) for Cs2Cu2PzSe6 with Standard Deviations in Parentheses ............... 377 6-19. Selected Distances (A) and Angles (deg) for CszAg2P28e6 with Standard Deviations in Parentheses .............. 378 6-20. Selected Distances (A) and Angles (deg) for KzAgzPZSe6 with Standard Deviations in Parentheses .............. 379 6-21. Selected Distances (A) and Angles (deg) for K4Au2( P28e6)2 with Standard Deviations in Parentheses ......... 380 6-22. Magnetic parameters for (I-IV) ...................................................... 382 LIST OF FIGURES 1-1: Projection of BizTe3 looking down the a-axis, with labeling................................................................................... 5 (A) ORTEP representation of szS3 viewed down the b-axis with labeling scheme. The long Sb---S contacts are highlighted by dashed lines. (B) A single infinite ribbon of szS3 .......................................... 8 1—3: Packing diagrams of (A) CsBi3Ss and (B) Cs3Bi7Se12 looking down the b-axis. Shaded spheres represent S(Se) atoms and small open spheres designate Bi atoms ............................................................................................. 12 1-4: Packing diagram of Sr4Bi65e13 and labeling scheme with a view down the b-axis ............................................ 13 1-5: (A) Packing diagram of SrBiSe3 viewed down the c-axis. Dark spheres indicate Bi atoms and large open spheres are Se atoms. (B) Interacting (Se3)2' anions with dashed lines representing close Se---Se contacts .................................... 14 1-6: Packing diagram of BaBiSe3 viewed down the c-axis. Dark spheres indicate Bi atoms and large open spheres are Se atoms. The square planar Se atoms are highlighted with the arrow ................................................................................ 15 1-7: Equilibrium phase diagram of the BizS3— [(28 system ............... 17 1-8: Projection of CsszgS 13 viewed down the c—axis with labeling ...................................................................... 22 xxi 1-9: (A) Packing diagram of Pd3(PS4)2 viewed down the a-axis with labeling. (B) ORTEP representation of a single Pd3(PS4)2 layer ................................ 29 1-10: (A) Schematic representation of NiszS6 viewed down the a-axis. (B) ORTEP representation of a single NiszS6 layer. The hexagonal Ni array is outlined with dashed lines ...................................................................................... 32 1-11: (A) Packing diagram of RbVPzS7 viewed down the a-axis. (B) ORTEP representation of the [VPzS7]nn‘ anionic layer with labeling ........................ 41 1-12: A schematic representation of a single V2P4Sl3 layer. The open spheres are S atoms and dark lines highlight the bonding in the [P4813]6' ligand ................................................... 45 1-13: (A) One-dimensional structure of [AuSe5]nn'. Au---Au contacts are indicated by dashed lines. (B) Structure of the [AuSe13]n3n' chain ................................... 53 1-14: (A) A single anionic layer in NaAuSez. (B) The one-dimensional chains in KAuSez............................... 54 1-15: ORTEP representation and labeling scheme of the layered structure of [Sn288]n2n'. The dashed lines indicate the shortest nonbonding Sn---S contacts (2.934(5) A) ..................................... 55 1-16: ORTEP representation and labeling scheme of K2Au28n286. View down the b-axis. Au---Au contacts are shown by dashed lines ............................... 56 .. ORTEP representation of a single (BiSz)nn‘ chain with labeling scheme .......................................................... 9) 2-2: ORTEP packing diagrams of the (Bi82)nn‘ chains along the a-axis, (A) and b-axis, (B). Dashed lines in (A) indicate Bi---S contacts. The shaded ellipsoids are Bi atoms, and the Cs atoms are represented by open circles ............................ 91 ORTEP representation of the packing diagram of KzBi38e13 down the c-axis. The shaded ellipsoids are the Bi atoms, and the open ellipsoids are the K atoms .............................................................. 92 2-4: Projections of the structures of (A) BizSe3 (BizTes-type) (B) BiZSe3 (SbZS3-type) (C) CdIz-type and (D) [Bi38e13]2' framework. The three structure types found in this framework are designated with dashed lines ............................. 93 ORTEP representation of the Bi(4) coordination in KzBi38e13. Dashed lines designate long contacts .................................................................. 94 (A) DSC thermogram of the Bi/CszS3 mixture. The cooling and reheating curves are shown above the heating thermogram for clarity (Peaks F and G). Peak temperatures (° C): A (97), B (113),C (170),D(270), E(274), F (176) and G(180). (B) DSC thermogram of the Cs28/ S mixture. Peak temperatures (° C):A (92), B (112), C (170) and H (178) ........................................................................................ 101 2-7: (A) DSC thermogram of the Bi/KzSeg mixture. Peak temperatures (° C): 1(160), 11(169), 111 (269) and IV (152). (B) DSC thermogram of the Kzse/ Se mixture. Peak temperatures (0 C): 1(161), II (170) andV (123) ............................................................ 102 2-8: Variable temperature electrical conductivity data for a single crystal of y-CsBiSZ ...................................................... 106 Variable temperature electrical conductivity data for a pressed pellet of KzBigse 13 .................................................. 107 2-10: Thermoelectric power of a polycrystalline aggregate of KzBigSe13 as a function of temperature ................................. 108 2-11: (A) Optical absorption spectrum of B—CsBiSz. (B) Optical absorption spectrum of y-CsBiSZ .............................. 110 2-12: Optical absorption spectrum of KzBigSe13. The absorption maximum occurs at 1.2 eV.................................. 111 2-13: Optical absorption spectrum of KBi3Ss. Inset shows the (a/S)2 vs. Energy plot ........................................ 131 2-14: The three-dimensional structure of KBi3Ss viewed down the b-axis. The shaped ellipsoids are Bi atoms and the open circle represent K atoms ................. 135 2-15: (a) Polyhedral representation of the [Bi3Ss]nn' framework projected in the b-direction. The octahedra that line the tunnel are shaded. (b) Adapted polyhedral represention of the hollandite anion framework projected in the c-direction showing the 2x2 tunnel structure-30¢ .................... 136 2-16: DTA thermogram of KBi3Ss ............................................................ 139 xxiv Comparison of the X-ray powder diffraction patterns of: (a) KBi3Ss with selected h01 peaks labeled. (b) (H20)Bi384.5-(88)0.06 and (c) BizS3 ............................................................................................. 145 2-18: SEM photographs of: (a) KBi3S5 before and (b) after reaction with aqueous HCl (66 hours). The white bar at the bottom of the micrographs represents a mum scale. The magnification is x1600 ............................................................ 146 2-19: (a) Thermogravimetric analysis plot of (HzO)Bi3S4,5-(Sg)0,()6 with weight 96 plotted against temperature (0 C). Performed under flowing N2. Step I (3.82% weight loss starting at 140 °C.) Step II (3.84% weight loss starting at 176 °C). (b) DSC thermogram of (H20)Bi3S4,5-(88)0_06 ............................ 148 2-20: Powder XRD pattern of (H20)Bi3S4,5-(Sg)(),06 measured as a function of temperature. The transformation to BizS3 is complete at 124 °C ..................... 149 2-21: Mid-IR diffuse reflectance spectra of (a) KBi6,33810 and (b) K2Bi3813 .................................................... 172 2-22: ORTEP representation of the packing diagram of KBi6,3381() down the b-axis with labeling .............................. 178 2-23: Projections of the structures of (A) BizTe3 (B) Cdlz and (C) [Bi6,33810]' framework. Both structure types found in this framework are designated with dashed lines ............................................................................. 179 2-24: ORTEP representation of the packing diagram XXV of KzBigS 13 down the b-axis with labeling ................................. 180 2-25: Variable temperature electrical conductivity data for polycrystalline chunks of: (a) KBi6,33S 10 and (b) K2BigSl3 .............................................................................. 187 2-26: Thermoelectric power as a function of temperature for polycrystalline chunks of: (a) KBi6,3381() and (b) KzBi8813 .............................................................................. 188 3-1: (A) Optical absorption spectrum of Cssz4Sg. (B) Optical absorption spectrum of CsSbS6. ................................. 203 3-2: Two-dimensional structure of the [Sb4Sg]nn' anionic framework with labeling as drawn by ORTEP. The dashed lines represent Sb—S long interactions in the range of 3.022(2) to 3.277(2) A. The shaded area outlines the 14—membered Sb/S rings in the structure ................................................................................ 207 (A) Stereoview of Cssz4Sg viewed down the [001] direction. (B) Packing diagram of Cssz4Sg showing a projection in the [100] direction. The dashed lines represent long Sb-S interactions underscoring the pseudo—3D character of the structure. The shaded ellipsoids are Sb atoms and the open ellipsoids are Cs atoms............................................ 208 (A) Structure of a single (SbS6)nn' chain with labeling. (B) Packing diagram of CsSbS6 looking down the a-axis (chain axis). The shaded ellipsoids are Sb atoms and the open ellipsoids are Cs atoms ............................ 209 3-5: (A) Differential thermal analysis (DTA) data for xxvi Cssz483 showing the first heating and cooling cycle. Heat is absorbed at 365 °C as the material melts but is not released upon cooling. (B) Data for reheating showing the crystallization of Cssz4S7 at 291 °C, followed by melting of the excess glass at 364 and 375 °C ................................................................................... 214 3-6: X-ray diffraction patterns of: (A) Cssz4Sg single crystals. (B) material heated to 400 °C then cooled (glassy state). (C) material after reheating to 310 °C (crystallization of Cssz4S7) ............................................. 216 3-7: DTA of CsSbS6 shows the endothermic peak at 232 °C corresponding to the decomposition of CsSbS6 to Cssz4Sg........................................................................... 217 4—1: Powder X-ray diffracton patterns of: (A) A3Sb(PS4)2 and (B) A3Bi(PS4) 2 ............................................. 241 4—2: Optical absorption spectrum of: (A) KBiP2S7 and (B) RbBiPzS7............................................................................. 247 4—3: Optical absorption spectra of: (A) K3Sb(PS4)2 (B) Rb3Sb(PS4)2(C)Cs3Sb(PS4)2.................................................. 248 4—4: Optical absorption spectra of: (A) K3Bi(PS4)2 (B) Rb3Bi(PS4)2(C)Cs3Bi(PS4)2................................................... 249 4—5: Optical absorption spectrum of Na0.16BiP286. ........................... 250 4—6: Thermograms of: (A) RbBiPzS7 and (B) Rb3Bi(PS4)2 ............... 251 4—7: ORTEP packing diagram of KBiPzS7 looking down the c-axis with labeling ................................................................. 259 4-8: ORTEP representation and labeling of the KBiPzS7 layer looking down the a-axis ................................................................. 260 xxvii 4—9: ORTEP representation and labeling of the Bi-S coordination site. The polyhedron is outlined with dotted lines for clarity ........................................................... 261 4-10: ORTEP representation and labeling of the BiS7 polyhedron viewed from the top of the capped trigonal prism. The probable location of the Bi3+ lone pair is shown.................................................................. 262 4—11: A single [Bi(PS4)2]3‘ anionic chain ............................................ 263 4-12: ORTEP packing diagram of K3Bi(PS4)2 looking down the b-axis with labeling ................................................................. 264 4—13: ORTEP packing diagram of Cs3Bi2(PS4) 3 looking down the b-axis with labeling ................................................................. 265 4—14: (A) The Bi2(PS4)2 dimeric unit. The Bi(2)-Bi(2') distance is 4.519(3) A. (B) A single [Bi2(PS4) 31n3n- layer..... 266 4—15: ORTEP packing diagram of Na0,1f,BiP286 looking down the a-axis with labeling. The Bi-S bonds are omitted for clarity ......................................... 267 4—16: ORTEP representation and labeling of the Bi-S coordination site. The polyhedron is outlined with dotted lines for clarity ........................................... 268 5-1: Solid state optical absorption spectra of (A) KBiPzSe6 and (B) KSbPzSe6. ................................................... 296 5-2: (A) Optical absorption spectrum of Cs88b4( P28e6) 5. (B) Optical absorption spectrum of ngBi4( P28e6)5 .................. 297 5-3: The extended structure of KBiPzSe6 looking down the b-axis with labeling ........................................................................ 305 (A) Structure and labeling of a [BiP28e6]‘ chain. xxviii The bridging Se atoms connecting the chains into layers are highlighted. (B) The Bi-Se coordination site with labeling ............................................................................ 306 5-5: Packing diagram of ngBi4( P28e6)5 viewed down the [010] direction. The open circles represent Cs atoms. Dark solid lines highlight the layered nature of the staircase framework ......................... 307 5-6: Structure of the [M2(P28e6)4] dimeric unit showing the M---M interaction ...................................................... 308 (A) Association of two [M2( P28e6) 4] "dimers" by a [P28e6]4‘ unit. (B) Further sharing of [P28e6]4' units among "dimers" results in a one-dimensional chain-like structure ....................................................................... 309 5-8: Two-dimensional structure of the [M4( P28e6)5]n8n' anionic framework with labeling as drawn by ORTEP ............. 310 The Cs+ coordination environments. The open circles represent Cs atoms. The open square in the Cs(1) coordination site represents the vacant corner of the square prism ......................................................................... 311 5-10: Variable temperature electrical conductivity data for single crystals of ngBi4( P28e6)5 .................................. 318 Far-IR spectra of: (A) K2MnP2$e6 (B) CszCuszSe6 and (C) K2Ag2P28e6......................................................................... 352 6-2: 6-3: Optical absorption spectra of (I ), (I I), and (I I l) ...................... 354 Optical absorption spectra of (IV) and (V) ................................ 355 6-4: Optical absorption spectra of (V 1), (VII), and (V I I I) ............. 356 6-5: Optical absorption spectrum of (IX) ............................................ 357 xxix Thermogram of szManSe6. Melting is observed at 781 °C, followed by recrystallization at 751 °C ...................... 358 (A) ORTEP representation of a single [ManSe6]n2n' chain. (B) Packing diagram of K2MnP28e6 viewed down the [100] direction with labeling .................................................. 362 (A) A single layer of the anPzQ, framework. The dashed lines indicate the hexagonal arrangement of Mn2+ ions. (B) Removal of one-half of the Mn2+ ions results in a single layer of the hypothetical [MnP2%]n2n' anion. Dashed lines highlight the possibility of chain formation ............................................................................... 363 (A) A single [CuszSe6]n2n' chain. (B) A single [AgszSe6]n2n‘ chain.............................................. 372 6-10: Energy levels of the d block of Cuz2+ at a separation of 2.58 A. Mixing of filled d orbitals with empty 3 and p orbitals results in a lowering of the antibonding (0*) and bonding (0) energy levels which gives rise to a net bonding interaction. Adapted from reference 26 .................... 373 6-11: Packing diagram of the three-dimensional structure of K2Ag2P28e6 viewed down the [100] direction with labeling ......................................................... 374 6-12: ORTEP representation of a fragment of the [AgszSe6]n2n' anion showing the local AgSe4 coordination environment with labeling ................................... 375 6-13: A single [Au2( P28e6)2]n2n' chain ............................................... 376 6-14: Plots of l/XM vs. T taken at 2000 G of applied field over 2-300 K for: (A) CszanPzSee, and XXX (B) K2FeP28e6. The inset graphs show expanded views of the region 2-30 K ........................................... 383 CHAPTER 1 Introduction The development of useful technologies often relies on the availability of solid-state materials with appropriate physico chemical properties. The exploratory synthesis of these new materials is often accomplished by traditional usually high- temperature fusion reactions to form thermodynamically stable products. This empirical method has led to several discoveries with great potential impact on our society ranging from the high- temperature superconductors and advanced ceramics to next- generation nonlinear optical solids. However, new synthetic methodology is needed to advance solid-state chemistry and insure its continued growth. The study of new metal Chalcogenides is particularly attractive given the staggering structural diversity and broad range of interesting electrical, optical, and catalytic properties that are characteristic of these compounds. In the past decade, metal Chalcogenide chemistry has experienced rapid growth due to an expanding arsenal of synthetic methodology available to the 1 2 inorganic chemist. In this dissertation, we have employed and developed the now proven molten alkali polychalcogenide salt method at intermediate temperatures (150-500 °C) for the synthesis of new ternary antimony and bismuth Chalcogenides. In addition, a significant variation of this technique has led to the discovery of alkali polychalcophosphate fluxes for synthesis of new quaternary metal thio( seleno)phosphate compounds. 1. A Review of Antimony and Bismuth Chalcogenides The first antimony and bismuth sulfides were formed in the earth's crust by nature through the reaction of metal sulfides with Group 15/ 16 sulfides at elevated temperatures or through hydrothermal reaction in superheated water. These compounds belong to a large class of Group 15/16 compounds, known as sulfosalt minerals. Bismuth compounds constitute about 20% of these compounds.1 These minerals have the general formula MxEySZ (M=Ag, Cu, Pb, Hg; E=As, Sb, Bi) where the trivalent E atom is coordinated by three S atoms to form a trigonal E833- pyramid.2 Examples of these formulas include Cu128b4813 (tetrahedrite),3 Hng483 (livingstonite),4 PbCuBiS3 (aikinite),5 and PbBi284 (galenobismutite).6 Another class of Group 15/ 16 compounds are the binary Chalcogenides, Eng (E=As, Sb, Bi; Q=S, Se, Te) which have received considerable attention, due to their potential application as thermoelectric cooling materials. During the past 30 years, there have been hundreds of papers describing various solid solutions of 3 Eng (M=As, Sb, Bi; Q=S, Se, Te) compounds.7 The best thermoelectric cooling materials near room temperature are BizTe38a and related compounds such as BizTe2_3SSeo,15.9 Candidates for thermoelectric cooling materials must have a high figure of merit (Z) which is defined by the following equation: eq. (1) where a=thermoelectric power (Seebeck coefficient), o=e1ectrical conductivity and x=therma1 conductivity. A useful thermoelectric material must possess high electrical conductivity and thermoelectric power while maintaining low thermal conductivity. The bismuth telluride compounds have large Z values (2.5-3.3 x 103 K1) near room temperature. 36‘ BizTe3 possesses a unique layered structure and belongs to the space group R-3m (see Figure 1-1). The layered nature of BizTe 3 has a strong influence on the properties resulting in highly anisotropic transport phenomena such as resistivity, Hall coefficient, magnetoresistivity, and thermal conductivity.9 The framework can be viewed as a defect NaCl structure by removing every third layer of Na atoms, which leads to the A283 stoichiometry. Each Bi atom is surrounded by six Te atoms to form an octahedral coordination geometry with Bi-Te bond distances of 3.071 and 3.247 A. These long distances could be a result of electronic repulsions between the localized spherical and stereochemically inactive 6S7- orbitals and the Te atoms. As a result, the 6s2 electrons are thought to be delocalized throughout the solid. Interestingly, weak interlayer Te---Te 4 interactions are also present (3.647 A). BizTe3 transforms to the a- In28e3 structure type at high temperature and pressure,10 while Bi28e3 undergoes an irreversible phase transition to the Sb283 (stibnite) structure type at high pressure.11 szTe312 and Bi28e313 are isomorphous with BizTe3 and possess E-Qbond distances of 2.974 and 3.168 A for szTe3 and 2.851 and 3.075 A for Bi28e3. Bismuth telluride is a narrow band gap semiconductor or semi- metal with an E3 of 0.17 eV and a room-temperature conductivity ranging from 102 to 103 S/ cm. Typical bandgap and conductivity values for 13ng compounds are given in Table 1-1. Table 1-1. Properties of Binary Group 15 Chalcogenides. Compound Eg (eV), r.t. Ref. 0 (S/cm), r.t. Ref. szS3 1.63 1.7 szSe3 1.19 1.2 szT63 0.21 BizS3 0.3 1.3 1.3 B12383 0.55 0.35 BizTe3 0.17 0.15 25 27 25 27 25 27 26 27 26 27 26 27 1.7 X 10'8 25 1 x 10'5 25 3 X 103 1 x 104 3 x 10'8 7 X 10'2 2.0 2.2 x 103 25 8d 26 26 29 28 6.3 x 102 8d =. Te2 s, Figure 1-1: Projection of BizTe3 looking down the a-axis, with labeling. The crystals of BizTe3 are plate-like with a shiny, metallic luster and can be made in either the n- or p-type forms. This latter feature is important for the fabrication of thermoelectric cooling devices. Large ingots of the material can be made from the elements by the Bridgman method at 800 °C. Crack-free slices along the cleavage planes of these ingots are suitable for detailed charge transport studies.8d Although this peculiar compound has attracted considerable attention, only a few theoretical calculations of the electronic band structure have been performed.14 Recently, the first density of states (DOS) calculations have been accomplished using the empirical tight binding method. 15 This study was undertaken to explain recent photoemission results and provide some insight into the bonding of BizTe 3. The photoemission results and the calculations indicate that s-p hybridisation is weak and imply that p orbital bonding, like in elemental Bi, dominates. The 90° Te-Bi-Te angles allow for maximum overlap of all three p orbitals. In addition, the DOS plots illustrate that the s-orbitals of Bi are low lying and reside in the conduction band. Information on the charge transfer of this highly covalent network was obtained as follows: Te(1) -0.33, Te(2) -0.19, and Bi +0.36. The second structure type that is prevalent in binary Group 15 Chalcogenides is the Sb253 (stibnite) structure type16 which is isomorphous with Sb28e317 and Bi283 (bisrnuthite).18 Sb283 and Bi283 are frequently found naturally as sulfosalt minerals. The crystal structure of SbZS 3 contains two infinite ribbons (Sb4S6)n along the b- 7 axis, which are weakly bonded in the a (3.335 A) and c (3.148-3.196 A) directions as shown in Figure 1-2a. A single ribbon is shown in Figure 1-2b. Sb(1) is surrounded by three sulfur atoms to form the familiar SbS33' trigonal pyramidal unit. The bonds for Sb(1) are presumably of p character with Sb-S distances of 2.58 and 2.57 A while the mean S-Sb-S bond angle is 91.43°. Sb( 2) is found in a square pyramidal geometry with five bonds to sulfur and the lone pair occupying the sixth sp3d2 orbital at the base of the pyramid. The distances from Sb(Z) to two basal sulfur atoms are 2.68 A while the other two basal atoms are found at 2.82 A. The apical sulfur is shorter at 2.49 A and the mean bond angle is 89.19°. Binary bismuth and antimony sulfides and selenides have been studied for possible applications in many areas. Sb283 and Bi283 are known photoconductors with band gaps in the range of 1.3 to 1.5 eV (see Table 1-1) which make them good candidates for certain photovoltaic applications. 19 Amorphous chalcogenide glasses such as A82383/szse3 have been investigated for possible memory switching devices because of their ability to oscillate between crystalline and amorphous states.20 Asng/Sb283 glass fibers have been used to increase resolution in X-ray camera tubes.7-1 Eng (E=Sb,Bi; Q=S, Se) compounds prevent deterioration of aluminosilicate cracking catalysts due to Ni and Fe deposits.22 Antimony thioantimonate (Sb284) has been shown to be a more effective antiwear additive for lubricating greases than M08223 while Sb28e3 is a promising material in CdSe-based thin-film solar cells.24 b'aXIs # Figure 1-2: (A) ORTEP representation of Sb2S 3 viewed down the b-axis with labeling scheme. The long Sb---S contacts are highlighted by dashed lines. (B) A single infinite ribbon of Sb283. In addition to the naturally occurring mineral sulfosalts and the binary Group 15 Chalcogenides, the ternary AxEyQ (A=alkali metal, alkaline earth metal, Cu, Tl; E=Sb, Bi; Q=S,Se,Te) compounds display remarkable structural diversity and potentially interesting charge transport properties. The earliest members of this family are the ABiSz (A=Ii,Na,K; Q=S,Se) compounds which were uncovered in the 1940's and 1950's.30 Later, the isomorphous ABiTez (A=Ii,Na) compounds were also discovered.31 The units cells were determined by powder diffracton data and revealed simple NaCl-type cubic structures (Fm3m space group) where the Bi and A atoms are disordered. Further investigation with rubidium cations resulted in the formation of RbBin (Q=S,Se) with a Bi-S distance of 2.953 A.32 The larger alkali cation leads to a NaCl-type superstructure (NaCrSz-type) with an ordered cation arrangement. Conductivity studies on polycrystalline samples of ABin compounds confirm their semiconductor nature. The conductivities of ABiSez compounds near room temperature range from 1.9 x 101 (KBiSez) to 1.6 x 10'2 S/cm (CsBiSez) while those of ABiSz are lower and vary between 4.55 x 10-3 (IiBiSz) and 8.33 x 103 S/cm (CsBiSz).33 Other synthetic investigations with alkali metal counterions uncovered two new compounds with unique tunnel structures, RbBi3Ss34 and CsBi3SS.35 They are reminiscent of the well-known MnOz hollandite frameworks.36 CsBi385 was synthesized by direct combination of Bi283, K2(D3 and S at 500 °C. The structure is shown in Figure 1-3a and consists of [B1385]' block-like chains that share 10 comers to form Cs-filled channels parallel to the crystallographic b- axis. Substitution of S by Se cannot stabilize the isostructural compound. The reaction of cesium acetate, bismuth and selenium in a 1:1:4 ratio at 500 °C gives rise to a new structure type, (153Bi7Se 12.37 The layers (Figure 1-3b) are assembled from the same block-like chains as discussed above. These chains are bridged by BiSe6 octahedra in an edge-sharing manner to form sheets that are separated by Cs+ ions. Upon changing from alkali metals to the main group Tl+ ion, 'l‘l4Bi285 was isolated.33 This compound is intriguing because the octahedrally coordinated Bi3+ and the four- and five-coordinate Tl+ ions each possess stereochemically active 6s2 electron pairs that manifest themselves structurally. The 8ng octahedra are distorted with an axial Bi-S distance of 2.64(2) A that is trans to a long distance of 3.05(2) A while the two different T1+ ions have trigonal pyramidal (including the lone pair) and square pyramidal coordination. In solid state chemistry, the size and charge of the counter cation can have a dramatic effect on the structure of the compound. Schafer and coworkers have utilized alkaline earth cations such as Sr2+ and Ba2+ to stabilize an intriguing set of three unusual layered [BixSey]nn- anionic frameworks. Sr4Bi65e1339 contains two- dimensional sheets made up of edge-sharing CdIz- and BizTeg-type bismuth selenide fragments. The BiSe6 octahedra share edges to create one-dimensional sixfold chains that extend along the b- direction and separate these layers. Sr2+ ions are inserted between the chains that separate the layers in this highly charged 1 1 [Bi6Se13]n3n' framework (see Figure 1-4). The structure of SrBiSe3,4O is shown in Figure 1-5a. It is composed of [BigSe13]n 1211- chains that interact with (Se3)7-' fragments to construct pseudo two-dimensional layers, between which Sr2+ ions are located. The (Se3)2- units form long contacts with the [BigSe13]n12n- chains as shown in Figure 1-5b. The formula can be represented as SrgBi38e13(Se3)2. In BaBiSe3,41 BiSe6 octahedra are connected by common edges to form fourfold one-dimensional chains that are bridged by unprecedented (Sex)x' polymeric units. These units contain unusual square planar Se atoms. Both types of anionic chains extend along the c direction and are bonded in the b direction to give rise to sheets, between which Ba2+ ions reside (see Figure 1-6). A more useful formulation is Ba4Bi4$e3(Se4), where the Se atoms in the [Se4]4‘ unit each have a formal charge of 1-. The shiny appearance of the compound coupled with its electron deficiency strongly suggest that the compound is a metal but electrical conductivity studies have not been reported. The metallic character and the presence of polyselenide units in Ba4Bi4Se3(Se4) are very unusual in bismuth chalcogenide chemistry. The isomorphous BaSbTe3 compound has also been synthesized.“1 Combination of Bi283 and BaS at high temperatures yielded two forms of BaBi284.42 a-BaBiZS4 contains a slight excess of Bi283 and 6- BaBiZS4 a slight excess of BaS in relation to the stoichiometric formula. SrBi254 was also prepared and found to be isotypic to B- BaBiZS4. Both structures consist of edge- and comer-sharing BiS6 distorted octahedra that form narrow Ba2+-filled tunnels that run along the c-axis. 12 (A) (3) w, 2.‘:_~.‘(,‘L./HJ “‘37- -' I ‘ HQ?!" ’ . ‘4 ”3L .- 1-.”!l____' JLU/ Jill" "V‘w. Figure 1-3: Packing diagrams of (A) CsBigSs and (B) CS3Bi7Se12 looking down the b-axis. Shaded spheres represent S(Se) atoms and small open spheres designate Bi atoms. l3 g n i l e b a l d n a 3 1 e $ 6 i B 4 r S f o P m a r g a i d g n i k c a e m e h c s : 4 - 1 e r u g i F . s i x a - b e h t n w o d w e i v a h t i w l4 (B) Figure 1-5: (A) Packing diagram of SrBiSe3 viewed down the c- axis. Dark spheres indicate Bi atoms and large open spheres are Se atoms. (B) Interacting (Se3)2' anions with dashed lines representing close Se-«Se contacts. 15 Figure 1-6: Packing diagram of BaBiSe3 viewed down the c-axis. Dark spheres indicate Bi atoms and large open spheres are Se atoms. The square planar Se atoms are highlighted with the arrow. 16 The incorporation of transition metals into bismuth chalcogenide compounds has also been investigated. Many examples exist in the Cu/ Bi/ Q (Q=S,Se) family that are either naturally occurring or synthetic including Cu1+3xBi5-ng (Q=S,Se),43 CuBiSz,44 and Cu3BiS3.45 Other transition metal systems include Mn1-xBi2+yQ4 (Q,=S,Se)46 and HgBiZS4.47 The Cu1+3xBi5-x(b alloy was prepared by high temperature fusion of Cu/Bi/Q with a ratio of 1:3:5 and possesses extensive non-stoichiometry on two of the Cu sites and Cu/Bi disorder on the third site. The structure is very similar to that of the stoichiometric compound, CuBi588, where the Bi atoms are found in three different coordination geometries: perfect octahedral, distorted octahedral and square pyramidal.48 The Cu atoms are found in a distorted tetrahedral site with an unusually short Cu+---Cu+ distance of 2.56 A. The authors suggest that Cu—Cu pairs may exist in the structure. The polyhedra of the three independent Bi atoms are linked together to form [Bi553]nn- sheets. These sheets are crosslinked by CuS4 tetrahedra to assemble the three-dimensional framework. Mn1-xBi2+yQ4 are isostructural to HgBiZS447 and synthesized by direct combination of the elements in the 600-1000 °C temperature range. Vapor transport techniques for crystal growth formed BizQ3 which implies that the ternary phases decomposed under these conditions. Interestingly, single crystals were grown with a low temperature flux consisting of a eutectic mixture of 45% IiCl and 55% RbCl (mp. 312 °C). It should be pointed out that the use of alkali metal halide fluxes to grow chalcogenides is somewhat uncommon. Such solvents are generally reserved for oxide crystal growth49a and for hard cation phases such as 112182.49b 17 In addition to the previously mentioned synthetic work, there have been several investigations of the high temperature AzQ/BizQ; (A = Li, Na, K, Rb, Cs; Q= 8, Se, Te) phase diagrams.50 However, the structural information only consists of unindexed powder X-ray diffraction patterns. An example of the K28/ 81283 phase diagram is shown in Figure 1-7. It becomes apparent that further work in the high temperature synthesis and structural characterization of new AxBin compounds is warranted. K.:I.;51 K815; $101.51 K0135: . t,'5 :00 t I . m I . I I | I I ° 3% ' r 5% g 5 5 it I) i. . E W ' II - II " Z 2 5 I 5 5” :i__o ‘ ° ‘ °/ : 5; z: ,.. " t . 2 m 2.7 5‘0 7.7 ms, 31.253, mole % Figure 1-7: Equilibrium phase diagram of the 81283-K28 system. 18 The synthetic solid state chemistry of ternary AbeyQ compounds is much more extensive than that of the bismuth chemistry due to the successful implementation of hydro(methanolo)thermal reactions in addition to the traditional solid state methods. In general, the structural chemistry of the antimony compounds is different from that of bismuth presumably because of the preference of the Sb?)+ ion for lower coordination numbers such as trigonal pyramidal (CN=3) and trigonal bipyramidal (including the lone pair) (CN=4). These trigonal pyramids and bipyramids can share edges and/ or corners to generate a large array of chain-like and layered structures. In comparison, the Bi3+ ion shows a greater tendency to expand its coordination sphere to include square pyramidal (CN =5) and perfect or distorted octahedral (CN=6) geometries which share edges to form layered and three- dirnensional tunnel frameworks. In addition, Abesz systems have the tendency to form amorphous glasses under certain high temperature conditions51 while bismuth compounds will crystallize under these conditions. Another difference between antimony and bismuth results from the reluctance of Bi to assume the 5+ oxidation state. Tetrahedral SbQ43' (Q=S,Se) anions, with Sb in the 5+ oxidation state, have been crystallized with a variety of counter cations whereas the bismuth analog is unknown.52 This phenomenon is sometimes called the "inert pair effect" and refers to the resistance of the inert pair of s electrons to participate in covalent bond formation. Thus, the trivalent state becomes progressively more stable as the group is descended.53a Recent theoretical work shows that 19 relativistic effects make an important contribution to the inert pair effect53b,c As in the case of bismuth, the isomorphous ASsz (A=Ii,Na,K; Q=S,Se,Te)54 family of layered compounds were synthesized along with the lower symmetry RbSsz and CsSsz members. Electrical conductivity and thermopower measurements on ASbTez compounds confirm the p-type semiconducting nature with room temperature conductivities that decrease upon moving from Ii (4 x 102 S/ cm) to Cs (2.5 x 10‘3 S/cm), while the Seebeck coefficients (a) are positive and increase from 465 (Ii) to 1440 nV/K (Cs) (see Table 1-2).55 The observed trend in electronic properties is dependent on the increase in alkali cation radii. The layers are separated by a larger distance in CsSbTez so there is less interlayer orbital overlap which gives rise to a larger anisotropy and larger band gap. Table 1-2. Properties of ASbTeZCompounds.56 Compound 0 (S/ cm), r.t. a (uV/ K), r.t. m. p. (°C) LiSbTez 4.0 x 10-2 NaSbTez 1.4 x 10-2 KSbTez 4.0 x 103 465 660 840 CsSbTe_2_ 2.5 x 10-3 1440 941 910 728 725 One-dimensional ASsz (A=Na,K,Rb,Cs; Q=S,Se) isomers exist where the electron lone pair of Sb3+ is stereochemically active.56 It has been reported that a-NaSbsz (one dimensional) undergoes a reversible structural transformation to B-NaBiSZ (two-dimensional) at 580 °C.56a The or --> 6 conversion is directly related to the loss of 20 the lone pair stereochemical activity of Sb3+. a-NaSbSZ is made by direct combination of Nazs and Sb283 at low temperature followed by slow cooling. It consists of trigonal bipyramidal SbS4 units that share opposite edges to form a chain-like polymeric structure. The Sb-S bond distances are 2.431(2) A (equatorial) and 2.773(2) A (axial). B— NaSsz is made by high temperature reaction of NaZS and Sb283 followed by quenching in water. A longer Sb-S bond distance of 2.888 A is observed due to the perfect octahedral coordination geometry. The K+, Rb+ and NH4+ analogs possess the same structural motif as a-NaSbsz but crystallize in different space groups.57 Increasing the counterion size (Cs+) results in a different structure consisting of SbS3 trigonal pyramids that share opposite corners to form one-dimensional [Ssz]n 11' chains.56d Schafer and coworkers have prepared a host of ternary alkali metal antimony chalcogenides using hydrothermal techniques with a variety of starting materials. The dimensionality of these structurally fascinating compounds extends from molecular to three- dirnensional. Two examples, K3Sb83-3Sb20358 and C528b3513,59 demonstrate this incredible structural diversity. K3SbS3-3Sb203 is prepared hydrothermally at 180 °C and possesses a novel macromolecular structure that is composed of Sb203 tubes that encapsulate K+ ions. These tubes are held together by molecular Sb833- trigonal pyramids that act as bridges by coordinating to K+ ions from three different tubes. CSsz3313 forms under hydrothermal conditions at 160 °C. SbS3 trigonal pyramids link via comers to construct large fragments consisting of six-, eight- and fourteen-membered rings. These fragments are bridged by Sb286 2 1 units to build two-dimensional sheets. The Sb256 units are composed of two edge-sharing SbS4 trigonal bipyramids (see Figure 1-8). This framework is representative of the types of Sb-S coordination environments that commonly prevail in this chemistry. As another illustration of the broad structural flexibility possible in this system, the isoelectronic compounds K28b4S7,60 KZSb4S7-H2Q61 afi-szSb4S7,62 (NH4)ZSb4S7,63 and Cssz4S7,64 all have different structures. Other examples of ternary antimony chalcogenides include CS6Sb10813-1.2H20,65 Tle3SS,66 and a-(B-)TleSe2.67 Upon increasing the charge on the counterion with the use of alkaline earth metal cations (Ca2+, Sr2+, Ba2+), Schafer and coworkers have synthesized molecular,68~ one-dimensional,69 and two- dimensional70 compounds with the use of hydrothermal and high temperature solid state reactions. Ca28b285 consists of isolated [SbS3]3' units and trans-[Sb284]2' dimers that are made up of two edge sharing Sb83 trigonal pyramids.68a Upon changing the counter cation to Ba2+, Ba4Sb4Se11 is formed.68b It is composed of four different molecular species: Sezz‘, SbSe33-, cis- and trans-[Sb28e4]2- dimers. Bagsbgs 17 contains SbS33' units and the only example of an isolated Sb3Sg7- anion. This anion is assembled from a central SbS4 trigonal bipyramid that bridges two SbS 3 groups by sharing opposite comers.63€ Two different one-dimensional chains are observed in Sr38b459.69a Two (SbS2)nn' chains made up of comer-sharing Sb83 trigonal pyramids and a double chain, (Sb285)n4n', consisting of edge sharing SbS 5 square pyramids, make up the formula unit. The chains are oriented in a sheet-like manner and the Sb lone pairs are 22 Figure 1-8: Projection of C523b3$13 viewed down the c-axis with labeling. 23 directed into a nonpolar domain that is segregated from the polar domain of the Sr+ cations. Edge- and corner-sharing of SbQ3 and SbQ4 units are also found in the layered structures of Ser4S7-6H20 and BaSb25e4.70 Sheldrick and coworkers have shown that a number of ternary phases can be easily prepared by the simple reaction of an alkali metal carbonate with the binary antimony chalcogenide under hydro(methanolo)thermal conditions. Examples include the layered RbSb3Se5,71 and the three-dimensional CS3Sb5®72 compounds. Also, CSZSb4Se6(Se2)73 is the only known example of a covalently bonded antimony polyselenide. The replacement of alkali metal cations with large organic counter cations leads to the formation of more open frameworks such as IN(CH3)4]Sb3Ss, [N(C3H7)4]Sb3Ss and (N2C4H8)Sb4S7-74 [N(C3H7)4]Sb3Ss and (N2C4H3)Sb4S7 are layered compounds that crystallize in a noncentrosymmetric space group (AmaZ), however, the weak second harmonic generation (SHG) signal is only four times that of quartz. In the case of [Sb385]nn-, a counterion size effect is observed. Substitution of [N(C3H7)4]+ by [N(CH3)4]+ would bring the [Sb3Ss]nn' layers close enough that considerable coulombic repulsions might develop, thus destabilizing the structure. These repulsions are overcome by combining the layers into a three- dimensional framework with intersecting channels. These compounds may serve as forerunners to a new class of microporous solids. Table 1-3 lists many of the known ternary antimony chalcogenides. Table 1-3. Ternary Antimony Chalcogenide Compounds. 24 Formula Synthesis Structure Ref. ASbQ2* direct combination a-NaSbSZ direct combination B-KSsz hydrothermal B—ASsz (A=Rb,Cs) direct combination (NH4)Ssz hydrothermal a-KSbSez direct combination B—KSbSez direct combination 2D 1D 1D 1D 1D 1D 1D K 331,3 3. 3Sb203 hydrothermal psuedo 3D 03sz38 13 hydrothermal K28b4S7 hydrothermal K28b4S7-H20 hydrothermal a-Rb28b4S7 hydrothermal B—szSb4S7 methanolothermal (NH4) 28b4S7 hydrothermal Cssz4S7 hydrothermal CS6Sb10318'1-2H20 hydrothermal Tle385 hydrothermal a-(B—)TleSe2 direct combination RbSb3Se5 methanolothermal Cs3Sb5® (Q=S,Se) methanolothermal Cssz4Se6(Se2) hydrothermal [N(C3H7)4]Sb355 hydrothermal [N2C4H3]Sb4S7 hydrothermal [N(CH3)4]Sb3Ss hydrothermal 2D 3D 3D Z) 2D 1D 2D 20 2D 2) 2D 2D 2D 2D 2) 31) CaszSs hydrothermal molecular BaSb4Se11 direct combination molecular Bang6S 17 direct combination molecular Sr3Sb489 direct combination Ba(en)4(SbSe2)2 room temp. soln. Ser4S7-6H20 hydrothermal BaszSe4 hydrothermal 1D 1D 21) 21) K3SbTe3 BaSbTe3 direct combination molecular hydrothermal ZD 54 56a 56b 56c,d 57 56e 56f 58 59 60 61 62a 62b 63 64 65 66 67 71 72 73 74a 74a 74b 68a 68b 68c 69a 69b 70a 70b 70c 40 * A=alkali metal; Q=S, Se, Te 2. Review of Metal Thio(seleno)phosphates 25 The chemistry of metal thio(seleno)phosphate compounds features broad structural diversity and interesting physical properties. These compounds can be grouped into four main classes: A. Orthothio(seleno)phosphate compounds with tetrahedral [PQ4]3' (Q=S,Se) ligands. Examples include BPS4, AlPS4, GaPS4, InPS4, Pb3(PS4)2,CU3PQ_4, Pd3(PS4)2, and LnPS4 (Ln=La, Ce, Pr etc.).75»76 The structures of these dense phases can be described as substituted metal chalcogenides with P occupying tetrahedral sites and M residing in tetrahedral, square planar (Pd2+), and square anti-— prismatic (Ln3+) sites. B. Thio(seleno)hypodiphosphates with {PZQE,]4- (Q=S,Se) ligands. These compounds fall into three subclasses depending on the metal coordination geometry. The largest and most important subclass is the M2P205 (M=Zn, Cd, Fe, Mn, Co, Ni, V, Mg; Q=S, Se) family with M2+ in an octahedral coordination environment.77 They possess layered structures related to Cdlz or CdClz in which P-P pairs and M2+ cations reside in the octahedral sites of the framework. The second involves divalent main group metals (Sn,78 Pb79) with trigonal prismatic coordination to form dense three-dimensional MszQ, lattices. The third consists of metals in tetrahedral environments to give compounds such as a-Ag4PZS6, B-Ag4PZS6, Ag4P28e6, and a- AgGaPzSe6.30 C. Compounds with pyrothiophosphate [P287]4- ligands. This ligand is not very common and only occurs in the molecular 26 AszP287,81 and solid state compounds such as Ag4PZS7, A87(PS4)(P287), H82P287, and RbVP237-82 D. Thiophosphates with Group 5 transition metals (V,33 Nb,84 Ta35). These compounds generally consist of a series of face-sharing trigonal prismatic dimers (M283), linked by various [PxSyJIl- anions ranging from [P83]3' to [P4313]6'. The isolated [PS4]3' anion has been crystallized with Ii+ and K+ cations.86 Recently, a-Na3PS4 has been isolated in pure form by the reaction shown below: 3Nazs + P285 ——7 2Na3PS4 eq. (2) 550°C AC-conductivity measurements show a-Na3PS4 to be a good ionic conductor with values between 4.17 x 10'6 (50 °C) and 8.51 x 10'2 S/ cm (500 °C).87 This compound also undergoes an a ---> B phase transition at 261 °C. Above 490 °C, there is evidence for a second high-temperature disordered phase, causing a steep increase in the conductivity. Several main group metal thiophosphates have many interesting physical properties. InPS475d crystallizes in a noncentrosymmetric tetragonal space group (I-4) which gives rise to a high non-linear susceptibility and piezo-coefficients.88 The structure can be viewed as a defect zinc-blende (ZnS) derivative, in which P and In atoms occupy half of the tetrahedral sites. Pb3(PS4)2 also crystallizes in a noncentrosymmetric space group P213 (cubic).75€ Second harmonic generation (SHG) measurements 27 revealed a weak signal about three times stronger than that of quartz. Pb3(PS4)2 was synthesized by direct combination of the elements at 600 °C. The Pb2+ is coordinated by eight S atoms in a truncated tricapped trigonal prismatic geometry with Pb-S distances between 2.88(1) and 3.47(1) A. In addition, crystals of GaPS475C exhibit a considerable birefringence.89 Another main group thiophosphate, BiPS4, features Bi3+ (stereochemically active lone pair) in two coordination environments.90 Bi( 1) is coordinated by six 8 atoms and occupies the center of a nearly square planar assembly of four 8 atoms with a dovetail-like arrangement of two further S atoms extending downward. The Bi(1)-S distances vary between 2.78(1) and 3.11(1) A. Bi(2) is coordinated to eight S atoms and is surrounded by four S atoms in a nearly planar rectangular geometry. In addition, two further pairs of S atoms, also in a dovetail-like configuration, are located above and below this plane. The Bi(2)-S distances vary between 2.68(1) and 3.30(1) A.20 Other main group compounds include BPS4,752l AlPS4,7Sb and T13PQ4.91 Several examples with transition metals have also been synthesized in this family of compounds. CrPS4 is the only example of an MPS4 compound with the metal in an octahedral coordination environment.92 It was synthesized from the elements in an evacuated quartz ampoule at 700 °C with a small amount of iodine added as a transport agent. Chains of edge-sharing CrS6 octahedra run along the b direction and are connected in the a direction by PS4 tetrahedra via corner- and edge-sharing to form the layered structure. High-temperature synthesis with Cu metal uncovered the isotypic Cu3PQ4 compounds, which have the wurtzite-related enargite 28 structure type.7Sf Electrical measurements on single crystals of Ou3PS4 and Cu3PS3Se show room temperature conductivities of 0.2 to 1.0 S/cm and carrier concentrations on the order of 1017 cm'3. Both materials are active as cathodes for the photoelectrolysis of water. In addition, optical spectroscopic measurements demonstrate that substitution of Se for S lowers the optical band gap from 238(5) to 2.06(4) eV.93 The use of a metal cation that prefers a different coordination geometry, such as Pd2+ (square pyramidal), can result in a new structure type. Purple-red crystals of Pd3(PS4)2 were obtained by reaction of the elements at 600 °C.758 Each [PS4]3' uses three S atoms to triply bridge three square planar Pd2+ centers. The fourth S atom is nonbonding and protrudes into the van der Waals gap (see Figure 1-9b). The Pd3(PS4)2 layers are not flat and form interdigitated sheets that stack in phase in the c direction as shown in Figure 1-9a. Photoelectrochemical characterization of this compound was performed to measure such properties as band gap, quantum yield for electron flow, and stability in a photoelectrochemical cell. This study was undertaken to discover possible semiconductor electrodes with built-in catalytic properties.94 The photoconductivity data suggest an indirect band gap of Eg=2.54 eV and possibly a direct gap of 2.89 eV. This compares to 2.15 eV determined by optical spectroscopy. A low quantum yield of about 1% at 3.1 eV was also measured. The small quantum yield and the large difference between the onset of optical absorption and the photoelectrochemically measured indirect transition can be ascribed to the low mobility of carriers in the material.94 29 (A) (B) 6:, Figure 1-9: (A) Packing diagram of Pd3(PS.4)2 viewed down the a-axis with labeling. (B) ORTEP representation of a single Pd3(PS.i)z layer. 30 Interestingly, extension to Au results in the formation of AuPS4, whose structure remains elusive. Considering the tendency of Au3+ to form square planar species, a polymeric chain structure consisting of alternating edge-sharing PS4 and AuS4 units has been proposed, based on vibrational spectroscopy.95 Quaternary mixed metal phases such as ZnAgPS4 and TlSnPS4 have also been investigated and structurally characterized.96 The reaction of f-block metals with phosphorous and sulfur at high temperature yields a large number of isostructural compounds (tetragonal, space group I41/acd, Z=16) with interesting luminescence properties (Nd, Tb).76 The compounds are air-sensitive because of the oxophilic Ln3+ metal center. The Ln3+ metal center resides in an eight-coordinate anti-square prismatic coordination site and probably prefers the [PS4]3- anion because of the charge balance with its stable trivalent oxidation state. An exception to this trend occurs with Eu. The Eu3+ analog could not be synthesized but instead formed Eu2P286 (Eu2+, trigonal prismatic environment), which is isostructural to szPZQE,.79 A large number of compounds of the formula M3(PS4)2 (M2+ = Fe, Ni, Zn, Cd, Cu) were reported but their XRD patterns are consistent with MzP 2%.97 Thus they probably do not exist. The largest class of compounds are the thio(seleno)hypodiphosphates which feature the versatile [P2Q,]4- ligand. The six Q atoms provide many potential bonding modes to metals. The structure of the [P2Qg]4' ligand resembles that of a staggered ethane molecule and contains a P-P bond in which the 31 formal oxidation state of P is 4+. Alkali metal salts of the [P286]4- anion have been synthesized for infrared and Raman spectroscopic studies.93 The divalent transition metals form a large subclass of compounds with the formula MzPZQ, The first members of this family of compounds were prepared and characterized in 1965 by Hahn and coworkers.77al These compounds are structurally related to the CdIz or CdClz structure types in which the M2+ cation and P-P pairs reside in the octahedral sites of the framework. It has been observed that the P-P bond lengthens to accomodate larger M2+ cations, thus providing a flexible lattice. The stacking of these layers most commonly gives rise to a monoclinic unit cell with space group C2/ m, though hexagonal forms are also known. The lamellar structure of N12P286 is shown in Figure 1-10a. Figure 1-10b displays a single NizPZS6 layer with the hexagonal arrangement of the Ni2+ centers. Optical transmission measurements on M2P286 (M=Mn, Zn, Cd) reveal that they are transparent, slightly colored insulators with band gaps of 3.0, 3.4, and 3.5 eV, respectively, whereas M2P2$6 with M=Fe, Ni are dark semiconductors with gaps of 1.5 and 1.6 eV. The M2P25e6 analogs with M=Mn, Fe have band gaps of 2.5 and 1.3 eV, respectively.99 The magnetic behavior of MszQg, compounds containing magnetic ions such as Mn2+ (d5), Fe2+ (d6), and Ni2+ (d8) have been thoroughly investigated.99,100,105a In the case of anPZS6, it is best described as a two dimensional Heissenberg antiferromagnet, with a Neel temperature (TN) of 78 K. The d5 electrons are in the high-spin 32 (A) (B) _ . . h $ . w . a . ~ . . . . n a « » . s » . r e v a w H _ - . fl ”v fi fi f fi v u e i l f fi w u o fl o W o N A . » . a » . w e fl Figure 1-10: (A) Schematic representation of NizP286 viewed down the a—axis. (B) ORTEP representation of a single NizP286 layer. The hexagonal Ni array is outlined with dashed lines. 33 state. Neutron diffraction studies have determined the magnetic structure, which consists of a two dimensional honeycomb lattice in which each Mn2+ ion is antiferromagnetically coupled to its three nearest in-plane neighbors, and in which all the spins are perpendicular to the layer plane.101 The antiferromagnetic interactions are due to super-exchange through Mn-S-Mn bridges. Table 1-4 shows magnetic data for anPzQé, FeszQg, and Ni2P286. Table 1-4. Magnetic Data for MzP 2Q, Compounds. Compound ueff (BM) 6 (K) TN (K) M-M (A) Ref. anPzS6 5.97 -160 78 3.50 105a anPzSe6 6.1 -201 85 3.69 99 FeszS6 5.44 5.0 FeszSe6 5.09 5.0 +15 +65 +37 -4 NiszS6 3.68 -S 59 3.9 -7 12 126 133 123 1 18 254 300 3.42 100C 99 3.62 100b 99 3.36 100C 99 This class of compounds exhibits rich intercalation chemistry. Niszse and Fe2P2(25 have been shown to intercalate lithium either chemically,99 upon reduction with butyl-lithium, or electrochemically. 102 These compounds are of potential importance as low-dimensional cathode materials for secondary lithium batteries.103 In 1980, Clement and coworkers discovered one of the 34 most unusual reactions found in intercalation chemistry.104 They observed that treatment of Mn2P286 with aqueous solutions of a number of ionic salts resulted in a removal of Mn2+ from inside the layers, and concomitant insertion of the guest cations in the inter- layer galleries. A wide variety of organometallic (cobalticenium), ammonium, and alkali metal cations intercalate into the host lattice. It has also been shown that pyridine and various polyethers can also be intercalated.10S Recently, Clement and coworkers have reported that ion-exchange intercalation of M2P2S6 (M=Cd, Mn) with a cationic organic dye induces a large second-order optical nonlinearity and permanent magnetization (Mn2+).106 These reactions are very general and can be written as in equation (3) (G+=guest monocation). MnPs3 + 2xG‘“aq ——> Mn1-xPS3,G2x,(HZO)y + an2+aq eq.(3) This process occurs by a different mechanism than is commonly found in traditional ion exchangers, where ions already in the interlayer gallery region are exchanged with guest ions. This behavior is observed for other labile M2+ cations such as Zn2+ and Cdz+, but more rigorous conditions are required for Fe2+ and Ni2+ which induce a crystal field stabilization. Interestingly, the ease of this chemistry has been related to the thermal parameters of the transition metal ion as determined by their crystal structure. 107 The labile Mn2+ ion has a high temperature factor so that it appears to rattle in the loose cage of S atoms. In addition, it has also been shown that intercalation transforms the antiferromagnetic host into a ferro or ferrimagnetic compound. 103 35 The remarkable structural flexibility of the host lattice has been demonstrated by the existence of compounds such as V1.56P286109 and In1,33P2Q,75d»110 which contain metal deficient layers. V1.56PZS6 is a mixed valence (V2+/V3+) compound as shown by the formula, V2+Q63V3+Qggfl 0,44P2S6 (II=vacancy). In1,33P2$675d is a metal-deficient threefold superstructure of the FezPZS6 type formed by the ordering of metal vacancies. It is an electrical insulator with an optical absorption edge at 3.1( 1) eV. The selenium analog crystallizes in two forms. In a-In1,33PZSe6,110 the metal vacancies are disordered while those of the B-phase are ordered and form a sixfold superstructure of the Fe2P28e6 structure type.75d Single crystals of a-In1,33P28e6 are highly resistive (>108 Q-cm) with an optical band gap of 1.9 eV.75d It is noteworthy that a-In133P28e6 may be suitable for photovoltaic devices.111 Substitution of the M2+ cations, via direct combination of the elements at high temperature, has been demonstrated for M2P286 to form MIM'IIIPsz,112 and MIZXM'112-xP286113 (MI=Ag, Cu; M'I I=Mn, Cd, Zn; M'I 11=V, Cr, In, Sc), which retain the stable Cdlz (CdClz) structure- type and subsequently enlarge this family of compounds. Recent investigations have focused on the possible existence of Cu+-Cu+ or Ag+---Ag+ dimers located in octahedral sites in CuCrPZS6,114 CuVPzS6,115 Cuo,szMn1,74PzS6113b and AgzManS(,.116 The crystal structure of CuCrP286 reveals a Cu electron density cloud with three extrema. One of them is in the middle on a pseudo-octahedral site and the other two are near the trigonal sulfur planes which suggests that CuS6 octahedral sites, Cu---Cu pairs and vacancies all exist at room temperature in the disorder model. Complimentary EXAFSll3b 36 studies have confirmed this disorder model but recent low- temperature neutron powder diffraction studies have disputed the existence of bimetallic entities.117 The structure of CuVP286 features the same Cu positions as found in CuCrPZS6 plus a third tetrahedral site located in the van der Waals gap. The location of copper atoms in 010,52Mn1,74P286 remains uncertain due to some inconsistencies between X-ray diffraction and EXAFS data. It appears that the occurrence of copper pairs in these copper substitution phases is still an open question. The structure determination of AgzMnPZS6 confirmed the presence of Ag+---Ag+ pairs with the Ag atoms found in trigonal planar coordination. Kniep and coworkers have extended this substitution chemistry to include the MzPZSee analogs and form layered quaternary selenodiphosphates for possible solar-cell applications.80d Large single crystals were prepared from the elements at 750 °C followed by quenching or by cooling from the melt (Bridgman technique). The crystal structures of CuInPZSea, AgInPZSe6, and B-AgGaPZSee are related to the CdIz structure type whereas the CuCrPZSe6, AgCrPZSe6, and AgAlPZSee. are related to the CdClz structure type. The most promising material in this group of compounds was found to be AgInPZSe6. The congruent melting behavior and perfect layer structure make it possible to grow thin fihns with crystallographic orientation directly from the melt. The crystals are dark red with a metallic luster and display a sharp band gap at Eg=1.79 eV. The second subclass involving [P286]4' ligands are the MszQ, (Q=S,Se) compounds with Sn2+ and Pb2+. Sn2+ and Pb2+ are found in distorted trigonal prisms of S(Se). Three additional Q atoms are 37 located above the centers of the vertical prism faces to form a tricapped trigonal prism so the effective metal coordination is nine. The P-P pairs reside in the octahedral sites to form the dense three- dirnensional network. The first representative of this group was Sn2P2S6, prepared in single-crystalline form by iodine transport.78a Several research groups synthesized and characterized Pb2P286, SnzPZSe6, and Pb2P28e6 independently.77ci78»79 A detailed structural analysis showed Sn2P285 to crystallize in the acentric space group Pc whereas the other members belong to the centric space group P21/c. It has been shown that Sn2P2S6 undergoes a second order, exothermic phase transition from ferroelectric (Pc) to paraelectric (P21/c) at 60 °C.118 A crystal structure at 110 °C confirmed that the transition results from the movement of Sn atoms within the rigid [P286]4- lattice such that they become related by a center of symmetry. It is possible that the stereochenrically active Sn2+ lone pair of electrons causes a distortion of the Sn—S coordination sphere during this transformation. This compound has been shown to be a promising ferroelectric material for use in memory devices.119 The optical band gaps and melting points of these compounds are shown in Table 1-5. Table 1-5. Optical Properties of SnszQ, and szP 2%.733 Compound E2 (eV), 295 K E; (eV), 77 K m. p. (°C) SnszS6 2.34 2.49 SnszSe6 1.79 szPzS6 2.56 szPzSeg 2.10 2.12 2.77 2.23 775 670 914 775 38 The third subclass of [P2Q,]4- compounds consist of metals in tetrahedral coordination. In or-Ag4P286,8oa the [P256]4' groups form ahnost planar layers that are connected with AgS4 tetrahedra via common edges and comers to form the three-dimensional structure. B-Ag4PZS680b is isostructural with Ag4PZSe680C (space group P212121) and possesses a higher symmetry version of the a-phase. The crystal structure of the mixed metal compound, a-AgGaP28e6, contains P28e6 octahedra which share common edges with GaSe4 tetrahedra, to give polyhedral chains running parallel to the [100] direction.80d In the [010] direction, the chains are interconnected by sharing common edges and corners with AgSe4 tetrahedra. The resulting layer structure is related to the three-dimensional structure of Ag4PZSe6 by similar principles of polyhedral condensation. Extension of this chemistry to Zn, results in the formation of Zn4( P286) 3,120a which contains one [P286]4' and two [P286]2- ligands. In [P286]2-, the two P atoms are linked by two bridging S atoms (two PS4 tetrahedra sharing an edge) so P is in the 5+ oxidation state. A second example of a compound with [P256]2- ligands is AgzP 286.1201) It should be noted that reaction with Ti metal does not result in the formation of an MszQ5 compound but instead forms TiPZS6, a new three-dimensional structure type.17-1 Although Ti prefers an octahedral environment, the stable 4+ oxidation state prevents crystallization of an MzP 286 derivative. In TiPZS6, the Ti atoms attain a distorted octahedral coordination with Ti-S distances ranging from 2.433(3) to 2.454(5) A. 39 Metal pyrothiophosphates feature a [P287]4- ligand formed by the comer-sharing of two PS4 tetrahedra. The strong IR active S-P-S stretching vibration (466 cm'l) is a useful diagnostic for indentification of this ligand in a compound.122 The Se analog is unknown. The [P287]4- ligand is not very common and only occurs in a few compounds. It has been noted that complicated Lewis acid- base equilibria exist among [P286]4-, [P286]2-, [P287]4', and [PS4]3- ligands in melts at high temperatures.123 The relative scarcity of compounds with the [P287]4' ligand as compared to those of [PS4]3- and [P286]4-, seems to suggest that the complex equilibria favors the formation of the latter two species. The AszPZS7 molecule is formed at 350 °C and is isotypic to P487 (sz symmetry).81 An As2+-As2+ dimer is chelated by four terminal S atoms of the [P287]4‘ ligand to form two six-membered rings that are joined by the As-As bond and the bridging S-P-S unit. The terminal P-S bond distances are shorter due to their partial double bond character. 124 Reaction of P and S with Ag metal at 600 °C resulted in Ag4PZS7, a compound with a three-dimensional structure consisting of edge- and corner-sharing of AgS4 tetrahedra and P257 polyhedra.82a A second Ag containing phase, Ag7(PS4)(PZS7), was isolated and features [P257]4' and [PS4]3* ligands. The Ag+ ions are disordered among the interstices of the anionic network which suggests that Ag7(PS4)(PZS7) could exhibit ionic conductivity.82b Reaction of P and S with HgS at 240 °C, results in the formation of Hg2P287.32C The Hg2+ ions reside in distorted tetrahedra that connect layers of P287 groups to form a three- 40 dimensional structure. Extension to quaternary systems involving alkali metals has recently been accomplished with the preparation of RbVPZS7.82d This phase was formed as a result of a failed attempt to synthesize a new substituted M2P286 compound. The acentric structure is shown in Figure 1-11. Each V86 octahedron of RbVP287 shares two edges with the PS4 tetrahedra of two [P287]4- ligands and two corners with the PS4 tetrahedra of a third [P287]4- ligand to form flat sheets. The Rb+ cations reside in between the layers. Since the early 1980's, a new and exciting family of transition metal thiophosphates has emerged involving the Group 5 transition metals (V, Nb, Ta). Thiophosphate and disulfide ligands provide either octahedral or bicapped prismatic coordination to the metal. These compounds are found in a variety of structures which vary from low dimensional linear chains and layers to three-dimensional helical tunnel frameworks. The crystal structures of these compounds deviate sharply from those found for late transition metals and feature new [PxSy]n- ligands that have not been previously reported. These compounds are synthesized by direct combination of the elements in the 400-700 °C range. This class of compounds has recently been reviewed by Brec and coworkers.125 In addition, the lithium intercalation chemisty of these compounds has been explored for possible battery cathode applications. Band structure calculations show that low-lying d-block acceptor orbitals are conducive for lithium intercalation.125 Table 1-6 lists these new ligands along with examples of previously known thiophosphate (B) Figure 1-11: (A) Packing diagram of RbVPZS7 viewed down the a-axis. '(B) ORTEP representation of the [VP257]nn‘ anionic layer with labeling. Table 1-6. Types of Thiophosphate ligands 42 _l_igand - Examles Ref. ' .13 3' S ’4 \ S S 1 ' 3- s | P s /l \s S _ P0,2vs2 (2D) 83a TaP86 (3D) 85a S\ x S .\ /3 NbPzSg (2D) 84a 2_ P P s/ \s/ \s S\ :8; 4” Mn2P256 (2D) 107a IR— P\ 1 SS 3 2 3 S \ /S\;/ IS ‘4' H P s 82 2 7 (31)) 82c .2? s s \ —s )1) S I S S \ x S A '4- IS ) , ’s ‘s—P \ s .1 s (,5 P—S S—P s 3' s VzPSlo (1D) 83C Nb4P2321 (2D) 84b . Table 1-6. (contd.) 43 ligand Example Reference ' _ _ s s s s?» ‘6- f s V2P4Sl3(2D) 83b \ _ S/ \S /s_P .4? S \P 81.8 \ ‘ _ _ 4_ S f P S/ \S NbPzSg (3D) 84d s.\ / s7 P \S \,,s P s/ ‘s _ \P/ \ 5 ‘s _ units. [P286]4' and [P287]4' are the only two thiophosphate anions that are not found in this Group 5 chemistry. The thiophosphate chemistry of vanadium is characterized by octahedral or bicapped trigonal prismatic coordination and an oxidation state of 3+ or 4+ to form one- or two-dimensional compounds. The structure of PQZVSZ is closely related to the layer compound Cdlz33a and possesses mixed v3+/4+ valency. In each layer, approximately every fifth V86 octahedra is capped by a P atom extending into the van der Waals gap. It represents the only example of a [PS3]3' ligand (P3+). The report of this ligand in this 44 non-stoichiometric compound seems dubious. Only polycrystalline samples of PQZVSZ could be prepared so the structure was analyzed using powder X-ray and neutron diffraction techniques. The calculated P-P' bond distance is 2.70 A which is much longer than a single P-P bond (2.20 A). This distance raises the question of whether weakly bonded P---P pairs exist in the compound or the distance is only an artifact of the partial P occupancy. Another interesting compound is V2P4513 which possesses the unprecedented [P4513]6- polymeric ligand.33b The V3+ octahedra share edges to form dimers. Each ligand connects to four different dimers to construct the V2P4513 sheets (Figure 1-12). The one-dimensional VzPSm is also known.83C Further studies with niobium revealed several ternary metal thiophosphate compounds with layered or three-dimensional structures. The preferred coordination geometry of Nb in this thiophosphate chemistry is bicapped trigonal prismatic. This coordination is very common in the binary MQ3 (M=Ti,Zr,Hf,V,Nb,Ta; Q=S,Se) layered chalcogenides. In MQ3, the bicapped trigonal prisms share triangular faces to form chains whereas the prisms in Nb/ P/ S compounds share rectangular faces through two S-S bonds to form [NbZS 12] dimeric building blocks as shown in scheme (A). Scheme (A) 45 Figure 1-12: A schematic representation of a single V2P4813 layer. The open spheres are S atoms and dark lines highlight the bonding in the [P4513]6' ligand. 46 These building blocks feature NbIV-NbI V bonding through the rectangular face with a typical distance of 2.869(1) A. The layered NbP28334a phase is assembled through linking of [NbZS 12] units by [P286]2' ligands. Two other layered compounds are obtained by end to end edge-sharing of [Nb2812] units to form [Nb289] chains. These chains can be connected by [P289]4- ligands (Nb4P2821)34b or [PS4]3- units (szPS 10184C A three-dimensional tunnel framework, NbP288, is obtained by bridging of four [NbZS 12] units with an unusual cyclic [P4512]4- ligand. The formula can be represented as Nb2(P4$12)(32)2.84d Similar investigations with the more electropositive Ta results in three-dimensional tunnel networks with Ta in the 5+ oxidation state. Closed-tube phase iodine transport of TaS6 (Ta(PS4)(Sz)) yielded gray single crystals.8521 The compound crystallizes in a tetragonal unit cell with an acentric space group (I41/acd). The familiar [Tazs 12] units are connected by [PS4]3' units to form small channels along the c-direction. A structurally related phase was synthesized at 500 °C. The black needle-like crystals of Ta4P4829 also have a tetragonal unit cell but crystallize in a different acentric space group (P435212).85b The compositional difference between the two compounds results from right-handed, helical, polymeric S 10 chains inserted in the tunnels of Ta4P4829 which are stabilized by van der Waals forces. High temperature chalcogen substitution reactions with Se yields a mixed Segsz chain in the tunnels.85C A minor phase was also characterized in these reactions and found to be TazP 2811, a tunnel structure with two different size channels.85d 47 It has been noted that niobium is a frontier element separating more covalent 1D and 2D vanadium derivatives from the more ionic 3D tantalum phases. There is a scarcity of quaternary compounds with alkali metals. Such compounds could exist and would be very interesting as pseudo-ternary A+/Mn+/[Pny]Z' compounds. The alkali metal not only donates electrons but also would act as a structure-directing counterion to generate new anionic frameworks. 48 3. Alkali Metal Polychalcogenide Fluxes About two decades ago, Scheel recognized the potential of alkali metal polysulfides as reaction media while concentrating on the recrystallization of known materials at high temperatures. Materials such as ZnS, CdS, MnS, PbS, NaCrSz, KCrSz, NaInSz, KFeSz, FeSz, NiSz, 0082, M082, Nsz, and HgS were grown successfully from sodium polysulfide fluxes at high temperatures (>700 °C) by Scheel and others.126 Sodium polysulfides were used because of the relative inertness of Na+ ions toward formation of ternary compounds. Ironically, while Scheel tried to avoid alkali metal incorporation, research in the Kanatzidis group over the last six years has focussed on trying to promote it with the use of various low temperature (150-500 °C) molten alkali metal polychalcogenide fluxes. This methodology has led to the discovery of nearly one hundred novel ternary and quaternary compounds with exotic structures and desirable physical properties. Alkali metal polychalcogenide fluxes are attractive for the synthesis of new ternary chalcogenide compounds for a variety of reasons. First, these fluxes have a wide range of melting points so that intermediate temperatures can be explored (see Table 1-7). It is necessary to remain in this temperature regime (150-500 °C) so that known thermodynamically stable compounds may be avoided. Second, the fluxes not only act as a solvent but also as sources of the elemental components of the products. Third, the polychalcogenide fluxes solidify upon cooling to a glassy solid. This glassy matrix can be removed by dissolution in water or polar organic solvents so that pure products may be obtained. Finally, the flux is a highly oxidizing medium as well as a basic (nucleophilic) one. The basicity of the Table 1 -7 . Melting Points (°C) of Some Known Alkali Metal Polychalcogenides. LizS H28 2 900-9 7 5 3 70 Na28 Na282 Na28 3 Na284 Na28 5 1180 490 229 275 252 NaZSe Na25e2 NaZSe 3 NaZSe4 >875 495 313 290 NazTe NazTez 95 3 348 Na28e6 258 NazTe6 43 6 K28 K282 K28 3 K284 K28 5 K286 840 470 252 145 206 189 K28e K28e 2 K28e 3 K28e4 K28e5 460 3 80 205 190 Rb 28 Rb 282 Rb 28 3 Rb 2S4 Rb 28 5 Rb 286 530 420 213 160 225 201 C828 C5282 C828 3 C8284 C8285 C8286 460 2 1 7 160 2 10 1 86 50 medium can be controlled by varying the AzQ/Q ratio as shown below. AzQ + (x-1)Q ------> Asz eq.(4) A short chain has strong Lewis basicity while a long QXZ- chain is less basic because the 2’ charge is dispersed over more Q atoms which reduces the negative charge on the terminal atoms. All of the Q atoms, especially the negatively charged terminal atoms, provide effective binding sites to the metal. The metal is oxidized and then attacked by the negatively charged species as shown in scheme (B). _ Q \0. - ..__ — Q Q " 7 2- Q’Q / M° + | —> M“ + Q —> M11 \ _ Q-/ —Q""""Q. Scheme (B) b Q\Q ‘ The resulting Mn+ cations, alkali metal cations and polychalcogenides associate to form various structures, depending on the coordination preference of the metal, the size of the alkali metal cation, the concentrations of the various ligands in the flux, and the reaction temperature. In 1987, Ibers and coworkers reported the first examples of the use of alkali metal polychalcogenides to synthesize new ternary metal chalcogenide compounds. K4Ti3814128 and NazTi28e3129 were 5 1 isolated from K28x and NaZSex fluxes, respectively, in the 375-470 °C temperature range. The structures of these new compounds consist of one-dimensional polymeric chains of [Ti3(Sz)6(S)2]n4n' and ['l‘i2(Se2)3(Se)2]n2n- and contain octahedrally coordinated Ti4+ centers bonded to Q22- and Q2- ligands. The full potential of this synthetic method was not realized because of the relatively high temperatures used in these reactions. In our laboratory, Y. Park has thoroughly explored the chemistry of late transition metals (Cu, Au, Hg) in the lower temperature regime (215-450°C). He has isolated a number of new metal (poly) chalcogenides with long QXZ- (x=2,3 ,4,5) ligands.130 The A/Au/ Se (A=Na,K,Cs) system is particularly interesting because of its structural diversity associated with various polychalcogenides from Se3Z' to Sesz- and the novel Au1+/ 3+ redox chemistry. KAuSes was prepared from the reactant ratio Au/KZSe/ Se of 1/ 2/ 8 at 250 °C This is a Au+ compound, composed of noncentrosymmetic one-dimensional [Au(Se5)]nn‘ chains running parallel to the c-axis as shown in Figure 1-13a. The [Au(Se5)]nn- chains are composed of Sesz- ligands bridging two Au atoms via terminal Se atoms. These chains interact with each other to form dimers through close Au---Au d10-d1O contacts (2.950(3) A). A slight reduction in the K28e ratio from 2 to 1.8 yielded the novel K3AuSe13 compound, the most Se rich compound known to date. The [Au(Se3)(Se5)2]n3n' chains can be described as a one dimensional assembly of square plane Au3+ centers that have two arm-like trans Sesz- ligands and are bridged by Se 32' ligands, counterbalanced by K+ ions. (see Figure 1-13b). It is remarkable that a slight change in the 52 Lewis basicity from KZSes to K28e5,4 can result in the formation of two completely different compounds with unusual structures. By doubling the amount of Au metal and increasing the reaction temperature to 290 °C, KAuSez was isolated. The structure consists of AuIIISe4 square planes that share opposite edges to form one- dimensional chains. By reducing the size of the alkali counter cation from K+ to Na+ and increasing the basicity of the flux, the one- dirnensional chains are condensed into [AuSez]- layers, separated by the Na+ cations. AuSe4 square planes share edges and corners to form the anionic sheets. The structures of the two anions are shown in Figure 1-14. The layered versus one-dimensional structure is a clear manifestation of a couterion size effect. J.-H. Liao expanded the scope of these reactions by exploring the polysulfide chemistry of the main group tin metal. Several ternary tin polysulfides such as K28n283, a-Rb28n283, B-Rb28n253, KZSnZSS, C528n286, and CstnS 14 were synthesized.13la He extended this chemistry to mixed metal systems using [SnQ4]4' and [Sn2Q,]4- building blocks. Mixed metal phases such as KzAuZSnS4, KzAuZSnZS6, and Rb2CqunS4 all possess novel structures.131b In the reaction of Sn with alkali metal polysulfides, it was observed that basic fluxes favor formation of molecular compounds. The reaction of Sn with less basic fluxes such as K285 at 275 °C yielded KZSnZS 3. The unique anionic structure contains tetrahedral and octahedral Sn4+ centers and is shown in Figure 1-15. These Sn centers are linked via 52' and (S4)2' so the layered anionic 53 (A) (B) . ’ O o O . . Sci o . o o , ‘ III Sela) - . Sofa) . 0 30H! o . . o o . A. . A. g 3“” O . . SQ!” Figure 1-13: (A) One-dimensional structure of [AuSe5]nn'. Au---Au contacts are indicated by dashed lines. (B) Structure of the [AuSe131n3 11‘ chain. 54 (A) (B) Figure 1-14: (A) A single anionic layer in NaAuSez. (B) The one- dimensional chains in KAuSez. 55 0 q ' . o . . . s O . ‘ W. 0 . . . 0 s 0 i4 . \‘I " SH) ‘ 50(2),‘8 5(5) b - 1 5(7) 5(4) 1? Figure 1-15: ORTEP representation and labeling scheme of the layered structure of [SnzsshZDz The dashed lines indicate the shortest nonbonding Snu-S contacts (2.934(5) A). Figure 1-16: ORTEP representation and labeling scheme of KzAuZSnzse. View down the b-axis. Au-«Au contacts are shown by dashed lines. 57 framework can be expressed as [Sn284(S4)]2-. This is one of the few examples of a tin polysulfide. The addition of Au in Sn/KZSx mixtures at 350 °C can provide structural flexibility due to the preference of Au+ for linear coordination. The KzAu28n256 structure is one-dimensional, featuring edge-sharing bitetrahedral [Sn286] units connected by linear Au+ atoms to form infmate chains. The [Au28n286]2' chains lie parallel to the crystallographic c-axis and are separated by potassium cations as shown in Figure 1-16. There is also a Au---Au short contact at 3.010(2) A. The Au---Au contact occurs inside the Sn(SAuS)ZSn eight-membered ring. Inspired by the enormous success of Park and Iiao, we decided to explore the chemistry of two Group 15 main group metals, antimony and bismuth. We have isolated a number of interesting compounds with novel structural architectures. In addition, high temperature reactions with K28 and Bi283 have resulted in compounds with interesting semi-metallic electrical conductivity and thermoelectric power. The results will be described in Chapters 2 and 3. During our studies with polychalcogenides, we noticed that the chalcophosphates [Pnyln ' were also rather uncommon in their occurrence in solid state compounds. We utilized our experience with alkali polychalcogenide fluxes to develop a general methodology by which new thio(seleno)phosphate (or chalcophosphate) compounds can be consistently obtained. These polychalcophosplrate fluxes, Ax[Psz], provide highly charged [PXQyP' anions that coordinate to the MM center. We have synthesized several bismuth 58 and antimony thiophosphates with [P287]4-, [PS4]3-, and [P286]4- ligands (Chaper 4). Further exploration using alkali metal polyselenophosphate fluxes with metals such as Sb, Bi, Mn, Fe, Cu, Ag, and Au resulted in the formation of a variety of new quaternary compounds with novel structure types and interesting physical properties which are also the subject of this thesis. The quaternary selenophosphate compounds will be discussed in Chapters 5 and 6. 59 List of References " ! N “ 5 5 : 9 9 1 > Takeuchi, Y.; Sadanaga, R. Z. Kristallogr., 1969, 130,346-368. Berry, L. G.; Mason, B. Mineralogy, W. H. Freeman and Co.: San Francisco, 1959. Wuensch, B. J. Z. Kristallogr., 1964, 125,459-488. Niizeki, N.; Buerger, M. J., Z. Kristallogr. 1957, 109,129-157. Kohatsu, 1.; Wuensch, B. J. Acta Cryst. 1971, 827,1245-1252. Iitaka, Y.; Nowacki, W. Acta Cryst. 1962, 15, 691-698. (a) Smith, M. J.; Knight, R. J.; Spencer, C. W. J. Appl. Phys. 1962, 33(7), 2186-2190. (b) Testardi, L. R.; Bierly, J. N. Jr.; Donahoe, F. J. J. Phys. Chem. Solids, 1962, 23, 1209. (c) Champness, C. H.; Chiang, P. T.; Parekh, P. Can. J. Phys. 1965, 43, 653-569. ((1) Yim, W. M.; Fitzke, E. V. J. Electrochem. Soc. 1968, 115,556-560. (a) Rowe, D. M.; Bhandari, C. M. Modem Thermoelectrics, Holt, Rinehart and Winston: London, 1983; p. 103. (b) Borkowski, K.; Przyluski, J. J. Mat. Res. Bull. 1987, 22, 381-387. (c) Ibuki, S; Yoschimatsu, S. J. Phys. Soc. Japan, 1955, 10, 549-554. ((1) Jeon, H-W.; Ha, H-P.; Hyun, D—B.; Shim, J-D. J. Phys. Chem. Solids 1991, 4,579-585. Kaibe, H.; Tanaka,Y.; Sakata, M.; Nishida, l. J. Phys. Chem. Solids 1989, 50, 945-950. 10. Atabaeva, E. Ya; Itskevich, E. S.; Mashkov, S. A.; Popova, S. V.; Vereshchagin, L. F. Sov. Phys-Solid State, 1968, 10,43-46. ll. Atabaeva, E Ya; Mashkov, S. A.; Popova, S. V.; Sov. Phys.- Crystallography, 1973, 18, 104-105. 12. 13. 14. 15. 16. Anderson, T. L.; Krause, H. B. Acta Cryst. 1974, 30B, 1307-1310. Nakajima, S. J. Phys. Chem. Solids, 1963, 24,479-485. Katsuki, S. I. J. Phys. Soc. Japan, 1969, 26, 53. Pecheur P.; Toussaint, G. Phys. Lett. A, 1989, 135,223-226. (a) Hofmann, W. Z. Kristallogr. 1933, 86, 225. (b) Scavnicar, S. Z. Kristallogr. 1960, 114, 85-97. 17. Tideswell, N. W.; Kruse, F. H.; McCullough, J. D. Acta Cryst. 1957, 10, 99- 102. 18. Kanishcheva, A. 8.; Mikhailov, Yu. N.; Trippel, A. F. lnorg. Mater. 1981, 17,1466-1468. 19. 20. Ibuki, S; Yoschimatsu, S. J. Phys. Soc. Japan, 1955, 10,549-554. Platakis, N. S. J. Non-Crystalline Solids, 1978, 27,331-346. 21. Chikawa, J.; Sato, 8.; Kawamura, T.; Goto, N. Proc. Plan Meet. X-ray lnstrum. Photon Fact, 1979, 87-88. 60 22. McKay, D. L. Belg. 866,332, US. Appl. 819,027, 1977. 23. King, J. P.; Asmerom, Y. ASLE Trans. 1981, 24, 497-504. 24. Bonnet, D. Photovoltaic Sol. Energy Confi, 1979,387—395. 25. Ibanez, A.; Olivier-Fourcade, J.; Jumas, J. C.; Phillippot, E.; Maurin, M. Z. Anorg. Allg. Chem. 1985, 540, 106-116 and references therein. 26. 27. Ismail, F. M.; Hanafi, Z. M. Z. Phys. Chem., Leipzig, 1986, 267,667-672. Bube, R. H. Photoconductivity of Solids, John Wiley and Sons, Inc.: New York, 1960; p. 234. 28. Smith, M. J.; Knight, R. J.; Spencer, 8. W. J. Appl. Phys., 1962, 33,2186- 2130. 29. 30. Gobrecht, H.; Boeters, K.-E.; Pantzer, G Z. Phys., 1964, 177, 68—83. (a) Boon, J. W. Rec. Trav. Chim. Pays-BaS, 1944, 63, 32. (b) Glemser, O.; Filcek, M. Z. Anorg. Allg. Chem., 1955, 279, 321-323 (c) Gattow, G.; Zemann, J. Z. Anorg. Allg. Chem., 1955, 279,324—327. 31. Trippel, A. F.; lazarev, Berul, S. 1. Russ. J. Inorg. Chem., 1978, 23,390- 391. 32. Voroshilov, Y. V.; Peresh, E. Y.; Golovei, M. I. Inorg. Mater., 1972, 8,777- 778. 33. (a) Peresh, E. Yu.; Golovei, M. 1.; Berul, S. I. Inorg. Mater. 1971, 7, 27-30. (b) Golovei, M. 1.; Berul, S. I.; Luzhnaya, N. P.; Peresh, E. Yu. Inorg. Mater. 1970, 6, 961-963. 34. 35. Schmitz, D.; Bronger, W. Z. Naturforsch., 1974, 29b, 438-439. Kanishcheva, A. S.; Mikhailov, J. N.; Lazarev, V. B.; Trippel, A. F. Dokl. Akad. Nauk, SSSR (Kryst), 1980, 252, 96—99. 36. R. G. Burns, M. B. Burns Manganese Dioxide Symposium, Tokyo, 1981; Vol. 2, Ch 6. 37. G. Cordier, H. Schafer, C. Schwidetzky, Rev. Chim. Miner., 1985, 22, 676- 683. 38. Julien-Pouzol, M.; Jaulmes, S.; Laruelle, P. Acta. Cryst. 1979, B35 , 1313- 1315. 39. G. Cordier, H. Schafer, C. Schwidetzky, Rev. Chim. Miner., 1985, 22, 631- 638. Cook, R.; Schafer, H. Rev. Chim. Miner. 1982, 19, 19-27. 61 41. 42. 43. K. Volk, G. Cordier, R. Cook, Schafer, H. Z. Naturforsch., 1990, 35b , 136- 140. B. Aurivillius, Acta Chem. Scand. 1983, A37 399-407. Iiautard, B.; Garcia, J. G.; Brun, G.; Tedenac, J. C.; Maurin, M. Eur. J. Solid State Inorg. Chem. 1990, 27,819-830. (a) Hofmann, W. Z. Kristallogn, 1933, 84, 177-203. (b) Kupcik, V. Diskussionstagung der Sektion fur Kristallkunde (D.M.G.) Marburg, 1965,16-17. 45. Matzat, E. Min. Petr. Mitt, 1972, 18,312-316. Lee, S.; Fischer, E; Czerniak, J.; Nagasundaram, N. J. Alloys and Compounds 1993, 197, 1-5. 47. Mumme, W. G.; Watts, J. A. Acta Cryst. 1980, 363,1300-1304. Ohmasa, M.; Nowacki, W. Z. Kristallogr., 1973, 137,422-432. 49. (a) Elwell, D.; Scheel, H. J., Crystal Growth from High temperature Solutions, Academic Press, London, 1975, Chapter 10. (b) Dugue', J.; Carre', D.; Guittard, M. Acta Cryst., 197 8, 34B, 403. 50. (a) Lazarev, V. B.; Trippel', A. F.; Berul', S. 1. Russ. J. Inorg. Chem., 1977, 22, 1218-1220. (b) Trippel', A. F.; Berul', S. I.; Lazarev, V. B. Russ. J. Inorg. Chem., 1980, 25, 1545-1547. (c) Berul', S. 1.; lazarev, V. B.; Trippel', A. F.; Buchikhina, O. P. 1977, 22, 1390-1393. ((1) lazarev, V. B.; Trippel', A. F.; Berul', S. 1. Russ. J. Inorg. Chem., 1980, 25,1694-1697. 51. Bazakutsa, V. A.; lazarev, V. B.; Gnidash, N. 1.; Salov, A. V.; Kulchitskaya, A. K. Russ. J. Inorg. Chem. 1988, 33,1204—1207. 52. (a) Grund, A.; Preisinger, A. Acta Cryst. 1950, 3, 363-366. (b) Mereiter, K.; Preisinger, A.; Guth, H. Acta Cryst. 1979, B35, 19-25. (c) Eisenmann, B.; Zagler, R. Z. Naturforsch., 1989, 44b, 249-256. ((1) Graf, H. A.; Schafer, H. Z. Anorg. Allg. Chem. 1976, 425, 67-80. 53. (a) Cotton, F. A.; Wilinson, G. Advanced Inorganic Chemistry (5th Edition), John Wiley and Sons: New York, 1988; pp 208-209. (b) Pitzer, K. 8. Acc. Chem. Res., 1979, 12, 272-276. (c) Pyykko, P.; Desclaux, J.-P. Acc. Chem. Res., 1979, 12,276-281. 54. (a) Bazakutsa, V. A.; Gnidash, N. I.; Lazarev, V. B.; Rogacheva, E. I.; Salov, A. V.; Sukhorukova, L. N.; Vasileva, M. F.; Berul, S. 1. Russ. J. Inorg. Chem. 1973, 18, 1722-1725. (b) Kovba, L. M.; Lazarev, V. B.; Moshchalkova, N. A.; Salov, A. V. Russ. J. Inorg. Chem. 1978, 23, 279-280. (c) Kovba, L M.; Lazarev, V. B.; Moshchalkova, N. A.; Salov, A. V. Russ. J. Inorg. Chem. 1976, 21, 857-858. (d) Kovba, L. M.; Lazarev, V. B.; Moshchalkova, N. A.; Salov, A. V. Russ. J. Inorg. Chem. 1978, 23, 426-430. (e) Berul, lazarev, V. B.; Salov, A. V. S. 1. Russ. J. Inorg. Chem. 1971, 16,1779—1781. 62 55. Bazakutsa, V. A.; lazarev, V. B.; Gnidash, N. 1.; Salov, A. V.; Podyachaya, E. N.; Zozuya; Moshchalkova, N. A. Russ. J. Inorg. Chem. 1979, 24, 1601- 1603. 56. (a) Olivier-Fourcade, E.; Phillippot, J.; M. Maurin Z. Anorg. Allg. Chem., 1978, 446, 159-168. (b) Graf, H. A.; Schafer, H. Z. Anorg. Allg. Chem., 1975, 414, 211-219. (c) Kanishcheva, A. S.; Kuznetsov, V. G.; Lazarev, V. B.; Tarasova, T. G. Zh. Strukt. Khim., 1977, 18, 1069. (d) Kanishcheva, A. S.; Mikhailov, Y. N.; Kuznetsov, V. G.; Batog, V. N. Dokl. Akad. Nauk SSSR , 1980, 251, 603-605. (e) Kanishcheva, A. S.; Palkina, K. K.; Kuznetsov, V G.; Lazarev, V. B.; Tarasova, T. G. Inorg. Mater. 1976, 12, 465-466. (f) Dittmar, V. G.; Schafer, H. Z. Naturforsch. 1977, 32b, 1346-1348. 57. Volk, K.; Bickert, P.; Kolmer, R.; Schafer, H. Z. Naturforsch. 1979, 34b, 380-382. 58. 59. Graf, H. A.; Schafer, H. Z. Anorg. Allg. Chem. 1975, 414,220—230. Volk, K.; Schafer, H. Z. Naturforsch. 1979, 34b,1637—1640. Eisenmann, B.; Schafer, H. Z. Naturforsch., 1979, 34b,383-385. 61. Graf, H. A.; Schafer, H. Z. Naturforsch. 1972, 27b, 735-739. 62. Dittmar, V. G.; Schvfer, H. Z. Anorg. Allg. Chem. 1978, 441, 93-97. (h) Sheldrick, W. S.; Hausler, H.-J. Z. Anorg. Allg. Chem. 1988, 557,105-111. 63. Dittmar, G.; Schafer, H. Z. Anorg. Allg. Chem. 1977, 437, 183-187. Dittmar, G.; Schafer, H. Z. Anorg. Allg. Chem. 1978, 441,98-102. 65. Parise, J. B. J. Chem. Soc., Chem. Commun., 1990, 22,1553-1554. Gostojic, M.; Nowacki, W.; Engle, P. Z. Kristallogr. 1982, 159,217-224. 67. Wacker, K.; Salk, M.; Decker-Schultheiss, G.; Keller, E. Z. Anorg. Allg. Chem. 1991, 606, 51-58. 68. (a) Cordier, G.; Schafer, H. Revue de Chemie Minerale, 1981, 18, 218—223. (b) Cordier , G.; Cook, R; Schafer, H. Revue de Chemie Minerale, 1980, 17, 1-6. (c) Dorrscheidt, W.; Schafer, H. Z. Naturforsch., 1981, 36b, 410- 414. 69. (a) Cordier, G.; Schwidetzky, C.; Schafer, H. Revue de Chemie Minerale, 1982, 19, 179-186. (b) Konig, T.; Eisenmann, B.; Schafer, H. Z. Anorg. Allg. Chem. 1982, 488,126-132. 70. (a) Cordier, G.; Schafer, H.; Schwidetzky, C. Z. Naturforsch., 1984, 39b, 131-134. (b) Cordier, G.; Schafer, H.; Z. Naturforsch., 1979, 34b, 1053- 1056. (c) Jung, J.-S.; Wu, 8.; Stevens, E. D.; O'Connor, C. J. J. Solid State Chem. 1991, 94,362-367. 71. Sheldrick, W. S.; Hausler, H.-J. Z. Anorg. Allg. Chem. 1988, 557, 98-104. 63 72. 73. 74. Sheldrick, W. S.; Hausler, H.-J. Z. Anorg. Allg. Chem. 1988, 561,149-156. Sheldrick, W. S.; Kaub, J. Z. Anorg. Allg. Chem. 1986, 536,114—118. (a) Parise, J. B.; Ko, Y. Chem. Mater. 1992, 4, 1446-1450. (b) Parise, J. B. Science , 1991, 251, 293-294. 75. (a) Weiss, A; Schafer, H. Z. Naturforsch. 1963, 18b, 81-82. (b) Weiss, A; Schafer, H. Naturwissen. 1960, 47, 495. (c) Buck, P.; Carpentier, C.-D. Acta Cryst., 1973, 829, 1864—1868. ((1) Diehl, R.; Carpentier, C.-D. Acta Cryst., 1978, B34, 1097-1105. (e) Post, F.; Kramer, V. Mat. Res. Bull, 1984, 19, 1607-1612. (f) Garin, J.; Parthe, E. Acta Cryst., 1972, 828,3672-3674. (g) Simon, A.; Peters, K.; Peters, E.-M.; Hahn, H. Z. Naturforsch. 1983, 38b, 426-427. 76. Le Rolland, B.; McMillan, F.; Moline, P.; Colombet, P. Eur. J. Solid State Inorg. Chem. 1990, 27, 715-724 and references therein. 77. (a) Hahn, H.; Klingen, W. Naturwiss. 1965, 52, 494. (b) Klingen, W.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem., 1973, 401, 97-112. (c) Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst., Sect. B, 1978, 34,1779- 1781. (d) Klingen, W.; Ott, R.; Hahn, H. Z. Anorg. Allg. Chem., 1973, 396, 271-278. (e) Jandali, M. Z.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem., 1978, 447, 105-118. 78. (a) Carpentier, C. D.; Nitsche, R. Mater. Res. Bull. 1974, 9, 401-410. (b) Carpentier, C. D.; Nitsche, R. Mater. Res. Bull. 1974, 9, 1097-1100. 79. (a) Becker, R.; Brockner, W.; Schafer, H. Z. Naturforsch. 1983, 38a, 874- 879. (b) Becker, R.; Brockner, W.; Schafer, H. Z. Naturforsch. 1984, 3%, 357-361. (c) Yun, H.; Ibers, J. A. Acta Cryst. 1987, 043, 2002-2004. (a) Toffoli, P.; Michelet, A.; Khodadad, P.; Rodier, N. Acta Cryst. 1982, B38, 706-710. (b) Toffoli, P.; Khodadad, F.; Rodier, N. Acta Cryst. 1983, C39, 1485-1488. (c) Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst. 1978, B34, 1779-1781. (d) Pfeiff, R.; Kniep, R. J. Alloys and Compounds, 1992, 186, l 1 1-133. 81. Hdnle, W. Z. Naturforsch. 1984, 39b,1088-1091. 82. (a) Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst. 1977, 333,1492-1494. (b) Toffoli, P.; Khodadad, F.; Rodier, N. Acta Cryst. 1977, 838,2374—2378. (c) Jandali, M. Z.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem., 1978, 445, 184—192. (d) Durand, E.; Evain, M.; Brec, R. J. Solid State Chem., 1993, 102, 146-155. 83. (a) Brec, R.; Ouvrard, G.; R. Freour, R.; Soubeyroux, J. L.; Rouxel, J. Mater. Res. Bull. 1983, 18, 689-696 . (b) Evain, M.; Brec, R.; Ouvrard, G.; Rouxel, J. J. Solid State Chem. 1985, 56, 12-20. (c) Brec, R.; Ouvard, G.; Evain, M.; Grenouilleau, P.; Rouxel, J. J. Solid State Chem., 1983, 7, 174-184. 64 (a) Grennouilleau, P.; Brec, R.; Evain, M.; Rouxel, J. Rev. Chim. Miner., 1983, 20, 628-635. (b) Brec, R.; Evain, M.; Grenouilleau, P.; Rouxel, J. Rev. Chim. Miner. 1983, 20, 283-294. (c) Brec, R.; Grenouilleau, P.; Evain, M.; Rouxel, J. Rev. Chim. Miner., 1983, 20, 295-304. ((1) Evain, M.; Brec, R.; Ouvard, G.; Rouxel, J. Mater. Res. Bull, 1984, 18, 41-48. 85. (a) Fiechter, S.; Kuhs, W. F.; Nitsche, R. Acta Cryst. , 1989, B36, 2217-2220. (b) Evain, M.; Queignec, M.; Brec, R.; Rouxel, J. J. Solid State Chem., 1985, 56, 148-157. (c) Evain, M.; Queignec, M.; Brec, R.; Sourisseau, C. J. Solid State Chem., 1988, 75, 413-431. ((1) Evain, M.; Lee, S.; Queignec, M.; Brec, R. J. Solid State Chem. 1987, 71,139-153. 86. (a) Mercier, R.; Malugani, J.-P.; Fahys, B.; Robert, G. Acta. Cryst. 1982, B38, 1887-1890. (b) Schafer, H.; Schafer, G.; Weiss, A. Z. Naturforsch. 1965, 20b, 811. 87. 88. 89. Jansen, M.; Henseler, U. J. Solid State Chem. 1992, 99,110-119. Bridenbaugh, P. M. Mat. Res. Bull, 1973, 8,1055-1060. Buck, P.; Nitsche, R. A. Z. Naturforsch., 1971, 26b, 731. Zirnmermann, H.; Carpentier, C.-D.; Nitsche, R. Acta Cryst. 1975, B31, 2003-2006. 91. (a) Toffoli, P.; Khodadad, P.; Rodier, N. Bull. Soc. Chim. Fr. 1981, 11,429- 432. (b) I. J. Fritz, T. J. Isaacs, M. Gottlieb and B. Morosin, Solid State Commun, 1978, 27, 535. 92. 93. Diehl, R.; Carpentier, C.-D. Acta Cryst. 1977, B33, 1399-1404. Marzik, J. V.; Hsieh, A. K.; Dwight, K.; Wold, A. J. Solid State Chem. 1983, 49, 43-50. Folrner, J. C.; Turner, J. A.; Parkinson, B. A. J. Solid State Chem. 1987, 68, 28-37. 95. Patzmann, U.; Brockner, W.; Cyvin, B. N.; Cyvin, S. J. J. Raman Spectroscopy, 1986, 17,257-261. (a) Toffoli, F.; Rouland, J. C.; Khodadad, P.; Rodier, N. Acta Cryst. 1985, 041, 645-647. (b) Becker, R.; Brockner, W.; Eisenmann, B. Z. Naturforsch., 1987, 423,1309-1312. 97. 98. Soklakov, A. 1.; Nechaeva, V. V. Inorg. Mater., 1970, 6, 873-874. Cyvin, S. J.; Cyvin, B. N.; Wibbehnann, C.; Becker, R.; Brockner, W.; Parensen, M. Z. Naturforsch., 1985, 40a, 709-713. Brec, R.; Schleich, D. M.; Ouvrard, G.; Louisy, A.; Rouxel, J. Inorg. Chem., 1979, 18,1814—1818. 100. (a) Odile, J.-P.; Steger, J. J.; Wold, A. Inorg. Chem. 1975, 14,2400-2402. (b) Taylor, B. E.; Steger, J. J.; Wold, A.; Kostiner, E. Inorg. Chem. 1974, 13, 2719-2421. (c) Taylor, B. E.; Steger. J. J.; Wold, A. J. Solid State Chem. 1973, 7,461-467. 65 101. Kurosawa, K.; Saito, S.; Yamaguchi, Y. J. Phys. Soc. Japan, 1983, 11,3919- 3926. 102. (a) Le Mehoute, A.; Ouvrard, G.; Brec, R.; Rouxel, J. Mater. Res. BuH., 1977, 12, 1191. (b) Thompson, A. H.; Whittingham, M. S. Mater. Res. Bull., 1977, 12,741. 103. Thompson, A. H.; Whittingham, M. S. U. S. Patent 4,049,879 1977. (b) Brec, R.; Le Mehaute', A. Fr. Patents 7,704,519 1977. 104. Clement, R. J. Chem. Soc., Chem. Commun., 1980, 647-648. 105. (a) Joy, P.; Vasudevan, S. J. Am. Chem. Soc. 1992, 114, 7792-7801. (b) Lagadic, 1.; Leaustic, A.; Clement, R. J. Chem. Soc., Chem. Commun., 1992, 1396-1397. 106. Iacroix, P. G.; Clement, R.; Nakatani, K.; Zyss, J.; Ledoux, 1. Science, 1994, 263, 65 8-660. 107. (a) Ouvrard, G.; Brec, R.; Rouxel, J. Mater. Res. Bull, 1985, 20,1181-1189. (b) Prouzet, F.; Ouvrard, G.; Brec, R. Mater. Res. BuH., 1986, 21 , 195-200. 108. Clement, R.; Audiere, J.-P.; Renard, J.-P. Rev. Chim. Miner. 1982, 19,560- 571. 109. Ouvrard, G.; Brec, R.; Rouxel, J. Compes Rendus Acac. Sci. Paris, 1982, 294,971. 110. Katty, A.; Soled, S.; Wold, A. Mater. Res. Bull., 1977, 12, 663-666. 111. Etrnan, M.; Katty, A.; Levy-Clement, C.; Lemasson, P. Mater. Res. Bull., 1982, 17,579-584. 112. (a) Ouvrard, G.; Brec, R.; Rouxel, J. Mater. Res. BuH., 1985, 20,1181-1189. (b) Lee, S.; Colombet, P.; Ouvrard, G.; Brec, R. Inorg. Chem., 1988, 27, 1291-1294. (c) Lee, S.; Colombet, P.; Ouvrard, G.; Brec, R. Mater. Res. Bull., 1986, 21, 917-928. ((1) Lee, S. J. Am. Chem. Soc. 1988, 110, 8000-8006 and references therein. (e) Ouili, Z.; LeBlanc, A.; Colombet, P. J. Solid State Chem., 1987, 66, 86-94. 113. (a) Lee, S.; Colombet, Ouvrard, G.; Brec, R. Inorg. Chem. 1988, 27,1291- 1294. (b) Mathey, Y.; Michalowicz, A.; Toffoli, P.; Vlaic, G. Inorg. Chem. 1984, 23, 897-902. 114. Colombet, P.; Leblanc, A; Danot, M.; Rouxel, J. J Solid State Chem. 1982, 41, 174—184. 115. Durand, E; Ouvrard, G.; Evain, M.; Brec, R. Inorg. Chem., 1990, 29,4916- 4920. 116. Leblanc, A.; Ouili, Z.; Colombet, P. Mat. Res. Bull., 1985, 20,947-954. 66 117. Mathey, Y.; Mercier, H.; Michalowicz, A.; Leblanc, A. J. Phys. Chem. Solids, 1985, 46,1025-1029. 118. Scott, B.; Pressprich, M.; Willet, R. D.; Clearly, D. A. J. Solid State Chem. 1992, %,294—300. 119. Arnautova, E; Sviridov, E.; Rogach, E. Savchenko, E.; Grekov; A. In tegra ted Ferroelec trics, 1 992, 1, 147-150. 120. (a) Bouchetiere, M.; Toffoli, F.; Khodadad, P.; Rodier, N. Acta Cryst. 1978, B34, 384—387. (b) Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst. 1978, 834,3561-3564. 121. Jandali, M. Z.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem., 1980, 470, 39—44. 122. (a) Queignec, M.; Evain, M.; Brec, R.; Sourisseau, C. J. Solid State Chem. 1986, 63, 89-109 (b) Andrae, H.; Blachnik, R. J. Alloys and Compounds 1992, 189,209-215. 123. Menzel, F.; Ohse, L; Brockner, W. Heteroatom Chem. 1990, 1,357—362. 124. Vos, A. ; Wiebenga, E. H. Acta Cryst. 1955, 8,217-223. 125. Evain, M.; Brec, R.; Whangbo, M.-H. J. Solid State Chem., 1987, 71,244— 262. 126. (a) Scheel, H. J. J. Cryst. Growth 1974, 24/25, 669-673. (b) Sanjines, R.; Berger, H.; Levy, F. Mater. Res. Bull. 1988, 23, 549-553. (c) Garner, R. W.; White, W. B. J. Cryst. Growth 1970, 7,343-347. 127. Sunshine, S. A.; Kang, D.; Ibers, J. A. J. Am. Chem. Soc. , 1987, 109,6202— 6204. 128. Kang, D.; Ibers, J. A. Inorg. Chem., 1989, 27,549-551. 129. (a) Gmelin's Handbuch der Anorganischen Chemie; Verlag Chemie: Weiheim/Brgstr.,FRG, 1966; Sodium, Suppl. Part 3, pp. 1202-1205 and references therein. (b) Pearson, T. G.; Robinson, P. L. J. Chem. Soc. 1931, 53, 1304—1314. (c) The Sodium-Sulfur Battery; Sudworth, J. 1.; Tilly, A. R., Eds.; Chapman & Hall: London; New York, 1985. (d) Fischer, W. Mat. Res. Soc. Symp. Proc. 1989, 135, 541-551. (e) Powers, R. W.; Karas, B. R. J. Electrochem. Soc. 1989, 136, 2787-2793. (f) Mathewson, G. H. J. Am. Chem. Soc. 1907, 29, 867-880. (g) Klemm, W.; Sodomann, H.; Langmesser, P. Z. Z Anorg. Allg. Chem., 1939, 241,281-304. 130. (a) Kanatzidis, M. G. Chem. Mater., 1990, 2, 353-363. (b) Kanatzidis, M. G.; Park, Y. J. Am. Chem. Soc., 1989, 111, 3767-3769. (c) Kanatzidis, M. G.; Park, Y. Chem. Mater., 1990, 2, 99-101. ((1) Park, Y.; Kanatzidis, M. G. Angew. Chem. Int. Ed. Engl., 1990, 29, 914—915. (e) Park, Y. Ph.D. dissertation, 1992, Michigan State University. 131. (a) Iiao, J.-H.; Varotsis, C.; Kanatzidis, M. G. Inorg. Chem., 1993, 32, 2453- 2462. (b) Iiao, J.-H.; Kanatzidis, M. G., Chem. Mater., 1993, 5,1561-1569. CHAPTER 2 (Part 1) Molten Salt Synthesis and Properties of Three New Solid State Ternary Bismuth Chalcogenides, B-CsBiSZ, y—CsBiSZ and K231138813 1 . Introduction To date, there are only a few structurally characterized ternary alkali metal bismuth chalcogenides, outside of the well-known NaCl- type ABin (A=li, Na, K; Q=S, Se, Te) compounds.1i2 To the best of our knowledge, the only other structurally characterized phases are CsBi3SS,3 RbBi3Ss,4 CS3Bi7Se 12,5 and Sr4Bi6Se13.6 This relative scarcity of ternary compounds provides a further impetus for exploratory research in this area. In particular, Bi is very attractive for study because of its inert 6s2 lone pair of electrons which may or may not be manifested structurally in a given compound. Whether the lone pair is stereochemically active or not affects both the lattice structure and the properties of the resulting compounds and thus exploration of the solid state chemistry of Bi is warranted. This issue is related to the larger question of stereochemical activity of a lone pair in compounds with elements in a 52 configuration. 67 68 Group 15 chalcogenide compounds have received considerable attention, due to their potential application as non-linear optical materials,7 photoelectrics,8 and thermoelectrics.9 The most common application that is unique to Group 15 chalcogenides is in the area of thermoelectric cooling materials. To date, the most investigated systems are various solid solutions of Mng (M=As, Sb, Bi; Q=S, Se, Te) compounds. 10 These materials possess high electrical conductivity and thermoelectric power and low thermal conductivity and are excellent materials for thermoelectric applications near room temperature.11 With these properties in mind, we pursued the synthesis of new ternary alkali metal bismuth chalcogenide compounds using the now proven polychalcogenide flux method.12,13 The latter involves the reaction of Bi in molten salts of AzQx (A=alkali metal) of varying chemical composition. We employed thermal analysis techniques to study the metal/ alkali polychalcogenide flux reaction. Namely, differential scanning Calorimetry (DSC) was used to monitor the various thermal events that occur during reaction and determine formation and crystallization of the products. This paper describes the synthesis, structural characterization, optical and infrared spectroscopic, and charge transport properties of three new compounds, B-CsBiSz, y—CsBiSZ and K2Bi38e13. The thermochemical characterization of the Bi/Cszsx and Bi/KZSex systems is also reported. The former is a new polymorph of CsBiSz while the latter represents a new structure type. 69 2. Experimental Section 2.1. Reagents Chemicals in this work were used as obtained: (i) bismuth powder, 99.999+% purity, -100 mesh; selenium powder, 99.5+% purity, -100 mesh, Aldrich Chemical Co., Inc., Milwaukee, WI. (ii) sulfur powder, sublimed, J. T. Baker Chemical Co., Phillipsburg, NJ. (iii) potassium metal, analytical reagent, Mallinckrodt Inc., Paris, KY; cesium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH (iv) methanol, anhydrous, Mallinckrodt Inc., Paris, KY; DMF, analytical reagent, diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. 2.2. Synthesis Synthesis: All manipulations were carried out under a dry nitrogen atmosphere in a Vacuum Atmospheres Dri-lab glovebox. For the preparation of C528 and K28e we used a modified literature procedure. 14 Cesium Sulfide, CszS. 4.80 g (0.036 mol) of slightly heated (‘30 °C) cesium metal was pipetted into a 250 ml round-bottom flask. A 150-ml volume of liquid ammonia was condensed into the flask at -78 °C (dry ice/acetone bath) under nitrogen to give a dark blue solution. 0.579 g (0.018 mol) of sulfur and a Teflon coated stirbar were added and the mixture was stirred for one hour to give 70 a light blue solution. The NH 3 was removed by evaporation under a flow of nitrogen as the bath slowly warmed to room temperature. The pale yellow solid (97% yield) was dried in vacuum overnight, flame-dried, and ground to a fine powder with a mortar and pestle in the glovebox. Caution: Cesium metal will react with Teflon and sulfur vigorously upon contact in the solid state. Potassium Selenide, KZSe. A 10.098 g (0.128 mol) sample of selenium was combined with 10.00 g (0.256 mol) of freshly sliced potassium metal and the reaction was carried out as above. The orange product (99% yield) was dried in vacuum overnight, flame- dried, and ground to a fine powder with a mortar and pestle in the glovebox. B-Cesium Disulfido-bismuthate(III), B-CsBiSz (I). An amount of 0.052 g (0.25 mmol) Bi, 0.335 g (1.125 mmol) CszS and 0.064 g (2.00 mmol) of S were mixed together with a spatula in a glass vial. The mixture was transferred to a 6 ml pyrex tube and was subsequently flame-sealed in vacuum (~103 torr). The reaction was heated to 290 °C over a 12 hour period in a computer controlled furnace, then isothermed at 290 °C for 96 hours, followed by cooling to 100 °C at a rate of 2 °C/hr, then to 50 CC in one hour. This gave a red-orange glassy flux. The tube was opened with a glass cutter and the excess Cszsx was dissolved in 150 ml of degassed methanol to give a clear, yellow solution. The insoluble product was washed with methanol repeatedly until the washings were colorless and then with ether. 0.085 g (83% yield) of transparent, dark, red needle-like crystals were obtained along with a small amount of pale yellow 71 impurity material. The red crystals and the amorphous yellow powder were insoluble in ethylenediamine, CH 3CN , DMF, ethanol, and (82. The mixture was quickly washed with 5 ml of water on a filter frit which removed the yellow impurity, followed by 20 ml of methanol. This final step was done quickly as the red crystals are slightly moisture sensitive. Prolonged exposure to water (several hours) resulted in the formation of a dark purple fihn on the surface of the crystals. Quantitative microprobe analysis on the red single crystals gave CsLoBimSlg (average of four data acquisitions). It is important that the C528 be made fresh using the method described here. Use of commercially available C528 (e.g. Cerac Inc.) more often than not did not yield B-CsBiSz but resulted in Bi283. The difference in the two starting materials was examined using SEM. The particle size of the C528 (Cerac) was considerably larger than that of C528 (section 2.2), which could account for the difference in reactivities. Reactions using increased amounts of starting materials yielded larger crystals. A mixture containing 0.105 g (0.500 mmol) of Bi, 0.670 g (2.25 mmol) of C528, and 0.128 g (4 mmol) of 8 under the same conditions afforded 3 mm sized crystals suitable for single crystal charge transport studies. y-Cesium Disulfido-bismuthate(III), y—CsBiS2. An amount of 0.042 g (0.20 mmol) Bi, 0.238 g (0.80 mmol) C528 and 0.051 g (1.60 mol) 8 were thoroughly mixed and the reaction was carried out as above. The mixture was heated to 350 °C over a 12 hour period, then isothermed for 6 days, followed by cooling to 80 °C at 2 °C/hr, then to 50 °C in one hour. The excess C528x was removed by 72 dissolving in degassed methanol. The product was then washed with ethyl ether (20 ml). 0.060 g (74% yield) of black hexagonal-shaped crystals were obtained. Quantitative microprobe analysis gave C51,()Bi1,082,2 (average of three acquisitions). Powder XRD peaks (d- spacings, A) 16.780 (vs), 12.468 (5), 8.268 (w), 6.267 (vw), 5.484 (w), 4.104 (vs), 3.592 (m), 3.410 (m), 3.314 (m), 3.231 (m), 3.025 (m), 2.927 (5), 2.868 (m), 2.730(vw), 2.302 (vw), 2.266 (vw), 2.181 (vw), 2.150 (vw), 2.071 (m), 2.039 (m), 1.970 (vw), 1.909 (vw), 1.862 (vw), 1.813 (w), 1.761 (vw), 1.716 (vw), 1.634 (vw), 1.632 (vw). D i p o t a s s i u m Tridecaselenido-octabismuthate(III) , K2Bi38e13 (II). An amount of 0.209 g (1.00 mmol) Bi, 0.079 g (0.500 mmol) K28e and 0.316 g (4.00 mmol) 8e were thoroughly mixed and the reaction was carried out as above. The mixture was heated to 330 °C over a 12 hour period, then isothermed for 10 days, followed by cooling to 70 °C at 2 °C/ hr, then to 50 °C in one hour. The excess K28ex was removed by dissolving in degassed DMF. The product was washed with ether (20 ml) and ethylenediamine (20 ml) to remove any trace amounts of Se. 0.306 g (88% yield) of thin black needles with black microcrystalline powder were obtained. Quantitative microprobe analysis gave K2,0Bi3,08e12,3 (average of four acquisitions). The homogeneity of B-CsBiS2 and K2Bi38e13 was confirmed by comparing the observed and calculated X-ray powder diffraction patterns. The dhkl spacings observed for the bulk materials were compared, and found to be in good agreement with the db k1 spacings calculated from the single crystal data using the program POWD 10.15 The results are summarized in Tables 2-1 and 2-2. 73 2.3. Physical Measurements. FT-IR spectra of the B-CsBiSz compound were recorded as a solid in a C51 matrix. The sample was ground with dry CsI into a fine powder, and a pressure of about seven metric tons was applied to the mixture to make a translucent pellet. The spectra were recorded in the far-IR region (600-100 cm‘l, 4 cm'1 resolution) with the use of a Nicolet 740 FT-IR spectrometer equipped with a TGS/PE detector and silicon beam splitter. Quantitative microprobe analysis of the compounds was performed with a JEOL JSM-3 SCF scanning electron microscope (SEM) equipped with a Tracor Northern Energy Dispersive Spectroscopy (EDS) detector. Single crystals of each sample were mounted on an aluminum stub wich was coated with conducting graphite paint to avoid charge accumulation on the sample surface under bombardment of the electron beam during measurements. A 1.8 correction factor was applied to the Se atomic percent in this system as a result of examination of several other Se containing crystals. Energy Dispersive Spectra (EDS) were obtained by using the following experimental set-up: X-ray detector position: 55 mm Working distance: 39 mm 74 Accelerating voltage: 20 kV Take-off angle: 27 deg Beam current: 200 picoamps Accumulation time: 60 seconds Window: Be Optical diffuse reflectance measurements were made at room temperature with a Shimadzu UV-3101PC double beam, double- monochromator spectrophotometer. The instrument was equipped with an integrating sphere and controlled by a personal computer. The measurement of diffuse reflectivity can be used to obtain values for the band gap which agree rather well with values obtained by absorption measurements from single crystals of the same material. The digitized spectra were processed using the Kaleidagraph'm software program. BaSO4 powder was used as reference (100% reflectance). Absorption data were calculated from the reflectance data using the Kubelka-Munk function:16 75: a (l-R)2 2R R is the reflectance at a given wavelength, or is the absorption coefficient and S is the scattering coefficient. The scattering coefficient has been shown to be practically wavelength independent for particles larger than 5 nm which is smaller than the particle size of the samples used here.16a,b 75 Differential Scanning Calorimetry was performed with a computer-controlled Shimadzu DSC-50 thermal analyzer under a nitrogen atmosphere at a flow rate of 35 ml/ min. The appropriate reactant mixtures (~18.0 mg total mass), as mentioned above, were crimped in an aluminum pan inside a nitrogen-filled glove box. The pan was placed on the sample side of the DSC-50 detector and an empty aluminum pan of equal mass was crimped and placed on the reference side. The samples were heated to the desired temperature at 3 °C/min, then isothermed for 800 min. followed by cooling at -0.5 ° C/min. to 75 °C. The reported DSC temperatures are peak temperatures with a standard deviation of 0.2 degrees. The adopted convention in displaying data is: exothermic peaks occur at positive heat flow while endothermic peaks occur at negative heat flow. Dc electrical conductivity and thermopower measurements were made on single crystals and polycrystalline compactions of the compounds. Conductivity measurements were performed in the usual four-probe geometry with 60- and 25mm gold wires used for the current voltage electrodes, respectively. Measurements of the pellet cross-sectional area and voltage probe separation were made with a calibrated binocular microscope. Conductivity data were obtained with the computer-automated system described elsewhere.17 Thermoelectric power measurements were made by using a slow ac technique18 with 60-mm gold wires serving to support and conduct heat to the sample, as well as to measure the voltage across the sample resulting from the applied temperature gradient. In both measurements, the gold electrodes were held in place on the sample with a conductive gold paste. 76 Conductivity specimens were mounted on interchangeable sample holders, and thermopower specimens were mounted on a fixed sample holder/ differential heater. Mounted samples were placed under vacuum (10-3 Torr) and heated to 320 K for 2-4 h to cure the gold contacts. For a variable-temperature run, data (conductivity or thermopower) were acquired during both sample cooling and warming to check reversibility. The temperature drift rate during an experiment was kept below 1 K/min. Typically, three to four separate variable-temperature runs were carried out for each sample to ensure reproducibility and stability. At a given temperature, reproducibility was within i596. Thermoelectric power results collected by the slow-ac technique require the production of a slowly varying periodic temperature gradient across the samples and measuring the resulting sample voltage. Samples were suspended between the quartz block heaters by 60-mm gold wires thermally grounded to the blocks with GE 7031 varnish. The magnitude of the applied temperature gradient was generally 1.0 K. Smaller temperature gradients gave essentially the same results but with somewhat lower sensitivity. 2.4. X-ray Crystallography The compounds were examined by X-ray powder diffraction for the purpose of phase purity and identification. Accurate dhkl spacings (A) were obtained from the powder patterns recorded on a 77 calibrated (with FeOCl as internal standard) Phillips XRG—3000 computer-controlled powder diffractometer with Ni filtered Cu Ka radiation operating at 35 kV and 35 mA. The data were collected at a rate of 0.12°/min. Structure solution of B-CsBiSz. A crystal with dimensions 0.50 x 0.20 x 0.10 mm was sealed in a glass capillary. Intensity data were collected using the (1)-20 scan mode on a Rigaku AFC6S four- circle automated diffractometer equipped with a graphite-crystal monochromator. The stability of the crystal was monitored by measuring three standard reflections periodically (every 150 reflections) during the course of data collection. No crystal decay was detected. An empirical absorption correction based on 1): scans was applied to the data, followed by a DIFABS correction to the isotropically refined data. 19 The structure was solved by direct methods using SHELXS86206 and refined by full-matrix least-squares techniques of the TEXSAN package of crystallographic programs.20b All calculations were performed on a VAXstation 3 100/ 76 computer. Structure Solution of K2Bi38 e13. A crystal with dimensions 0.10 x 0.20 x 0.50mm was mounted on a glass fiber. Intensity data for the crystal were collected using the (1)/20 scan mode on a Rigaku AFC6S four-circle automated diffractometer. The stability of the crystal was monitored by measuring three standard reflections periodically (every 150 reflections) during the course of data collection. No crystal decay was detected. Absorption corrections were applied as above. The structure was solved by direct methods using SHELXS8620a and refined by full-matrix least-squares 78 techniques of the TEXSAN package of crystallographic programs.20b A careful inspection of the final structure revealed the presence of pseudo mirror symmetry perpendicular to the c-axis. Since the or and B angles of the unit cell are very close to 90° we attempted to refine the model at the higher symmetry monoclinic space group C 2/m (#12) where a'=2a, b'=c, c'=-b, B=87.96°. The results where characterized by divergent refinement and an unstable structure. Thus, we chose the lower symmetry triclinic space group. All calculations were performed on a VAXstation 3100 Model 76 computer. The complete data collection parameters, details of the structure solution and refinement for both compounds are given in Table 2-3. The coordinates and average temperature factors (Beq) of all atoms and their estimated standard deviations for both compounds are given in Tables 2-4 and 2-5. Table 2-1. Calculated and Observed X-ray Powder Diffraction Patterns for B-CsBiSz. h k 1 deand. A debs. A I/ Imax, (obs) 100 1 1 o 0 1 1 -1 1 1 1 1 1 1 2 0 0 2 1 7.62 5.97 5.74 5.01 4.25 4.06 3.99 7.7 6.01 5.77 5.03 4.25 4.07 3.99 -1 0 2, 2 1 0 3542,3541 3.549 -2 1 1 1 2 1,01 2 -1 1 2 1 0 2 2 1 1, 1 3 0 0 2 2, 1 1 2 -13 1, -2 1 2 2 2 1 1 2 2 3.45 238,236 3.32 3.01 295,294 2.872,2.870 2.812,2.808 2.607 2.549 3 1 0, 2 3 0 2455,2452 2 0 2 -3 0 2 -14 1 1 3 2 -1 2 3 1 41 2 2 2 1 1 3 2.372 2.315 2.224 2.192 2.168 2.143 2.127 2.101 -3 2 2, 3 2 1 2.085,2.037 -3 3 1 40 0,2 41 4 1 0, 1 5 0 -1 0 4 -4 2 1 1 3 3 -3 4 1 0 5 2,-1 5 2 1 1 4,-2 4 3 -314,2 3 3 -1 6 0,0 3 4 2.005 1905,1900 1.868,1.864 1.832 1.803 1.787 1.755 1694,1689 l.643,1.641 1612,1605 1.578,1.567 3.46 3.36 3.30 3.01 2.96 2.869 2.811 2.605 2.542 2.455 2.367 2.331 2.222 2.189 2.167 2.147 2.127 2.100 2.038 1.993 1.908 1.874 1.830 1.806 1.785 1.757 1.692 1.641 1.606 1.570 70 20 8 4 20 18 36 86 46 21 61 22 51 61 34 9 26 29 13 14 4 10 18 17 19 23 17 7 100 28 20 17 9 8 15 13 6 5 Table 2-2. Calculated and Observed X-ray Powder Diffraction 80 Patterns for K28138e13. h K 1 dcglcd, A dobsd, A I/Imgx (ObSd) l 0 0 0 1 0 1 l 0 -1 l 0 -2 1 0 1 2 0 3 0 O 1 3 0 13.58 12.08 9.20 8.87 5.83 5.60 4.53 3.90 12.4 9.33 5.72 4.56 3.99 1-1 l,-2 l 1 3.595,3.593 3.616 -3 2 0 -2 3 0 4 0 0, -l 2 1 2 01, 410 3.56 3.41 3.40,3.39 3.31,3.30 2 l 1, -3-1 1 3.21,3.20 O 4 0 -3 3 0 2 2 1 -3-2 1 -4 2 0 2 4 0 3-1 1 -5 1 0 2-3 1 -3 4 0 3.02 2.96 2.93 2.92 2.89 2.80 2.75 2.63 2.53 2.472 -l 4 1, 0-4 1 2.429,2.428 4-1 1 2 5 0 -2 5 0 2-4 1 5 0 l, -1 0 2 5 4 0, 5-2 1 2 6 0 -7 l 0 0 3 2 -2-3 2 -1 6 1 -4 2 2 -5-1 2 -4 6 0 2-3 2 8 1 0 8 2 O -7 4 0 0 7 1 2-4 2 1 5 2 2.383 2.303 2.251 2.2% 2.125,2.123 2.081 2.057,2.042 1.985 1.950 1.905 1.837 1.835 1.805 1.797 1.760 1.705 1.702 1.690 1.650 1.607 1.595 1.592 1.545 3.58 3.45 3.42 3.32 3.23 3.08 2.98 2.953 2.952 2.85 2.80 2.76 2.63 2.55 2.490 2.451 2.380 2.312 2.278 2.218 2.127 2.083 2.053 1.986 1.948 1.907 1.847 1.834 1.798 1.782 1.763 1.728 1.707 1.693 1.658 1.629 1.607 1.596 1.550 11 8 16 13 11 36 33 8 9 10 45 5 a) 95 67 5 10 10 4 6 11 35 13 16 8 8 42 100 20 16 18 11 11 14 15 13 8 11 33 28 6 9 8 6 13 81 Table 2-3. Summary of Crystallographic Data and Structure Analysis for B-CsBiSz and KzBi38e 13 Formula B-CsBiSz K2Bi85e13 FW a, A b, A c, A 6:, deg 6, deg 7, deg 406.01 2776.52 7.794(5) 13.768(2) 9.610(6) 12.096(3) 7.329(4) 4.1656(6) 90.000 89.98( 1) 102.16(5) 98.64( 1) 90.000 87.96( 1) Z; V, A3 4; 537 1; 685 2. 0.71073 (Mo Ka) 0.71073 (Mo Ka) space group P21/C (NO- 14) P-I (No. 2) Dcalcn g/cm3 5.03 6.73 n, cm- 1 400 (Mo Ka) 684.2 (Mo Ka) ZBmax, deg 50 (Mo Ka) 50 (Mo Ka) Temp., 0C -3 6 23 Final R/Rw, % 7.4/ 10 8.2/ 8.9 Total Data Measured 165 2 Total Unique Data 15 5 1 Data with F62>30(F67-) 1053 No. of Variables 38 2443 2406 1 171 107 82 Table 2-4. Fractional Atomic Coordinates and Beq Values for B- CsBiS2 with Estimated Standard Deviations in Parentheses. atom X y 2 Beq,a A2 CS Bi 8(1) 8(2) -0.1383(4) -0.43 14(3) -0.7879(4) 2.16(5) 0.4115(2) -0.1657(2) -0.1440(2) 1.48(2) 0.186(1) -0.256(1) -0.461(1) 1.8(2) 0.319(2) 0.084(1) -0.199(2) 3.4(3) Table 2-5. Fractional Atomic Coordinates and Beq Values for K2Bi38e 13 with Estimated Standard Deviations in Parentheses. atom X y z Begha A2 K 0.831(1) 0.051(2) 0.415(4) 2.6(8) Bi(l) 0.8809(2) 0.3803(3) 0.4408(6) 1.6(1) Bi(2) 0.3787(2) 0.5464(3) 1.1887(7) 1.6(1) Bi(3) 1.1510(2) 0.2940(3) 0.0758(6) 1.6(1) Bi(4) 0.5628(2) 0.8671(3) 0.2822(7) 1.8(1) Se(l) 0.2625(5) 0.4267(7) 0.632(2) 1.6(3) Se(2) 1.0370(5) 0.1977(7) 0.516(2) 1.7(3) Se(3) 0.7788(5) 0.2707(7) 0.887(2) 1.7(3) Se(4) 0.5094(5) 0.3536(7) 1.254(2) 1.6(3) 8e(5) 1.2811(5) 0.1169(7) 0.140(2) 1.5(3) Se(6) 0.5837(5) 1.0886(7) 0.293(2) 1.4(3) Se(7) 1.0000 0.5000 0.0000 1.2(4) ameq = (4/3)[a2B(1,1) + 628(2,2) + c28(3,3) + ab(cosY)B(l,2) + ac(cosB)B(1,3) + bc(cos500 °C), but no structural characterization is available.21 To the best of our knowledge only the unit cell parameters of a-CsBi82 have been reported but no structural details have been found.22 Another new ternary phase, y-CsBiS2 was obtained by dissolving Bi in a C528,, (x=3.0-3.7) flux at 350 °C. Stacks of black hexagonal plates were obtained from these reactions. SEM photographs showed a very thin plate-like morphology for these 84 crystals which made data collection difficult. A rhombohedral unit cell with a=4.166(4)A and c=48.77(5)A, was found but the structure could not be refined to give a reasonable model. The unit cell of this compound is closely related to that of RbBiS223 in which CdCl2-type (Bi82)- layers (perpendicular to the c-axis) alternate with Rb+ ions. The length of the c-axis suggests six (BiS2)' layers in the unit cell. The coordination sphere of Bi is perfect octahedral. Similar investigations of Bi reactions in Li28x and Na28x (x=2.2- 5) fluxes in the temperature range of 290 °C to 345 °C gave liBiSz (68% yield) and NaBi82 (72% yield), which possess NaCl-type structures.1‘il Far-IR spectroscopy of B-CsBi82 shows two absorptions at 286 cm’ 1 (s) and 226 cm' 1 (m) which are similar to those of Bi2S3 at 290 crn' 1 (w) and 220 cm'1 (s).24 The two absorbancies are due to Bi-S stretching vibrations, but at this stage we cannot assign which modes since systematic IR spectroscopic data in this area are lacking. A solid state far-IR of K2Bi38e13 shows no peaks, the only feature being a broad absorbance beginning at 200 cm'1. The presence of this sort of feature is common in electrically conducting materials. The synthesis of K2Bi88e13 can be accomplished by dissolving Bi in K2Se9 flux over a wide range of molar ratios at 330-370 °C followed by isolation in DMF. Similar investigations using Na2Se9 flux at 320 °C gave only NaBiSe21C along with a small amount of elemental Se, as confirmed by powder X-ray diffraction. 85 B-CsBi82 and K2Bi38e13 belong to the (A2Q)n(Bi2Qg)In (A=alkali metal; Q=S, Se) general family of compounds with n=1 and m=1, 4 respectively. Other members of this family include CsBi385 (n=1, m=3),3 Cs3Bi7Se12 (n=3, m=7),5 and K2Bi4S7 (n=1, m=2).16 The latter phase has an unknown structure. The synthesis of new ternary bismuth chalcogenides with various n and m values may be possible e.g. ABng (n=1, m=5). Recently, another ternary bismuth sulfide KBi6,33Slo has also been synthesized which could belong to this family of compounds.25 3.2 Description of Structures Structure of p-CsBiS2 (I). This compound possesses the C58sz structure type.26 The Bi lone pair is stereochemically active in this compound. The anion has a one-dimensional, polymeric structure with charge balancing Cs+ cations found between the chains. Selected bond distances and angles for (I) are given in Table 2-6. The structure can be described as Bi83 trigonal pyramids connected at two vertices to form fully extended (BiS2)nn- chains along the a-axis. The two chains per unit cell are related by a center of symmetry. Figure 2-1 shows a view of an individual (Bi82)nn- chain. Along the chain backbone we can observe an alternation of long and short Bi-S(1) bonds. Figure 2-2 shows two views of the unit cell along the a and b axes, respectively. The chains of B-CsBiS2 are oriented in a sheet-like manner and the Bi lone pairs are directed into a nonpolar domain that is segregated from the polar domain of the C5 atoms. The [BiS2J- framework can be thought of as a 86 derivative of the well-known Bi2S3 compound and generated by successively dismantling the B1283 framework by incorporation of C528. The 8/ Bi ratio increases from 1.50 to 2.00 upon going from Bi283 to [Bi82]- as the structure is broken up into chains. Since the shortest distance between chains is 3.00(1) A, the structure can be alternatively viewed as layers of (BiS2)nn' chains separated with layers of Cs+ ions. The terminal sulfur atoms in the chain are pointed toward the Cs+ ions. The Bi-S bond distances in B- CsBiSz are comparable to those found in many bismuth sulfosalts, such as PbCuBiS3.-7-7 Many of these sulfosalts have a fourth and fifth sulfur atom at greater distances from the Bi to form a square pyramidal coordination. In KBiSz, the lone inert pair is not stereochemically active so the bismuth atom is found in a perfect octahedral environment with much longer Bi-S bonds at 3.02( 1) A.13 The same situation is thought to be present in y-CsBiS 2. These longer bonds are due to the electrostatic repulsion between the lone pair around the Bi nucleus and the negatively charged 8 atoms. The influence of the lone pair on Bi coordination is demonstrated through the inspection of the S-Bi-S angles, shown in Table 2-7. This effect is consistent with predictions based on V.S.E.P.R. theory.28 The angles of B-CsBi82, in Table 2-6, are compared to the S-M-S angles found in the Group 15 chain-like structures, CsSbS226 and NaA58229 (Table 2-7). The angles decrease going from As to Bi, as the lone pair orbital becomes larger and more diffuse. The long and short bond alternation along the chain (2.54( 1) 87 and 2.733(8) A) and the short terminal bond of 2.531(1) A are features that also appear in CsSbS2 and NaAsS2. Structure of K2Bi38 e13 (II) . This compound has a complicated three-dimensional structure made up of BiSe6 octahedra and BiSes square pyramids, which form tunnels filled with K+ cations. Selected bond distances and angles for (I I) are given in Table 2-8. The BiSe6 octahedra form NaCl-type layers that are linked by BiSe5 distorted square pyramids. Figure 2-3 shows the packing diagram of the extended structure down the c-axis. The [Bi38e13]2' framework can also be thought of as a hybrid of three different layered structure types interconnected to form a 3-D network. Structural features from the Bi2Te3,3O Sb283,31 and Cdl232 lattices are represented in this framework. Figure 2—4 shows the structures of the three layered materials (a-c). The features of the three structure types found in the [BigSe13]2' framework are highlighted in Figure 2-4d. The Bi2Te3-type fragments are linked by CdI2—type octahedra to form sheets that are connected in the b-direction by BiSes square pyramids (Sb283-type) to form the framework. Alternatively, the [Bi38e13]2' framework can also be viewed as a derivative of the well-known Bi28e3 compound generated by breaking down the Bi28e 3 framework by incorporation of K2Se in the molar ratio of 4:1 (i.e. [K2Se][Bi2Se3]4). The Se/ Bi ratio increases from 1.50 in Bi28e3 to 1.63 for the [Bi38e13]2- anion. K2Bi38e 13 is structurally related to C53Bi78e125 and Sr4Bi6Se13.6 In C53Bi7Se12, the [Bi7Se12]3‘ anion is layered. It contains the Cd12 and Bi2Te3-type fragments that were mentioned above but no Sb2S 3 88 characteristics. The CdI2-type octahedra link the Bi2Te3-type (or NaCl-type) blocks in an edge-sharing manner to form a lamellar structure. The highly charged [Bi6Se13]8' anion has a very interesting structure. It contains two-dimensional sheets made up of edge- sharing CdI2- and Bi2Te3-type fragments. One-dimensional chains, comprised of Bi2Te3-type blocks, extend along the b-direction and separate these layers. Atoms Bi(l) and Bi(2) possess slightly distorted octahedral coordination as evidenced by the average bond distances and bond lengths in Table 2-8. The corresponding Bi-Se bond distances range from 2.847(9)A to 3.038(3)A for Bi(l) and 2.869(9)A to 3.021(9)A for Bi(2). These distances are similar to those reported for, C53Bi78e 12.5 The Bi(3) coordination environment is distorted with a Bi(3)-Se(5) bond distance of 2.730(8)A which is trans to a long Bi(3)- Se(7) distance of 3.172(3)A. This same distorted coordination is found in C53Bi78e 12.5 This distorted octahedral coordination has been seen in several other bismuth compounds, including Sr4Bi6Se136 and in several bismuth sulfosalts, including PbBi2S4 (galenobismuthite)33 and PbCu4Bi5811.34 Bi(4) is the only atom in which the inert lone pair of electrons is stereochemically active as revealed by the distorted square pyramidal coordination shown in Figures 2-3 and 2-5. This coordination site consists of four equatorial bonds of 2.95(3) A (ave.) and a short axial distance of 2.704(9) A. This short bond is trans to two Bi(4)-Se(4) interactions at 3.552(1) (ave.) which are shorter than the approximate sum of the van der Waals radii (4.4 A).35 The Se(4,4')-Bi(4)-Se(6) angles are much less than 180° at 141.8(2)° and 141.9(2)° and the Se(6)-Bi(4)-Se(6', 6", 5, 5') angles are less than 90° 89 at 83(2)° (ave.). This type of arrangement is found in the mixed metal compound, Cu1,6Bi4,3Se3, where the Bi(3) has a short axial bond of 2.71(1)A that is trans to a Bi-Se distance of 3.54(1)A with an angle of 139.4(4)A.36 90 82 SI 31' Bi Figure 2-1. ORTEP representation of a single (8182):)!" chain with Iabeling scheme. 91 Figure 2-2. ORTEP packing diagrams of the (Bi82)nn' chains along the a-axis, (A) and b-axis, (B). Dashed lines in (A) indicate Bi-«S contacts. The shaded ellipsoids are Bi atoms, and the C5 atoms are represented by open circles. 92 . a . 2 / a. - 9 ~. /" // 8:4 563 1', ii§///J‘ ‘ 1 / \1; e- - B / 4: ’71; / ‘33.- __ ‘ .3 56627“ Sc2 1:“ " ScS 33:“ -~ _. “5/ "2 ‘ A - ~— Figure 2-3. ORTEP representation of the packing diagram of KzBi38e13 down the c-axis. The shaded ellipsoids are the Bi atoms. and the open ellipsoids are the K atoms. 93 (A) (B) Tc (D) \ ‘\ ‘- I I szsr-type --------- >‘- I ‘. “a. — __... \ ’ I l. - -"‘ O _ 0". '0‘. ‘ 0, 9)- i ‘9... Figure 2-4. Projections of the structures of (A) Bizse3 (BizTe3- tYpe) (B) 812563 (Sb283-type) (C) CdIz-type and (D) [Bi38e13]2' framework. The three structure types found in this framework are designated with dashed lines. 94 Se A S Z ( , “c4' % 2 \1 \I Scé'g 86.5 “ Bi4 s... l 8:6 3... Figure 2-5. ORTEP representation of the Bi(4) coordination in K28l38e13. Dashed lines designate long contacts. 95 Table 2-6. Selected Distances (A) and Angles (deg) in the B-CsBiS2 with Standard Deviations in Parentheses.a Bi-S(1) 2.531(9) S(l)-Bi-S(1') 88.1(3) Bi-S(1') 2.733(8) S( 1 )-Bi-S(2) 100.3(4) Bi-S(2) 2.505(9) S( l ')-Bi-S(2) 93.5(3) Bi-8(2') 3.00(1) S(2)-Bi-S(2') 90.0(3) S( 1 ')-Bi-S(2') 93.8(3) Bi-S(1)-Bi 99.0( 3) Cs-S(1) 3.533(9) Bi-Bi' 3.909(3) Cs-8(2') 3.589(9) Bi-Bi" 4.004(4) Cs-S(1") 3.54( 1) Cs-S(1"') 3.598(9) Cs-S(2') 364(1) Cs-S(2") 4.13(l) Cs—S(2"') 3.81(1) Cs-S (mean) 3.7(2) a)The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations ol={zn(1n- l)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. Table 2—7. Comparison of Bond Distances (A) and Angles (deg) in Selected Group 15 Sulfides (M=As,Sb,Bi). Formula Sl-M-S 1 ' S 1 -M-82 81 '-M-82 M-Sl M-82 M—Sl ' NaA 582 95.5( 1) 103.6(1) 102.0(1) 2.31( 1) 217(1) 2.33( 1) C58 b82 8906(5) 100.73(6) 9528(6) 2.450(2) 2.366(2) 2.583(3) B—CSB iSz 88.1(3) 100.3(4) 93.5(3) 2.531(9) 2.505(9) 2.733(8) 96 Table 2-8. Selected Distances (A) and Angles (deg) in the [BigSe13]7-' Framework with Standard Deviations in Parentheses.a Bi( 1 )-Se( 1") 2.988(9) 8e(l")-Bi(1)-Se(2) 175.9(3) Bi(1)-Se(2) 3.009(8) Se(3)-Bi( 1 )-Se(7) 176.4(2) Bi( 1 )-Se( 3) 2.847(9) Se(3')-Bi(1)-Se(7') 176.1(2) Bi( 1 )-Se(3') 2.867(7) 8e( 1" )-Bi( 1 )-Se( 3) 92.88(2) Bi(1)-Se(7) 3.038(3) Se(l")-Bi(1)-Se(3 ') 92.6(2) Bi(1)-Se(7') 3.037(3) 8e( 1")-Bi( l )-Se(7) 88.3(2) Bi( l)-Se (mean) 2.97( 9) Se(l")-Bi(1)-Se(7') 88.5(2) Bi(2)-Se( l) 3.012(7) Se(2)-Bi( l )-Se(3') Bi(2)-Se(1') 3.021(9) Se(2)-Bi(1)-Se(7) Se(2)-Bi( l )-Se(3) 90.3(2) 89.9(2) 88.4(2) Bi(2)-Se(3') 3.024(8) Se(2)-Bi(1)-Se(7') 88.7(1) Bi(2)-Se(4) 2.881(8) Se-Bi( 1 )—Se (mean) 90(2),176.1(2) Bi(2)-Se(4') Bi(2)-Se(4") 2.869(9) 2.878(7) Bi(2)-Se (mean) 2.95(7) Bi(3)-8e( 1) 3.062(9) Se-Bi( 2)-Se (mean) 90(3), 175(1) Bi(3)-8e( 1') 3.067(7) Bi(3)-Se(2) 2.860(9) 8e( 1 )-Bi(3 )-Se( 2) 172.4(3) Bi(3)-Se(2') 2.878(7) 8e( 1 ')—Bi(3 )-Se(2') 172.3(3) Bi(3)-Se(5) 2.730(8) Se(5)-Bi(3)-Se(7) 179.9(2) Bi(3)-Se(7) 3.172(3) Se-Bi(3)-Se (mean) 90(4), 175(4) Bi(3)-Se (mean) 3.0(2) Se(4)-Bi(4)-Se(4') 71.8(2) Bi(4)-Se(4) 3.552(8) Se(4)-Bi(4)-Se(6) 141.9(2) Bi(4)-Se(4’) Bi(4)-Se(5)eq 3.551(9) 2.988(8) 8e(4')-Bi(4)-Se(6) 141.8(2) Bi(4)-88(5')eq 2.987(7) Se(5)-Bi(4)-Se(6) 80.5(2) Bi(4)-Se(6) 2.704(9) Se(S')-Bi(4)-Se(6) 80.5(2) Bi(4)-88(6')eq 2.926(6) Se(6')-Bi(4)-Se(6) 85.1(2) Bi(4)-86(6")eq 2.913(7) 8e(6")-Bi(4)-Se(6) 85.0(3) Bi(4)8eeq (mean) 2.95(3) Se-Bi(4)-8e (mean) 83(2) ”The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations ol={>:n(ln- l)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 97 3 .3 . Thermal Analysis Thermal analysis has been used previously to investigate the synthesis of solid state compounds,37 the mechanism of crystallization in chalcogenide glasses38 and the high temperature phase diagrams of many A2Q/M2Q3 (A= alkali metal; Q= 8, Se, Te; M= As, Sb, Bi) systems.39 In order to study the reactions of Bi in the C528x and K2Sex fluxes we employed differential scanning calorimetry. The goal was to gain further insight into the thermal events that lead to the new compounds and the temperature at which they occur. The thermogram of a Bi/C528/nS mixture that leads to the synthesis of B-CsBiS2 is given in Figure 2-6A and shows four exothermic peaks. To ascertain the origin of each peak, various control experiments were performed using C528/n8 flux, Bi/C528 and Bi/S mixtures under identical conditions. Analysis of the C528/S reaction thermogram shown in Figure 2-6B revealed two exothermic peaks (A,B) that are a result of atomic diffusion and reaction of C528 with S to form a polysulfide flux. The flux reacts further with excess 8 to form a new polysulfide flux thereby producing exotherm C. Upon cooling, the excess flux solidifies at 178 ° C (peak H). The same exothermic peaks are observed in the Bi/C528/nS mixture (Figure 6A) and thus are assigned to the same processes. The C528x flux then reacts with Bi at 274 °C to produce B-CsBiSz giving rise to peak E in Figure 2-6. Endotherm D, at 270 ° C, is assigned to the melting of Bi metal just prior to reaction with Cs2Sx. All of the melted Bi immediately reacts at 274 °C (peak E), which is supported by the 98 absence of a Bi recrystallization exotherm peak upon cooling. Peak E is assigned to the reaction of Bi and initial crystal nucleation. At 176 °C, exotherm F is observed on cooling which can be attributed to crystallization of the excess C528x flux. A melting endotherrn at 180 °C (peak G) was observed as the mixture was heated to 310 °C a second time. This result verified that peak F was due to excess flux crystallization and not due to product crystallization. To verify that peak F is not due to the crystallization of B-CsBi82, a bulk synthesis was carried out under the same heating and isotherm conditions in a sealed pyrex tube in a computer-controlled furnace. The reaction was quenched from 275 ° C (after exotherm E) to room temperature. The product was isolated and powder XRD confirmed that the red crystals were single phase B-CsBiSz. It should be noted that C528 purchased commercially yielded only Bi2S 3, instead of B-CsBiS2, under the conditions described here. This was confirmed with a bulk synthesis reaction in the furnace. Furthermore, DSC therrnograms obtained from the C528/n8 mixtures using this C528 were completely different from that shown in Figure 2-6B. The small particle-size C528 synthesized from the liquid ammonia method is very reactive and provides the best starting material for the synthesis of B-CsBiS 2. Care was taken to prevent any premature reaction at room temperature during the DSC experiments. The thermogram for the synthesis of K2Bi38e13 is shown in Figure 2-7A. As in the B-CsBiSz synthesis described above, therrnograms from K2Sex, Bi/K2Se and Bi/ Se reactions were obtained 99 for reference purposes. Clearly the exothermic peaks (I, II) are due to the reaction between K2Se and Se to form a polyselenide flux as suggested by the control-reaction of K28e/Se whose DSC thermogram is given in Figure 2-7B. The redox reaction of Bi with the formed K2Sex flux occurs at 268 °C shown by the sharp exotherm III. However, the broad rise of the specific heat preceding peak 111 could also be attributed to product formation and reactant diffusion. DSC of the Bi + 88e system produced a sharp endotherm at 226 °C corresponding to the melting of Se followed by a very sharp exotherm at 268 ° C due to the formation of B128e3. The formation of Bi28e3 and K2Bi38e13 occurs at the approximate melting point of Bi metal which is masked in these systems. Exotherm IV in Figure 2- 7A upon cooling, can be attributed to the crystallization of the excess K28ex. Exotherm IV is at a different temperature than exotherm V (Figure 2-7B) because the excess K2Sex flux is of different chemical composition following the reaction with Bi metal. Bulk synthesis in the furnace at 330 °C, under DSC conditions, gave black needles of K2Bigse13 upon quenching from 275 °C to room temperature, supporting our assignment of exotherm IV. As in B-CsBi82, crystal growth of K2Bi38e 13 is not instantaneous, rather it is a gradual process which begins with nucleation at the reaction exotherm followed by growth at the heating and isotherm stages and briefly thereafter upon cooling. By studying these reactions with the DSC technique, we have identified the important reaction steps preceding the reaction of the metal with the flux and determined the exact reaction temperatures. Based on these results 100 we were able to reduce the total time of synthesis of these compounds from the original 10 days to only 24 hours. 101 (A) Y ' F ' Bi/CszS/S :IMS'S E110 3"” ?‘N ‘1 , 1 G (l t j) C t” / ” ‘. I 1 . . j E A i‘ \ P . )- L I l, Endo ,3) mw _ I ‘x . l/ W ‘4 I _ D ‘4 . -‘ 0 b 0 I00 :00 300 400 Temp. 1°C) I T Cs:S/S .rt 4.5:8 0 Endo ” 1’ i I L r j ‘ * mW )- .. ,, c a l 3 1 ./\I (f I 1 . A. 4' \5- b ‘1 TP $\5 %\:11 H \ ‘5. . I l Elle1 OF - . . 0 I00 200 300 400 Temo. (°C) Figure 2-6. (A) DSC thermogram of the Bi/C5283 mixture. The cooling and reheating curves are shown above the heating thermogram for clarity (Peaks F and G). Peak temperatures (°C): A (97), B (113), C (170), D (270), F. (274), F (176) and G(180). (B) DSC thermogram of the C528/S mixture. Peak temperatures (°C): A (92), B (112), C (170) and H (178). 102 (A) r . ' f r ' ' T Exo 2 0L. BileSe/Se at 1:1:8 _ III - MW L . L L . [I A [V ‘ l,/ ,/ " \ - ‘ /'/ ,' ‘- /' ’ l -"I___’ o 100 25b :60 40 0 Temp. NC) K:Sc/Se at l;8 .4 4 . 4 1 . . 4 I 1 . i ! I .3 I g 3 v I -\<', ! ' ' l . "N A .1" / [ft/"o \" mW f r" i i f Endo -LO} 1L . i) 100 200 300 400 Temp. (°C) Figure 2-7. (A) DSC thermogram of the Bi/K-28e9 mixture. Peak temperatures (°C): 1 (160), II (169), III (269) and IV (152). (B) DSC thermogram of the Kzse/Se mixture. Peak temperatures (°C):I (161), II (170) andV (123). 103 3.4. Electrical Conductivity and Thermoelectric Power Measurements. Four-probe electrical conductivity measurements on single crystals of B-CsBiSZ showed that the material is a semiconductor with room temperature conductivity o~10‘5 S /cm which drops quickly to 10'12 S/cm at 200 K. K2Bi3$e13 is also a semiconductor but with significantly more facile charge transport as evidenced by both conductivity and thermoelectric power measurements as a function of temperature. In contrast, the room-temperature conductivity of single crystals of the layered y-CsBiSZ is ~10'2 S/ cm and exhibits a thermally activated temperature dependence, which is shown in Figure 2-8. Thermoelectric power data for y-CSBiSz show a very large negative Seebeck coefficient (-5 80 mV/K at 120 K and -450 mV/K at 300 K) suggesting a n-type semiconductor. Figure 2-9 shows the log conductivity as a function of temperature for a pressed polycrystalline pellet of K2Bi88e13. This study was performed on several bulk samples of the compound with similar results. The magnitude and temperature dependence of the conductivity suggest that the compound is a semiconductor. The conductivity of ~102 S/cm at room temperature is compared to the values 1.6-2.0 x 103 S/cm for Bi28e3 (Bi2Te3-type),40 and 0.6-0.8 S/cm for Bi28e3 (BiZS3- type) obtained from single crystals of these materials.41 The lower conductivity of K2Bi38e13 and the deviation from linearity particularly at low temperatures in the logo vs. 1/T plot (see Figure 2-9) may be due to the effects of grain boundaries in the pressed 104 pellet. It should be noted that typically for three-dimensional materials the conductivities of polycrystalline compactions are two to three orders of magnitude lower than the corresponding values from single crystals. Efforts are continuing, to obtain larger single crystals that are suitable for conductivity measurements. The conductivity measurements alone can not unequivocally characterize the electrical behavior of KzBigse 13. A complementary probe to address this issue is thermoelectric power (TP) measurements as a function of temperature. TP measurements are typically far less susceptible to artifacts arising from the resistive domain boundaries in the material because they are essentially zero— current measurements. This is because temperature drops across such boundaries are much less significant than voltage drops. Figure 2-10 shows typical TP data of a polycrystalline aggregate of K2Bi3$e13 as a function of temperature. The TP is negative throughout the temperature range studied (80450 °C) temperature by direct combination of the elements or alkali carbonates with Bi and S. In addition, mixed Bi/ transition metal/Q (Q=S, Se) solid solution systems have been investigated including 011+3xBi5-x%10 and Mn1-xBiz,.yQI.11 Several naturally occurring bismuth sulfosalts, including PbBiZS4 (galenobismuthite)12 and PbCu4Bi5811,13 have also been observed. When we consider the structural diversity that can result from the manifestation of the 116 117 inert lone pair of electrons on Bi3+ and from the variable coordination preference of this ion, it becomes apparent that further work in this area is warranted. Recently, we have synthesized three ternary bismuth chalcogenides using AzQx (A=K, Cs; Q=S, Se) fluxes below 400°C that display a wide range of structural diversity.14 5- CsBiSz possesses a linear chain structure with comer sharing BiS3 trigonal pyramids while y-CsBiSZ forms a layered superstructure of RbBiSz. In K2Bi3$e13, layers of BiSe6 octahedra are connected by BiSes square pyramidal units to form a three-dimensional network with K+ ions located in the channels. Both y-CsBiSZ and KzBi3$e13 exhibit high thermopower and reasonably high n-type electrical conductivity. In an effort to expand this very interesting chemistry, we investigated the reactivity of Bi in K28X fluxes. The alkali metal polychalcogenide flux method has proven to be an extremely useful low temperature route to access new, often metastable compounds with novel structures.15,16 The possibility of open metal chalcogenide framework structures that combine the utility of microporous oxides with the useful electronic properties of metal chalcogenides to form a new class of microporous chalcogenides is intriguing. Bedard and coworkers reported the hydrothermal synthesis of several tin sulfides that display microporoous activity after partial removal of the occluded template. 17 Another Open chalcogenide framework was reported in the antimony sulfide system using the hydrothermal method. 13 Recently, a novel molten salt approach using (Ph4P)zSex fluxes at 200 °C was shown to give (Ph4P)[M(Se6)2] (M=Ga, In, T1).19 The open polymeric framework in this compound consists of tetrahedral M3+ centers and bridging Se6z‘ ligands that form an extended structure in two dimensions with Ph4P+ cations residing 118 within the layers. In this paper, we report on KBi385, a novel metastable compound with a three-dimensional structure with surprisingly large tunnels for a chalcogenide compound, and with topotactic ion- exchange properties. 2 . Experimental Section 2.1. Reagents Chemicals in this work were used as obtained: (i) bismuth powder, 99.999+% purity, -100 mesh, Aldrich Chemical Co., Inc., Milwaukee, WI. (ii) sulfur powder, sublimed, J. T. Baker Chemical Co., Phillipsburg, NJ. (iii) potassium metal, analytical reagent, Mallinckrodt Inc., Paris, KY. (iv) methanol, anhydrous, Mallinckrodt Inc., Paris, KY; diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. (v) hydrochloric acid, 365-3 8% volume solution, EM Science, Gibbstown, NJ. (vi) sodium acetate, Fischer Scientific Co., Fairlawn, NJ. (vii) cesium acetate, technical grade, Penn Rare Metals Division Kawecki Chemical Co., Revere, PA. (viii) sodium chloride, EM Science, Gibbstown, NJ. (ix) rubidium chloride, cesium choride, 99.9% pure, Cerac, Inc., Milwaukee, WI. 119 2.2. Synthesis All manipulations were carried out under a dry nitrogen atmosphere in a Vacuum Atmospheres Dri-Iab glovebox. For the preparation of K28 we used a modified literature procedure.20 Potassium Sulfide, K28. An amount of 2.460 g (0.077 mol) sulfur and 6.00 g (0.153 mol) freshly sliced potassium metal were added to a 250 ml round-bottom flask. A 150 ml volume of liquid ammonia was condensed into the flask at -78 °C (dry ice/ acetone) under nitrogen to give a light blue solution. The NH3 was removed by evaporation under a flow of nitrogen as the bath slowly warmed to room temperature. The pale yellow solid (99% yield) was dried under vacuum overnight, flame-dried, and ground to a fine powder with a mortar and pestle. Potassium Pentasulfido-tribismuthate(III), KBi38 5. The reaction of 0.052 g (0.25 mmol) Bi, 0.072 g (0.65 mmol) K28 and 0.096 g (3.0 mol) 8 in an evacuated pyrex tube at 300 °C for 5 days, followed by cooling at 4 °C/hr to 120 °C then to room temperature in one hour, afforded small dark gray needles (94% yield, based on Bi metal). The product, stable with respect to water and air, was isolated by dissolving the excess K28), with methanol under inert atmosphere. Semi-quantitative microprobe analysis on single crystals gave K1,4Bi2,385,2 (average of three data acquisitions). Far-IR (CsI matrix) gave a broad absorbance from 260 to 158 cm-1. Raman sprectroscopy gave a strong, broad shift at 267 cm-1. (H20)Bi3S4_5-(Sg)0,06. The Stirring Of 0.200 g (0.242 mmol) of finely ground KBi385 in a solution of 0.5M HCl (30 ml), at room 120 temperature in air, afforded a microcrystalline product with the same dark gray color as the starting material. The product was filtered and washed with distilled water, ethanol and ether. Varying the reaction time (0.5 - 66 hours) and lowering the temperature to 0 °C did not influence the final result. Extensive semi-quantitative (EDS) microprobe analyses were performed on the crystallites from several reaction products, giving a K/ Bi ratio of 0.01 to 0.03,21 suggesting that virtually all K+ was removed. B-Rubidium Pentasulfido-tribismuthate(III), B-RbBi385. An amount of 0.042 g (0.051 mmol) KBi385 was heated with excess RbCl in a 1:100 ratio in an evacuated pyrex tube at 350-380 °C for 2 days. After cooling to room temperature over 6 hours, the product was washed with distilled water, MeOH and ether. Semi-quantitative (EDS) microprobe analysis on single crystals gave Rb1,0Bi3,586_0 (average of three data acquisitions) suggesting that virtually all K+ was removed. Solid State Reactions with Alkali Metal Chlorides. 0.042 g (0.05 1 mmol) KBi385 was heated with excess ACI (A=Na, Rb, Cs) in a 1:100 ratio in an evacuated pyrex tube at 200-400 °C for 2 days. After cooling to room temperature over 6 hours, the product was washed with distilled water, MeOH and ether. Details of the reaction conditions are given in Table 2-9. The homogeneity of KBi385 was confirmed by comparing the observed and calculated X-ray powder diffraction patterns. The (1th spacings observed for the bulk materials were compared, and found to be in agreement with the (1th spacings calculated from the single crystal data using the program POWD10.22 The results are shown in 1 2 1 Table 2-10. 2.3. Physical Measurements The FT-IR spectrum of KBi385 was recorded as a solid in a CsI matrix. The sample was ground with dry CsI into a fine powder, and a pressure of about seven metric tons was applied to the mixture to make a translucent pellet. The spectra were recorded in the far-IR region (600-100 cm' 1) with the use of a Nicolet 740 FT-IR spectrometer. The Mid-IR spectra (4000-3 50 cm'l) were recorded as a solid in a KBr matrix. Raman spectra were recorded at room temperature with a Nicolet FT-Raman 950 spectrometer. The instruments and experimental setups for optical diffuse reflectance measurementsZ3 and quantitative microprobe analysis on SEM/EDS are the same as those in Section 2.3 in Chapter 2 (Part 1). Thermogravimetric analysis (TGA) was performed with a computer-controlled Shimadzu TGA-50. The samples were heated in a quartz cup from room temperature to 330 °C at a heating rate of 3 °C/min. under a nitrogen flow rate of 57 ml/min. Differential Scanning Calorimetry (DSC) was performed with a computer-controlled Shimadzu DSC-50 thermal analyzer under a nitrogen atmosphere at a flow rate of 35 ml/min. The samples were crimped in an aluminum pan inside a nitrogen-filled glove box. The pan was placed on the sample side of the DSC-50 detector and an empty aluminum pan of equal mass was crimped and placed on the 1 2 2 reference side. The samples were heated to the desired temperature at 5 ° C/ min. The reported DSC temperatures are peak temperatures with a standard deviation of 0.2 degrees. The adopted convention in displaying data is: exothermic peaks occur at positive heat flow while endothermic peaks occur at negative heat flow. Differential Thermal Analysis (DTA) was performed with a computer-controlled Shimadzu DTA-50 thermal analyzer. The ground single crystals (~10.0 mg total mass) were sealed in quartz ampules under vacuum. An empty quartz ampule of equal mass was sealed and placed on the reference side of the detector. The samples were heated to the desired temperature at 20 °C/min, then isothermed for 10 min followed by cooling at 10 ° C/ min to 200 °C and finally by rapid cooling to room temperature. Indium (reported m.p. 156.6 °C) and antimony (reported m.p. 630.9 °C) served as calibration standards. The reported DTA temperatures are peak temperatures. The DTA samples were examined by powder X-ray diffraction after the experiments. Mass Spectrometry (MS) was performed with a VG Instruments Trio-1000 Mass Spectrometer. The solid samples were heated at 5 °C/min. to 300 °C with a solid probe. 2.4. X-ray Crystallography KBi385 was examined by X-ray powder diffraction for the purpose of phase purity and identification. Accurate <1th spacings (A) were obtained from the powder patterns recorded on a Rigaku 123 Rotaflex Powder X-ray Diffractometer with Ni filtered Cu Ka radiation operating at 45 kV and 100 mA. The data were collected at a rate of 1.0 deg/min. The variable temperature powder XRD experiment was performed with the instrument described above. The instrument was equipped with a temperature controller and the experiment was carried out under nitrogen flow with an enclosed sample compartment. The data were collected at a rate of 2.4 deg/ min. The sample was heated to 65 °C and was allowed to equilibrate for 30 minutes before the XRD was taken. The sample was then heated in 20 °C increments with 30 minute equilibration times at each step. The final XRD was taken at 145°C. Structure solution of KBi385. A crystal with dimensions 0.50 x 0.10 x 0.10 mm was mounted on a tip of a glass capillary. Intensity data were collected using the (1)-20 scan mode on a Rigaku AFC6S four-circle automated diffractometer equipped with a graphite-crystal monochromator. The stability of the crystal was monitored by measuring three standard reflections periodically (every 150 reflections) during the course of data collection. No crystal decay was detected. An empirical absorption correction based on «p scans was applied to the data. The space group was determined to be ana (#62), and a solution was found using the automated direct methods function in the SHELX886 program24 and refined by full-matrix, least-squares techniques of the TEXSAN package of crystallographic programs.Z'5 Three bismuth atoms, five sulfur atoms and a potassium atom were 124 located. Each of the nine atoms reside on a mirror plane. Inspection of the difference Fourier map revealed a significant electron density peak at a distance from 8 reasonable for K atoms. This peak was assigned to a second K atom which also resided on a mirror plane. Assuming Bi3+ and 82', electroneutrality dictates that only one K+ atom should be found. After least squares refinement, the isotropic temperature factor for K(2) was rather high at 14.0 A2. The occupancy and the temperature factor of this atom were refined to give values of 0.23 and 8.4 A2 respectively. At this stage, a disorder model K( 1) 1-x K(2)x B1385 seemed to be reasonable. The occupancies of K(1,2) were constrained to 0.5 then refined. Least squares refinement of this model resulted in a occupancy value of 0.39 and a temperature factor of 1.67 for K( 1) which decreased from 2.48. K(2) gave a occupancy of 0.1 1 and a temp. factor of 1.62 which decreased from 8.4. A DIFABS correction was applied to the isotropically refined data.26 K(2) was refined isotropically because of its small occupancy while the rest of the atoms were refined anisotropically to give a final R/RW = 5.1/6.7. All calculations were performed on a VAXstation 3 100/ 76 computer. The complete data collection parameters, details of the structure solution and refinement for KBi385 are given in Table 2-1 1. The atomic coordinates, average temperature factors and their estimated standard deviations for KBi3S 5 are given in Table 2-12. Table 2-9. Solid State Reactions of KBi385 with Alkali Metal Halides. 125 Temp. (°C) NaCl RbCl CsCl 1 97 no reaction no reaction no reaction 297 NaBi82 + BizS 3 beK1-xBi3SSa y-CSBiSz 347 384 400 NaBi82 + 31283 B-RbBi3Ssb y-CsBi82 B-RbBi3Ssb y-CsBiS2 + phaseC a-RbBi3Ss y-CsBi82 + phase a EDS analysis gave Rb0,54K0,46Bi3,185,6. b Topotactic ion-exchange C EDS analysis gave Csl_oBi3,385,1 (needles) and Csl,0Bi2,583,5 (hexagonal plates). Table 2-10. Calculated and Observed X-ray Powder Diffraction 126 Patterns for KBi38 5. h kl d( c210), A dlobs.).A 1/ Imgtx (ObS-) 12.15 12.13 101 0 0 2 2 0 0 2 0 2 1 0 3 3 0 1 3 0 3 3 0 3 40 2 8.68 8.51 6.08 5.48 5.39 4.13 4.05 3.82 1 12, 2 1 1 3.61,3.60 3 0 4 1 0 5 0 1 3 1 1 3 3 l l 2 0 5 5 0 2 3 1 2 4 0 4 41 0 3.45 3.40 3.33 3.27 3.25 3.22 3.17 3.09 3.04 2.94 41 1, 0 0 6 2.90, 2.89 1 0 6 5 1 l 2 1 5 5 1 4 3 1 6 108 8 0 1 5 l 5 4 1 6 0 2 0 6 0 6 5 0 7 3 1 7 6 1 5 7 1 4 2 l 8 2.85 2.58 2.52 2.238 2.179 2.153 2.111 2.088 2.063 2.038 2.025 2.005 1.985 1.934 1.881 1.869 5 1 7, 4 2 2 1.799,1.798 4 2 3 1.752 1 1 9, 9 0 4 1.735,1.733 6 0 8 l 2 6 2011 1.724 1.658 1.552 8.70 8.50 6.07 5.48 5.38 4.13 4.04 3.81 3.60 3.45 3.40 3.34 3.28 3.25 3.22 3.16 3.09 3.04 2.94 2.90 2.85 2.58 2.52 2.238 2.179 2.153 2.106 2.087 2.063 2.039 2.023 2.005 1.986 1.933 1.881 1.869 1.799 1.752 1.734 1.723 1.660 1.552 78 4 10 51 5 13 4 8 14 11 42 37 14 6 12 25 11 5 6 13 35 100 4 4 4 6 10 6 12 6 6 4 4 10 5 8 5 4 4 9 8 8 4 Table 2-11. Summary of Crystallographic Data and Structure 127 Analysis for KBi 385 Formula KBi385 FW a, A b, A c, A a, deg 6. deg v. deg 826.34 17.013(5) 4.076(2) 17.365(4) 90.000 90.000 90.000 Z; V, A3 4; 1204(2) ). 0.7 1073 (Mo Ka) space group ana (No. 62) Dcalc., g/cm3 4.56 0, cm:1 447 (Mo Ka) 29max, deg 50 (Mo Ka) Temp., °C 23 Final R/Rw,% 5.1/6.7 Total Data Measured 13 1 1 Total Unique Data 13 1 1 Data with F02>30(F02) 782 No. of variables 59 *R=E(IFQI'IFc1)/EIF01 Rw={2w(IFol-IFCI)2/2wlFo|2}1/2 128 Table 2-12. Fractional Atomic Coordinates and Beq Values for KBi385 with Estimated Standard Deviations in Parentheses. atom x y z Beg, A2 Bi(l) 0.2380(1) 1/4 0.7111(1) 1.40(8) Bi(2) 0.0831(1) -1/4 0.5717(1) 1.25(8) Bi(3) 0.3433(1) -1/4 0.8743(1) 1.76(9) K(1)b 0.5019(7) -3/4 0.7041(8) 1.2(5) K(2)C 0.148(3) 1/4 0.018(3) 1.7(9) S(l) 0.3535(7) -1/4 0.7162(7) 1.6(6) 5(2) -0.0363(8) -3/4 0.6022(7) 1.9(6) 8(3) 0.1235(6) -1/4 0.7194(6) 1.1(5) S(4) 0.2325(6) 1/4 -0.1369(6) 1.3(5) 5(5) 0.1894(7) 1/4 0.5393(7) 1.7(5) a)Beq = (4/3)[aZB(1,1) + bZB(2,2) + c2B(3,3) + ab(c05y)B(1,2) + ac(cosB)B(1,3) + bc(cos:n(1n- l)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 138 3.3. Thermal Analysis The thermal behavior of KBi385 was investigated with differential thermal analysis (DTA). Figure 2-16 shows that upon heating KBi38 5, a broad endotherm occurs at 520 °C. Upon cooling, no corresponding exothermic peak is observed suggesting an irreversible change in the structure even though no change in the physical appearance of the compound was observed. Powder X-ray diffraction (XRD) of the heated sample gave a new XRD pattern, perhaps forming p-KBi3S 5. The low angle (1 0 1) peak of KBi385 shifts from 12.13 to 11.33 A. The XRD peaks from this apparently new phase match those from the minor impurity phase found in the Bi + 2.8 K2853, reaction mentioned above but do not match those found for the high temperature K28/Bi283 phases.28 Further work is needed to characterize this compound. 139 520 I I I I I ‘20 0 100 200 300 400 500 600 Temp. (°C) Figure 2-16. DTA thermogram of KBi385. 140 3.4. Ion-Exchange Studies The relatively open structure of KBi385 encouraged us to explore possible ion-exchange behavior. For example, ion-exchange studies of KBi385 were performed in a dilute aqueous HCl solution with the goal of synthesizing a proton-exchanged material. Our first attempt resulted in an air oxidation product (H20)Bi3S4,5(Sg)0,06 instead of a truly topotactic ion-exchanged product. Figure 2-17 shows the powder X-ray diffraction (XRD) patterns of KBi385 and (H20)Bi384,5-(83)0,06. The low angle peaks of KBi385 remain after the reaction, indicating that the framework is more or less intact. However, a number of additional peaks appear which reveal a possible reduction in symmetry, and suggest that the reaction is not really topotactic. Examination of the product with scanning electron microscopy (SEM) shows no change in crystallite size (1-30 m) and shape suggesting at least a pseudomorphic reaction (see figure 2-18).39 A Thermogravimetric Analysis (TGA) diagram of (H20)Bi384,5-(83)0,05 up to 330 °C is shown in Figure 2-19a. Step I (3.82% weight loss), starting at 140 °C, is due to the loss of one H2O as confirmed by mass spectrometry. This is immediately followed by Step II (3.84% weight loss), starting at 176 °C, which corresponds to the loss of 1/ 16 mole of 83. The formation of Bi283 after step 11 was confirmed by XRD (vide infra). Since the H20 is not removed until 140 °C, this suggests that H20 is located inside the tunnels and not on the surface. As a control experiment, KBi385 was stirred in H20 for one hour. The TGA of the product did not show any weight loss up to 14 1 400 °C. The assignment of Step II in figure 2-19a is supported by the mass spectrum which shows m/z peaks corresponding to Sx+ (x=1-8). The dominant peak is found at m/z 64 (82+). The presence of a large intensity 82+ peak during this decomposition is corroborated by the large body of evidence that supports this observation. For example, complex sulfur vapor equilibria have been well studied and at temperatures higher than ~3 20°C and at pressures lower than ~1 Torr, the vapor phase of sulfur is dominated by 82+."O The heating of HgS to 308 °C under low pressure gave 82+ as the major ion peak.41 To determine if the S3 fragment was located on the surface of the material or inside the tunnels, the material was washed with (:82 then ether. TGA of this CSz-washed product gave the same result as shown in Figure 2-19a. The mass spectrum of this product up to 300 °C also confirmed the loss of H20 and 8,, fragments which confirms that 83 is not located on the surface of the particles. Differential Scanning Calorimetry (DSC) of (H20)Bi384,5-(S3)0,06 to 4000C gave a large, sharp endothermic peak at 146 ° C followed by a small, broad exothermic peak at 156 °C (see Figure 2-19b). These thermal events correspond to dehydration, loss of S3, and subsequent formation of Bi283. For comparison, the DSC of KBi385 performed under identical conditions and featureless. In order to further characterize the properties of (H20)Bi384,5-(S3)0,06 the powder XRD pattern (Figure 2-20) was measured as a function of temperature. Peak broadening was observed as the temperature was increased and at 125 °C, a mixture 142 of (H20)Bi3S4,5-(S3)0,06 and Bi283 was detected. At the final temperature of 145°C, only broad peaks assignable to Bi283 were observed. The crystallite size was calculated to be ~160 A.42 Heating (H20)Bi3S4,5-(S3)0,06 to 80 °C under vacuum for 11 hours, revealed a mixture of Bi283 (major product) and (H20)Bi3S4,5-(S3)0,06 by XRD. These data suggest that the framework collapses upon removal of the water and then expels sulfur to form Bi28 3. The Mid-IR spectrum of (H20)Bi384,5-(S3)0,06 did not show any S-S stretching vibrations. Since the proposed concentration of S3 in the compound is very dilute, it is possible that the intensities of these vibrations are very weak. Since the ion-exchange chemistry conducted in air gave an oxidation product, the same experiment was carried out under nitrogen atmosphere. XRD of the product gave the same pattern as KBi3S 5 with broadened peaks and EDS confirmed the presence of K+ indicating no appreciable ion-exchange. Based on the data presented, a scheme regarding the reaction of KBi38 5 with aqueous HCl is proposed: [Bi385]' + 1/4o2 + H3o+ , (H20)Bi3,s4,5°(38)“)6 + 1/2HZO I A -H2o Bi334.s'(38)0.06 3/2 B‘ s 12 3 ‘ A -1/16 58 Scheme (A) 143 In summary, these results suggest that the reaction of KBi385 with an aqueous HCl solution in air results in the formation of an interesting metastable oxidation product, (H20)Bi384,5-(S3)0,06. We also explored cation-exchange reactions with alkali and ammonium ions. Room temperature reactions with aqueous solutions of various salts (NaCl, Rb2CD3, CsCl, NH41\D3) and methanol solutions of cesium acetate did not show ion-exchange. Hydrothermal treatment of KBi3S 5 at 130°C with alkali metal salt solutions of RbCl and CsCl and gave Bi283, while methanothermal treatment of KBi385 at 85 °C with Cs and Na acetate salts did not show ion-exchange. 3.5. Solid State Ion-exchange Reactions with Halide Salts. The complications arising from the ion-exchange attempts using solvents prompted us to investigate "dry" reactions with solid alkali halide salts according to eq.1. 1031355 + AC1 , ABi385 + KCl eq.(1) A A= Na, Rb, Cs Solid state ion-exchange is a very useful synthetic method and has been demonstrated in several systems, including A2Mo6(25 (A=Na, K, Rb, Cs, Cu, Ag; Q=S, Se),43 Abe6Qg (A=Na, Rb, K, transition metals; Q=S, Se, Te),44 AV5($45 and AV6Q (A=Na, K, Rb, Cs; Q=S, Se, Te).45 In these systems, the halide salt is heated above its melting point. 144 Because KBi385 is metastable, the reaction cannot be heated above the melting point (>700 °C) of either NaCl, RbCl or CsCl. It is therefore remarkable that in the RbCl case complete, topotactic ion-exchange occurs below 400 °C forming the isostructural B-RbBi385 (see Table 2- 9). At or above 400 °C the known denser phase a-RbBi3854 forms, confirming that both KBi385 and p-RbBi385 are metastable structures. The fact that this relatively facile ion-exchange reaction is a solid- solid reaction suggests that the exchanging ions possess high ionic mobilities promoted by the large tunnels present in the [Bi3S5I structurefl6 This suggests that the ABi3S5 may be good ionic conductors. The B-phase possesses higher symmetry because the split peaks that are seen at low angles in KBi385 are single peaks in B- RbBi385 suggesting a possible transformation to a tetragonal structure. Interestingly, B-RbBi385 cannot be synthesized by the molten alkali polysulfide approach. Reactions with NaCl (200-400 °C) resulted in a mixture of NaBi82 and Bi283. The Na+ ion may be too small to support the [Bi385]- framework at this temperature so it converts to the thermodynamically stable layered NaBi82?- Reactions with CsCl gave a mixture of the layered y-CsBiS219 and a minor impurity phase. EDS on several 103m needles gave CsBi3,1Ss.2 and analysis on 103m plates resulted in a ratio of CsBi2,5S3,4. XRD of the mixture did not reveal the presence of CsBi385.3 Ion-exchange reactions with solid NH4C1 and NH4I (150—400 °C) gave only Bi28 3. 145 202 303 404 101 y t i s n e t n i (A) (B) (C) 5. 10. 15. 20. 25. 30. 35. 40. 45. 50. 55. so. (29) Figure 2-1 7. Comparison of the X-ray powder diffraction patterns of: (a) KBi385 with selected hOI peaks labeled. 0)) (H20)313$45-(Ss)o.06 and ' (C) 81283 145 101 202 J 5‘ '5; 5 E 5. (A) I (B) (C) t ‘ l l I I l I 11‘ I), I1 ILL. ..1A. .A.. 10. 15. 20. 25. 30. 35. 40. 45. 50. 55. 60. (29) Figure 2-17. Comparison of the X-ray powder diffraction patterns of: (a) [(81385 with selected h01 peaks labeled. (b) (H20)BI3S4.5’(58)0.06 and ' (C) 31233 146 Figure 2-18: SEM photographs of: (a) KBi385 before and (b) after reaction with aqueous HCl (66 hours). The white bar at the bottom of the micrographs represents a 103m scale. The magnification is x1600. (A) 147 (B) 148 100.8! T _ 91.81 fi Temperetere ('C) no 1‘6 IL - I i , 0 250 Temperature ('C) -. 4 500 Figure 2-19. (a) Thermogravimetric analysis plot of (H20)Bi3S4,5-(S3)o,06 with weight 96 plotted against temperature (0 C). Performed under flowing N2. Step I (3.82% weight loss starting at 140 °C.) Step II (3.84% weight loss starting at 176 °C). (b) DSC thermogram of (H20)B13845-(83)o,06. 149 I i (H20)BlsS4s-(Ss)o.06 23°C 3‘ '3 t: 9 E Wand—NM Figure 2-20. Powder XRD pattern of (H20)Bi3S45-(S3)o,06 measured as a function of temperature. The transformation to Bi28 3 is complete at 124 °C 150 3.6. Conclusion The isolation of KBi385 from a polysulfide flux at a relatively low temperature confirms the usefulness of molten salts in accessing low enough temperatures for the stabilization of reactive solid state materials. Although the void space within this metastable material is not substantial, KBi385 represents a rare example of a chalcogenide with an open framework and ion-exchange properties. This and other materials provide the basis for further exploratory work to identify not only new open framework, more stable, chalcogenides but also a reliable general synthetic methodology to these materials. Such a methodology could be greatly aided by drawing lessons from the recent preparative success of the more familiar zeolites,47 aluminosilicates (MGM-41),48 and pillared clays.49 151 List of References (a) Rowe, D. M.; Bhandari, C. M. Modern Thermoelectrics, Holt, Rinehart and Winston: London, 1983; p. 103. (b) Borkowski, K.; Przyluski, J. J. Mat. Res. Bull. 1987, 22,381-387. (a) Boon, J. W. Rec. Trav. Chim. Pays-Has, 1944, 63, 32. (b) Glemser, 0.; Filcek, M. Z. Anorg. Allg. Chem., 1955, 279, 321-323 (c) Gattow, G.; Zemann, J. Z. Anorg. Allg. Chem., 1955, 279,324—327. N S - F I ’ S ‘ Q N W O F Kanishcheva, A. S.; Mikhailov, J. N.; Lazarev, V. B.; Trippel, A. F. Dokl. Akad. Nauk, SSSR (Kryst), 1980, 252, 96-99. Schmitz, D.; Bronger, W. Z. Naturforsch., 1974, 29b, 438-439. Julien-Pouzol, M.; Jaulmes, S.; Laruelle, P. Acta. Cryst. 1979, B351313- 1315. Aurivillius, B. Acta Chem. Scand. 1983, A37 399-407. G. Cordier, H. Schafer, C. Schwidetzky, Rev. Chim. Miner., 1985, 22 676- 683. G. Cordier, H. Schafer, C. Schwidetzky, Rev. Chim. Miner., 1985, 22 631- 638. K. Volk, G. Cordier, R. Cook, H. Schafer, Z. Naturforsch., 1990, 35b 136- 140. Liautard, B.; Garcia, J. C.; Brun, G.; Tedenac, J. C.; Maurin, M. Eur. J. Solid State Inorg. Chem. 1990, 27,819-830. 11. Lee, S.; Fischer, E; Czerniak, J.; Nagasundaram, N. J. Alloys and 12. 13. 14. Compounds 1993, 197, 1-5. Iitaka, Y.; Nowacki, W. Acta Cryst., 1962, 15, 691-698. Kupcik, V.; Makovicky, E. N. Jb. Miner. A01, 1968, 7,236-237. McCarthy, T. J.; Ngeyi, S.-P.; Liao, J.-H.; DeGroot, D.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Mater., 1993, 5,331-340. 15. (a) Kanatzidis, M. G. Chem. Mater. 1990, 2, 353-363. (b) Kanatzidis, M. G.; Park, Y. J. Am. Chem. Soc. 1990, 111, 3767-3769. (c) Kanatzidis, M. G.; Park, Y. Chem. Mater. 1990, 2, 99-101. ((1) Park, Y.; Kanatzidis, M. G. Angew. Chem. Int. Ed. Engl. 1990, 29,914-915. 16. (a) Sunshine, 8. A.; Kang, D.; Ibers, J. A. J. Am. Chem. Soc. 1987, 109, 6202-6204. (b) Kang, D.; Ibers, J. A. Inorg. Chem. 1989, 27, 549-551. 17. R. L Bedard, S. T. Wilson, L. D. Vail, E. M. Bennett, E M. Flanigen, Zeolites: Facts, Figures, Future (P. A. Jacobs, R. A. van Sonten, Eds.) (1989) 275. 18. Parise, J. B. Science, 1991, 251,293-294. Dhingra, S.; Kanatzidis, M. G. Science, 1992, 258, 1769-1772. 20. Feher, F. Handbuch der Praparativen Anorganischen Chemie: Brauer, G., Ed.; Ferdinand Enke: Stuttgart, Germany, 1954; pp. 280—281. 21. Also confirmed by Atomic absorption analysis performed by Oneida Research Services, Inc. on one of our reaction products as a check for our SEM-EDS elemental analysis results. 22. D. K. Smith, M. C. Nichols, M. JE. Zolensky, POWDlO: A Fortran IV program for Calculating X-ray Powder Diffraction Pattern, version 10. Pennsylvania State University, 1983. 23. (a) Wendlandt, W. W.; Hecht, H. G. "Reflectance Spectroscopy", Interscience Publishers, 1966 (b) Kotiim, G. "Reflectance Spectroscopy", Springer Verlag, New York, 1969 (c) Tandon, S. P.; Gupta, J. P. Phys. Stat. Sol. 1970, 38,363-367. 152 24. G. M. Sheldrick, In Crystallographic Computing 3; G. M. Sheldrick, C. Kruger, R. Doddard, Eds.; Oxford University Press: Oxford, England, 1985; pp 175-189. 25. TEXSAN: Single Crystal Structure Analysis Software, Version 5.0, (1989). Molecular Structure Corporation, The Woodlands, TX 77381. 26. Walker, N.; Stuart, D. Acta Cryst. 1983, 39A 158-166. This phase is isostructural to RbBiSz as evidenced by powder X-ray 28. 29. 30. 31. 32. 33. diffraction. Berul, S. 1.; Lazarev, V. B.; Trippel, A. F.; Buchikhina, O. P. Russ. J. Inorg. Chem. 1977, 22(9), 1390-1393. Bube, R. H. "Photoconductivity of Solids", John Wiley and Sons, Inc., 1960; pp. 233-235. (a) Fanchon, E.; Vicat, J.; Hodeau, J.-I.; Wolfers, P.; Tran Qui, D.; Strobel, P. Acta Cryst. 1987, 43B , 440-448. (b) Fanchon, E.; Vicat, J.; Hodeau, J.-L.; Levy, J.-P.; Wolfers, P. Acta Cryst. 1987, Sect. A 43(Suppl.) C-129. (c) Bursill; L A. Acta Cryst. 1979, B35, 530-538. Torardi, C. C. Mater. Res. Bull. 1985, 20, 705-713. Torardi, C. C.; McCarley, R. E. J. Solid State Chem. 1981, 37,393-397. R. G. Burns, M. B. Burns Manganese Dioxide Symposium, Tokyo, 1981; Vol. 2, Ch 6. 34. (a) Khanna, S. K.; Gruner, G.; Orbach,R.; Beyeler, H. U. Phys. Rev. Lett. 1981, 47, 255-257. (b) Beyeler, H. U.; Bernasconi, J.; Strassler, 8. Fast Ion Transport in Solids (P. Vashishta, J. N. Mundy, and G. K. Shenoy, Eds.), North-Holland, New York (1979) p. 503. Clearfield, A. J. Chem. Rev. 1988, 88, 125-148. Shen, X.-M.; Clearfield, A. J. J. Solid State Chem. 1986, 64, 270-282. Bystrdm, A.; Bystrom, A. M. Acta Cryst. 1979, B35,530—538. Wadsley, A. D. Acta. Cryst. 1953, 6, 433-438. In a pseudomorphic reaction the product retains the morphology of the starting material but (in contrast to a topotactic reaction) its internal crystal and framework structure has been altered. Iiao, C. L.; Ng, C. Y. J. Chem. Phys. 1986, 84(2), 778-782. J. Berkowitz, in Elemental Sulfur, edited by B. Meyer (Interscience, New York, 1965). 42. Debye-Scherrer formula: D: (0.9 * 1. * 57.3)/(81/2 * cose) (D=ave. crystallite size (A); A=rad. wavelength; 81/2=peak width at half height; 0=Bragg angle). 43. Tarascon, J. M.; Hull, G. W.; DiSalvo, F. J. Mat. Res. Bull. 1984, 19,915-924. (a) Huan, G.; Greenblatt, M. Mat. Res. Bull. 1987, 22, 505-512. (b) Huan, G.; Greenblatt, M. Mat. Res. Bull. 1987, 22, 943-949. 45. Ohtani, T; Sano, Y.; Kodama, K.; Onoue, S.; Nishihara, H. Mat. Res. Bull. 1993, 28, 501-508. Even though this reaction appears to proceed from solid to solid at such a low temperature, one may also imagine the presence of a liquid phase covering the surface of the KBi385 particles. However no eutectic composition between KCl and RbCl is known below 700 °C. 47. Ozin, G. Adv. Mater., 1992, 4,612-646. Kresge, C. T.; Leonowicz, M. E; Roth, W. J.; Vartuli, J. C.; Beck, J. 8. Nature, 1993, 359,710-712. 49. Pinnavaia, T. J. Science, 1983, 220,365-371. CHAPTER 2 (Part 3) High Temperature Synthesis and Properties of KBi6,33810 and K2Bi3813 1 . Introduction Group 15 chalcogenide compounds have received considerable attention, due to their potential application as non-linear optical materials,1 photoelectrics,2 and thermoelectrics.3 The most common application that is unique to Group 15 chalcogenides is in the area of thermoelectric cooling materials. The most investigated systems over the past 30 years are various solid solutions of M2Qg (M=As, Sb, Bi; Q=S, Se, Te) compounds.4 These materials possess high electrical conductivity and thermoelectric power and low thermal conductivity and are excellent materials for thermoelectric applications near room temperature.5 With these properties in mind, it is surprising that very little exploratory synthesis of new ternary bismuth chalcogenide phases has been reported. Bi is very attractive for study because of its inert 632 lone pair of electrons which may or may not be manifested structurally in a given compound. Whether the lone pair is stereochemically active or 153 154 not affects the lattice structure, the electronic structure and thus the properties of the resulting compounds and thus exploration of the solid state chemistry of Bi is warranted. This issue is related to the larger question of stereochemical activity of a lone pair in compounds with heavy main-group elements in a 32 configuration. Outside of the well-known NaCl-type ABiQ2 (A = Li, Na, K; Q: 8, Se)6 compounds, the only other phases that have been fully characterized structurally are RbBiQ2 (Q= 8, Se),7 CsBi3,85,8 RbBi385,9 Tl4Bist,10 a-(B-)BaBi284,11 CS3Bi7Se12,12 Sr4Bi68e13,13 and BaBiSe3.14 These compounds have been prepared at high (>450 °C) temperature by direct combination of the elements or alkali carbonates with Bi and S(Se). In addition, mixed Bi/ transition metal/Q (Q=S, Se) solid solution systems have been investigated including Cu1+3xBi5-x%15 and Mn1-xBi2+y(24.16 Bismuth compounds constitute ~20% of the known naturally occurring sulfosalts, including PbBi2S4 (galenobismuthite),17 PbCU4BiSSll,18 CuBi583,19 and PbCuBiS3.20 Recently, we have synthesized four new ternary bismuth chalcogenides, B-(y-)CsBi82,21 K2Bi3Se13,7-1 and KBi385,-7-2 using A2Qx (A=K, Cs; Q=S, Se) fluxes at intermediate temperatures below 400°C that display a wide range of structural diversity. K2Bi3Se13 possesses interesting thermoelectric properties with an electrical conductivity of 102 S/cm and a Seebeck coefficient ranging from -210 to -260p.V/ K at room temperature.21 During our investigation, we noticed that there are several high temperature A2Q/Bi2Qg (A = Li, Na, K, Rb, Cs; Q = 8, Se, Te) phase diagrams reported in the literature.23 However, the structural information on the possible 1 5 5 phases only consists of unindexed powder X—ray diffraction patterns. It becomes apparent that further work in the high temperature synthesis and structural characterization of new AxBinz compounds is warranted. This paper describes the high temperature synthesis, structural characterization, optical, and charge transport properties of two new compounds, KBi6,33810 (I) and K2Bi3813 (I 1). Both of these compounds possess new three-dimensional tunnel frameworks based on Bi2Te3- and Cd12-type fragments with K+ ions found in the channels. 2. Experimental Section 2. 1 . Reagents Chemicals in this work were used as obtained: (i) bismuth powder, 99.999+% purity, -100 mesh, Aldrich Chemical Co., Inc., Milwaukee, WI. (ii) sulfur powder, sublimed, J. T. Baker Chemical Co., Phillipsburg, NJ. (iii) potassium metal, analytical reagent, Mallinckrodt Inc., Paris, KY (iv) methanol, anhydrous, Mallinckrodt Inc., Paris, KY; DMF, analytical reagent, diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. 156 2.2. Synthesis All manipulations were carried out under a dry nitrogen atmosphere in a Vacuum Atmospheres Dri-Iab glovebox. For the preparation of K28 we used a modified literature procedure.24 The preparation of K28 is reported in Section 2.2 of Chapter 2 (Part 2). Bismuth Sulfide, Bi2S 3. 2.090 g (10.0 mmol) Bi and 0.496 g (15.5 mol) 8 were ground thoughly with a mortar and pestle. The mixture was transferred to a 6 ml pyrex tube and was subsequently flame-sealed in vacuum (~10'3 torr). The reaction was heated to 500 °C over 24 hours in a computer controlled furnace, then isothermed at 500 °C for 4 days, followed by cooling to 100 °C at a rate of 4 °C/hr, then to 50 °C in one hour. The product was ground into a fine powder and stored in the glove box. KBi6,33S 10 (I). 0.040 g (0.363 mmol) K28 and 0.746 g (1.45 mmol) Bi283 were mixed together with a spatula in a glass vial. Several drops of acetone were placed in a quartz tube (9 mm diameter, 6 ml volume) and it was heated with a flame to create a carbon film on the inside surface of the tube. The mixture was transferred to the carbon-coated quartz tube and was subsequently flame-sealed under diffusion pump vacuum (~10'5 torr). The reaction was heated to 750 °C over a 48 hour period in a computer controlled furnace, then isothermed at 750 °C for 6 days, followed by cooling to 550 °c at a rate of 10 °C/hr, then to 500C in 10 hours. The product was washed with degassed water (50 ml), methanol (20 ml) 157 and ether (20 ml) to remove any trace amounts of K28. 0.733 g (92% yield, based on Bi283) of shiny silver polycrystalline material was obtained. Quantitative microprobe analysis gave K1,0Bi4,185,6 (average of four acquisitions). D i p o t a s s i u m Tridecasulfido-octabismuthate( III) , K2Bi3813 (II) . The reaction of 0.015 g (0.138 mmol) K28 and 0.283 g (0.550 mmol) Bi283 was prepared as above. The mixture was heated to 725 °C over a 48 hour period, then isothermed for 4 days, followed by cooling to 525 °C at 10 °C/hr, then to 50 °C in 10 hours. The product was isolated as above to give 0.294 g (99% yield) of shiny silver crystals and agglomerates. Qnantitative microprobe analysis gave K1,oBi4,6S7_3 (average of four acquisitions). The homogeneity of (I) and (I I) was confirmed by comparing the observed and calculated X-ray powder diffraction patterns. The dhkl spacings observed for the bulk materials were compared, and found to be in good agreement with the dth spacings calculated from the single crystal data.25 The results are summarized in Tables 2-14 and 2-15. 2.3. Physical Measurements Mid-IR diffuse reflectance spectra of (I) and (I I) were recorded as a solid. The sample was ground into a powder prior to data acquisition. The spectra were recorded in the mid-IR region (4000-400cm‘ 1, 4cm-l resolution) with the use of a Nicolet 740 FT- IR spectrometer equipped with a diffuse reflectance attachment. 158 The instruments and experimental setups for optical diffuse reflectance measurements, quantitative microprobe analysis on SEM/ EDS, and charge-transport measurements are the same as those in Chapter 2 (Part 1) (Section 2.3). The experimental setup for differential thermal analysis (DTA) is the same as that in Chapter 2 (Part 2) (Section 2.3). 2.4. X-ray Crystallography Both compounds were examined by X-ray powder diffraction for the purpose of phase purity and identification. Accurate dhkl spacings (A) were obtained from the powder patterns recorded on a Rigaku Rotaflex Powder X-ray Diffractometer with Ni filtered Cu Ka radiation operating at 45 kV and 100 mA. The data were collected at a rate of 1.0 deg/ min. Structure solution of KBi6,33810. A crystal with dimensions 0.02 x 0.05 x 0.25 mm was mounted on a glass fiber. Intensity data were collected using the (1)-20 scan mode on a Rigaku AFC6S four-circle automated diffractometer equipped with a graphite-crystal monochromator. The stability of the crystal was monitored by measuring three standard reflections periodically (every 150 reflections) during the course of data collection. No crystal decay was detected. The structure was solved by direct methods using SHELX88626a and refined by full-matrix least-squares techniques of the TEXSAN package of crystallographic programs.26b An empirical absorption correction based on 1)) scans was applied to 159 the data. Seven bismuth atoms, ten sulfur atoms and a potassium atom were located on mirror planes. After least squares refinement, the isotropic temperature factor for Bi(7) was very high at 12.18 A2 with R/RW = 7.4/ 9.9. A refinement of its multiplicity indicated a lower occupancy on this site. The occupancy and temperature factor of this atom were refined to give values of 0.263 and 3.945 A2 respectively (R/RW = 5.7/ 6.9). The coordination environments of the other six bismuth atoms were octahedral while that of Bi(7) was an unusual eight-coordinate site consisting of bond distances ranging from 2.79 to 3.48 A. At this stage, a model with bismuth atom vacancies seemed to be reasonable. Since the isotropic temperature factor for Bi(6) was also significantly higher (2.707 A2) than those of the other five bismuth atoms (ave.= 1.16 A2), the occupancy and the temperature factor of Bi(6) were also refined to give values of 0.407 and 1.561 A2 respectively (R/Rw = 5.1/ 6.0). The occupancies were fixed at 0.263 (Bi(7)) and 0.407 (Bi(6)). DIFABS correction was applied to the isotropically refined data (R/Rw = 4.8/5.3).27 All atoms were refined anisotropically to give a final R/RW = 4.3 /4.7. All calculations were performed on a VAXstation 3 100/ 76 computer. As a check for Bi/K disorder, the occupancy of the K site was refined and revealed no change which suggests that the site is fully occupied with K. Structure Solution of K2Bi3813. A crystal with dimensions 0.03 x 0.05 x 0.30 mm was mounted on a glass fiber. Intensity data collection and structure solution methods are the same as above. Friedel pairs were collected due to the possibility of an acentric structure. Eight bismuth atoms, thirteen sulfur atoms and two 160 potassium atoms were located on mirror planes. After least squares refinement, the isotropic temperature factor for Bi(8) was rather high at 6.50 A2 with R/RW = 13.2/14.3. The occupancy and temperature factor of this atom were refined to give values of 0.33 1 and 3.31 A2 respectively (R/Rw = 12.5/ 13.4). The coordination environments of the other seven bismuth atoms were octahedral while that of Bi(8) was the same unusual eight-coordinate site as seen in KBi333810 consisting of bond distances ranging from 2.77 to 3.48 A. The K(1) and K(2) coordination sites were very similar to that of Bi(8). Refinement of the occupancies of K( 1) and K(2) showed large increases to 0.815 for K( 1) and 0.814 for K(2). This suggested that more electron density was required at these sites, therefore, a model with Bi and K disordered over all three sites seemed reasonable. Successive refinements taking into account this hypothesis resulted in a formula of K2,1Bi7,9813 where an additional +0.2 charge is needed for electroneutrality (R/Rw=12.3/ 13.0). The occupancies were rounded off and fixed and refinement gave K2Bi3813 with no change in the R value. The atomic composition of the Bi-rich site A was found to be 60% Bi and 40% K while sites B and C gave 20% Bi and 80% K. DIFABS correction was applied to the isotropically refined data (R/Rw=7.5/8.5).27 All atoms were refined anisotropically except Bi(8), K( 1) and K(Z) (R/Rw=7.3/ 8.2). Averaging the data did not affect the R value. K2Bi3813 is found with positional disorder between Bi and K with a bias toward K2_1Bi7,9813 suggesting a Bi3+ or K+ deficiency. Non-stoichiometry represented by K2+3xBi3. X813 could be present. 161 The complete data collection parameters, details of the structure solution and refinement and for both compounds are given in Table 2-16. The coordinates and average temperature factors (Beq) of all atoms and their estimated standard deviations for both compounds are given in Tables 2-17 and 2-18. Table 2-14. Calculated and Observed X-ray Powder Diffraction 162 Patterns for KBi6,33810. h k l dcatlcd" A (1on A I/Imgx, (0138) 2 0 l 0 0 2 1 0 2 1 0 3 2 0 3 4 0 2 0 0 4 2 0 4 3 0 4 6 0 0 6 0 1 4 0 4 1 1 2 2 0 5 2 1 2 3 0 5 5 0 4 7 0 1 2 l 3 4 0 5 7 0 2 4 1 2 2 0 6 5 1 1 2 l 4 5 0 5 5 l 2 0 6 5 l 3 4 l 4 6 1 2 5 1 4 9 0 2 4 0 7 2 1 6 6 1 4 7 l 3 1 0 8 3 0 8 0 1 7 8 1 3 10 0 3 3 1 7 7 1 5 8 l 4 1004 10.23 10.20 9.72 9.01 6.26 5.71 5.11 4.86 4.51 4.16 4.01 3.93 3.78 3.73 3.70 3.60 3.50 3.42 3.38 3.33 3.28 3.24 3.20 3.13 3.08 3.03 3.02 2.97 2.84 2.81 2.78 2.75 2.63 2.58 2.52 2.487 2.469 2.440 2.418 2.326 2.3(X) 2.271 2.255 2.211 2.181 2.170 2.156 9.71 9.0) 6.24 5.70 5.11 4.85 4.50 4.15 4.(X) 3.92 3.77 3.74 3.69 3.60 3.49 3.41 3.38 3.33 3.28 2.189 3.20 3.12 3.08 3.03 3.02 2.97 2.84 2.81 2.78 2.75 2.62 2.57 2.52 2.486 2.466 2.437 2.415 2.321 2.297 2.270 2.252 2.208 2.179 2.165 2.149 13 8 26 5 64 3 6 14 9 l3 7 22 14 16 20 1(1) 52 55 50 5 7 8 10 33 33 16 79 41 33 5 17 5 3 22 4 8 8 3 12 21 5 10 8 9 16 15 Table 2-14. (cont'd) 163 h k l dCQlCdH A dObS’ I/Imgk (ObS) 2 0 9 9 l 3 l 1 8 2 1 8 9 1 4 3 1 8 2.126 2.116 2.083 2.060 2.034 2.023 101 3, 41 8 1.976,1.975 9 1 5 5 1 8 1 1 9 9 1 6 8 l 7 4 2 4 12 1 0 12 1 1 5 1 9 1 0 11 7 2 1 6 0 10 9 1 7 2 1 10 14 0 2 4 2 6 14 0 3 4 2 7 1.940 1.918 1.905 1.842 1.827 1.802 1.801 1.791 1.776 1.763 1.753 1.749 1.743 1.738 1.692 1.665 1.661 1.591 2.123 2.112 2.081 2.055 2.029 2.021 1.971 1.935 1.916 1.904 1.839 1.824 1.807 1.799 1.791 1.771 1.760 1.754 1.747 1.742 1.736 1.688 1.667 1.662 1.591 5 5 18 33 25 12 10 17 9 5 6 10 6 9 10 15 15 16 12 8 5 10 5 5 5 Table 2-15. Calculated and Observed X-ray Powder Diffraction 164 Patterns for K2Bi3813. 11 R] (1 CM A d0b§d, A I/Imbx(0b8d) 0 0 1 1 0 0 1 0—1 0 0 2 l 0 2 2 0 1 2 0 2 1 0-3 3 0 1 3 0-2 3 0 2 1 0—4 4 0—1 3 0 3 2 0-4 4 0-2 1 1-2 0 0 5 17.80 16.82 12.27 8.90 7.84 7.58 6.09 5.61 5.33 4.76 4.73 4.31 4.10 4.06 3.95 3.81 3.62 3.56 1 0-5, 3 0 4 3.49,3.47 4 0 3 5 0 0 1 1-3 3 1 1 5 0—2 2 1 3 3 1-2 4 0 4 41 0 3 1-3 2 0-6 4 1 2 4 0 5 3 1-4 1 0-7 4 0 6 1 1 6 3 0 7 4 1-5 6 0—5 3 1 6 2 0-8 l 17 5 1-5 8 0 1 0 2 0 3 1 7 3.42 3.36 3.30 3.24 3.15 3.11 3.10 3.04 2.93 2.89 2.80 2.77 2.71 2.65 2.52 2.414 2.372 2.308 2.267 2.211 2.200 2.155 2.138 2 103 2 085 2 038 2 009 17.45 16.79 12.23 8.87 7.82 7.57 6.08 5.60 5.32 4.76 4.72 4.30 4.10 4.06 3.94 3.81 3.61 3.56 3.48 3.41 3.36 3.29 3.23 3.15 3.11 3.09 3.04 2.92 2.88 2.80 2.77 2.70 2.65 2.51 2.414 2.368 2.307 2.260 2.208 2.198 2.151 2.134 2.098 2.082 2.032 2.007 4 7 14 28 5 5 29 47 S S 5 20 9 31 6 25 8 14 100 43 17 24 16 17 8 11 15 43 44 16 8 29 5 20 10 19 8 6 15 10 9 11 10 13 15 11 165 Table 2-15. (cont'd) h k 1 d calcd, A d obsd, A I/Imgx (ObSd) 5 1-6 7 1 3 1 1 8 6 1 5 2 0 9 7 1 4 8 1-2 3 2 4 6 0 8 8 0-6 1.959 1.950 1.938 1.931 1.922 1.871 1.831 1.757 1.736 1.72 1.955 1.948 1.937 1.930 1.919 1.868 1.827 1.754 1.736 1.721 10 11 8 10 11 6 6 7 17 4 166 Table 2-16. Summary of Crystallographic Data and Structure Analysis for KBi6,33810 and KzBi3$13. I II Formula KBi6.33810 K2Bi8S 13 FW a, A b, A c, A 0:, deg (3, deg Y, deg 1682.54 2 166.82 24.05(1) 16.818(2) 4.100(2) 4.074(5) 19.44(1) 17.801(3) 90.000 90.000 90.000 90.48( 1) 90.000 90.000 2; v, A3 4; 1917 2; 1220(2) ). 0.71073 (Mo Ka) 0.71073 (Mo Ko) space group ana (No. 62) P21/m (No. 11) Dcalcs g/cm3 5.828 5.900 p, cm" 1 590 (Mo K(1) 588 (Mo K01) ZBmax, deg 50 (Mo Ka) 50 (Mo Ka) Temp., °c 23 23 Final R/Rw, % 4.3/4.7 7.3/ 8.2 Total Data Measured 20 15 Total Unique Data 2015 3 888 3 792 Data With Foz>30(Fo2) 862 1924 (averaged) No. of Variables 108 13 1 *R=2(lFol‘ch1)/21Fol RW=12W(lFol'chl)2/ZWIF012}1/2 167 Table 2-17. Fractional Atomic Coordinates and Beq Values for KBi333810 with Estimated Standard Deviations in Parentheses. atom x y z Beq,a A2 Bi(l) Bi(2) Bi(3) Bi(4) 31(5) 0.2995(1) 0.25 0.0913(1) 1.2(1) 0.17185(9) 0.75 0.1898(1) 1.3( 1) 0.4327(1) 0.75 0.0364(1) 1.5( 1) 0.0435(1) 0.25 0.2754(1) 1.6( 1) 0.1502(1) 0.25 -0.0143(1) 2.0(1) Bi(6)b 0.5090(1) 0.75 0.4015(1) 1.6( 1) Bi(7)C 0.2219(3) 0.25 0.3830(3) 4.0(3) K 5(1) 5(2) 5(3) 5(4) 5(5) 5(6) 5(7) 5(8) 5(9) 0.3705 0.25 0.2834(8) 2.6(7) 0.0878(6) 0.25 0.1390(9) 2.4(8) 0.2227(7) 0.75 0.0332(8) 2.3(8) 0.3594(6) 0.75 0.1482(7) 1.9(7) 0.5000 0.25 0.0897(8) 1.9(7) 0.2375(6) 0.25 0.2294(7) 1.7(7) 0.4732(5) 0.75 0.2716(7) 1.5(6) 0.4287(8) 0.25 0.4452(8) 3.0(8) 0.3621(6) 0.25 -0.0166(7) 1.6(7) 0.6227(6) 0.75 0.1930(8) 1.5(7) 5(10) 0.7998(7) 0.75 0.1299(7) 2.0(7) aBeq = (4/3)[a2B(1,1) + b2B(2,2) + c2B(3,3) + ab(cosY)B(1,2) + ac(cosB)B( 1,3) + bC(COSa)B(2,3). bThis site is 81% occupied. CThis site is 53% occupied. 168 Table 2-18. Fractional Atomic Coordinates and Beq Values for K2Bi3813 with Estimated Standard Deviations in Parentheses. atom x y z Becfi‘ A2 Bi(l) Bi(2) Bi(3) Bi(4) Bi(5) Bi(6) Bi(7) 0.9176(1) 1.1729(1) 0.6731(1) 0.25 0.25 0.25 0.5774(1) 127(9) 0.6245(1) 1.3(1) 0.5194(1) 1.3(1) 0.8930(1) -0.75 0.9783(1) 1.4( 1) 0.6929(1) -0.25 0.9550(1) 1.3(1) 1.0134(1) -0.25 0.7540(1) 1.4(1) 0.4917(1) -1.75 0.8849(1) 1.5(1) Bi(8)(K)b 0.2908(3) -2.25 0.8195(3) 3.5(1) K(1)(Bi)C 0.7268(7) -2.25 0.7188(8) 5.9(3) K(2)(Bi)C 0.5219(8) -1.25 0.6542(8) 6.7(3) 5(1) 5(2) 5(3) 5(4) 5(5) 5(6) 5(7) 5(8) 5(9) 5(10) 5(11) 5(12) 5(13) 0.978(1) -0.25 0.9073(9) 2.0(7) 0.8982(8) 0.25 0.7227(8) 1.2(6) 0.8082(8) -0.25 0.559(1) 1.8(7) 0.3889(9) -1.75 0.7739(9) 1.8(7) 0.807(1) -1.25 1.0574(9) 1.8(7) 0.7676(8) -0.75 0.8767(7) 1.5(6) 1.0418(8) -0.75 0.5966(8) 1.5(6) 1.131(1) 0.629(1) -0.75 -1.75 0.7654(9) 1.7(6) 0.661(1) 2.3(8) 0.5784(7) -0.25 0.4890(9) 1.5(6) 0.573(1) -1.25 0.820(1) 2.5(8) 0.7258(8) 0.610(1) 0.25 0.25 0.352(1) 1.7(7) 0.039(1) 2.7(8) aBeq = (4/3)[a2B(1,1) + b2B(2,2) + c2B(3,3) + ab(c05y)B(1,2) + ac(cosB)B(1,3) + bc(c05a)B(2,3)]. bThis site contains 60% Bi and 40% K. CThis site contains 80% K and 20% Bi. 169 3. Results and Discussion 3.1. Synthesis, Spectroscopy and Thermal Analysis The synthesis of pure KBi6,33810 (I) can be accomplished by reacting K28 and Bi283 (1:4) at 750 °C. By lowering the reaction temperature to 725 °C, pure K2Bi3813 (I I) is obtained. This result suggests that I is more thermodynamically stable than (I I). The known ternary phase, oz-KBiSz,6 was obtained as a pure phase by reacting K28 and Bi28 3 (1:1) at 725 °C. Surprisingly, instead of a-KBiS 2, B-KBi82 was observed as the major product in the reaction of K28 and Bi283 (1:2) at 725 °C along with KBi6,33810 as a minor product. Increasing the amount of Bi283 to (1:3) resulted in the formation of KBi6,33810 with B-KBiSz as the minor impurity phase. B- KBi82 is isostructural to RbBi827 in which CdCl2-type (Bi82)' layers (perpendicular to the c-axis) alternate with Rb+ ions. Three (BiS2)- layers are found in the unit cell of this compound. The coordination sphere of Bi is perfect octahedral. The results of several direct combination reactions are shown in Table 2-19. KBi6,3351(), K2Bi3513, and B-KBiSz belong to the (A2Q)n(Bi2Q3)m (A=alkali metal; Q=S, Se) general family of compounds with n=1 and m=6.33, 4, 1 respectively. The synthesis of new ternary bismuth chalcogenides with various n and m values may be possible e.g. ABism (n=1, m=5). 170 The optical properties of (I) and (I I) were assessed by studying the UV-visible-near IR reflectance spectra of the materials. The compounds absorbed light in the range of 0.5 to 6.2 eV. The spectra confirm that the band gaps, Eg, of these two compounds are less than 0.5 eV (the detection limit of the instrument). Diffuse reflectance Mid-IR spectroscopy was used to probe the small band gap of the compounds. The absorption edges in both compounds were found to be virtually identical in the range of 0.06-0.25 eV (see Figure 2-21). The small band gaps for (I) and (I I) could result from the existence of a narrow d0pant level band just below the conduction band which would also give rise to the n-type conduction behavior observed for (I) and (I I) (vide infra). The thermal behavior of (I) and (I I) was investigated with differential thermal analysis (DTA). KBi6_33810 melts congruently at 710 °C while K2Bi3813 melts incongruently to give KBi333810 as evidenced by powder XRD. Table 2-19. High Temperature K2S/ Bi28 3 Reactions. 1 7 1 K28 Bi283 Temp. (°C) Product 1 1 1 1 1 1 1 1 1 1 1 2 3 4 4 4 5 675 a-KBiSz + impurity 725 a-KBi82 725 B-KBiSz +1 (minor) 725 I + B-KBiSz (minor) 675 725 750 I I I I 725 II +B1253 6.33 725 II +Bi283 172 aitoaaozfizfiozbsotioim mien new E l _ _ o ' ) s t i n u . b r a ( e c n a t c e fl e R e s u f f i D ) s t i n u . b r a ( e c n a t c e l f e R e s u f f i D 0 o . (aatoaaozauasonIhoimaeoieo mm Figure 2-21. Mid-IR diffuse reflectance spectra of (a) K31633810 and (b) K2813813. 173 3.2. Description of Structures. Structure of KBi6,33810 (I). This compound has a three- dimensional structure made up of Bi2Te3-type (NaCl-type) blocks and Cd12-type fragments that connect to form tunnels filled with eight- coordinate K+ cations (K-Save, = 3.3(1) A). Selected bond distances and angles for (I) are given in Tables 2-20 and 2-21. Figure 2-22 shows the packing diagram of the extended structure down the b-axis. The [Bi6,33810]' framework can be thought of as a hybrid of two different layered structure types interconnected to form a 3-D network. Structural features from the Bi2Te328 and Cd1229 lattices are represented in this framework. Figures 2-23a and 2-23b show the structures of the two layered materials. The features of both structure types found in the [Bi6,33810]- framework are highlighted in Figure 2-23c. The Bi2Te3- type fragments are linked by Cdlz-type octahedra to form the channel framework. Another interesting feature found in this structure is the presence of small triangular-shaped empty channels that are lined by Bi(7)-S(8)-Bi(4)-S(4)-Bi(3)-S(9). The same type of channels have been observed in fi-BaBi284.11 The existence of Bi2Te3- and CdIz-type fragments with Bi in an octahedral coordination site is a common structural motif that runs through much of the known bismuth chalcogenide chemistry. The structure of CsBi3853 is comprised of Bi2Te3-type single chains that run along the b-direction and share corners in the a-c plane to form 174 a three-dimensional tunnel structure filled with Cs+ ions. In CS3Bi7Se12,12 the [Bi7Se12]3' anion is layered. It contains Cd12 and Bi2Te3-type fragments that connect in an edge-sharing manner to form a lamellar structure. In Sr4Bi6Se13,13 the highly charged [Bi68e13]3' anion has a very interesting structure. It contains two- dirnensional sheets made up of edge-sharing Cdl2- and Bi2Te3-type fragments. One-dimensional double chains, comprised of Bi2Te3- type blocks, extend along the b-direction and separate these layers. In BaBiSe3,14 one-dimensional single chains, comprised of Bi2Te3-type blocks, are linked by unusual (Sex)X' chains to form a layered structure. The bonding in this compound is not valence precise so delocalized arguments must be invoked. This compound has a metallic luster. Bi(3) and Bi(4) possess regular octahedral coordination with Bi- 8 bond distances ranging from 2.80(1) A to 2.94(1) A for Bi(3) and 2.812(9) A to 2.86(1) A for Bi(4). These distances are similar to those reported in CsBi385 (Bi3).8 The octahedral coordination environments in Bi(1), Bi(2), Bi(5), and Bi(6) are distorted with a short bond that is trans to a long bond but with normal octahedral angles. This type of coordination environment is very prevalent in bismuth chalcogenide chemistry and results from the influence of the non-bonded, stereochemically active 652 electron lone pair. For example, the Bi(2)-8(9) bond distance of 2.57(2) A is trans to a long Bi(2)-8(1) distance of 3.28(2) A. This same type of distorted octahedral coordination, bordering on square pyramidal coordination, is found in many compounds including CsBi385 (Bi1,2),3 T14Bi28510 and a-(B-)BaBi284.11 The sites of Bi(6) and Bi(7) are partially 175 occupied at 81% and 53%, respectively. Bi(7) possesses a distorted trigonal bipyramidal coordination (including the lone pair) with four normal Bi-S bonds ranging from 2.79( 1) to 3.21(2) A. Also, the lone pair is directed at four sulfur atoms with long distances of 3.47(1) and 3.48(1) A. The axial S(2")-Bi(7)-S(5) angle is much less than 180° at 148.3(5)° and the equatorial 8(10')-Bi(7)-S(10") angle is 94.5(5)°. These angles are influenced by the stereochemically active lone pair. Structure of K2Bi3813 (II). This compound also possesses a three-dimensional structure made up of Bi2Te3- and Cd12-type fragments that connect to form a structure with tunnels. The K+ cations are disordered with one of the Bi3+ ions over three distinct crystallographic sites. This, although not expected, can be rationalized by the similar sizes of Kr and Bi3+. This structure is quite different from that of its selenium analog K2Bi3Se13.21 Selected bond distances and angles for (I I) are given in Tables 2-22 and 2-23. Figure 2-24 shows the packing diagram of the extended structure down the b-axis. As in (I), the [B13813]2‘ framework is based on structural features from the Bi2Te3 and CdI2 lattices and contains the small triangular-shaped empty channels. The tunnels of (I I) are more open than those of I because more K28 is present per formula unit. (I I) can be viewed as a derivative of the well-known Bi283 compound generated by breaking down the Bi283 framework by incorporation of K28 in the molar ratio of 1:4 (i.e. [K28][Bi2S3]4). The corresponding [K28]/[Bi283] ratio in I is 1:6.33 so less K+ ions are present resulting in smaller channels. 176 Bi(4) and Bi(6) possess regular octahedral coordination with Bi- 8 bond distances ranging from 2.77(1) A to 2.96(2) A for Bi(4) and 2.80(2) A to 2.86(1) A for Bi(6). As in I, the octahedral coordination environments of Bi(1), Bi(2), Bi(3), Bi(5), and Bi(7) are distorted with a short bond that is trans to a long bond but with normal octahedral angles. For example, the Bi(2)-8(8) bond distance of 2.61(2) A is trans to a long Bi(2)-S(3") distance of 3.29(2) A. Bi(8) is disordered over three sites with K(1) and K(2). The Bi(8) site contains approximately 60% Bi and 40% K while the other two sites contain mainly K (~80%). Bi(8) possesses the same type of distorted trigonal bipyramidal coordination (including the lone pair) as found in Bi(7) (KBi6,33810), with four normal Bi-S bonds ranging from 2.75(1) to 3.06(2) A and four longer distances from 3.43(1) to 3.50(1) A. The two K sites have distances ranging from 2.81(1) to 3.54(2) A for K(1) and 2.72(2) to 3.66(2) A for K(2). These distances are similar to those found for the predominantly Bi(8) site. The K-8 distances below 3.0 A are unusual and presumably a result of averaging over the mixed K/ Bi sites. There are no known systems that have alkali metal/ Bi disorder so we could find no precedence in the literature. The closest example is the Cu1+3xBi5-xS315 alloy, where Cu+ and Bi3+ ions are disordered over one site. Close examination of the known ternary Bi/Q compounds reveals an interesting fact. The e.s.d.'s of the Bi-S and Bi-Se bond lengths are quite high for such structures. For example, values for I and H range from 0.006 to 0.02. Other bismuth chalcogenide crystal structures show high e.s.d.'s for these bonds as follows: CsBi385 177 (0.01),8 T14Bi285 (0.01-0.02),10 ot-(B-)BaBi284 (0.01-0.03),11 C53Bi78e12 (0.011),12 Sr4Bi68e13 (0.011),13 BaBiSe3 (0.014),14 and B-CsBiS2 (0.008-0.01).21 As the diffracting quality of these crystals seems quite good, no obvious explanation for this has been advanced. 178 . , 1 ’ 1 £ 4 9 5 9 5 ' 0 ' “ 0 w : ‘ "1 e I t ) “ 3 ? . “ ‘ 3 0 ‘ " ' v " 2 ‘ v i ‘ l “ % . ‘ é § 3 . 4 1 ‘ 2 4 . “ _ . I 4 . “ . ‘ A ‘ . t i f m ‘ 4 t “ ‘ i fl . “ , 0 / “ “ " \ W ‘ ’ 7 \ \ F I Figure 2-22. ORTEP representation of the packing diagram of KB1633810 down the b-axis with labeling. 179 (B) (C) Figure 2-23. Projections of the structures of (A).B12Te3 (B) CdIz and (C) [Bi633810]' framework. Both structure types found in this framework are designated with dashed lines. 180 Figure 2-24. ORTEP representation of the packing diagram of K2Bi3813 down the b-axis with labeling. Table 2-20. Selected Distances (A) in KBi333810 with Standard 181 Deviations in Parenthesesa Bi(1)-S(2) Bi(1)-S(2') Bi(l)-S(3) Bi(1)-S(3') Bi(1)-S(5) Bi(1)-S(8) 298(1) 298(1) 274(1) 274(1) 3.07(1) 258(1) Bi(5)-S(1) Bi(5)-S(2) Bi(5)-S(2') Bi(S)-S(7') Bi(5)-S(7") Bi(5)-S(10) 334(2) 284(1) 284(1) 290(1) 290(1) 255(1) Bi(1)-S (mean) 2.8(2) Bi(5)-S (mean) 2.9(3) Bi(2)-8(1) Bi(2)-S(l') Bi(2)-8(2) Bi(2)-8(5) Bi(2)-S(5') Bi(2)-8(9) 3.04(1) 3.04(l) 3.28(2) 270(1) 270(1) 257(2) Bi(6)-8(1) Bi(6)—S(1") Bi(6)-8(6) Bi(6)-8(7) Bi(6)-S(7'") 290(1) 290(1) 2.67(1) 294(1) 294(1) Bi(6)-S(7"") 334(2) Bi(2)-S (mean) 2.9(3) Bi(6)-S (mean) 2.9(2) Bi(3)-8(3) 280(1) Bi(7)-S(2") Bi(3)-8(4) 2.810(6) Bi(7)-8(5) Bi(3)-S(4') 2.810(6) Bi(7)-8(8) Bi(3)-S(4") Bi(3)-8(8) Bi(3)-S(8') 294(1) 285(1) 285(1) Bi(7)-S(8') Bi(7)-S(9') Bi(7)-8(9) 3.21(2) 3.01(2) 3.48(1) 3.48(1) 3.47(1) 3.47(1) Bi(3)-S (mean) 284(5) Bi(7)-S(10') 2.79(1) Bi(4)-8(1) Bi(4)-S(4'") 2.86(2) 282(1) Bi(7)-S(10") 2.79(1) Bi(7)-S (mean) 3.2(3) Bi(4)-S(6') 2.812(9) K-S(3) Bi(4)-S(6") 2.812(9) K-S(3') Bi(4)-S(9') Bi(4)-8(9) 2.86(1) 2.86(1) K-S(5) K-S(6) Bi(4)-S (mean) 284(3) K-S(6"') K-S(7) K-S(10') K-S(10") 334(2) 334(2) 3.37(2) 3.22(1) 3.22(l) 344(2) 315(1) 3.15(1) K-S (mean) 3.3(1) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations o1={2n(ln- 1)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. Table 2-21. Selected Angles (deg) in KBi6,33810 with Standard 182 Deviations in Parentheses. S(2)—Bi(1)-S(2') 86.9(4) S(2)-Bi(S)-S(2') 922(5) S(2)-Bi(1)-S(3) 88.0(4) S(2)-Bi(S)-S(7") 88.8(3) S(3)-Bi(1)-S(3') 96.9(5) S(2)-Bi(5)-S( 10) 898(4) S(3)-Bi(1)-S(8) 91.3(4) S(7')-Bi(S)-S(7") 89.9(5) S( 1 )-Bi(2)-S( 1') 84.7(4) S( 1 )—Bi(6)-S( 1") 89.9(5) S( 1 )-Bi(2)-S(S) 883(3) S( l )-Bi(6)-S(6) 873(4) 8( l )-Bi(2)-S(9) 89.0(4) S(6)-Bi(6)-S(7) 93.5(4) S(S)-Bi(2)-S(9) 90.9(4) S(7)-Bi(6)-S(7"') 88.4(5) S(3)-Bi(3)-S(4) 94.4(3) S(S)-Bi(7)-S( 10') 80.0(4) S(3)-Bi(3)-S(8) 84.6(4) S(S)-Bi(7)-S(10") 80.0(4) S(4)-Bi(3)-S(4') 93.7(3) S(lO')-Bi(7)-S(10") 94.6(5) S(4)-Bi(3)-S(8) 87.2(2) S(2")-Bi(7)-S(5) 148.3(5) S(l)-Bi(4)-S(6') 855(4) 8( 1)-Bi(4)-S(9) 87.2(4) S(4)-Bi(4)-S(6') 94.5(3) S(4)-Bi(4)-S(9) 927(3) Table 222 Selected Distances (A) in K2Bi3813 with Standard 183 Deviations in Parentheses.a Bi(1)-S(mean) 29(2) Bi(7)-S(11) Bi(1)-S(2) Bi(1)-S(3) Bi(1)-S(3') Bi(1)-S(7) Bi(1)-S(7') Bi(1)-S(7") 2.611(1) 2.764(9) 2.764(9) 293(1) 293(1) 3.l8(2) Bi(2)-S(3") Bi(2)-8(7) Bi(2)-S(7') Bi(2)-8(8) Bi(2)-S(12') Bi(2)-S(12") 3.29(2) 3.04(1) 3.04(1) 261(2) 2.688(9) 2.688(9) Bi(6)-8(7) Bi(6)-8(8) Bi(6)-S(8') 284(2) 284(1) 284(1) Bi(6)-S(mean) 284(2) Bi(7)-8(4) Bi(7)-S(11'") Bi(7)-S(13") Bi(7)-S(13'") Bi(7)-S(13"") Bi(8)-8(4) Bi(7)-S(mean) 29(3) 261(1) 271(1) 271(1) 338(2) 299(1) 299(1) 275(1) 275(1) 3.43(1) 3.43(1) 3.50(1) 3.50(1) 306(2) 3.01(2) 3.71(1) 3.71(1) 272(1) 272(1) 3.10(2) 3.66(1) 3.66( 1) 307(2) Bi(2)-S(mean) 29(3) Bi(8)-S(4‘) Bi(3)-8(3) Bi(3)-S(3') Bi(3)-S(9') Bi(3)-S( 10) Bi(3)-S(10') 3.13(1) 3.13(1) 263(2) 2.639(8) 2.639(8) Bi(8)-S(5") Bi(8)-S(5'") Bi(8)-S(8") Bi(8)-S(8'") Bi(8)-S( 12'") Bi(8)-S(13'") Bi(3)-S( 12) 3.12(2) Bi(8)-S(mean) 3.2(3) Bi(3)-S(mean) 29(3) Bi(4)-S( 1) Bi(4)-S(1') Bi(4)-S(1") Bi(4)-8(5) Bi(4)-S(5') Bi(4)-8(6) 280( 1) 280(1) 296(2) 287(1) 287(1) 2.77(1) K(1)-S(2") K(1)-S(2'") K(1)-S(3") K(1)-S(6") K(1)-S(6'") K(1)-8(9) K(1)-S(9") Bi(4)-S(mean) 285(7) K(1)-S(11'") 353(1) 333(1) 3.16(2) 3.54(2) 3.54(2) 281(1) 281(1) 3.17(2) K(1)-S(mean) 33(3) Bi(5)-S(5') Bi(5)-S(6) Bi(5)-S(6') Bi(5)-S(11") Bi(5)-S(13) Bi(S)-S(13') 264(2) 278(1) 278(1) 3.12(2) 289(1) 289(1) K(2)-S(4) K(2)-S(4") K(2)-S(9) K(2)-S(9'") K(2)-S(10") Bi( 5 )-S(mean) 29(2) K(2)-S(10'") Bi(6)-S( 1) Bi(6)-8(2) Bi(6)-S(2') 280(2) 2.86(1) 2.86(1) K(2)—S(10"") K(2)-S(11) K(2)-S(mean) 33(4) aol=i2n(1n-l)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 184 Table 223. Selected Angles (deg) in K2Bi3813 with Standard Deviations in Parentheses. S(2)-Bi(1)-S(3) 91.6(4) S(S')-Bi(5)-S(6) 91.0(4) S(2)-Bi( 1 )-S(7) 88.8(4) S(S')-Bi(S)-S( 13) 89.6(4) S(3)-Bi(1)-S(3 ') 95.0(4) S(6)-Bi( 5)-S(6') 94.4(4) S(3)-Bi(1)-S(7) 88.6(3) S(13)-Bi(5)-S(13') 89.6(5) S(3")-Bi(2)-S(7) 85.1(3) S( 1)-Bi(6)-S(8) 94.7(4) S(3")-Bi(2)—S(12) 95.2(4) S(2)-Bi(6)-S(7) 85.8(3) S(7)-Bi(2)-S(8) 87.4(4) S(2)-Bi(6)-S(8') 88.5(3) S(8)-Bi(2)-S(12') 91.4(4) S(7)-Bi(6)-S(8) 87.0(4) S(3)-Bi(3)-S(3') 81.3(3) S(4)-Bi(7)-S(11) 90.7(4) S(3)-Bi(3)-S(9) 89.7(4) S(4)-Bi(7)-S(13'") 88.0(4) S(9')-Bi(3)-S(10) 91.2(4) S(l 1)-Bi(7)-S(13"") 88.4(4) S(10)-Bi(3)-S(12) 88.8(4) S(13")-Bi(7)-S(13'") 88.2(4) S(l)-Bi(4)-S(1 ') 93.4(5) S(4)-Bi( 8)-S(4') 95.7(5) S( 1)-Bi(4)-S(6) 95.5(4) S(4)-Bi(8)-S(5") 157.4(4) S(5)-Bi(4)-S(5') 90.4(4) S(4)-Bi(8)-S(12"') 75.9(4) S(S)-Bi(4)-S(6) 86.4(4) S(S'")-Bi(8)-S(12) 126.7(3) 185 3.3. Electrical Conductivity and Thermoelectric Power Measurements. Four-probe electrical conductivity measurements on polycrystalline chunks of KBi6,33810 showed that the material is a semiconductor with room temperature conductivity o~102 S/cm which drops to 10'4 S/cm at 5 K. Figure 2-25a shows the log conductivity vs. temperature plot for KBi6,33810. The data can be fit to the equation shown below suggesting an activation energy of Ea=0.045 eV. E, _ k T O—Ooe B eq. (2) Where o=electrical conductivity (S/ cm), Ea=activation energy, kB=Boltzmann constant, and T=temperature (K). Figure 2-25b shows the log conductivity as a function of temperature for K2Bi3$13. The conductivity of ~102 S/cm (at 300 K) and the weak temperature dependence between 5-300 K coupled with the IR-optical data (vide supra) suggests that this compound is a semi-metal or a narrow band gap semiconductor. The conductivity of ~102 S/ cm is compared to the values of other semi-metals such as BizTe3 (2.2 x 103 S/cm)3O and Bi28e3 (1.6-2.0 x 103 S/cm, Bi2Te3-type)31 obtained from single crystals of these materials. The conductivity measurements alone cannot unequivocally characterize the electrical behavior of (I) and (I I). A complementary probe to address this issue is thermoelectric power (TP) 186 measurements as a function of temperature. TP measurements are typically far less susceptible to artifacts arising from resistive domain boundaries in the material because they are essentially zero- current measurements. This is because temperature drops across such boundaries are much less significant than voltage drops. Figure 2-26a shows typical TP data of KBi6,33810 as a function of temperature. The TP is negative throughout the temperature range studied (80 30(Foz) 1889 No. of Variables 65 2826 2609 1926 74 Crystal Dimensions 0.20x0.50x0.70 mm 0.12x0.20x0.30 mm * R=Z(|Fol-|Fcl)/ZIFOI Rw={2‘.w( IFol-IFCI)Z/ZWIFOIZ}1/2 Table 3-4. Fractional Atomic Coordinates and Beq Values for C528b483 with Estimated Standard Deviations in Parentheses. 200 atom x y z B(eq) 03(1) Sb(1) Sb(2) 5(1) 5(2) 5(3) 5(4) 0.16496(7) 0.06678(5) 1.32325(8) 1.83(2) -0.18018(7) 0.45220(5) 0.66230(8) 151(2) 0.40888(7) 0.32190(5) 0.92176(8) 142(2) -0.1566(3) 0.4058(2) 0.2934(3) 153(7) 0.5781(3) 0.0751(2) 1.0945(3) 1.65(7) 0.0745(3) 0.2366(2) 0.8152(3) 1.49(6) 0.5233(3) 0.2837(2) 0.5893(3) 1.73(7) Table 3-5 . Fractional Atomic Coordinates and Beq Values for CsSbS6 with Estimated Standard Deviations in Parentheses. atom x j z B(eq) Cs Sb -0.0826( 1) 0.40998(4) 0.18541(5) 2.82(2) 0.02834(8) 0.09589(4) 0.11379(4) 1.62(2) S( 1) -0.0049(4) 0.0694(1) -0.1242( 2) 205(7) S(Z) 0.1085(4) 0.0673(2) 0.3633(2) 2.37(7) S(3) 0.0089(3) 0.1923(2) 0.4244(2) 2.28(7) 8(4) —0.3439(4) 0.1818(2) 0.4124(2) 2.37(8) S(5) 0.5085(3) 0.2270(1) 0.2373(2) 2.10(7) S(6) 0.4558(3) 0.1142(1) 0.1209(2) 1.90(7) ameq = (4/3)[a2B(1,1) + b2B(2,2) + c28(3,3) + ab(cosv)B(1,2) + ac(cosp)8(1,3) + bc(cosa)B(2,3)]. 201 3. Results and Discussion 3.1. Synthesis and Spectroscopy The syntheses of C525b483 and CsSbS6 were a result of a redox reaction in which the Sb metal is oxidized by polysulfide ions to give Sb3+ species. CsSb56 was formed from a Sb/CSZS/S ratio of 1/ 1/ 9 at 260°C while C525b483 can be obtained at 280°C with a ratio of 1/0.9- 1.1/8. C525b483 appears to be more thermodynamically stable than CsSb86 under similar acidic flux conditions. This is supported by thermal analysis data which reveal that CsSbS6 decomposes to C525b453 upon heating (vide infra). The IR spectroscopic data for CsSbS6 show absorptions at 484, 473 and 448 cm'l, which we assign to 5-5 stretching vibrations. Similar absorbances were not observed in C528b453 because the S-S bond resides on a center of symmetry. However, the S-S stretching mode is Raman-active and occurs at 474 cm-1. The absorbances found at lower energy for CsSbS6 (365-195 cm-l) and C525b453 (341- 23 3 cm- 1) have been tenatively assigned as Sb-S vibrations. The optical properties of (I) and (I I) were assessed by studying the UV/near-IR reflectance spectra of the materials. The spectra confirm the semiconductor nature of the materials by revealing the presence of sharp optical gaps as shown in Figure 3-1. Both compounds exhibit steep absorption edges from which the band-gap, Eg, can be assessed at 2.05 and 2.25 eV respectively. By comparison, the corresponding band-gap of Sb283 is 1.7 eV.22 A linear relationship is found in both compounds if the square of the 202 absorption coefficient, (a/S)2 is plotted vs. (E'Eg) suggesting that the band gaps are direct in nature.23 However, independent confirmation is needed to verify this conclusion. The relatively good air and moisture stability of CsSb284 and its excellent optical transparency below 2.05 eV make it a good candidate for exploration of its linear and non-linear optical properties. The optical absorption spectrum of (I I) also features absorbances at higher energy which are due to other electronic transitions in the solid, perhaps within the 852' fragment. 203 , ) * 1 1 b l 3 7 2 “ ) s t i n u . b r a ( . f f e o C n o i t p r o s b A S / a 5,-2.05w _ - ~ 1 '1 r- E"‘3183'tev) A 6 L fl"; L 5 a ' 3 8 . I ' 5 ‘1 4 - ”8' o 3 ~ g .9. E." 2 2 3 a: 2.92 A. 3.36 . i 4 73 6.26 , g i I CsSbSo / E.-2.2sev I f I 1 r" ‘ — z L ' I" . - < 1 4 , 0 -——-$-='—J 1 I n a 0 1 2 3 4 5 6 7 Energy teV) Figure 3-1. (A) Optical absorption spectrum of C525b453. (B) Optical absorption spectrum of CsSbSe. 204 3.2 Description of Structures Structure of C528b483 (I). (I) is isostructural to the hydrothermally prepared, C528b48e3.16 At the most fundamental level this structure is made up of condenced 8b833' pyramids and 822- units. Although, overall this structure is three-dimensional, it is assembled by cross-linking one-dimensional [8b486] parallel chains with disulfide ligands to form sheets that are parallel to the (011) crystallographic plane. Figure 3-2 shows a view of an individual layer of [8b48 3]nn -. The layer can alternatively be described as being composed of 14-membered rings of alternating Sb-S atoms and 82 units. The rings are connected in two dimensions by Sb-S-Sb linkages to form the layer. The layers stack one on top of the other with their 14-membered rings in registry so that they form channels running down the c-axis. The channels are clearly visible in Figure 3-3A. Long Sb-S interactions (3.456(2) A) between sheets give rise to a second set of channels running down the a-axis that are filled with ten-coordinate Cs+ ions (Cs-S mean=3.62(4) A) (Figure 3-3B). Due to these long Sb-S interactions between the sheets, the structure possesses pseudo three-dimensional character. Selected bond distances and bond angles for (I) are given in Table 3-6. Sb( 1) is found in a distorted square pyramidal coordination due to the stereochernically active lone pair at the base of the pyramid (Figure 3-3 B). This type of coordination is found in 8b283 (stibnite).24 The four Sb(1)-8 bonds range from 2.418(2) to 2.783(2) A with a fifth coordination site occupied by a weakly interacting 8(4) that has a distance of 3.277(2) A (Figure 3-3B). This distance is 205 considerably shorter than the sum of the Sb—S Van der Waals radii (4.05A).25 These distances, ranging from the Sb-S bond at 2.418(2) A to the interaction at 3.277(2) A, are very comparable to those found in other ternary antimony (IH) sulfides, such as C538b5895 and B-szSb4S7.3 The coordination environment of Sb(2) could be described as a trigonal pyramid with three bonds ranging from 2.440(2) to 2.592(2) A (Figure 3-4B). The Sb(2) is also interacting with two sulfur atoms along the top of the trigonal pyramid. A strong interaction is found at 3.022(2) A with 8(1) and a weak interaction is observed with S( 1") at 3.456(2). This type of coordination is observed in 8b283 (8b 1)24 and CsSsz.ld Structure of CsSb86 (II). CsSb86 is the first example of a solid state antimony polysulfide compound. The only other known antimony polysulfide is the molecular dimeric antimony polysulfide complex [8b2815]2' that has been recently prepared by Rauchfuss et al.26 The novel one-dimensional structure of (I I) is composed of szsz rhombi that are linked along the a-axis by pentasulfide ligands to form chains (Figure 3-4A). The formula can alternatively be expressed as CsSb8(85). These chains are separated by nine- coordinate Cs+ ions (mean Cs-S distance = 3.69(3) A). Figure 3-4B shows the packing diagram of CsSb86, along the a-axis (chain axis). Selected bond distances and bond angles for (I) are given in Table 3- 7. The 8b is bonded to four sulfur atoms With bond distances ranging from 2.390(2) to 2.737(2) A. The coordination around the 8b atom (including the lone pair) is trigonal bipyrarnidal with the lone 206 pair at the equatorial position. The repulsion by the lone pair gives rise to a nearly axial 8(2)-8b-8(1) angle of 162.10(6)° and an 8(1')- 8b-8(6) angle of 99.81(6)° in the trigonal plane. The two axial 8b-8 bonds (2.635(2) A, 2.737(2) A) are longer than the two equatorial bonds (2.390(2) A, 2.5 15(2) A). This combination of 8b coordination and variation in bond length has been seen in Sb(3) of C528b4877-7 where the axial bond lenths of 2.637(4) A and 2.739(5) A and equatorial bond lengths of 2.456(5) A and 2.500(5) A were observed. Several other phases have demonstrated these characteristics including B-szSb4S7 (Sb 1,2)3, RbSsz (Sb 1,2)1C, Ser4S7-6H20 (Sb 4) 12, and Cssz3513 (Sb 4).28 The bridging mode of the 112-8 5 ligand has been observed in the one-dimensional solid state compound, KAu8529 and the basket-like dimer, [M02(NO)2(82)3(85)(OH)]3-.30 Normal S-S bond distances are observed in the 852- ligand (2.058(5)A mean). Solid state compounds containing sz' ligands with x > 3 are rare. Some examples include NH4CuS431, afi-KCu8432, CsCu8633, (NH4)2Pd81134,K2Pd8e1035, (Ph4P)Agse436. (Me4N)AgSes36, KAu36537. A K3Au8e1337 and K28n283.33 207 Figure 3-2. Two-dimensional structure of the [8b483]nn‘ anionic framework with labeling as drawn by ORTEP. The dashed lines represent Sb-S ion interactions in the range of 3.022(2) to 3.277(2) . The shaded area outlines the 14-membered Sb/S rings in the structure. 208 (A) (B) Figure 3-3. (A) Stereoview of C528b483 viewed down the [001] direction. (B) Packing diagram of C528b483 showing a projection in the [100] direction. The dashed lines represent long Sb-S interactions underscoring the pseudo-3D character of the structure. The shaded ellipsoids are Sb atoms and the open ellipsoids are Cs atoms. 209 (A) (B) Figure 3-4. (A) Structure of a single (8b86)nn' chain with labeling. (B) Packing diagram of C58b86 looking down the a-axis (chain axis). The shaded ellipsoids are 8b atoms and the open ellipsoids are Cs atoms. Table 36. Selected Distances (A) and Angles (0) in C528b483 with 210 Standard Deviations in Parentheses.a Sb(1)-S(1) 2.418(2) S(l')-Sb(1)-S(3) 87.79(6) Sb( 1)-S( 1') 2.783(2) S( 1')-Sb( l)-S(4) l71.50(6) Sb(1)-S(3) 2.501(2) S(3)-Sb(1)-S(4') 173.65(6) Sb( 1 )-S(4) 2.712(2) S(1)-Sb(1)-Sl') 8647(6) Sb(1)-S(4’) 3.277(2) S(1)-Sb(1)-S(3) 9385(7) Sb(2)-S( 1) 3.022(2) S( l')-Sb(2)-S( 1") 107.81(4) Sb(2)-S( 1") 3.456(2) S( 1')-Sb(2)-S(2) l72.22(5) Sb(2)-8(2) 2.592(2) S( 1')-Sb(2)-S(3) 8355(6) Sb(2)-S( 3) 2.449(2) S( 1')-Sb(2)-S(4') 85.67(6) Sb(2)-S(4) 2.430(2) S(2')-Sb(2)-S(2) 7918(5) S(2')-Sb(2)-S(3) 154.03(6) S(2)-S(2') 2.092(4) Sb(1)-Sb(1') 3.798(1) S(2')-Sb(2)-S(4') 103.23(6) Sb(1)-Sb(2) 3.898(1) S(2)-Sb(2)-S(4') 96.18(7) Sb(2)-Sb(2') 3.833(1) S(3)-Sb(2)-S(4') 100.83(7) Cs-Cs' 3.913(1) Cs-S( 1) Cs-S(2) 3.576(2) Cs-S(3) 3.527(2) 3.470(2) Cs-S(3') 3.632(2) Cs-S(2') 3.839(2) Cs-S(3") 3.534(2) Cs-S(2") 3.648(2) Cs-S(4") 3.518(2) Cs-S(2'”) 3.850(2) Cs-S(4"') 3.632(2) Ave. Cs—S 3.62(4) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations ol={2n(1n- DZ/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. Table 3-7. Selected Distances (A) and Angles (0) for CsSbS6 with Standard Deviations in Parentheses.a 211 Sb-S(6) 2.515(2) S(2)-S(3) 2.052(3) Sb-S(2) 2.737(2) S(3)-S(4) 2.060(3) Sb-S(1) 2.635(2) S(4)-S(S) 2.063(3) Sb-S(l') 2.390(2) S(S)-S(6) 2.057(2) Ave. Sb-S 2.57(8) Ave. S-S 2.058(2) Sb-Sb' 3.711(2) Cs-S( 1) Cs—S(6) 3.582(2) Cs-S( 5) 3.693(2) 3.629(2) C5-S(2) 3.718(2) Cs—S(1') 3.630(2) Cs-S(2') 3.786(2) Cs-S(3) Cs-S(2) 3.676(2) Cs-S(4) 3.822(2) 3.679(2) Ave. Cs-S 3.69(3) S(6)-Sb-S(2) 90.87(6) Sb-S(6)-S(5) 97.23(8) S(6)-Sb-S( 1) 8531(6) Sb-S(2)-S(3) 101.41(9) S(6)-Sb—S( 1') 99.8l(6) S(2)-S( 3 )-S(4) 105.4(1) S( 2)-Sb-S( 1) 162.lO(6) Sb-S( l)-Sb 9508(6) S(2)-Sb-S( 1') 7847(6) S(6)-S( 5 )-S(4) 108.7(1) S(1)-Sb-S(1 ') 84.91(6) S(3)-S(4)-S(5) 105.3(1) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01={2n(1n- Dz/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 212 3 .3 Thermal Analysis The thermal behavior of (I) and (I I) was investigated with differential thermal analysis (DTA). Figure 3-5A shows that upon heating compound I, a sharp melting endotherm occurs at 365 °C. Upon cooling, no corresponding exothermic peak is observed, indicating that no crystallization takes place. Red glassy material was observed in the quartz ampule. Powder X-ray diffraction (XRD) of the sample at room temperature gave an amorphous pattern which is characteristic of a glassy material. Upon subsequent reheating a broad exotherm at 291 °C was observed, followed by two endotherms at 364 and 375 °C (see Figure 3-5B). We speculate that occasionally C528b483 does not fully convert to glass due to small but significant differences in thermal conditions from run to run. XRD studies show that the exotherm at 291 °C is the result of crystallization of C528b48723’39 from the glassy matrix (see Figure 3- 6). Thus, upon heating, C528b483 transforms to Cssz487 via an intermediate glassy state. The small endotherms at 364 and 375 °C are probably due to melting of excess glassy material and perhaps Cssz4S7. Figure 3-7 shows that upon heating compound (I I), a sharp endotherm occurs at 232 °C. Upon cooling, no exothermic peak is observed indicating the absence of crystallization. XRD of the sample at room temperature gave C528b483 and C528. The expelled sulfur was observed to deposit on the opposite end of the ampule from the sample. The endothermic peak is due to the decomposition of CsSbS6 2 1 3 to Cssz4S 3. 214 (A) i n 0-1' / . 51 endo 2 -1 O<% . 0 100 200 300 400 500 TinILCWD (B) 2!) cxo 1 5 - 291 /f\\ 1 0 ‘1 mx ,— L - - _ P d -375 364 _ - V > . 1 0 ~ 5“ Jl' :ndo - 5 0 100 200 300 400 500 Temp. 1°C) Figure 3-5. (A) Differential thermal analysis (DTA) data for C528b483 showing the first heating and cooling cycle. Heat is absorbed at 365 °C as the material melts but is not released upon cooling. (B) Data for reheating showing the crystallization of C528b487 at 291 °C, followed by melting of the excess glass at 364 and 375 °C. 215 Figure 3—6: X-ray diffraction patterns of: (A) C528b483 single crystals. (B) material heated to 400 °C then cooled (glassy state). (C) material after reheating to 310 °C (crystallization of C528b487). 216 ) A ( ) B ( Intensity ) C ( W ' U ' U ' U I U ' U I U I V I ' I V ' I I U ' V W I U ' V V I U I U I U I ' W ' U I V I U I V I U I U ' U l U ' t ' V ' V ' V ' I V W T I V ' U I U ' V ' U " l ' l U ' U ' U ' U ' U I U I U I U I U I U I U ' I ' I [ . 0 6 . 5 5 . 0 5 . 5 4 . 0 4 . 5 3 . 0 3 . 5 2 . 0 2 . 5 1 . 0 1 . 5 0 2 217 -4- -6n _ .. i} v -8— endO-to— -12 . 232 ._ I l l l .. O 50 100 150 200 .250 300 Temp.(°C) Figure 3-7. DTA of CsSb86 shows the endothermic peak at 232 °C corresponding to the decomposition of C58b86 to C525b4s8- 218 3.4. Conclusion In summary, we have synthesized the first two examples of polysulfide ligands in solid state antimony compounds using C528x fluxes at low temperature (< 300 °C). (1) and (I I) possess abrupt absorption edges with bandgaps, E3, of 2.05 and 2.25 eV respectively. (I I) decomposes to (I) at 232 °C and (1) forms a glass at 365 °C which transforms to C528b487 upon reheating. It is remarkable that two new phases could be uncovered in a somewhat crowded antimony/ sulfide field. Since it appears that ternary antimony (III) sulfides can form many different structure types, we are expanding our effort in the search for new ternary antimony (III) chalcogenides while focusing on relatively low reaction temperatures. The participation of polychalcogenide ligands in solid state frameworks promises to yield new solids with relative open or flexible structures. This work provides concrete evidence that temperatures in combination with alkali polychalcogenide fluxes are excellent reaction media for exploration of new metastable 1 compounds. 219 List of References (a) Olivier-Fourcade, E.; Phillippot, J.; M. Maurin Z. Anorg. Allg. Chem., 1978, 446, 159-168. (b) Graf, H. A.; Schafer, H. Z. Anorg. AIIg. Chem., 1975, 414, 211-219. (c) Kanishcheva, A. 8.; Kuznetsov, V. G.; Lazarev, V. B.; Tarasova, T. G. Zh. Strukt. Khim., 1977, 18, 1069. (d) Kanishcheva, A. 8.; Mikhailov, Y. N.; Kuznetsov, V. G.; Batog, V. N. Dok1.Akad. Nauk SSSR , 1980, 251, 603-605. Cordier , G.; Schafer, H. Revue de Chemie Minerale, 1981, 18, 218-223. Dorrscheidt, W.; Schafer, H. Z. Naturforsch., 1981, 36b, 410-414. Cordier, G.; Schwidetzky, C.; Schafer, H. Revue de Chemie Minerale, 1982, 19,179-186. Parise, J. B.; K0, Y. Chem. Mater. 1992, 4,1446-1450. Sheldrick, W. 8.; Hausler, H.-J. Z. Anorg. Allg. Chem. 1988, 561,149-156. Dittmar, V. G.; Schafer, H. Z. Anorg. Allg. Chem. 1978, 441, 93-97. Sheldrick, W. 8.; Hausler, H.-J. Z. Anorg. Allg. Chem. 1988, 557,105-111. Parise, J. B. J. Chem. Soc., Chem. Commun., 1990, 22,1553-1554. Eisenmann, B.; Schafer, H. Z. Naturforsch., 1979, 34b,383-38S. Graf, H. A.; Schafer, H. Z. Naturforsch. 1972, 27b, 735-739. Gostojic, M.; Nowacki, W.; Engle, P. Z. Kn’stallogr. 1982, 159,217-224. Dittmar, V. G.; Schafer, H. Z. Anorg. AHg. Chem., 1977, 437, 183-187. Cordier, G.; Schafer, H.; Schwidetzky, C. Z. Anorg. AHg. Chem., 1984, 39b, 131-134. (a) Kanatzidis, M. G. Chem. Mater. 1990, 2, 353-363. (b) Kanatzidis, M. G.; Park, Y. J. Am. Chem. Soc. 1990, 111, 3767-3769. (c) Kanatzidis, M. G.; Park, Y. Chem. Mater. 1990, 2, 99-101. ((1) Park, Y.; Kanatzidis, M. G. Angew. Chem. Int. Ed. Engl. 1990, 29,914-915. (e) Sunshine, 8. A.; Kang, D.; Ibers, J. A. J. Am. Chem. Soc. 1987, 109, 6202-6204. (f) Kang, D.; Ibers, J. A. Inorg. Chem. 1989, 27, 549-551. 16. 17. 18. Sheldrick, W. 8.; Kaub, J. Z. Anorg. Allg. Chem. 1986, 536,114-118. Feher, F. Handbuch der Praparativen Anorganischen Chemie: G. Brauer; Ed.; Ferdinand Enke: Stuttgart, Germany, 1954. (a) Wendlandt, W. W.; Hecht, H. G. "Reflectance Spectroscopy", Interscience Publishers, 1966 (b) Kotiim, G. "Reflectance Spectroscopy", Springer Verlag, New York, 1969 (c) Tandon, S. P.; Gupta, J. P. Phys. Stat. Sol. 1970, 38,363-367. 19. Sheldrick, G. M. In Crystallographic Computing 3; Sheldrick, G. M., Kruger, C., Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985; pp 175-189. TEXSAN: Single Crystal Structure Analysis Software, Version 5.0, (1981). Molecular Structure Corportion, The Woodlands, TX 77381. Walker, N.; Stuart, D. Acta Crystallogr., 1983, A39,158-166. Bube, R. H. "Photoconductivity of Solids", John Wiley and Sons, Inc., 1960; PP. 233-235. Pankove, J. I. in "Optical Processes in Semiconductors" Dover Publications, New York, 1975. Hofmann, W. Z. Kn'stailogr. 1933, 86,225. Pauling, L. The Nature of the Chemical Bond, 3rd Edition, New York: Cornell University Press, 1966, p. 260. Partha, P. P.; Rauchfuss, T. B.; Wilson, 8. R. J. Am. Chem. Soc. 1993, 115, 3316-3317. Dittmar, V. G.; Schafer, H. Z. Anorg. AUg. Chem. 1978, 441 ,98-102. Volk, K.; Schafer, H. Z. Naturforsch. 1979, 34b,1637-l640. 20. 21. 22. 23. 24. 25. 26. 27. 28. 220 29. 30. 31. 33. 34. 35. 37. 38. 39. Park, Y.; Kanatzidis, M. G. To be submitted for publication. Muller, A.; Eltzner, W.; Bogge, H.; Krickemeyer, E. Angew. Chem. Int. Ed. Eng]. 1983, 22,884-885. Burschka, C. Z. Naturforsch. 1980, 35B,1511-1513. Kanatzidis, M. G.; Park Y. J. Am. Chem. Soc. 1989, 111,3767-3769. McCarthy, T. J.; Zhang X.; Kanatzidis, M. G. Inorg. Chem., 1993, 32, 2944- 2948. Haradem, P. 8.; Cronin, J. L.; Krause, R. A.; Katz, L. Inorg. Chim. Acta 1977, 25,173-179. Kim, K.-W.; Kanatzidis, M. G. J. Am. Chem. Soc. 1992, 114,4878-4883. Huang, S.-P., Kanatzidis, M. G. Inorg. Chem. 1991, 30,1455-1466. Park, Y.; Kanatzidis, M. G. Angew. Chem. Int. Ed. Engl. 1990, 29,914-915. Liao, J.-H.; Varotsis, C.; Kanatzidis, M. G. Inorg. Chem., 1993, 32, 2453- 2462 Indexed C528b487 using "Powder Diffraction File-Inorganic Phases", vol. 32, pg. 214, (1987). Five additional low intensity peaks were also observed which could be due to the presence of a small amount of impurity. CHAPTER 4 New Quaternary Bismuth and Antimony Thiophosphates ABiP287 (A=K,Rb), A3M(PS4)2 (A=K,Rb,Cs; M=Sb,Bi), CS3Bi2(PS4)3, and Nao,16Bi1,23P286 in Molten Alkali Metal Polythiophosphate Fluxes. 1 . Introduction Recent advances in the development of alkali polychalcogenide fluxes as reaction media at intermediate temperatures (200 300°C We view the Ax[PySZ] fluxes as a significant variation over the A28x fluxes in that they provide not just P atoms but excess [PxSy]n- anions which act as rnineralizers. In other words, the acid/ base characteristics of the Ax[PySZ] fluxes are very different from those of the A28x fluxes in that they tend to be more basic.1 The chemical properties of these melts can be controlled by the ratios of their constituent elements. The relatively good solubility properties of Ax[Py8fl salts in water and organic solvents allow for easy isolation 224 of products. 2. Experimental Section 2.1. Reagents Chemicals in this work were used as obtained: (1) antimony powder, 99.999+% purity, -200 mesh, Cerac Inc., Milwaukee, WI. (ii) bismuth powder, 99.99996 purity, -100+200 mesh, Johnson Matthey/AESAR Group, Seabrook, NH; (iii) potassium metal, Aldrich Chemical Co., Inc., Milwaukee, WI; (iv) rubidium metal and cesium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (v) phosphorous pentasulfide, 99% purity, Aldrich Chemical Co., Inc., Milwaukee, WI; (vi) sulfur powder, sublimed, J. T. Baker Chemical Co., Phillipsburg, NJ. (vii) methanol, anhydrous, Mallinckrodt Inc., Paris, KY. (viii) diethyl ether, AC8 anhydrous, EM Science, Inc., Gibbstown, NJ. 2.2. Synthesis All manipulations were carried out under a dry nitrogen atomosphere in a Vacuum Atmospheres Dri-lab glovebox. For the preparation of K28, Rb28, and C528 we used a modified literature procedure.15 The preparation of K28 is reported in Section 2.2 of Chapter 2 (Part 2) and that of C528 is reported in Section 2.2 of Chapter 2 (Part 1). The preparation of Rb28 is the same as that of C825. 225 Na28. An amount of 2.440 g (0.076 mol) sulfur and 3.50 g (0.152 mol) freshly sliced sodium metal were added to a 250 ml round-bottom flask. A 150 n11 volume of liquid ammonia was condensed into the flask at -78 °C (dry ice/ acetone) under nitrogen to give a light blue solution. The NH3 was removed by evaporation under a flow of nitrogen as the bath slowly warmed to room temperature. The pale yellow solid (98% yield) was dried under vacuum overnight, flame-dried, and ground to a fine powder with a mortar and pestle in the glovebox. KBiP287 (I). An amount of 0.031 g (0.15 mmol) Bi, 0.100 g (0.450 mmol) P285, 0.033 g (0.30 mmol) K28, and 0.019 g (0.60 mol) 8 were thoroughly mixed and transferred to a 61111 pyrex tube which was subsequently flamed sealed in vacuo (~10;3 torr). The reaction was heated to 400 °C over 12 hours in a computer- controlled furnace. It was kept at 400 °C for 4 days, followed by cooling to 110 °C at a rate of 4 °C/ hr, then to room temperature in one hour. The excess yellow-colored KxPySz matrix was removed from the red crystals (68% yield) with dimethylformamide (DMF). The crystals are air and water stable. Semi-quantitative microprobe analysis on red crystals gave K1,0Bi1,4P2,4893 (average of three data acquisitions). Far-IR (CsI matrix) show absorptions at 600(5), 583(5), 576(ssh), 564(msh), 557(msh), 526(wsh), 464(vs), 412(w), 300(vw), 270(wsh), 250(msh), 247(m), 238(msh), 227(wsh), 208(wsh), 203(wsh), 192(vw), 174(vw), 169(vw), 157(vw), and 149(vw) cm'1. RbBiP287 (II). An amount of 0.031 g (0.15 mmol) Bi, 0.100 g (0.450 mmol) P285, 0.061 g (0.30 mmol) Rb28, and 0.029 g (0.90 226 , mol) 8 was mixed and heated as above except at 420 °C. Isolation as above gave 0.050 g of red crystals (57% yield, based on Bi). The compound was found to be isostructural to KBiP287 by powder X-ray diffraction. Semi-quantitative microprobe analysis on red crystals gave Rb1,0Bi13P1,987,4 (average of three data acquisitions). K3Sb(PS4)2 (III). An amount of 0.018 g (0.15 mmol) Sb, 0.067 g (0.30 mmol) P285, 0.033 g (0.30 mmol) K28, and 0.029 g (0.90 mol) 8 were thoroughly mixed and heated as in (1) above, except at 410 °C. The product was isolated by dissolving the flux with degassed DMF under inert atmosphere and then washing with anhydrous ether to give 0.045 g of yellow needle-like crystals (54% yield, based on Sb). Semi-quantitative microprobe analysis on single crystals gave K1,3Sb1,oP1,787,6 (average of three data acquisitions). Far-IR (CsI matrix) show absorptions at 620(m), 612(5), 579(5), 563(5), 546(w), 511(m), 494(m), 409(m), 325(m), 310(m), 287(w), 268(w), 253(w), 238(w), and 167(w) cm'1. Rb38b(P84)2 (IV). An amount of 0.036 g (0.30 mmol) Sb, 0.133 g (0.60 mmol) P285, 0.122 g (0.60 mmol) Rb28, and 0.038 g (1.20 mol) 8 were mixed and heated as in (1) above, except at 450 °C. The product was isolated by dissolving the flux with degassed DMF under inert atmosphere and then washing with anhydrous ether. Final washing with H2O, MeOH, and ether gave 0.090g of yellow needle-like crystals (43% yield, based on 8b). Semi- quantitative microprobe analysis on single crystals gave Rb1,78b1,oP1,5862 (average of three data acquisitions). Far-IR (CsI matrix) show absorptions at 632(5), 624(5), 611(5), 581(5), 564(5), 227 544(w), 535(w), 510(5), 493(w), 409(w), 341(m), 325(m), 315(m), 298(m), 266(w), 256(w), 240(w), 200(w), 183(w) and 149(w) cm'1. Cs3Sb(PS4)2 (V). An amount of 0.024 g (0.20 mmol) 8b, 0.067 g (0.30 mmol) P285, 0.119 g (0.40 mmol) C528, and 0.077 g (2.40 mol) 8 were mixed and heated as in (1) above. The product was isolated as in (IV) to give 0.107 g of yellow needle-like crystals (64% yield, based on Sb). Semi-quantitative microprobe analysis on single crystals gave C53,08b1,oP2,oSgg (average of three data acquisitions). Far-IR (CsI matrix) show absorptions at 621(m), 605(5), 577(5), 560(5), 508(m), 493(m), 408(w), 329(w), 316(5), 298(w), 266(w), 262(w), 238(w), 189(w), and 164(w) cm-1. K3Bi(P84)2 (VI). An amount of 0.031 g (0.15 mmol) Bi, 0.067 g (0.30 mmol) P285, 0.066 g (0.60 mmol) C528, and 0.058 g (1.80 mol) 8 were mixed and heated as in (IV). The product was isolated by dissolving the flux with degassed DMF under inert atmosphere and then washing with anhydrous ether to give orange needles and white powder. Final washing with H2O, MeOH, and ether gave 0.05 83 of orange-brown needle-like crystals (60% yield, based on Bi). The crystals appear to be air sensitive. Semi-quantitative microprobe analysis on single crystals gave K2,1Bi1,oP1,6S6,9 (average of three data acquisitions). Far-IR (CsI matrix) show absorptions at 616(m), 609(5), 571(5), 550(5), 516(m), 498(m), 412(w), 291(w), 266(w), 226(w), 201(w), 167(w), and 150(w) cm'1. Rb3Bi(PS4)2 (VII). An amount of 0.031 g (0.15 mmol) Bi, 0.067 g (0.30 mmol) P285, 0.122 g (0.60 mmol) Rb28, and 0.058 g ( 1.80 mol) 8 were heated to 450 °C over 12 hours then isothermed for 6 days, followed by cooling to 150 °C at a rate of 2 °C/hr, then to 228 room temperature in one hour. The product was isolated as in (IV) to give 0.048 g of orange-brown needle-like crystals (41% yield, based on Bi). The crystals appear to be air sensitive. Semi- quantitative microprobe analysis on single crystals gave Rb2,1Bi1,0P12S6,5 (average of three data acquisitions). Cs3Bi(PS4)2 (VIII). An amount of 0.031 g (0.15 mmol) Bi, 0.067 g (0.30 mmol) P285, 0.089 g (0.30 mmol) C528, and 0.019 g (0.6 mol) 8 were mixed and heated as in (I I 1) above. The product (air and water stable) was isolated as in (V I) to give 0.062 g of orange needle-like crystals (45% yield, based on Bi). Semi-quantitative microprobe analysis on single crystals gave C52,1B11,0P1,5863 (average of three data acquisitions). Far-IR (CsI matrix) show absorptions at 624(5), 616(5), 610(5), 570(5), 549(5), 513(m), 496(m), 412(w), 304(w), 293(m), 279(w), 264(m), 247(w), 231(m), 186(w), and 160(w) cm'1. C53Bi2(PS4)3 (IX). An amount of 0.031 g (0.15 mmol) Bi, 0.067 g (0.30 mmol) P285, 0.179 g (0.60 mmol) C528, and 0.058 g (1.80 mol) 8 were mixed and heated as in (V). The product was A isolated as in (V I) to give a mixture of red rectangular block-like crystals (C53Bi(PS4)2) and orange crystals (C53B12(P84)3). Semi- quantitative microprobe analysis on red single crystals gave C51,4BimP1385,5 (average of three data acquisitions). Far-IR (CsI matrix) of manually separated red crystals show absorptions at 585(m), 581(msh), 578(msh), 574(msh), 554(msh), 551(msh), 548(msh), 524(5), 516(5), 500(m), 413(w), 303(msh), 292(m), 261(m), 253(msh), and 239(w) cm-l. 229 Nao,15Bi1,23P2$6 (X). An amount of 0.062 g (0.30 mmol) Bi, 0.200 g (0.90 mmol) P285, 0.047 g (0.60 mmol) Na28, and 0.038 g (1.20 mol) 8 were mixed as above and heated to 400 °C over 12 hours then isothermed for 6 days, followed by cooling to 120 °C at a rate of 4 °C/ hr, then to room temperature in one hour. The product was isolated by dissolving the flux with degassed DMF under inert atmosphere and then washing with anhydrous ether to give 0.103 g of dark red chunky crystals (77% yield, based on Bi). Semi- quantitative microprobe analysis on single crystals gave Bi1,0P1,985,4 (average of three data acquisitions). Elemental analysis at Oneida Research Services gave an Na/ Bi ratio of 10:91. Far-IR (CsI matrix) show absorptions at 535(very broad), 434(5), 375(w), 280(5), 254(w), and 227(m) cm'1. 2.3. Physical Measurements The instruments and experimental setups for Infrared measurements, optical diffuse reflectance measurements, and quantitative microprobe analysis on SEM/EDS are the same as those in Chapter 2 (Part 1) (Section 2.3). The experimental setup for differential thermal analysis (DTA) is the same as that in Chapter 2 (Part 2) (Section 2.3). 2.4 X-ray Crystallography The compounds were examined by X-ray powder diffraction (XRD) for the purpose of phase purity and identification. Accurate €1th spacings (A) were obtained from the powder patterns recorded on a Rigaku rotating anode (Cu Ka) X-ray powder diffractometer, 230 Rigaku-Denki/RW400F2 (Rotaflex), at 45 kV and 100 mA with a 1 °/ min scan rate. The results are summarized in Tables 4-1 to 4-3. Single crystal X-ray diffraction data for K3Bi(PS4)2 were collected with a Nicolet P3 four-circle automated diffractometer equipped with a graphite-crystal monochromator at -60 °C. The data were collected with an on step scan technique because of the long a- axis of 23.894(8) A. Due to the acentric P212121 space group, Friedel pairs were collected so that the correct enantiOmer could be refined. Refinement of the right-handed enantiomer gave R/Rw=5.5 / 6.0 while that of the left-handed enantiomer gave 8.0/8.7. In the A3M(PS4)2 system, only the K3Bi(PS4)2 member could be characterized by single crystal X-ray diffraction studies. Crystals of the other analogs were found to be twinned. The compounds are isostructural as evidenced by the similarity of their respective powder diffraction patterns (Figure 4-1). Single crystal X-ray diffraction data for Cs3Biz(PS4)3 and Nao,16Bi1,23P286 were collected at room temperature on a Rigaku AFC6 diffractometer and the w/ 26 scan technique was used. None of the crystals showed any significant intensity decay as judged by three check reflections measured every 150 reflections throughout the data collections. The space groups were determined from systematic absences and intensity statistics. The structures were solved by direct methods of 8HELXS-8616 and refined by full-matrix least-squares techniques of the TEXSAN package of crystallographic programs. 17 An empirical absorption correction based on w-scans were applied to each data set, followed by a DIFABS 18 correction to the isotropically refined structure. All atoms were eventually 231 refined anisotropically. All calculations were performed on a VAXstation 3 100/ 76 computer. In the case of Nao,16Bi1,23P286, one bismuth atom, one phosphorous atom, and three sulfur atoms were located on general positions. After least squares refinement, the isotropic temperature factor for Bi was rather high at 4.6 A2 with R/RW = 20/25. The occupancy and temperature factor of this atom were refined to give values of 0.56 and 3.5 A2 respectively (R/RW = 15.6/ 18.2). In order to maintain electroneutrality (Bi1,12P286), +0.64 of charge must be compensated. At this stage of the refinement, it was apparent that Na must have been incorporated into the compound but no significant Na electron density peaks were observed. It was assumed that Na and Bi were disordered on the same site. Successive refinements taking into account this hypothesis resulted in a formula of Na0,64Bi1,1211 0,24PzS6 (fl = vacancy) with an R/ Rw=15.9/ 18.4. A final R/Rw of 7.6/8.4 was achieved after DIFABS correction and anisotropic refinement. Elemental analysis suggests that the Na/ Bi ratio is 1.0/9.1 which is not consistent with the above refinement. Refinement using this formula gave Nao,16Bi1,2gfl 0,56P286 (D = vacancy) and an increase in the R/Rw (8.3/9.3). The complete data collection parameters and details of the structure solution and refinement for (I), (V 1), (IX), and (X) are given in Table 4-4. The coordinates of all atoms, average temperature factors, and their estimated standard deviations are given in Tables 4-5 to 4-8. 232 Table 4-1. Calculated and Observed X-ray Powder Diffraction Patterns for KBiP287. h k l dcalcd. A dobsdr,K I/ Imax (obsd) 1 O O 1 1 O O 1 l l 1 1 0 2 1 2 0 0 1 2-1 2 1 O 0 1 2 1 0-2 1 1-2 1 3 0 9.50 7.52 7.31 5.77 5.10 4.75 4.50 4.43 4.27 4.12 3.91 3.77 1 3 l, 2 2 1 3.47,3.46 1 2 2 2 0-2 3 0 O O 4 O O 1 3 l 4 O 1 1-3 1 4 l O 2 3 3 2-1 l 2 3 3 1-2 3 1 2 2 1 3 1 4-2 O 3 3 3 3 -1 l 5 O 1 3 3 0 O 4 4 l 1 2 4 2 1 0-4 3 4 O 2 5 0 O 5 2 0 2 4 3 2-3 2 1 4 3 3 3 5 l O 3 0-4 5 1-1 l 4 4 5 O 2 3.40 3.30 3.17 3.08 2.94 2.93 2.82 2.78 2.72 2.70 2.61 2.55 2.53 2.491 2.464 2.438 2421 2.382 2.357 2.274 2.253 2.240 2.217 2.206 2.185 2.164 2.133 2.073 2.015 1.924 1.878 1.856 1.843 1.793 1.747 9.50 7.52 7.30 5.77 5.09 4.75 4.50 4.43 4.26 4.10 3.90 3.76 3.47 3.40 3.30 3.17 3.07 2.94 2.92 2.82 2.78 2.72 2.69 2.61 2.55 2.53 2.491 2.458 2.438 2.420 2.378 2.336 2.273 2.257 2.243 2.220 2.206 2.186 2.161 2.115 2.072 2.003 1.902 1.881 1.856 1.843 1.788 1.745 75 83 7 12 29 1(1) 25 53 11 6 9 16 51 4 3 32 9 16 19 11 9 5 18 7 8 16 5 4 4 12 60 3 5 10 7 11 14 12 22 4 5 4 5 9 5 5 9 3 233 Table 4—1. (cont'd) h k l dcgjcd, A dobsd_. A I/ Imgx (obsd) 0 7 1 4 S 0 0 6 3 5 3 -1 1 6-3 2 7-1 3 4-4 4 6 O 6 1-1 1.726 1.709 1.699 1.697 1.674 1.623 1.589 1.552 1.550 1.728 1.710 1.7(1) 1.694 1.673 1.623 1.588 1.554 1.550 10 4 7 9 3 3 13 11 11 234 Table 4-2. Calculated and Observed X-ray Powder Diffraction Patterns for K3Bi(PS4) 2. h k l dcalcd. A dam, A 1/193K (obsd) 2 00 1 O 1 2 O 1 3 O l l l 1 3 1 O 2 1 l 4 l 0 4 l l 3 O 2 5 1 O O 1 2 6 0 1 2 1 2 0 2 O 5 O 2 7 0 1 7 l O 8 O O 3 2 1 2 O 3 4 2 1 l 1 3 2 1 3 5 2 1 8 1 1 9 O l 6 2 1 6 O 3 9 1 1 5 2 2 O O 4 8 2 O 7 2 2 4 2 3 10 1 2 9 l 3 15 10 8 3 3 11.95 11.75 8.46 7.21 5.98 5.30 5.17 4.95 4.49 4.02 3.93 3.91 3.77 3.65 3.59 3.40 3.29 3.19 3.05 2.99 2.96 2.92 2.81 2.74 2.69 2.65 2.62 2.55 2.486 2.405 2.386 2.363 2.262 2.244 2.126 2.111 2.018 1.912 1.551 1.549 8.35 7.14 5.93 5.26 5.13 4.95 4.46 4.0) 3.96 3.90 3.74 3.63 3.58 3.39 3.27 3.19 3.04 2.98 2.94 2.91 2.80 2.74 2.68 2.64 2.61 2.54 2.474 24(1) 2.383 2 357 2 255 2 240 2 121 2 110 2 015 1 909 1 552 1 549 52 47 11 19 29 10 7 11 19 7 4 6 5 15 13 10 7 13 100 5 11 11 11 6 10 6 5 5 10 19 7 6 8 7 9 6 6 5 4 235 Table 4-3. Calculated and Observed X-ray Powder Diffraction Patterns for Na0,16Bi1,23P256. h k l dcglcd, A dobsfl, A l/lmgx (ObSd) 0 l 1 1 0-1 1 0 1 1 l 0 0 0 2 1 1-1 1 1 1 0 1 2 0 2 0 1 1-2 0 2 1 1 1 2 2 0 0 1 2 0 1 2-1 1 2 1 2 1-1 2 l 1 2 0-2 1 1-3 2 0 2 1 2 2 1 l 3 2 1-2 2 1 2 2 2 0 0 2 3 0 0 4 1 3 0 1 3-1 1 3 1 2 1-3 1 1-4 1 1 4 3 1 1 2 3 0 2 2-3 3 1-2 2 2 3 3 1 2 2 1-4 3 2-1 1 3 3 0 4 0 3 0-3 3 2-2 1 4-1 5.76 5.46 5.28 4.87 4.68 4.37 4.28 3.94 3.65 3.42 3.40 3.33 3.27 3.19 3.03 3.0) 2.87 2.82 2.73 2.66 2.64 2.61 2.60 2.56 2.482 2.437 2.372 2.341 2.280 2.22 2.209 2.194 2.132 2.089 2.027 1.953 1.946 1.935 1.897 1.886 1.873 1.848 1.831 1.824 1.820 1.758 1.730 5.73 5.44 5.26 4.86 4.67 4.36 4.27 3.93 3.64 3.42 3.39 3.33 3.27 3.18 3.03 2.99 2.87 2.81 2.72 2.65 2.63 2.61 2.59 2.55 2.477 2.433 2.367 2.336 2.276 2.217 2.205 2.189 2.132 2.086 2.021 1.951 1.944 1.931 1.895 1.883 1.868 1.848 1.829 1.823 1 817 1 756 1 728 38 24 5 19 28 22 64 7 4 13 83 H) 14 3 7 8 49 51 E 100 24 7 31 9 16 3 59 24 29 8 9 6 5 4 11 4 6 6 4 10 7 5 4 5 4 10 5 236 h kl dcalgd. A dobsd. A I/ 1max (obsd) 3 2 2 0 2 5 1 3-4 2 3 3 4 O O l 3 4 2 1-5 3 3-1 4 l 0 2 4 0 4 1-1 0 4 3 O O 6 1.721 1.666 1.644 1.640 1.638 1.623 1.611 1.608 1.598 1.594 1.584 1.575 1.561 1.719 1.664 1.642 1.638 1.637 1.622 1.609 1.609 1.596 1.591 1.582 1.574 1.560 3 4 4 8 4 4 8 5 8 9 4 4 7 Table 4-4. Crystallographic Data for 1, VI, IX, and X. I VI Formula KBiP287 K 3Bi( PS4) 2 FW a, A b, A c, A (1, deg 8, deg y, deg 542.57 644.70 9.500(3) 23.894(8) 12.303(4) 6.799(3) 9.097(3) 9.049(2) 90.0 9059(3) 90.0 90.0 90.0 90.0 2; v, A3 4; 1063.1(6) 4; 1470(1) A (Mo Ka) 0.71073 0.71073 space group P21/c (No. 14) P212121 (No. 19) Dcalc.. g/cm3 3.339 p, cm‘1 (Mo [(0) 184.85 Zemax, deg Temp., °C 50 23 2.913 140.81 50 -60 Final R/Rw, % 2.8/3.1 5.5/6.0 Total Data 2192 Total Unique (Ave.) 2060 Data F02>30( 1:02) 1672 No. of variables 1(1) 3997 1999 1745 128 Table 4—4. (cont'd) 238 D( X Formula Cs3Bi2(PS4)3 Na0,16Bi1,23P2S6 PW a, A b, A c, A 0., deg 8, deg y , deg 1294.32 525.47 18.091(5) 6.554(2) 6.791(2) 7.297(2) l8.723(3) 9.371(1) 90.0 90.0 97.95(2) 92.04(2) 90.0 90.0 2; v, A3 4; 2278(1) 2;447.8(2) 1. (Mo Ka) 0.71073 0.71073 space group P21/c (No. 14) P21/n (No. 14) peak, g/cm3 3.773 .1, cm-1 (Mo 1(a) 213.34 29max, deg Temp., °C 50 23 3.731 236.19 50 23 Final R/Rw, % 7.0/9.0 7.7/8.5 Total Data 4537 Total Unique (Ave.) 4375 Data F02>36(1=02) 2623 No. of variables 182 1514 1407 892 47 *R=2(|Fol-|Fcl)/2|Fol Rw={2w(lFol-IFCI)2/2wlFol2}1/2 239 Table 4-5. Fractional Atomic Coordinates and Beq Values for KBiP287 with Estimated Standard Deviations in Parenthesesfi1 atom x y z Beg? A2 Bi(l) K(1) S(l) 8(2) S(3) S(4) 8(5) 8(6) 8(7) P(l) P(2) 0.3419l(3) 0.13823(2) O.ll948(3) l.26(1) -0.1153(2) 0.1911(2) -0.1292(2) 2.26(7) 0.6536(2) 0.1937(1) 0.1485(2) 0.1161(2) 0.2928(1) 0.1101(2) 1.11(7) 1.20(7) 0.1655(2) 0.0297(2) -0.0738(2) 1.27(7) 0.4087(2) 0.1716(1) 0.4314(2) 1.16(7) 0.5057(2) -0.0554(1) 0.1860(2) l.25(7) 1.2472(2) 0.5509(1) 0.0802(2) 0.93(6) 0.1237(2) 0.0545(2) 0.2942(2) 0.6811(2) 0.0367(1) 0.1997(2) 1.18(7) 0.74(6) 0.2224(2) 0.103 1(1) 0.4811(2) 0.74(6) Table 4-6. Fractional Atomic Coordinates and Beq Values for K381(P84)2 with Estimated Standard Deviations in Parenthesesa atom x y z Beq,a A2 Bi( 1) 1((1) 1((2) 1((3) S( 1) 8(2) 8(3) 8(4) 8(5) 8(6) 8(7) 8(8) P( 1) P(2) 0.05174(3) 0.0597(1) 0.35126(6) 1.18(2) —0.0755(2) -0.0045(8) 0.7590(5) 0.2140(2) -0.499l(8) 0.4096(5) -0.3127(2) 0.5098(8) 0.9776(5) 0.0683(2) 0.0142(8) 0.6411(5) 0.1845(2) -0.2484(8) 0.7087(6) 0.1650(2) -0.031(1) 0.3843(4) 0.1918(2) 0.2370(8) 0.6906(6) 0.0790(2) 0.4370(7) 0.3701(4) 0.1735(2) 0.482(1) 0.0895(6) -0.0546(2) 0.2199(6) 0.4241(5) 0.0528(3) 0.2317(7) 0.0519(5) 0.1555(2) —0.0065(7) 0.6119(4) 2.9(2) 2.6(2) 2.8(2) 1.9(2) 2.1(2) 2.3(2) 2.2(2) 1.5(1) 2.5(2) 1.6(2) 2.0(2) l.2(2) 0.0929(2) 0.4659(7) 0.1418(5) 1.1(1) aBeq = (4/3)[a213(1,1) + bZB(2,2) + c213(3,3) + ab(cosY)B( 1,2) + ac(cosB)B( 1,3) + bc(c05a)B( 2,3). 240 Table 4-7. Fractional Atomic Coordinates and Beq Values for Q3Biz(PS4) 3 with Estimated Standard Deviations in Parenthesesa atom x y z Betta A2 Bi( 1) Bi(2) Cs(l) Cs(2) Cs(3) S( 1) 8(2) S(3) 8(4) S(5) 8(6) 8(7) 8(8) 8(9) S(10) S(l 1) S(lZ) P( 1) P( 2) P(3) 0.07105(7) 0.1923(2) 0.79059(6) l.24(5) 0.40215(8) 0.2964(3) 0.50535(7) 3.36(7) 0.1025(1) -0.2648(3) 0.9877(1) 0.3746(2) -0.2486(4) 1.3094(1) 2.6(1) 3.8( 1) 0.2373(1) 0.2794(3) 0.6310(1) 233(9) 0.1149(5) ‘0.176(1) 0.8046(4) 2.1(3) 0.2203(5) -0.221(1) 0.6669(5) 0.1119(4) 0.235(1) 0.9368(4) 0.0575(5) 0.023(1) 0.6495(4) 0.2739(5) 0.499(1) 0.4517(5) 0.2179(5) 0.192(1) 0.3104(4) 0.4196(6) -0.491(1) 0.6469(5) 0.4230(5) 0.253(1) 0.3684(4) 0.2775(5) 0.021(1) 0.4737(5) 0.4414(5) 0.736(2) 0.4951(4) 0.4229(6) 0.005(1) 0.6423(5) 0.0634(5) -0.467(1) 0.6633(4) 0.1193(4) -0.211(1) 0.6945(4) 0.2222(4) 0.244(1) 0.4178(4) 0.4623(5) 0247(1L 0.6072(4) 1.9(3) 1.6(3) 1.6(3) 1.9(3) 2.4(4) 2.8(4) 2.0(3) 2.1(4) 2.8(4) 2.8(4) 1.5(3) 1.3(3) 1.0(3) 1.2(3) Table 4-8. Fractional Atomic Coordinates and Beq Values for Nao,16Bil,23P286 with Estimated Standard Deviations in Parentheses.a atom x y z B egf A2 Bib 8(1) 8(2) 8(3) P( 1) NaC 0.0377(2) 0.6348(2) 0.2470(2) 4.47(7) —0.1147(5) 0.9985(5) 0.2589(4) 0.3455(5) 0.8074(4) 0.4350(4) -0.3192(5) 1.1981(5) -0.0417(4) —0.0644(5) 1.1103(4) 0.0664(3) 0.0377 0.6348 0.2470 1.3(1) 1.3(1) 1.6(1) 0.8( 1) 3.5 aBeq -.-. (4/3)[aZB(1,1) + b2B(2,2) + c213(3,3) + ab(cosY)B( 1,2) + ac(coss)B(1,3) + bc(c05a)B(2,3). b6496 occupied C 0.08% occupied 241 (A) AJSb(P84)2 K Rb ' >1 .17.." W _ 7“ A aW 3 ‘ :‘ZVAw—A‘.‘ A A a a—n C3 . . . . . 50 55 50 (B) AJBi(P84)2 K Rb >\ 3:. U) 1:: cu H E. Cs "'T'ITI'Y'I' ' "Horny"... [u q r “ I . I " v: 5. 10. 15. 20. 25. 30. 35. 40. 45. so. 55. 29 ' so. Figure 4-1. Powder X-ray diffracton patterns of: (A) A38b(PS4)2 and (B) A3Bl(P84)2 242 3. Results and Discussion 3.1. Synthesis, Spectroscopy and Thermal Analysis The syntheses were a result of a redox reaction in which the metal is oxidized by polysulfide ions in the Ax[PySZ] flux. The resulting metal ions are then coordinated by the highly charged [PS4]3-, [P287]4-, and [P286]4- ligands. The molten polythiophosphate flux method is very effective for crystal growth and isolation of pure products in this system. Thiophosphate ligand formation appears to be controlled by the metal to P285 ratio and the Lewis basicity of the alkali metal cation. Stabilization of [PS4]3' anions is observed in A3M(PS4)2 and (153Bi2(PS4)3 from a Bi/Pst ratio of 1:1.5-2.0. Several attempts to isolate pure CS3Bi2(PS4)3 resulted in a mixture of C53Biz(PS4)3 and C53Bi(PS4)2. Increasing the amount of P285 in the flux (Bi/P285 ratio of 1:3) yielded the [P287]4- ligand in KBiP287. In the Bi/sts/NaZS/S ( 1/ 3/ 2/ 4) system, the less basic Na+ counterion stabilizes a lower 4+ oxidation state for P, which leads to the formation of [P286]4' anions in Na0,16Bi1,23P286. A second phase was obtained in the Bi/sts/NaZS/S system using the respective ratio (1/2/2/4-8) at 400 °C. This reaction affords a homogeneous product consisting of dark red twinned crystals of a new compound (evidenced by powder X- ray diffraction). The far-IR spectrum of KBiPzS7 (Table 4-9) shows absorbances at 600(5), 583(s), 576(ssh), 564(msh), 557(msh), 526(wsh), 464(vs), 412(w) cm'l. The very strong absorbance at 464 cm'1 represents the characteristic P-S-P stretching vibration while the remaining 243 absorbances are due to -P83 stretching vibrations by analogy to Ag4P287.19.20 A second set of absorbances at 300(vw), 270(wsh), 250(msh), 247(m), 238(msh), 227(wsh), 208(wsh), 203(wsh), 192(vw), 174(vw), 169(vw), 157(vw), and 149(vw) cm'1 are assigned to Bi-S stretching vibrations and PS deformation modes.l9,20,21 The Far-IR spectra of A3M(PS4)2 (I I I-V I I I), C33Bi2(PS4)3 (IX), and Nao,16Bil,23P286 (X), shown in Table 4-9, are quite complex. Absorptions in the 400-650 cm-1 range for (I I H X), are tenatively assigned to P-S vibrational stretching modes by analogy with the vibrational spectra of other [PS4]3- containing compounds such as InPS4, GaPS4, and BiPS4.21 Absorptions below 400 cm'1 are assigned to S-P-S bending modes and M-8 vibrations.22 The far-IR spectrum of Nao,16Bi1,23P286 shows a broad absorption at 535 cm-1 which can be assigned to P83 stretching vibrations. An assignment of the strong absorption band at ~434 an1 in N ao,16Bi1,23P286 is not as straightforward. In Na4P286-6H20, an absorbance at 443 cm-1 was ascribed to an out-of-phase PS3 mode, corresponding to a P-P vibration.23 The P—P vibration is expected to be IR inactive because it resides on a center of symmetry. Surprisingly, it was not observed in the Raman spectrum. The spectra of Na4P286-6H20 were described in terms of internal modes of PS 3 groups and combinations of in- phase and out-of-phase translational and rotational motions.23 Since the P-P bond in Nao,16Bi1,23P286 also resides on a center of symmetry, a similar situation is assumed to exist in this compound. Thus, by analogy to Na4P286-6H20, the strong band at 434 cm-1 in N30JfiBiL28PZS6 can tentatively be ascribed to the out-of-phase 244 translational P83 mode. Below 400 cm'l, the bands are assigned to a combination of S—P-S bending modes and M-8 vibrations as above. The optical properties of (I-X) were assessed by studying the UV/near-IR reflectance spectra of the materials. The spectra confirm the semiconductor nature of the materials by revealing the presence of sharp optical gaps as shown in Figures 4—2 to 4-5. The A3M(PS4)2 (A=K, Rb, Cs; M=8b, Bi) compounds exhibit steep absorption edges from which the band-gap, Eg, can be assessed at 2.25 (I ), 2.25 (I I), 2.75 (I I I), 2.67 (IV), 2.80 (V), 2.26 (V I), 2.21 (V I I) 2.28 eV (VIII), respectively. Impurity bands are observed in the spectra of K3Bi(PS4)2 and Rb3Bi(PS4)2 shown in Figures 4-4a and 4-4b. These two compounds are air sensitive which results in some decomposition of the compounds. The band-gap, Eg, of Nao,16Bil,28P285 is 1.88 eV (X). The higher energy absorptions found in several of the compounds are readily resolved and are assigned to electronic S—>M charge transfer transitions. Differential thermal analysis (DTA) shows that KBiP287 and ' RbBiP287 melt at 5 1 1 and 488 °C, respectively, and upon cooling form amorphous glasses. Nao,16Bil,23P286 melts incongruently at 652 °C to form a mixture of N301631L28P286 and an impurity phase. The A 3M(PS4) 2 compounds melt congruently, suggesting that large single crystals or microcrystalline thin films can be grown from the melt. Table 4-10 summarizes optical and melting point data for all compounds. Typical thermograms for RbBiP287 and Rb3Bi(PS4)2 are shown in Figure 4-6. Table 4-9. Far-IR Spectra for (I -X). 245 KBi_PL287 K 3Sb(PS4); Rb3Sb(PS4)J2 Cs38b(PS4)2 620(8) 612(s) 579(s) 563(5) 546(w) 511(m) 494(m) 409(m) 325(m) 310(m) 287(w) 268(w) 238(w) 167(w) 600(8) 583(8) 576(ssh) 564(msh) 557(msh) 526(wsh) 464(vs) 412(w) 300(vw) 270(wsh) 250(msh) 247(m) 238(msh) 227(wsh) 208(wsh) 203(wsh) 192(vw) 174(vw) 169(vw) 157(vw) 149(vw) 621(m) 605(s) 577(s) 560(s) 508(m) 493(m) 408(w) 329(w) 316(s) 298(w) 266(w) 262(w) 238(w) 189(w) 164(w) 632(3) 624(s) 611(s) 581(3) 564(s) 544(w) 535(w) 510(s) 493(w) 409(w) 341(m) 325(m) 315(m) 298(m) 266(w) 256(w) 240(w) 200(w) 183(w) 149(w) K3Bi(PS4); Cs3Bi(PS4)2 Cs3Bi2(PS4)3 NaulgBi] Z§PZS§ 535(m, broad) 434(s) 375(w) 280(s) 254(w) 227(m) 616(m) 609(s) 571(s) 550(s) 516(m) 498(m) 412(w) 291(w) 266(w) 201(w) 167(w) 150(w) 624(s) 616(s) 610(s) 570(s) 549(s) 513(m) 496(m) 412(w) 304(w) 293(m) 279(w) 264(m) 247(w) 231(m) 186(w) 160(w) 585(m) 581(msh) 578(msh) 574(msh) 554(m) 551(msh) 548(msh) 524(3) 516(s) 500(m) 413(w) 303(msh) 292(m) 261(m) 253(msh) 239(w) Abbreviations: s=strong, m=medium, w=weak, sh=shoulder, v=very 246 Table 4-10. Optical Band Gaps and Melting Point Data for (I -X). Formula Eg (eV) M. P. (°C) KBiP287 RbBiP287 K3Sb( PS4) 2 Rb3Sb(PS4)2 C33Sb(PS4)2 K3Bi(PS4) 2 Rb3Bi(PS4)2 C53Bi(PS4)2 NaBiP;S(, 2.25 2.25 2.75 2.67 2.80 2.26 2.21 2.28 1.88 5 1 1 (incongruent) 488 (incongruent) 5 49 (congruent) 471 (congruent) 521 (congruent) 559 (congruent) 610 (congruent) 642 (congruent) 652 (incongruent) 247 g 8 1 1 l 1 1 1 5 7.) ‘E < 6‘ 51:. 55 8 4 ‘8’ 3: ta 3.47 (A) 1... 2.67 - )— KBiP287 Eg=2.ZSeV ‘3 2‘ '2 H 3’ o '- ‘ r 1 r r I I . 0 l 2 3 4 5 6 7 Energy(cV) L g 1 l 1 L 4 3.5- 3_ 2.5“ 21 1.54 2.61 3.28 (B) RbBiP287 E,=2.25ev - .. - L- )— _. 1- 0.5‘ J r- 1 I I l l l 0 ) s t i n U . b r A ( . f f e o C n o i t p r o s b A S / a 0 1 2 3 4 5 6 7 Energy (eV) Figure 4-2. Optical absorption spectrum of: (A) KBiPzS7 and (B) RbBiP287 ; 3 1 1 1 1 1 1 '_§ '1 .0 3 2 5 ~ 2- 5’3 1 s- ,q 3.20 K38b( 1 )-8(4) 8(2)-P( 1 )-8(3) S(2)-P(l)-S(4) S(3)-P( 1)-5(4) S(5)-P(2)-8(6) S(5)-P(2)-S(7) S(S)-P(2)-S(8) S(6)-P(2)-S(7) S(6)-P(2)-S(8) S(7)-P(2)-S(8) S-P-S (ave.) K(3)-S(2) K( 3 )-8( 3) K( 3 )-S(4) 1((3 )-S(6) 1((3 )-S(6') K-S (ave.) 73.8( 1) 91.3(1) 87.4( 1) 161.4(1) 88.5(1) 157.9(1) 100.4(1) 80.3(1) 70.7(1) 109.3(2) 103.5(2) 109.4(2) 109.5(2) 112.4(2) 112.3(3) 113.7(2) 107.1(2) 104.3(2) 108.4(2) 113.8(3) 109.2(2) 109(3) 3.287(5) 3.320(5) 3.438(5) 3.402(6) 3.536(6) 3.3( 1) Bi-S( 1) Bi-8(3) 131-5( 5) Bi-S(7) Bi-S(8) Bi-S (ave.) Rims”), “25(8), P( 1)-5( 1) P( 1)-5( 2) 1>( 1)-S( 3) P( 1)-5(4) P(2)-S(5) P(2)-S(6) P(2)-S(7) P(2)-S(8) P-S (ave.) K(l)-S(l) 1(( 1 )-S( 2) K( 1 )-8(4) K(1)-S(5) K( 1 )-S(7) 1(( 1 )-S(8) K(2)-S(2) 1((2)-5(2') K(2)-S(3) 1((2)-5(4) K(2)-8(4') K(2)-S(5) K(2)-S(6) K(2)-S(6') 2.673(3) 2.789(4) 2.649(4) 2.842(5) 2.956(3) 2.8(1) 3.399(4) 3.453(5) 2.108(5) 1.994(5) 2.070(4) 1.997(6) 2.091(5) 1.980(5) 2.044(5) 2.027(5) 204(5) 3.414(6) 3.133(6) 3.320(6) 3.390(5) 3.429(5) 3.374(5) 3.264(5) 3.490(6) 3.390(6) 3.163(5) 3.486(6) 3.283(5) 3.154(5) 3.062(5) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01={Zn (1n- I)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1the mean length (or angle), and n the number of bonds. 271 Table 4-13. Selected Distances (A) for C53B12(PS4)3 with Standard Deviations in Parenthesesa Bi(1)-S( 1) 2.626(8) Cs(1)-S( 1) Bi( 1 )-S( 3) 2.751(7) Cs( 1 )-S( 2) Bi(1)-S(4) 2.859(8) Cs( 1 )-S( 3) Bi( 1 )-S( 6) 2.746(9) Cs(1)-S(3') Bi(1)-S(12) 2.900(9) Cs(1)-S(4) Bi(1)-S (ave.) 2.8(1) Cs(1)-S(5) Cs( 1 )-S( 9) Bi( 1)...s(4') 3.528(9) Cs( 1)-S( 12) Bi( 1)...s(12') 3.312(8) Cs(2)-S(2) Cs(2)-S(7) Bi(2)-S( 5) 2.764(9) Cs(2)-S(8) Bi(2)-8(7) 300(1) Cs(2)-S(8') Bi( 2)-S( 8) 2.659(8) Cs(2)-S( 10) Bi(2)-8(9) 2.925(9) Cs(2)-S(11) Bi(2)-S(10) 3.08(1) Cs(3)-S( 2) Bi(2)-S(10') 2.841(9) Cs( 3)-S(2') Bi(2)-S (ave.) 2.9(2) Cs(3)-8(4) Bi(2)---S(l 1) 3.219(9) Cs(3)-8(7) Cs(3)-8(6) Cs(3)-8(9) Cs(3)-S( 1 2) Cs-S (ave.) Bi( 1)-Bi( 1') Bi(2)-Bi(2') P(1)-S(1) P(1)-S(2) P(1)-S(4) P(1)-S(12) P(2)-S(3) P(2)-S(5) P(2)-S(6) P(2)-S(9) P(3)-S(7) P(3)-8(8) P(3)-S( 10) P(3)-S(1l) P-S (ave.) 2.09( l) 1.97( 1) 2.06(1) 2.05(1) 2.08(1) 2.03( 1) 2.03( 1) 2.02( 1) 2.01(1) 2.06(1) 2.08(1) 2.00(1) 2.04(4) 3.520(8) 3.719(9) 3.537(9) 3.538(9) 3.688(9) 3.729(9) 365(1) 3.797(9) 3.590(9) 3.641(1) 3.631(9) 3.649(9) 3.522(8) 3.79(1) 3.485(9) 3.481(9) 3.746(9) 3.431(8) 3.62(1) 3.59(1) 3.710(9) 3.6(1) 4.401(2) 4.519(3) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations o)={zn(ln- I)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 272 , Table 4-14. Selected Angles (°) for CS3B12(PS4)3 with Standard Deviations in Parentheses.a S(1)-Bi(1)-S(3) 88.0(2) S(1)-P(1)-S(2) 115.2(5) S(1)-Bi(1)-S(4) 72.3(2) S(1)-P(1)-S(4) 102.9(5) S(1)-Bi(1)-S(6) 89.2(3) S(1)-P(1)-S(12) 107.0(5) S(1)-Bi(1)-S(12) 82.5(3) S(2)-P(1)-S(4) 113.0(5) S(3)-Bi(1)-S(4) 159.9(2) S(2)-P(1)-S(12) 109.6(5) S(3)-Bi(1)-S(6) 73.3(2) S(4)-P(1)-S(12) 108.6(4) S(3)-Bi(1)-S(12) 82.2(2) S(3)-P(2)-S(5) 107.4(5) S(4)-Bi(1)-S(6) 101.4(3) S(3)-P(2)-S(6) 106.0(5) S(4)-Bi(1)-S(12) 99.1(2) S(3’)-P(2)-S(9) 112.3(5) S(5)-P(2)-S(6) 114.3(5) S(5)-Bi(2)-S(7) 93.5(3) S(5)-P(2)-S(9) 107.6(5) S(5)-Bi(2)-S(8) 85.9(3) S(6)-P(2)-S(9) 109.3(5) S(5)-Bi(2)-8(9) 70.1(2) S(7)-P(3)-S(8) 109.2(5) S(5)-Bi(2)-S( 10) 71.6(2) S(7)-P(3)-S( 10) 107.4(5) S(7)-Bi(2)-S( 10) 65.8(2) S(7)-P(3)-S(11) 114.3(5) S(7)-Bi(2)-S( 10') 93.2(3) S(8)-P(3)-S( 10) 105.0(5) S(8)-Bi(2)-S(9) 86.7(3) S(8)-P(3)-S( 1 1) 109.7(5) 8(8)-Bi(2)-S( 10) 89.0(3) S(10)-P(3)-S(11) 110.9(5) S(10)-Bi(2)-S(10') 80.5(3) S-P-S (ave.) 109(3) alThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01=12n (In- I)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 273 Table 4-15. Selected Distances (A) and Angles (°) for Nao,16Bil,23P286 with Standard Deviations in Parentheses.a Bi-S( 1) 2.839(3) S(1)-Bi-S(1') 8906(6) Bi-S(1') 2.944(4) S(1)-Bi-S(2) 79.03(9) Bi-S(2) 2.917(3) S(1)-Bi-S(2') 77.45(9) Bi-S(2') 3.170(4) S(1)-Bi-S(3) 70.16(9) Bi-S(2") 3.048(3) S(1)-Bi-S(3") 8253(9) Bi-S(3) 3.190(4) S(1')-Bi-S(2') 71.62(9) Bi-S(3') 3.250(4) S(1')-Bi-S(2") 88.81(9) Bi-S(3") 2.974(4) S(1')-Bi-S(3) 67.28(9) Bi-S (ave.) 3.0(1) S(1')-Bi-S(3') 9053(9) P-S( 1) P-S( 1) P-S( 1) P-P' S(2)-Bi—S(3) 74.71(9) 2.018(4) S(2)-Bi-S(3') 77.11(9) 2.032(4) S(3)-Bi-S(3') 73.1( 1) 2.027(4) 2.219(6) S(1)-P-S(2) 115.5(2) S(1)-P-S(3) 114.6(2) S(2)-P-S(3) 110.0(2) S(1)-P-P' 106.6(2) S(2)-P-P' 103.1(2) S(3)-P-P' 105.8(2) 21The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01=12n (1n- I)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 274 3.3. Conclusions The synthesis of new quaternary thiophosphates with Ax[PySZ] molten salts provides a useful synthetic approach with broad scope. The Ax[PySZ] fluxes provide [PxSyPI' units that display remarkable versatility in terms of ligand binding to metals. The high negative charge of these units makes them hard to stabilize in conventional aqueous or organic solvents. The relatively low melting temperatures of the Ax[PSy] fluxes should allow for the isolation of new or metastable structures and also provide a reliable method for stabilization of [PxSy]n ' units. 275 List of References (a) Kanatzidis, M. G. Chem. Mater. 1990, 2, 353-363. (b) Kanatzidis, M. G.; Park, Y. Chem. Mater. 1990, 2, 99—101. (c) Zhang, X.; Kanatzidis, M. G. J. Am. Chem. Soc. 1994, 116,1890-1898. (a) Kanatzidis, M. G.; Park, Y. J. Am. Chem. Soc. 1989, 111, 3767-3769. (b) Park, Y.; Kanatzidis, M. G. Angew. Chem. Int. Ed. Engl. 1990, 29,914-915. (c) Iiao, J.-H.; Varotsis, C.; Kanatzidis, M. G. Inorg. Chem., 1993, 32, 2453- 2462. (a) Hahn, H.; Klingen, W. Naturwissenschaften 1965, 52, 494. (b) Hahn, H.; Ott, R.; Klingen, W. Z. Anorg. Allg. Chem. 1973, 396,271-278. An earlier claim of the existence of the M3(PS4)2 (M=first row transition metals) general class of compounds proved incorrect. The reported XRD powder pattern for these materials are identical to those of MPS3. (a) Diehl, R.; Carpentier, C.-D. Acta Cryst, 1978, B34,1097-1105. (b) Buck, P.; Carpentier, C.-D. Acta Cryst, 1973, 829, 1864-1868. (c) Zimmermann, H.; Carpentier, C.-D.; Nitsche, R. Acta Cryst. 1975, 831, 2003-2006. (d) Becker, R.; Brockner, W.; Eisenmann, B. Z. Naturforsch., 1987, 42a, 1309-1312. (e) Ferrari, A.; Cavalca, L. Gazz. Chim. Ital. 1948, 78, 283-285. (f) Diehl, R.; Carpentier, C.-D. Acta Cryst. 1977, B33, 1399- 1404. (g) Simon, A.; Peters, K.; Peters, E.-M.; Hahn, H. Z. Naturforsch. 1983, 38b, 426-427. (h) Jansen, M.; Henseler, U. J. Solid State Chem. 1992, 99,110-119. (a) Toffoli, P.; Rouland, J. C.; Khodadad, P.; Rodier, N. Acta Cryst. 1985, C41, 645-647. (b) Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst. 1982, B38, 2374-2378. (c) Toffoli, P.; Khodadad, P.; Rodier, N. Bull. Soc. Chim. Fr. 1981, 11,429-432. (a) Mercier, R.; Malugani, J.-P.; Fahys, B.; Robert, G. Acta. Cryst. 1982, B38, 1887-1890. (b) Fiechter, 8.; Kuhs, W. F.; Nitsche, R. Acta Cryst. 1980, B36, 2217-2220. (c) Schafer, H.; Schafer, G.; Weiss, A. Z. Naturforsch. 1965, 20b, 811. (d) Brec, R.; Evain, M.; Grenouilleau, P.; Rouxel, J. Rev. Chim. Min. 1983, 20, 283-294. (e) Brec, R.; Grenouilleau, P.; Evain, M.; Rouxel, J. Rev. Chim. Min. 1983, Z), 295-304. (f) Evain, M.; Lee, 8.; Qpeignec, M.; Brec, R. J. Solid State Chem. 1987, 71, 139-153. (g) Jandali, M. Z.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem. 1985, 530, 144-154. (h) Weiss, A; Schafer, H. Z. Naturforsch. 1963, 18b, 81-82. Evain, M.; Brec, R.; Whangbo, M.-H. J. Solid State Chem. 1987, 71, 244- 262. Thompson, A. H.; Whittingham, M. S. U. 8. Patent 4,049,879 1977. (b) Brec, K; Le Mehaute', A. Fr. Patents 7,704,519 1977. Bridenbaugh, P. M. Mat. Res. Bull, 1973, 8,1055-1060. 10. Buck, P.; Nitsche, R. A. Z. Naturforsch., 1971, 26b, 731. 11. Scott, B.; Pressprich, M.; Willet, R. D.; Clearly, D. A. J. Solid State Chem. 1992, %, 294-300. 276 12. 13. 14. Arnautova, E; Sviridov, E.; Rogach, E. Savchenko, E.; Grekov; A. Integrated Ferroelectrics, 1992,],147-150. Menzel, F.; Ohse, L; Brockner, W. Heteroatom Chem. 1990, 1(5),357-362. McCarthy, T. J.; Kanatzidis, M. G. manuscript in preparation. 15. Feher, F. Handbuch der' Praparativen Anorganischen Chemie: Brauer, G., Ed.; Ferdinand Enke: Stuttgart, Germany, 1954; pp. 280-281. 16. G. M. Sheldrick, In Crystallographic Computing 3; Sheldrick, G. M., Kruger, C., Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985; PP175-189. 17. TEXSAN: Single Crystal Structure Analysis Software, Version 5.0, (1981). Molecular Structure Corportion, The Woodlands, TX 77381. 18. Walker, N.; Stuart, D. Acta Cryst, 1983, A39, 158-166. 19. Menzel, F.; Ohse, L.; Brockner, W. Heteroatom Chem. 1990, 1,357-362. 20. (a) (beignec, M.; Evain, M.; Brec, R.; Sourisseau, C. J. Solid State Chem. 1986, 63, 89-109 (b) Andrae, H.; Blachnik, R. J. Alloys and Compounds 1992, 189,209-215. 21. D'ordyai, V. 8.; Galagovets, I. V.; Peresh, E. Yu.; Voroshilov, Yu. V.; Gerasimenko, V. 8.; Slivka, V. Yu. Russ. J. Inorg. Chem., 1979, 24,1603- 1606. 22. Patzmann, U.; Brockner, W.; Cyvin, B. N.; Cyvin, S. J.; J. Raman Spectr., 1986, 17,257-261. 23. Mathey, Y.; Clement, R.; Sourisseau, C.; Lucazeau, G. Inorg. Chem., 1980, 19, 2773-2779. 24. 25. 26. 27. 28. P. Toffoli, P. Khodadad, N. Rodier, Acta Cryst. 1977, 833,1492-1494. M. Z. Jandali, G. Eulenberger, H. Hahn, Z. Anorg. Allg. Chem. 1978, 445, 184-192. Durand, E; Evain, M.; Brec, R. J. Solid State Chem. 1992, 102,146-155. Kanishcheva, A. 8.; Mikhailov, Yu. N.; Trippel, A. F. Inorg. Mater. 1981, 17,1466-1468. McCarthy, T. J.; Ngeyi, S.-P.; Iiao, J.-H.; DeGroot, D.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Maren, 1993, 5,331-340. 277 29. Iiautard, B.; Garcia, J. C.; Brun, G.; Tedenac, J. C.; Maurin, M. Eur. J. Solid State Inorg. Chem., 1990, 27,819-830. 30. L. Pauling The Nature of the Chemical Bond, 3rd Edition, New York: Cornell University Press, 1966, p. 260. 31. Vos, A.; Wiebenga, E. H. Acta Cryst. 1955, 8217-223. 32. Simon, A.; Peters, K.; Peters, E.-M.; Hahn, H. Z. Naturforsch. 1983, 38b, 426-427. 33. McCarthy, T. J.; Kanatzidis, M. G. J. Chem. Soc., Chem. Commun., 1994, in press. 34. McCarthy, T. J.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Mater., 1994, in press. 35. McCarthy, T. J.; Kanatzidis, M. G. manuscript in preparation. 36. 37. 38. McCarthy, T. J.; Tanzer, T. A.; Chen, L-H.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. manuscript in preparation. Bronger, W.; Balk-Hardtdegen, H. Z. Anorg. Allg. Chem., 1989, 574, 89-98. Toffoli, P.; Michelet, A.; Khodadad, P.; Rodier, N. Acta Cryst. 1982, B38, 706-710. CHAPTER 5 Coordination Chemistry of [P28 e6]4' in Molten Alkali Metal Polyselenophosphate Fluxes. Isolation of KMP28 es and C58M4( P2866)5 (M=Sb,Bi) 1. Introduction Recently, we reported the synthesis of new quaternary metal thiophosphate compounds (ABiP287; A=K,Rb) using molten alkali polythiophosphate fluxes at intermediate temperatures. 1 We have extended this methodology to new quaternary metal selenophosphates, because the chemistry of selenophosphate ligands is not well developed. The rare [PSe4]3' ligand is found in solid state compounds such as Cu3PSe42 and Tl3PSe4,3 and in the unusual tungsten complex, [W(Se)(PSe2)(PSe4)]2-, which also contains the unprecedented heteroallylic [PSe2]- unit.4 Most of the known solid state selenophosphates contain the ethane-like [P28e6]4- ligand. Thesecompounds belong to an important M2P2Q. (Q=S,Se) family of compounds which are structurally related to Cd12.5)6 Thus far, transition metals have received the most attention in this system although a few examples with main group elements such as Sn5€,7 and Pb593 are known. Sn and Pb are found in trigonal prismatic sites instead of octahedral sites to form a three-dimensional 278 279 structure type. The acentric Sn2P286 is a promising ferroelectric material for use in memory devices.7C It is noteworthy that In1,33P28e6 may be suitable for photovoltaic devices9)1O while other members of the M2P2Cb family have been studied for rechargeable battery11 and ion-exchange applications.1 2 Recently, several mixed- metal selenophosphates of the M2P2Cx family have been prepared.13 These compounds are typically synthesized by direct combination of the elements in the 500-800 °C temperature range. Studies in various laboratories have shown that the M2P205 structural type is very thermodynamically stable. In order to explore new metal/selenophosphates at lower temperatures and to obtain new structure types we adopted the flux technique using the polyselenophosphate Ax[PySeZ] fluxes at <600°C. A key feature of these melts is that they are Se-rich. The Ax[PySeZ] fluxes provide excess [PySezln‘ anions which bind to metal ions and act as mineralizers. They also provide a strong basic medium which discourages the formation of M2P28e6. We highlight that Ax[PySeZ] fluxes behave significantly different than their sulfur analogs and ' stabilize readily [PzSe6]4- units. This results in solid state compounds with no sulfur analogs. We report here the synthesis, structural characterization, optical, thermal, and electrical properties of four new quaternary compounds. KBiP28e6 and KSbP28e6 are isostructural and feature a novel layered structure containing the [P28e6]4' building block in a remarkably complex bonding mode. Csssb4(P28e6)5 and C53814(P28e6)5 also form complex isostructural layered structures and exhibit a rare example of unusually close Sb---Sb and Bi---Bi 280 interactions, similar in magnitude to those found in the corresponding elements.. The ethane-like [P28e6]4- ligand is found in three unique bonding modes. 2 . Experimental Section 2.1. Reagents Chemicals in this work were used as obtained: (1) antimony powder, 99.999+% purity, -200 mesh, Cerac Inc., Milwaukee, WI; (ii) bismuth powder, 99.99996 purity, -100+200 mesh, Johnson Matthey/AESAR Group, Seabrook, NH; (iii) red phosphorus powder, Morton Thiokol, Inc., - 100 mesh, Danvers, MA; (iv) selenium powder, 99.S+% purity -100 mesh, Aldrich Chemical Co., Inc., Milwaukee, WI; (v) potassium metal, analytical reagent, Mallinckrodt Inc., Paris, KY; (vi) cesium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, (vii) DMF, analytical reagent, diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. 2.2. Synthesis Synthesis: All manipulations were carried out under a dry nitrogen atmosphere in a Vacuum Atmospheres Dri-Iab glovebox. For the preparation of K28e and C828e we used a modified literature procedure.14 The preparation of K28e is reported in Section 2.2 of Chapter 1 (Part 1). Cszse. A 5.00 g (38 mmol) aliquot of slightly heated (~30 °C) cesium metal was pipetted into a 250-ml round-bottom flask. A 281 ISO-ml volume of liquid ammonia was condensed into the flask at -78 °C (dry ice/acetone bath) under nitrogen to give a dark blue solution. 1.485 g (19 mmol) Se and a Teflon-coated stir bar were added, and the mixture was stirred for 1 h to give a light blue solution. The NH3 was removed by evaporation under a flow of nitrogen as the bath slowly warmed to room temperature. The pale orange solid (98% yield) was dried under vacuum overnight, flame- dried, and ground to a fine powder in the glove box. P28 e5. The phosphorous selenide glass "P28e5" was prepared by heating a stoichiometric ratio of the elements in an evacuated pyrex tube for 24 hours at 460 °C followed by slow cooling to room temperature. This was ground up and stored in a nitrogen glove box. KBiP28e6 (I). An amount of 0.031 g (0.15 mmol) Bi, 0.206 g (0.450 mmol) P28e5, 0.047 g (0.3 mmol) 1(28e, and 0.095 g (1.2 mmol) Se was sealed under vacuum in a pyrex tube and heated to 410 °C for 4 days followed by cooling to 1 10 °C at 4 °C/hr. The excess deep red KxPySez matrix was removed with DMF to reveal analytically pure dark red rectangular plate-like crystallites of KBiP28e6 (69% 5 yield). Single crystals suitable for data collection were synthesized at 450 °C and were accompanied by a K/P/Se impurity. The crystals are air and water stable. Quantitative microprobe analysis on single crystals gave K1,0B113P2,1Se63 (average of three data acquisitions). Far-IR absorbances are given in Table 5-1. KSbP28 e5 (11). An amount of 0.018 g (0.15 mmol) Sb, 0.206 g (0.450 mrnol) P28e5, 0.047 g (0.3 mmol) K28e, and 0.095 g (1.2 mmol) Se were mixed and heated as above, except at 450 °C for 6 days and cooled to 150 °C at 4 °C/hr (75% yield). The compound was found to 282 be isostructural to KBiP28e6 by powder X-ray diffraction. Quantitative micrOprobe analysis on red crystals gave K1,08b1,1P1,4Se52 (average of three data acquisitions). CsaSb4( P28e6)5 (III). An amount of 0.018 g (0.15 mmol) Sb, 0.206 g (0.450 mmol) P28e5, 0.103 g (0.30 mmol) Cszse, and 0.047 g (0.60 mmol) Se were thoroughly mixed and heated as in (I ), except as 460 ° C for 6 days. The product, which is stable in water and air, was isolated by dissolving the flux with degassed DMF under inert atmosphere and then washing with anhydrous ether to give 0.120 g (75% yield based on Sb) of pure dark red crystals. Quantitative microprobe analysis on single crystals gave C8238b1,0P23Se72 (average of three data acquisitions). Far-IR absorbances are given in Table 5-1. CssBi4( P2S e6)5 (IV). The reaction of 0.031 g (0.15 mmol) Bi, 0.206 g (0.450 mmol) P28e5, 0.103 g (0.30 mmol) Cszse, and 0.047 g (0.60 mmol) Se under the same conditions as (I I I) gave 0.155 g (90% yield based on Bi) of black crystals. Quantitative microprobe analysis on the single crystals gave Cs2,oBil_oP1,3Se6,5 (average of three data acquisitions). Far-IR absorbances are given in Table 5-1. The homogeneity of (I), (I I I), and (IV) was confirmed by comparing the observed and calculated X-ray powder diffraction patterns. The dnkl spacings observed for the bulk materials were compared, and found to be in good agreement with the tin k1 spacings calculated from the single crystal data.15 The results are summarized in Tables 5-2, 5-3, 5-4. 2.3. Physical Measurements 283 The instruments and experimental setups for optical diffuse reflectance measurements, far-IR spectroscopy, quantitative microprobe analysis on SEM/EDS, and charge-transport measurements are the same as those in Chapter 2 (Part 1) (Section 2.3). The experimental setup for differential thermal analysis (DTA) is the same as that in Chapter 2 (Part 2) (Section 2.3). 2.4. X-ray Crystallography All compounds were examined by X-ray powder diffraction for the purpose of phase purity and identification. Accurate dhkl spacings (A) were obtained from the powder patterns recorded on a Rigaku Rotaflex Powder X-ray Diffractometer with Ni filtered Cu Ka radiation operating at 45 kV and 100 mA. The data were collected at a rate of 1.0 deg/ min. Structural solution of CS38 b4( P28 e6)5. Single crystal X- ray diffraction data for C538b4(P28e6)5 were collected with a Nicolet P3 four-circle automated diffractometer equipped with a graphite- crystal monochromator. The data were collected with 6- 20 step scan technique. Structural solution of KBist e6 and CssBi4( P28 e6)5. Intensity data for CssBi4(P28e6)5 were collected with a Rigaku AFC6 diffractometer equipped with a graphite-crystal monochromator. The data were collected' with the 01/29 scan technique. None of the three crystals showed any significant intensity decay as judged by three check reflections measured every 150 284 , reflections throughout the data collection. The space groups were determined from systematic absences and intensity statistics. The structures were solved by direct methods of SHELXS-86l6 and refined by full-matrix least-squares techniques of the TEXSAN package of crystallographic programs.17 An empirical absorption correction based on w-scans was applied to each data set, followed by a DIFABSl8 correction to the isotropically refined structure. All atoms were eventually refined anisotropically. All calculations were performed on a VAXstation 3 100/ 76 computer. The complete data collection parameters and details of the structure solution and refinement for (I ), (I I I), and (I V) are given in Table 5-5 . The coordinates of all atoms, average temperature factors, and their estimated standard deviations are given in Tables 5-6, 5-7, and 5-8. Table 5-1. Far-IR Data for KMP28e6 and CS3M4(P28e6)5.a I I I I I V 490(8) 522 (m) 518 (m) 477 (s) 511 (m) 499(s, broad) 453 (w) 500 (m) 460(m, broad) 292 (m) 474 (m) 424 (w) 227 (w) 462 (m) 416 (m) 202 (w) 448 (w) 406 (m) 180 (m) 414 (m) 392 (m) 170 (m) 406 (s) 366 (wsh) 394 (s) 337 (vw) 297 (m) 293 (msh) 287 (s) 289 (m) 204 (m) 280 (msh) 188 (w) 253 (wsh) 177 (m) 249 (wsh) 149 (s) 226 (w) 221 (wsh) 201 (vw) 192 (m) 171 (m) .144 (m) aAbbreviations: s=strong, m=medium, w=weak, sh=shoulder, v=very. 286 Table 5-2. Calculated and Observed X-ray Powder Diffraction Patterns for KBiP28e6. h k 1 dc cd, A dObid) A I/Imgx (ObSd) 1 0 1 0 2 1 2 1 3 0 0 1 3 0 1 1 2 3 4 3 2 4 4 1 3 2 0 1 4 3 1 5 4 1 5 1 0 2 5 4 3 4 0 5 6 0 1 0 1 0 -2 0 2 0 -2 1 -2 1 1 0 2 2 0 2 0 0 1 2 -1 1 -2 1 2 3 1 2 -2 2 0 1 -3 0 -2 l 2 0 1 1 0 1 -2 2 -3 0 2 2 -3 1 0 4 4 0 -4 2 -3 2 3 0 -2 1 -4 3 1 1 -2 1 -5 2 2 1 1 4 3 0 2 3 -1 2 3 1 3 2 -2 0 -2 11.53 11.53 100 6.34 6.14 5.77 5.13 4.78 4.59 3.895 3.844 3.798 3.607 3.561 3.495 3.432 3.335 3.230 3.172 3.119 3.070 3.028 2.912 2.883 2.846 2.797 2.765 2.734 2.698 2.585 2.565 2.542 2.492 2.479 2.430 2.380 2.357 2.326 2.298 2.235 2.207 2.173 2.159 2.136 2 115 2.076 2.066 6.34 6.15 5.77 5.13 4.78 4.59 3.894 3.841 3.799 3.610 3.562 3.493 3.432 3.329 3.232 3.171 3.119 3.062 3.026 2.912 2.881 2.847 2.796 2.761 2.735 2.695 2.586 2.565 2.541 2.493 2.477 2.430 2.379 2.355 2.325 2.297 2.233 2.208 2.172 2.157 2.134 2.115 2.076 2.064 25 46 35 8 13 44 22 6 7 10 21 18 6 7 8 30 4 38 66 10 61 56 30 4 15 48 7 40 19 21 25 7 25 6 7 8 5 6 11 18 4 4 6 7 Table 5-2. (cont'd) 287 h k l dcglcd, A dobsd, A I/ Imgx (ObSd) 2 5 2 5 4 2 4 5 6 5 4 5 3 2 4 4 3 3 5 1 2 7 6 3 1 2 1 1 -6 1 -5 0 -6 2 4 3 -3 1 2 2 -3 3 -1 2 -6 3 -3 1 -7 4 -3 1 4 1 —7 1 5 4 -3 3 2 1 6 4 -4 2 -3 3 -1 2.049 2.040 1.995 1.981 1.969 1.947 1.898 1.855 1.802 1.757 1.748 1.744 1.723 1.710 1.700 1.689 1.677 1.660 1.624 1.622 1.615 1.603 1.579 2.049 2.036 1.992 1.981 1.971 1.949 1.899 1.852 1.799 1.758 1.750 1.745 1.723 1.711 1.697 1.688 1.677 1.662 1.624 1.620 1.616 1.603 1.579 9 16 6 5 8 6 5 13 7 8 12 16 14 11 10 5 5 6 7 5 4 8 7 288 Table 5-3. Calculated and Observed X-ray Powder Diffraction Patterns for ngSb4( P28e6)5. h k 1 dcglcd, A (10er 1/ [mg (ObSd) 1 1 2 1 1 0 3 1 1 2 3 1 0 2 3 2 2 3 4 0 2 0 1 2 2 0 1 3 6 4 2 5 2 6 7 8 6 6 2 2 0 6 9 4 8 0 1 1 -1 0 0 1 -2 1 1 0 2 3 1 2 -2 2 -3 1 -4 2 1 3 -1 3 2 3 -1 2 -3 3 -2 2 4 2 -4 1 4 3 0 0 -6 4 4 2 1 1 5 4 -1 3 5 0 -7 3 0 4 4 2 2 4 -5 4 1 4 -6 1 -6 0 -5 1 4 2 6 1 2 6 4 4 -2 2 10 5 -3 2 -1 4 8 4 -4 11.10 11.13 8.38 7.71 6.58 6.20 5.25 4.81 4.67 4.05 3.78 3.69 3.66 3.52 3.40 3.33 3.25 3.10 3.01 2.99 2.88 2.85 2.84 2.78 2.73 2.68 2.60 2.53 2.458 2.405 2.200 2.180 2.072 2.024 2.006 1.957 1.871 1.845 1.761 1.701 1.699 1.691 1.675 1.649 1.551 1.546 8.37 7.74 6.57 6.22 5.20 4.83 4.69 4.05 3.79 3.70 3.66 3.52 3.38 3.32 3.25 3.11 3.01 3.00 2.88 2.85 2.82 2.77 2.73 2.68 2.60 2.54 2.433 2.414 2.203 2.181 2.075 2.025 2.008 1.958 1.867 1.849 1.764 1.704 1.695 1.689 1.677 1.654 1.554 1.549 4 3 3 4 7 4 - 6 2 5 5 8 9 6 3 4 5 100 9 11 18 21 6 3 3 3 3 3 6 9 5 4 4 8 5 2 4 7 3 8 10 9 4 3 4 3 289 Table 5-4. Calculated and Observed X-ray Powder Diffraction Patterns for CssBi4( P28e6)5. h kl dcglcd. A dobs_¢_1‘, A I/Imgx (obsd) 1 1 1 2 1 2 1 O 3 1 1 1 2 3 1 0 2 1 3 0 2 3 4 3 4 2 0 1 2 2 1 3 6 4 0 3 5 7 2 6 8 0 6 4 8 O 1 1 -1 1 0 1 0 1 -2 1 1 1 0 0 2 3 1 2 —2 2 2 2 -3 1 -4 2 1 3 -1 3 2 3 -1 2 4 2 -3 3 2 3 4 2 -4 1 -4 3 -2 2 -3 0 -6 4 4 2 4 1 1 5 0 0 -7 3 0 4 2 1 -6 1 8 5 -1 4 1 2 -2 4 -6 4 2 0 -4 2 10 2 -8 4 8 4 -4 11.28 11.27 10 8.43 8.06 7.84 6.61 6.48 6.27 5.29 4.90 4.69 4.56 4.07 3.81 3.73 3.67 3.53 3.41 3.39 3.35 3.23 3.13 3.03 2.98 2.97 2.94 2.87 2.84 2.79 2.76 2.70 2.55 2.482 2.450 2.425 2.190 2.103 2.092 2.073 2.033 1.866 1.858 1.704 1.699 1.567 1.562 8.42 8.05 7.84 6.61 6.47 6.26 5.28 4.89 4.68 4.55 4.06 3.80 3.73 3.66 3.52 3.41 3.38 3.35 3.22 3.12 3.03 3.00 2.97 2.94 2.87 2.84 2.78 2.75 2.69 2.55 2.478 7 3 6 4 3 8 4 7 4 5 4 7 11 6 9 6 7 7 4 100 11 7 5 6 18 5 4 6 4 4 7 2.446 17 2.423 2.185 2.102 2.089 2.072 2.029 1.865 1.855 1.704 1.699 1.565 1.561 5 4 5 6 4 11 7 5 10 10 5 5 Table 5-5. Crystallographic Data for KBiPZSe6, CSng4(ste6)5 and 290 C$8Bi4(PZS€6)5- I I I I I V Formula KBiP2$e6 CS83b4( P28e6) 5 CS8Bi4( P28e6) 5 FW a, A b, A c, A 0:, deg B, deg 7, deg 783.79 4228.8 4577.7 12.403(5) 15.489(3) 15.752(4) 7.595(2) 11.505(2) 11.523(2) 12.412(3) 17.772(3) l7.916(2) 900(1) 900(1) 900(1) 111.58(3) 9559(2) 9520(2) 90.00) 90.00) 90.00) 2; v, A3 4; 1087(1) 2;3152 2;3241 A 0.71073 0.71073 0.71073 space group P21/C (NO. 14) P21/n (NO. 14) P21/n (NO. 14) Dcalc, g/cm3 4.788 4.455 4.694 p, cm-1 366.75 236.55 (Mo [(01) 321.67 (Mo [(01) Zemax, deg 50 (Mo 1(a) 50 (Mo Ka) 50 (Mo [(11) Temp., 0C 23 -80 23 Final R/Rw, 96 3.9/4.S 3.2/4.9 5.4/6.7 Total Data Measured 2170 Total Unique Data 2070 Data F02>36(1=62) 1228 No. of Variables 92 61(1) 5891 3544 236 6261 6028 2598 236 * R=Z(|Fol-|Fcl)/Z|Fol Rw={EW( lFol-IFCI)2/2wlFO|2}1/2 Table 5-6. Fractional Atomic Coordinates and Beq Values for KBiP28e6 with Estimated Standard Deviations in Parentheses. 291 atom x y 2 B(GQ) Bi 0.11808(7) 0.1546(1) 0.83190(7) 2.05(3) Se(l) 0.1627(2) 0.5064(3) 0.9237(2) 1.75(8) Se(Z) 0.3109(2) 0.0012(3) 1.0618(2) 227(9) Se(3) 0.3371(2) 0.2001(3) 0.8056(2) 1.91(8) Se(4) 0.0576(2) 0.3 1 18(3) 0.61 16(2) 1.79(8) Se(S) 0.1159(2) -0.1839(3) 0.6871(2) 1.73(8) Se(6) 0.4097(2) 0.7311(3) 0.8841(2) 1.87(8) K 0.4089(4) 0.5818(7) 1.1387(4) 2.6(2) P(1) 0.2377(4) 0.6271(8) 0.8037(4) 1.7(2) P(2) 0.2397(4) 0.4064(7) 0.6850(4) 1.5(2) 243...; = (4/3 )[a213( 1,1) + b2B(2,2) + c28(3,3) + ab(c08y)B(1,2) + ac(cosB)B(1,3) + bc(c03a)B(2,3)]. Table 5-7. Fractional Atomic Coordinates and Beq Values for C53$b4( P28e6)5 with Estimated Standard Deviations in Parentheses. 292 atom x y z B(eq) Cs(l) 0.4041(1) 0.2492(2) 0.2902(1) 1.54(7) Cs(2) 0.0910(1) 0.2202(1) 0.2101(1) 1.20(6) Cs(3) 0.6898(1) 0.2361(1) 0.4081(1) 1.12(6) Cs(4) 0.9859(1) 0.2475(1) 0.5015(1) 1.23(6) Sb(l) 0.7848(1) 0.1302(1) 0.15902(8) 0.59(6) Sb(2) 0.8176(1) 0.3399(1) 0.02670(8) 0.75(6) Se( 1) 0.8796(2) 0.2477(2) 0.2941(1) 0.9( 1) Se(2) 0.7798(2) 0.2365(2) 0.6027(1) 1.0( 1) Se( 3) 0.6700(1) 0.0065(2) 0.0655(1) 0.8( 1) Se(4) 0.7796(1) -0.0265(2) 0.2676(1) 0.8( 1) Se(S) 0.8790(1) 0.5053(2) -0.0537(1) 0.9(1) Se(6) 0.6707(1) 0.4752(2) 0.0412(1) 0.76(9) Se(7) 0.4881(2) 0.2660(2) 0.5043(2) 1.0( 1) Se( 8) 0.6179(2) 0.2145(2) 0.1974(1) 0.9( 1) Se(9) 0.57 14(1) -0.0364( 2) 0.3475(1) 1.0( 1) Se( 10) 0.2645(2) -0.0173(2) 0.2578(1) 1.2( 1) Se(ll) 1.2075(2) 0.2327(2) 0.4060(2) 1.1( 1) Se(12) 1.0322(1) 0.0122(2) 0.3613(1) 1.1(1) Se(13) 0.5661(1) 0.5026(2) 0.3666(1) 0.8(1) ‘ Se(14) 0.3297(1) 0.4880(2) 0.4408(1) 1.2( 1) Se( 15) 1.0502(1) 0.4977(2) 0.3695(1) 1.4( 1) P( 1) P(2) P(3) P(4) PLS) 1.1547(4) 0.0536(5) 0.4170(3) 0.8(2) 0.5823(3) 0.0399(5) 0.1556(3) 0.5(2) 0.4423(3) 0.4463(5) 0.5136(3) 0.4(2) 0.6380(4) -0.0775(5) 0.2488(3) 0.5(2) 0.2450(3) 0.4385(5) 0.1269(3) 0.6(2) aBeq = (4/3)[a2B(1,1) + b2B(2,2) + c2B(3,3) + ab(cosY)B(1,2) + ac(coss)B(1,3) + bC(COSa)B(2,3)]. Table 5-8. Fractional Atomic Coordinates and Beq Values for CSgBi4(P2Se6)5 with Estimated Standard Deviations in Parentheses. 293 atom x y z B(eq) Cs( 1) 0.4040(2) 0.243 1(3) 0.2888(2) 3.2( 1) Cs(2) 0.0916(2) 0.2203(3) 0.2102(2) 2.8( 1) Cs(3) 0.6903(2) 0.2363(3) 0.4086(2) 2.7( 1) Cs(4) 0.9861(2) 0.2478(3) 0.5019(2) 2.7( 1) Bi( 1) 0.7869(1) 0.1275(1) 0.16063(8) 1.94(6) Bi(2) 0.8180(1) 0.3368(2) 0.0215(1) 2.49(7) Se( 1) 0.8808(3) 0.2458(4) 0.2945(2) 2.1(2) Se(2) 0.7796(3) 0.2393(4) 0.603 1(2) 2.4(2) Se(3) 0.6654(3) 0.0033(4) 0.0650(2) 2.1(2) Se(4) 0.7768(2) -0.0336(4) 0.2697(2) 2.0(2) Se(S) 0.8763(3) 0.5093(4) -0.0605(2) 2.4(2) Se(6) 0.6694(3) 0.4765(4) 0.0393(2) 2.1(2) Se(7) 0.4936(3) 0.2662(4) 0.5090(3) 2.4(2) Se(8) 0.6171(3) 0.2128(4) 0.1974(2) 1.9(2) Se(9) 0.5726(3) -0.0395(4) 0.3472(2) 2.6(2) Se( 10) 0.2590(3) -0.02 16(4) 0.2558(2) 2.6(2) Se( 11) 1.2010(3) 0.2304(4) 0.4013(3) 2.5(2) Se(12) 1.0328(3) 0.0072(4) 0.3610(2) 2.5(2) Se( 13) 0.5595(3) 0.5008(4) 0.3647(2) 2.2(2) Se(14) 0.331 1(3) 0.4777(4) 0.4477(2) 2.9(2) Se(15) 1.0481(3) 0.4970(4) 0.3649(3) 2.8(2) P( 1) P(2) P(3) P(4) P(5) 1.1543(6) 0.0501(9) 0.4161(5) 1.6(4) 0.5831(6) 0.038(1) 0.1566(5) 1.7(4) 0.4467(6) 0.445 1(8) 0.5 174(5) 1.4(4) 0.6379(6) -0.0789(8) 0.2491(5) 1.3(4) 0.2475(6) 0.4363(9) 0.1292(5) 1.6(4) aBeq = (4/3)[a2B(1,1) + b2B(2,2) + c28(3,3) + ab(cosY)B(1,2) + ac(cos|3)B(1,3) + bc(cosa)B(2,3)]. 294 3. Results and Discussion 3.1. Synthesis, Spectroscopy and Thermal Analysis The syntheses of these compounds were a result of a redox reaction in which the metal is oxidized by polyselenide ions in the Ax[PySeZ] (A=K, Cs) flux to' give M3+ species. KMPzSe6 crystallizes from a Lewis acidic flux with a M/PzSes/KZSe/Se ratio of 1/ 3/ 2/ 8. In the case of C83M4(P2Se6)5, a more basic flux consisting of M/ PzSes/CsZSe/ Se (1/3/2/4) was used to obtain pure compounds. For CssBi4( P28e6)5, decrease in the amount of P2Se5 resulted in Bi28e3 under these conditions. The far-IR spectra of KMPZSe6 and C53M4(P28e6)5 are characteristically complex and show a large number of absorptions between 522 and 144 cm‘1 (see Table 5-1). Since systematic IR spectroscopic data for selenophosphate ligands are lacking, and thus we cannot assign P-P and -PSe3 vibrational frequencies. However, it is tempting to assign the first two or three higher frequency vibrations to modes due primarily to P-P stretching.19 The optical properties of (I )-(I V) suggest that they are wide band-gap semiconductors. The optical absorption spectrum shown in Figure S-la indicates that KBiPZSe6 is optically transparent below the band gap, E3, of 1.61 eV. The E8 of KSbPZSe6 is found to be 2.00 eV, see Figure 5-1b. The band-gaps, Eg, of CssSb4(P28e6)5 and C838i4(P2Se6)5 are 1.58 and 1.44 eV, respectively (see Figure 5-2). The slight decrease in band gap in going from Sb to Bi in C53M4(P25e6)5 suggests that these metals contribute comparatively 295 , little to the highest energy levels of the valence band, which is most likely dominated by Se-based orbitals. Three absorptions at 2.33, 3.67, 5.37 eV for (I I I) and two absorptions at 2.11 and 5.26 eV for (IV) are readily resolved and are assigned to electronic Se-->M charge transfer transitions. Based on the calculated solar conversion efficiency versus band gap plot,20 the band gaps of KBiPZSe6 and C53M4( P28e6) 5 (1.44-1.61 eV) would fall near the peak of conversion efficiency. Differential thermal analysis (DTA) shows that KBiPZSeé, KSbPZSe6, CS3Sb4( P28e6)5, and CssBi4( PZSe6)5 melt congruently at 554, 479, S32, and 497 °C respectively, suggesting that large single crystals or microcrystalline thin films can be grown from the melt. (A) ) s t i n u . b r a ( . f f c o C n o i t p r o s b A S / u 296 0.5~ .. 0 7 01 (B) 2.5 1 l L 1 l l .— _ __ 27 'E 3 :5 $- 5 2-1 3:; 1.5- 8 D g g 8 < 0.5" U: B O 1_ 856928.. lit-2.111%! f _ I I I T I O 1 2 3 4 5 6 7 Figure 5-1. Solid state optical absorption spectra of (A) KBIPZSee and (B) KSbPZSee. 297 (A) 5 l 1 1 1 1 1 ’23 '5 e' 4‘ S E 3— 8 E; 2 2_ Casmtrzseas E‘-1.58CV " _ _ r _ I I I I I o 1 2 3 4 5 6 7 Enersflev) E- 8 .0 < 1- <4 5 o (B) ) s t i n u . b r A ( . f f e o C n o i t p r o s b A S / a 6 L g 21! l 1 1 l E‘=l.44eV ) a ( l 2 "‘ r— T r 7 i _ 1 I 1 - 0 o 1 . 2 3 4 5 6 7 Energy (eV) Figure 5-2. (A) Optical absorption spectrum of C53Sb4(P2Se6)5. (B) Optical absorption spectrum of CSgBi4( PzSe5)5. 298 3.2. Description of Structures Structure of KMP28 e6. KMPZSe6 has the complicated layered structure shown in Figure 5-3. The layers are separated by seven- coordinate 10 ions (K-Se mean=3.36(2)A). The [MPZSe6]- slabs are assembled from chains of distorted corner-sharing BiSe6 octahedra along the b-axis (see Figure 5-4a). The [PZSe6]4- ligand trichelates to one Bi while a fourth Se atom acts as a bridge between two other Bi atoms in the chain. The fifth Se is non bonding while the sixth Se connects to a Bi on the neighboring chain to form a layer in the bc- plane (see Figure 5-3). For clarity, the unique connectivity of [P28e6l4' is illustrated separately in scheme (A). Scheme (A) Selected distances and angles for KMP2Se6 are found in Tables 5-9. Bi is coordinated by six Se atoms to form a distorted octahedron. The Bi-Se distances range from 2.821(2) to 3.194(3)A and compare well with those found in KzBi3Se1321 and C83Bi7Se12.22 The distortion of the octahedron can partly be explained by the stereochemically active 6s2 lone pair of Bi3+. The Bi-Se coordination sphere is shown in Figure S-4b. In accordance with the VSEPR model,23 lone pair-bond pair repulsions in the bismuth coordination 299 sphere result in longer bond lengths for those bonds adjacent to the lone pair (Bi-Se(2,5,5') 3.079(2)-3.194(3)A) and shorter lengths for those more remote (Bi-Se(1,3,4) 2.821(2)-2.878(2)A). In addition, these repulsions give rise to an opening of the bond angles straddling the lone pair (Se(S)-Bi—Se(5)' 118.32(5)°, Se(2)-Bi-Se(5) 126.07(6)°). The P(1)-P(2) bond in [PzSe6]4' is 2.238(7)A and the P-Se distances range from 2.145(5) to 2.223(S)A. Similar distances have been observed in ngP 28e6.24 Structure of C83M4( P28e6)5. C53M4(P28e6)5 possess a remarkably complicated layered structure shown in Figure 5-5 . The [M4( P28e6) 5],, 811' layers are composed of [M4(P28e6)4] zig-zag chains that are linked by additional [P28e6]4' ligands to form a staircase layered framework. The layers are interdigitated to form tunnels along the b-direction. There are two kinds of cesium ions, one that fills the tunnels and another located within the layers. There are three types of [PzSe6]4' ligands in this structure each possessing unique denticity. Type (1) is a capping ligand that trichelates to the metal while the other three Se atoms are ' nonbonding (terminal). Type (2) also trichelates to the metal while two of the remaining Se atoms bond to two separate metals and the sixth Se is terminal. Type (3) chelates to four metals leaving two nonbonding Se atoms. The P-P bond resides on a center of symmetry. For clarity, the connectivities of the [P2Se6]4° ligands are illustrated separately in scheme (B). It is clear that beyond the well known Mszcx, class of compounds and the few other examples that already exist, the [PZCLF' unit presents excellent new possibilities as a building block of new frameworks. These possibilities can be readily exploited by the type of molten salt reaction reported here. 300 Se Se Se M :DSe Se 4_/ / ix. \M// Sc 4_/ ,6/ i\. \M// Type 1 Type 2 M 5 e ‘ SC / e 5 M P—— P M Se/ S\e\ 33 M Type 3 Scheme (B) One of the most distinguishing features found in this structure are the rather short M---M interactions at 3.5 34(2) A for Bi and 3.441(2) A for Sb (see Figure 5-6). For comparison, the sum of the van der Waals radii for Bi and Sb are 4.8 and 4.4 A. Therefore, the obvious question to ask is are these significant bonding interactions or not? To the best of our knowledge, other examples of such remarkably short distances involving these elements are found only in the corresponding elemental structures. Bi metal has an intralayer bond distance of 3.06 A and an interlayer distance of 3.52 A. The 301 isostructural Sb metal has distances of 2.90 A and 3.34 A. The longer distances in these metals are considered weakly bonding in nature.7-5 We note however that the metals in these forms are in the zero oxidation state while (I I I) and (I V) are formally in the 3+ state. If a metal-metal bond exists in these compounds, it must form via the interaction of the lone pairs in Sb3+ and Bi3+. The M3+ (d10s2) ions with square pyramidal coordination exhibit a stereochemically active lone pair at the base of the pyramid. The bases of the square pyramids in C83M4(P2Se6)5 face each other in close proximity. A weak interaction of this sort is unusual since lone pair/ lone pair interactions should be non-bonding. Perhaps a closer analog to the case described here is the afi-PbO where weak Pb---Pb bonding interactions (interlayer Pb---Pb distance of 3.87 A) were proposed to exist by Hoffmann and coworkers.26 Closed-shell bonding interactions have also been observed in other systems such as 2,2',5,5'-tetramethylbistibole ((C6H3Sb)2) in which the lone pairs in each Sb interact with other dimers by Sb---Sb interactions (3.625 A) to form one-dimensional chains.27 This Sb-Sb bond length alternation is the result of a Peierls distortion as determined from electronic band structure calculations.28 Tetramethyldistibane exhibits similar behavior with an intermolecular Sb-Sb distance of 3.678(1)A29 while tetrakis(t:1imethylsily)dibismuthane shows a Bi-Bi distance of 3.804(3) A.30 Although these M---M distances are longer than those observed in (I I I) and (I V) they are similar in origin. It is noteworthy that the case of M---M interaction is unique in (111) and (IV) in the sense that they are the only examples where "inert" n82 lone pairs face directly at each other. Other closed-shell systems 302 . such as the classical d10--d10 (Cu+, Ag+, Au+) interactions have also been well documented.31:32 Extended Huckel calculations by Hoffmann and coworkers have shown that mixing of empty 4s and 4p orbitals with occupied 3d orbitals lowers the energy of the o and 0* orbitals for several copper complexes.33 A similar type of interaction, due to mixing with empty p orbitals, could exist in C53M4(P2Se6)5. It has also been proposed for several d10-d10 complexes that the bridging ligands bring the metals closer in order to maximize M-L interactions.34 Since C83M4(ste6)5 features two [P23e6]4- bridging ligands for each dimer, it is possible that the multidentate [P2Se6]4' coordination is playing an important role. It is not clear whether the interactions contribute significantly in determining the structure or the metals are simply forced to tolerate the short contact due to packing forces. The rest of the structure is built up as follows. Type 1 ligands cap a metal on each M---M dimeric unit (see figure 5-6) while one Type 2 ligand connects two dimeric units together (see Figure 5-7). These larger units are assembled into [M4(PZSe6)4] zig-zag chains by additional Type 2 ligands. The zig—zag chains are connected by Type 3 ligands above and below to form the [M4( P2Se6)5]n 311' staircase layers as shown in Figure 5-8. Selected distances and angles for ngM4(PZSe6)5 are found in Tables 5-10 and 5-11. M(1) and M(2) are found in two similar square pyramidal coordination sites (see Figure 5-6). The M atoms are slightly displaced from their respective square planes. The bond distances of CSng4(P25€6)5 range from the short axial bond distances of 2.648(3) and 2.613(3)A to 3.010(3) and 3.039(3) A for Sb(l) and 303 Sb(2), respectively. These compare well with the square pyramidal Sb-Se distances found in TISbSe2.3S The bond distances for the Bi analog range from 2.711(4) and 2.682(5) A to 3.027(5) and 3.037(5) A which compare well with those found in K2Bi3Se13.21 The P-P bonds of the [P2Se6]4- ligands are normal at 2.25( 1) A (average). The terminal Se atoms have significantly shorter P-Se distances due to partial double bond character.36 For example, the terminal P(1)-Se(12) distance is 2.108(7)A while the metal binding Se(6) and Se(ll) atoms have P(1)-Se(6) and P(1)-Se(11) distances of 2.222(6) and 2.233(6) A, respectively. In CssBi4(P2Se6)5, Cs(1,2,4) are located in the tunnels that are created by the interdigitated layers. Cs(l) and Cs(2) are ten- coordinate with average Cs-Se distances of 3.9(1) A and 3.8(1) A, respectively. The coordination environment of Cs( 1) resembles a tri- capped truncated square prism. Cs(2) possesses a bi-capped square prismatic coordination. The nine-coordinate Cs(3) (ave. 3.8(1) A) resides within the layers in the center of a mono-capped square prism. Cs(4) is found in a 12-coordinate tetra-capped square prism with an average Cs(4)-Se distance of 3.93(8) A. The Cs-Se coordination environments are shown in Figure 5-9. The difference between the sulfur and selenium chemistry in these quaternary systems is partly due to the difference in oxidizing ability of the excess polychalcogenide ligands in these fluxes. In this temperature range polythiophosphate fluxes yield [P2S7]4—1 and [PS4]3-37 ligands with P in the 5+ oxidation state whereas the polyselenophosphate fluxes stabilize the [PzSe6]4' unit with P in the 4+ oxidation state. This leads to entirely different structural 304 chemistry. 305 Figure 5-3. The extended structure of KBiPZSe6 looking down the b—axis with labeling. 306 ‘ /11° \ {57.4sax/(f: / bridging Se (B) Q Sc3 Se5' Se2 L... 3.. SeS , Figure 5-4. (A) Structure and labeling of a [BiPZSeel' chain. The bridging Se atoms connecting the chains into layers are highlighted. (B) The Bi—Se coordination site with labeling. 307 74:70}. 5.175" 9 - 55.. -“7'85.“ 8,! ;/.“ \ - 55.. ' gr Figure 5-5. Packing diagram of (153Bi4(P7_Se6)5 viewed down the [010] direction. The open circles represent Cs atoms. Dark solid lines highlight the layered nature of the staircase framework. 308 Sc3' Sc8' . P2! 9 SclSN 'Sc4' P4' ‘ Se9' Sci 8152 ‘s'. Sc8 8:9 “\Seé Sc15'/ p4 ____“ )qu— - — -1112 70- Ml ~52) Fly Sc4 \l 9Se3 57 Sell 3“ 5.13 6!: 5:7 3"” P3 5°7' 5:13' Figure 5-6. Structure of the [M2(P2Se6)4] dimeric unit showing the MmM interaction. 309 (A) Figure 5-7. (A) Association of two [M2(P25e6)4] "dimers" by a [P2Se614' unit. (B) Further sharing of [P28e614‘ units among "dimers" results in a one-dimensional chain- iike structure. 310 Figure 5-8. Two-dimensional structure of the [M4(P28e6)5]n8n- anionic framework with labeling as drawn by ORTEP. 311 Csl Figure 5-9. The Cs+ coordination environments. The open circles represent Cs atoms. The open square in the Cs(l) coordination site represents the vacant comer of the square prism. Table 5-9. Selected Distances (A) and Angles (°) in KBiP28e6 with Standard Deviations in Parentheses.a 312 Bi-Se( 1) 2.878(2) Se(1)-Bi-Se(2) 9075(6) Bi-Se(2) 3.194(3) Se(1)-Bi-Se(3) 83.17(6) Bi-Se( 3) 2.873(2) Se(1)-Bi-Se(4) 8626(6) Bi-Se(4) 2.82 1(2) Se(1)-Bi-Se(5) 72.48(6) Bi-Se(5) 3.131(2) Se(2)-Bi-Se(3) 7354(6) Bi-Se( 5') 3.079(2) Se(2)-Bi-Se(5) 9338(6) Bi-Se (ave.) 3.0(2) Se(2)-Bi-Se(5') 126.07(6) Se(3)-Bi-Se(4) 76.37(6) P(1)-Se( 1) 2.223(5) Se(3)-Bi-Se( 5) 8093(6) P(1)-Se(5) 2.198(6) Se(4)-Bi-Se(5) 8125(6) P(1)-Se(6) 2.145(5) Se(4)-Bi-Se(5') 81.39(6) P(2)-Se(2) 2.146(5) Se(5)-Bi-Se(5') 118.32(5) P(2)-Se(3) 2.197(6) P(2)-Se(4) 2.221(6) Se(1)-P(1)-Se(5) 110.8(2) P-Se (ave.) 2.19(3) Se(1)-P(1)-Se(6) 115.5(2) P(1)-P(2) 2.238(7) Se(2)-P(2)-Se(3) 116.9(2) Se(5)-P(1)-Se(6) 113.0(3) Se(2)-P(2)-Se(4) 115.9(2) K—Se(1) 3.283(5) Se(3)-P(2)-Se(4) 105.7(2) K-Se( 2) 3.418(6) Se(1)-P(1)-P(2) 104.0(3) K-Se(3) 3.320(5) Se(5)-P(1)-P(2) 103.1(2) K-Se(3') 3.398(5) Se(6)-P(1)-P(2) 109.4(3) K-Se(6) 3.362(5) Se( 2)-P( 2)-P( 1) 109.4(3) K-Se(6') 3.357(5) Se(3)-P(2)-P( 1) 102.8(2) K-Se(6") 3.356(5) Se(4)-P( 2)-P( 1) 104.9(2) K-Se (ave.) 3.36(5) alThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01={2n(1n- 1)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. Table 5-10. Selected Distances (A) in C53M4(P2Se6)5 with Standard Deviations in Parentheses.a C38Sb4(PZS€6)S C§8Bi4(£2$e6)5 M-M' 3.441(2) 3.534(2) M( 1 )-Se( 1) M(1)-Se(3) M( 1 )-Se(4) M( 1 )-Se( 8) M(1)-Se(13) 3.010(3) 2.716(3) 2.648(3) 2.905(3) 2.811(3) M(1)-Se (ave.) 2.8(1) M(2)-Se(5) M(2)-Se(6) M(2)-Se(7') M(2)—Se(9') M(2)-Se(11) 2.613(3) 2.789(3) 2.970(3) 3.039(3) 2.739(3) M(2)-Se (ave.) 2.8(2) P(1)-P(5) P(2)-P(4) P(3)-P(3') P-P (ave.) P(1)-Se(6) P(1)-Se(1 1) P(1)-Se(12) P(2)-Se(3) P(2)-Se(8) P(2)-Se(15') P(3)-Se(7) P(3)-Se(13) P(3)-Se(14) P(4)-Se(1') P(4)-Se(4) P(4)-Se(9') P(5)-Se(2) P(5)-Se(5) P(S)-Se(10) P-Se (ave.) 2.240(7) 2.244(8) 226(1) 225(1) 2.222(6) 2.233(6) 2.108(7) 2.230(6) 2.194(6) 2.114(6) 2.204(6) 2.225(6) 2.123(6) 2.158(6) 2.264(6) 2.171(6) 2.139(6) 2.305(6) 2.130(6) 2.19(6) 3.027(5) 2.838(4) 2.711(4) 2.978(5) 2.898(5) 2.9(1) 2.682(5) 2.883(5) 3.037(5) 3.132(5) 2.813(5) 2.9(2) 2.24(1) 225(1) 224(2) 224(1) 222(1) 2.23( 1) 2.13(1) 222(1) 220(1) 2.12(1) 220(1) 221(1) 2.15(1) 2.l8(1) 225(1) 2.16(1) 2.15(l) 2.30(1) 2.13(1) 2.19(5) Table 5-10. (cont'd) CS8§b4(P2$e6)5 g8Bi4(PZSe6)5 Cs(1)-Se(2) Cs(l)-Se(5) Cs(1)-Se(7) Cs(l)-Se(8) Cs(1)-Se(10) Cs(1)-Se(10') Cs(1)-Se(ll) Cs(l)-Se(13) Cs(1)-Se(14) Cs( 1 )-Se( 15) Cs(1)-Se (ave.) Cs(2)-Se(1) Cs(2)-Se(2) Cs(2)—Se(7) Cs(2)-Se(9) Cs(2)-Se(10) Cs(2)-Se(10') Cs(2)-Se(11) Cs(2)-Se(12) Cs(2)—Se(13) Cs(2)-Se(14) Cs(2)-Se (ave.) Cs(3)-Se(l) Cs(3)-Se(2) Cs(3)-Se(3) Cs(3)-Se(5) Cs(3)-Se(6) Cs(3)-Se(7) Cs(3)-Se(8) Cs(3)-Se(9) Cs(3)-Se(13) Cs(3)-Se(14) Cs(3)-Se (ave.) Cs(4)-Se(1) Cs(4)-Se(2) Cs(4)-Se(3) Cs(4)-Se(3') Cs(4)-Se(6) Cs(4)-Se(6') 3.687(3) 4.077(3) 3.904(4) 3.861(3) 3.763(3) 3.786(3) 3.842(3) 3.998(3) 4.079(3) 4.161(3) 3.9(2) 3.743(3) 3.679(3) 3.849(4) 3.836(3) 3.868(3) 3.769(3) 3.765(4) 3.775(3) 3.661(3) 4.059(3) 3.8(1) 3.730(3) 3.600(3) 3.797(3) 3.928(3) 3.757(3) 3.721(3) 3.809(3) 3.738(3) 3.653(3) 4.188(3) 3.8(2) 3.888(3) 3.815(3) 4.098(3) 3.946(3) 3.855(3) 3.991(3) 3.710(6) 4.019(6) 4.077(5) _ 3.883(5) 3.824(6) 3.769(6) 3.933(6) 4.008(5) 4.162(6) 4.071(6) 3.9(1) 3.782(6) 3.700(6) 4.172(5) 3.793(6) 3.864(6) 3.872(6) 3.807(6) 3.691(6) 3.827(5) 3.644(5) 3.8(1) 3.783(6) 3.638(5) 3.829(6) 3.983(5) 3.788(5) 3.737(6) 3.863(5) 3.792(6) 3.723(6) 3.8(1) 3.929(5) 3.866(6) 4.111(5) 3.908(5) 3.887(5) 3.999(5) Table 5-10. (cont'd) C88$b4( P236615 Q8314 P2366)5 Cs(4)-Se(8) Cs(4)-Se(11) Cs(4)-Se(12) Cs(4)-Se(12') Cs(4)-Se(15) Cs(4)-Se(15') 3.886(3) 3.978(3) 3.795(3) 3.884(3) 3.904(3) 3.795(3) Cs(4)-Se (ave.) 3.90(9) 3.928(5) 3.982(6) 3.865(6) 3.858(5) 3.957(6) 3.855(5) 3.93(8) 4The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01={2n(1n- DZ/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 316 Table 5-1 1. Selected Angles (0) in C88M4(P2$e6)5 with Standard Deviations in Parentheses. C88$b4(P2$e6)5 CssBi4(P2$e6)5 Se(1)-M(1)-Se(4) 76.78(8) Se(1)-M(1)-Se(8) 9221(8) 78.1( 1) 93.4(1) Se(1)-M(1)-Se(4) 91.41(8) 905( 1) Se(1)-M(1)-Se(4) 9729(8) 100.3(1) Se(5)-M(2)-Se(9) 81.48(8) Se(5)-M(2)-Se(11) 91.49(9) Se(6)-M(2)-Se(9) 9453(8) 832(1) 91.0(1) 93.4( 1) Se(7)-M(2)-Se(11) 10529(9) 112.0(1) M(1)-Se(3)-P(2) 83.1(2) M(1)-Se(13)-P(3) 982(2) M(1)-Se(1)-P(4') 95.8(2) 82.6(3) 96.7(3) 95.3(3) M(2)-Se(9)-P(4') 102.3(2) 104.1(3) M(2)-Se(7)-P(3) 945(1) M(2)-Se(6)-P( 1) 80.7(2) 92.7(3) 80.1(3) Se(6)-P(1)-Se(11) 102.6(3) 103.9(4) Se(6)-P(1)-Se(12) 116.2(2) 116.1(5) Se(3)-P(2)-P(4) 102.1(3) 103.2(5) I Se(8)-P(2)-Se(15') 119.2(3) 117.9(5) Se(13)-P(3)-P(3') 100.4(4) 101.9(6) Se(7)-P(3)-Se(13) 111.7(3) 111.9(4) Se(1)-P(4')-Se(9) 115.8(3) 1 15.3(5) Se(1)-P(4')-Se(4') 112.4(2) 111.96) 317 3.3. Electrical Conductivity Measurements. Four-probe electrical conductivity measurements on single crystals of C53Bi4(P28e6)5 showed that the material is a semiconductor with room temperature conductivity o~10‘9 S/ cm which drops to 10-10 S/cm at 253 K. Figure 5-10 shows the conductivity vs. 1000/T semilog plot for CS3Bi4( P2Se6) 5. The data can be fit to the equation shown below suggesting a gap of Eg=1.135 eV and an activation energy of Ea=05 67 eV. _ 0—006 Ell k T 8 eq. (1) The value of 1.135 eV compares to 1.44 eV determined by optical spectroscopy (vide supra). The reason for the discrepancy is not known but may be due to the fact that the electrical conductivity does not represent the intrinsic semiconducting regime but may arise from impurity levels in the gap. It must be noted that the optical measurement is the most reliable Eg indicator. The antimony compound is an insulator. 318 10'9 .. . E Q 2’, O 10'10111111111111L1111111|111 3.25 3.3 3.35 3.4 3.45 3.5 1000rr (K") Figure 5-10: Variable temperature electrical conductivity data for single crystals of CsaBi4( PzSe6)5. 3.4. Conclusions 3 19 In conclusion, the synthesis of new quaternary selenophosphates with A/ P/ Se molten salts provides a useful synthetic approach with broad scope. For example, the new family of AzMPzSe6 (A=K, Rb, Cs; M=Mn, Fe) compounds has been synthesized in this way.38 The advantage of Ax[PySeZ] fluxes is that they provide reliably [P2Se6]4‘ units which other conventional aqueous or organic solvents have trouble stabilizing due to the high negative charge. Thus, solid state and coordination chemistry with [PzSe6]4' and metal centers becomes tractable based on the method highlighted in this chapter. It would be interesting to see whether changes in the nominal stoichiometry of these fluxes can result in other [PxSeyPl' units as well. 320 List of References - : w a s - 4 . 10. 11. 12. 13. 14. 15. 16. l7. 18. 19. T. J. McCarthy and M. G. Kanatzidis, Chem. Mater. 1993, 5,1061-1063. Garin, J.; Parthe, E. Acta Crystallogn, 1972, 828,3672-3674. Fritz, I. J.; Isaacs, T. J.; Gottlieb, M.; Morosin, B. Solid State Commun., 1978, 27,535. O'Neal, S. C.; Pennington, W. T.; Kolis, J. W. Angew. Chem. Int. Ed. Engl. 1990, 29,1486-1488. (a) Klingen, W.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem., 1973, 401, 97-112. (b) Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst, Sect. B, 1978, 34, 1779-1781. (c) Klingen, W.; Ott, R.; Hahn, H. Z. Anorg. Allg. Chem., 1973, 396, 271-278. (d)Jandali, M. Z.; Eulenberger, G.; Hahn, H. Z Anorg. Allg. Chem., 1978,447,105-118. (a) Ouvrard G. Brec, R.; Rouxel, J. Mater. Res. Bull., 1985, 20, 1181- 1189. (b) Lee, S.;Colombet,P.;Ouvrard,G.;Brec,R. Inorg. Chem., 1988,27, 1291-1294. (c) Lee, 8.; Colombet, P.; Ouvrard, G. Brec, R. Mater. Res. Bull, 1986, 21, 917-928. (d) Durand E.; Ouvrard, G.; Evain, M.; Brec, R. Inorg. Chem., 1990, 29,.4916—4920 (a) Carpentier, C. D.; Nitsche, R. Mater. Res. Bull. 1974, 9, 401-410. (b) Carpentier, C. D.; Nitsche, R. Mater. Res. Bull. 1974, 9, 1097- 1100. (c) Arnautova, E; Sviridov, E.; Rogach, E. Savchenko, E.; Grekov; A. Integrated Ferroelectrics, 1992,],147-150. (a) Becker, R.; Brockner, W.; Schafer, H. Z. Naturforsch. 1983, 383, 874- 879. (b) Becker, R.; Brockner, W.; Schafer, H. Z. Naturforsch. 1984, 39a, 357-361. (c) Yun, H.; Ibers, J. A. Acta Cryst. 1987, C43, 2002-2004. (a) Katty, A.; Soled, S.; Wold, A. Mater. Res. Bull., 1977, 12, 663-666. (b) Diehl, R.; Carpentier, C. D. Acta Cryst, Sect. B, 1978, 34,1097-1105. Etman, M.; Katty, A.; Levy-Clement, C.; Lemasson, P. Mater. Res. Bull, 1982, 17579-584. Thompson, A. H.; Whittingham, M. S. U. S. Patent 4,049,879 1977. (b) Brec, K; Le Mehaute', A. Fr. Patents 7, 704,519 1977. (a) Clement R. J. Chem. Soc., Chem. Commun. 1980, 647-648. (b) Michalowicz, A.; Clement R. Inorg. Chem. 1982, 21 , 3872-3877. (c) Joy, P A.; Vasudevan, S. J. Am. Chem. Soc., 1981, 114,7792-7801. Pfeiff, R.; Kniep, R. J. Alloys and Compounds, 1992, 186,111-133. Feher, F. Handbuch der Praparativen Anorganischen Chemie: Brauer, G., Ed.; Ferdinand Enke: Stuttgart, Germany, 1954; pp. 280-281. CERTUS: Molecular Simulation Software, Version 3.0, (1992), Cambridge Molecular Design, Waltham, MA 02154. G. M. Sheldrick, In Crystallographic Computing 3; Sheldrick, G. M., Kruger, C., Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985; PP 175-189. TEXSAN: Single Crystal Structure Analysis Software, Version 5.0, (1981). Molecular Structure Corportion, The Woodlands, TX 77381. Walker, N.; Stuart, D. Acta Cryst., 1983, A39,158-166. The characteristic stretching frequency range for P-P interactions of similar interatomic distances is about 400-600 cm‘l. K. Nakamoto in "Infrared and Raman Spectra of Inorganic and Coordination Compounds", Wiley, New York, 1978. 20. Loferski, J. J. J. Appl. Phys., 1956, 27,777. 321 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. McCarthy, T. J.; Ngeyi, S.-P.; Liao, J.-H.; DeGroot, D.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Mater., 1993, 5,331-340. G. Cordier, H. Schafer and C. Schwidetzky, Rev. Chem. Miner., 1985, 22 ’ 676. (a) R. J. Gillespie, Can. J. Chem., 1960, 38, 818. (b) R. J. Gillespie, J. Chem. Educ, 1963,40, 295. (c) R. J. Gillespie, Molecular Geometry; Van Nostrand Reinhold; London, 1972; p. 6. M. Z. Jandali, G. Eulenberger and H. Hahn, Z. Anorg. Allg. Chem., 1978, 447,105. (a) Barrett, C. S.; Cucka, P.; Haefner, K. Acta Cryst. 1963, 16, 451-453. (b) Cucka, P.; Barrett, C. S. Acta Cryst. 1962, 15, 865-872. Trinquier, G.; Hoffmann, R. J. Phys. Chem., 1984, 88,6696-6711. Ashe III, A. J.; Butler, W.; Diephouse, T. R. J. Am. Chem. Soc., 1981, 103, 207—209. Hughbanks, T.; Hoffmann, R.; Whangbo M.-H.; Stewart, K. R.; Eisenstein O.; Canadell, E. J. Am. Chem. Soc. 1982, 104,3876-3879. Mundt, O.; Riffel, H.; Becker, G; Simon, A Z. Naturforsch. 1984, 39b, 317- 322. Mundt, 0.; Becker G.; Rt'issler, M.; Witthauer, C. Z Anorg. Allg. Chem. 1983, 506, 42—58. (a) Dance, 1. G. Polyhedron 1983, 2, 1031-1043. (b) Hollander, F. J.; Coucouvanis, D. J. Am. Chem. Soc. 1977, 99, 6268-6279. (c) Chadha, R.; Kumar, R.; Tuck, D. G. J. Chem. Soc., Chem. Commun. 1986, 188-189. (d) Coucouvanis, D.; Murphy, C. N.; Kanodia, S K. Inorg. Chem. 1980, 19, 2993—2998. 32. (a) Burschka, C.; Bronger, W. Z. Naturforsch. 1977, 328, 11-14. (b) Burschka, C. Z. Naturforsch. 1979, 348, 675-677. (c) Schils, H.; Bronger, W. Z Anorg. Allg. Chem. 1979, 456, 187-193. ((1) Savelberg, G.; Schafer, H. 2. Naturforsch.1978,33B,711-713. (a) Merz, K. M.; Hoffmann R. Inorg. Chem. 1988, 27, 2120-2127. (b) Mehrotra, P. K.; Hoffmann, R. Inorg. Chem. 1978, 17,2187-2189. (a) Cotton, F. A.; Feng, X.; Matusz, M.; Poll, R. J. Am. Chem. Soc. 1988, 110, 7077-7083. (b) Lee, S. W.; Trogler, W. C. Inorg. Chem. 1990, 29,1659-1662. Wacker, K.; Salk, M.; Decker-Schultheiss, G.; Kelley, E. Z. Anorg. Allg. Chem., 1991, 6%, 51-58. Vos, A.; Wiegenga, E. H. Acta Cryst., 1955, 8, 217-223. McCarthy, T. J.; Kanatzidis, M. G. work in progress. McCarthy, T. J.; Kanatzidis, M. G. manuscript in preparation. 33. 34. 35. 36. 38. CHAPTER 6 Synthesis in Molten Alkali Metal Polyselenophosphate Fluxes: A New Family of Transition Metal Selenodiphosphate Compounds, AzM P28e6 (A=K, Rb, Cs; M=Mn, Fe), AzM'2P28e6 (A=K, Cs; M'=Cu, Ag) and K4Au2(PZSe6)2 1 . Introduction Recently, we have used the chalcophosphate flux method successfully to synthesize several quaternary main group compounds, ABiP2S7 (A=K, Rb),1 KMPZSe6 (M=Sb, Bi)2 and C53M4(P28e6)5 (M=Sb, Bi).3 The Ax[PyQ] (Q=S, Se) fluxes provide - excess Ax[Psz]n ' anions which bind the metal ions and perhaps also act as mineralizers. Using polyselenophosphate Ax[Pysezl fluxes below 600°C, we wanted to explore the chemistry of first row transition metals to see if this strongly basic medium would discourage the formation of the well known M2P28e6 compounds. The MszQ, (M=Mn, Fe, Co, Ni, Zn, Cd, V, Mg; Q=S, Se) family of compounds are structurally related to CdIz.4»5 These compounds possess interesting magnetic6 and intercalation properties7 and transition metal thiophosphates are of potential importance as low- dimensional cathode materials for secondary lithium batteries.8 322 323 Substitution of M2+ cations by M+ (Ag,Cu) and M3+ (Al, In, Cr, V) has been demonstrated for M2P2(x compounds to form quaternary chalcophosphates which retain the stable Cdlz structure-type.9 Recently, Clement and coworkers have reported that ion-exchange intercalation of M2P2S6 (M=Cd, Mn) with a cationic organic dye induces a large second-order optical nonlinearity and permanent magnetization (Mn2+). 10 Although [PxSeyPl' units such as [PSe4]3' and [P25e6]4- can potentially exhibit many different bonding modes giving rise to a wealth of different compounds, relatively little is known about the chemistry of selenophophate ligands. Presumably, progress in this area has been hindered by the lack of suitable solvents for the dissolution of these highly charged ligands. The rare [PSe4]3' ligand is found in solid state compounds such as Cu3PSe411 and T13PSe4.17- Kolis and coworkers have isolated two transition metal complexes with unusual coordinating [PSezl' and [PSe5]3- ligands by using "P4Se4" glasses in DMF solutions.13 We have shown recently that the use of molten alkali selenophosphate fluxes is a convenient method for stabilizing highly charged [PxSeyPl' units. Here we report the synthesis, structural characterization, optical, thermal and magnetic properties of a new class of compounds, AzMPZSe6 (A=K, Rb, Cs; M=Mn, Fe), A2M'2P2Se6 (A=K, Cs; M'=Cu, Ag), and K4Au2(P2Se6)z. In the case of Mn and Fe, successful dismantling of the MszSe6 framework results in the formation of one-dimensional, polymeric anions that are reminiscent of the TiI3 structure type. Replacement of the octahedral M2+ metal centers with M'+---M'+ (M'=Cu,Ag) pairs gives rise to CszM'PZSee. In 324 A2M'2P28e6, a change in the counterion size (A) (substitution of Cs+ with K“) results in dramatic change in structure from a one- dimensional chain to a complex three-dimensional tunnel framework. The structure of K4Au2(PZSe6)2 features a one-dimensional chain with linear AuISez and square planar AuIIISe4 units bridged by [P25e6]4- ligands. It is a rare example of a mixed valent Au+/Au3+ compound. 2. Experimental Section 2.1. Reagents Chemicals in this work were used as obtained: (1) iron metal powder, 99.9% purity, -325 mesh, and gold metal powder, 99.9596 purity, -325 mesh, Cerac Inc., Milwaukee, WI; (ii) manganese metal powder, 99.9% purity -50 mesh, Aldrich Chemical Co., Inc., Milwaukee, WI; (iii) copper and silver metal powders, Fisher Scientific Co., Fair Lawn, NJ; (iv) red phosphorous powder, Morton Thiokol, Inc., -100 mesh, Danvers, MA; (v) selenium powder, 995+% purity -100 mesh, Aldrich Chemical Co., Inc., Milwaukee, WI; (vi) potassium metal, Aldrich Chemical Co., Inc., Milwaukee, WI; (vii) rubidium metal and cesium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (viii) DMF, analytical reagent, diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ; (ix) triethylphosphine, Aldrich Chemical Co., Inc., Milwaukee, WI. 325 22. Synthesis All manipulations were carried out under a dry nitrogen atomosphere in a Vacuum Atmospheres Dri-Lab glovebox. For the preparation of K2Se, RbZSe, and C825e we used a modified literature procedure.14 The preparation of K2Se is reported in Section 2.2 of Chapter 1 (Part 1). CszSe. An amount of 5.00 g (0.038 mol) of slightly heated (~30°C) cesium metal was pipetted into a 250-ml round-bottom flask. A ISO-ml volume of liquid ammonia Was condensed into the flask at 78°C (dry ice/acetone bath) under nitrogen to give a dark blue solution. Selenium (1.485 g, 0.019 mol) and a Teflon-coated stir bar were added, and the mixture was stirred for 1 h to give a light blue solution. The NH3 was removed by evaporation under a flow of nitrogen as the bath slowly warmed to room temperature. The pale orange solid (98% yield) was dried under vacuum overnight, flame- dried, and ground to a fine powder in the glove box. RbZSe. Same procedure as CszSe. P ZS e5. The amorphous phosphorus selenide glass, "PzSes", was prepared by heating a stoichiometric ratio of the elements in an evacuated pyrex tube for 24 h at 460 °C. The glass was ground up and stored in a nitrogen glove box. KzMnPZSe6 (I). An amount of 0.014 g (0.25 mol) Mn, 0.228 g (0.50 mmol) PZSes, 0.079 g (0.50 mmol) K2Se, and 0.197 g (2.50 mmol) Se were thoroughly mixed and transferred to a 6m] pyrex tube which was subsequently flame-sealed in vacuo (~10'3 torr). The reaction was heated to 450 °C over 12 hours in a computer- 326 controlled furnace. It was kept at 450 °C for 6 days, followed by cooling to 150 °C at a rate of 4 °C/ hr, then to room temperature in one hour. The moisture-sensitive product was isolated by dissolving the flux with degassed DMF under inert atmosphere and then washed with anhydrous ether to give orange needle-like crystals and a small amount of gray powder (crystallized flux). Semi-quantitative microprobe analysis on single crystals gave K2,2Mn1,0P2,5Se6,5 (average of three data acquisitions). These crystals were used for crystallographic studies. Pure material was obtained by heating Mn/ P/ KZSe/ Se in a stoichiometric ratio in a pyrex tube at 500°C for 4 days. The product was purified by washing with degassed DMF and ether to give 0.378 g (56% yield, based on Mn) of orange crystallites. Far-IR (CsI matrix) shows absorptions at 478(3), 463(s), 303(m) and 210-143(w) cm'l. szManSe6 (11). Pure material was obtained by heating a stoichiometric ratio of Mn/P/ Rb2Se/ Se as above to give 0.396 g (52% yield, based on Mn) of orange crystallites. The crystallites are moisture sensitive. Semi-quantitative microprobe analysis on single crystals gave Rb 1,3Mn1,0P 1,55e4,7 (average of three data acquisitions). Far-IR (CsI matrix) shows absorptions at 480(s), 465(s), 304(m) and 201-148(w) cm-l. CszMnPZSe6 (111). An amount of 0.014 g (0.25 mol) Mn, 0.228 g (0.50 mmol) PZSes, 0.157 g (0.46 mmol) CszSe, and 0.197 g (2.50 mmol) Se were heated as above. The product was isolated by . dissolving the flux with degassed DMF under inert atmosphere and then washed with anhydrous ether to give orange needle-like crystals and elemental Se. The crystals appear to be stable in water. 327 Senri-quantitative microprobe analysis on single crystals gave Csngn1,oP2,ZSe63 (average of three data acquisitions). These crystals were used for crystallographic studies. Pure material was obtained by heating Mn/P/Cs2Se/Se in a stoichiometric ratio as above to give 0.423 g (49% yield, based on Mn) of orange crystallites. Far-IR (CsI matrix) shows absorptions at 481(8), 464(5), 305(m) and 189—152(w) cm'l. KzFePZS e6 (IV). An amount of 0.014 g (0.25 mmol) Fe, 0.228 g (0.50 mmol) P2Se5, 0.079 g (0.50 mmol) KZSe, and 0.197 g (2.50 mmol) Se were heated as in (1) except for 4 days. The moisture sensitive product was isolated by dissolving the flux with degassed DMF under inert atmosphere and then washing with anhydrous ether to give red needle-like crystals, water soluble flux, and elemental Se. This product was allowed to stir in a solution of triethylphosphine (PEt3) (0.1 ml)/ether (20 ml) for 2 hours. Washing with ether gave red KzFePZSee but also black Fe2P25e6 as a minor impurity phase (~5- 10%). The presence of Fe2P2Se6 was detected by X-ray powder diffraction. Several attempts to obtain pure KzFePZSe6 failed. Serrri- ’ quantitative microprobe analysis on the moisture-sensitive red single crystals gave K1,3Fe1_0P1,3Se4,9 (average of three data acquisitions). These crystals were used for crystallographic studies. Pure sample for magnetic susceptibility measurements was obtained by manual separation of KzFeP28e6 crystals. Far-IR (CsI matrix) gave absorbances at 475(s), 466(s), and 303(m) cm-l. Direct combination reactions with Fe/ P/ KZSe/Se at 500 °C gave similar results. CsteP28e6 (V). Anamount of 0.014 g (0.25mmol) Fe, 0.343 g (0.75 mmol) P2Se5, 0.172 g (0.50 mmol) CszSe, and 0.197 g (2.50 328 mmol) Se were heated as in (IV). The product was isolated as in (IV) above to give 0.112 g of red needle-like crystals ( 52% yield, based on Fe). Semi-quantitative nricroprobe analysis on single crystals gave Cs1,9Fe1,oP1,6Se5,6 (average of three data acquisitions). Far-IR (C51 matrix) shows absorptions at 478(5) and 469(m) cm'l. C52Cu2P2Se5 (VI). An amount of 0.016 g (0.25 mmol) Cu, 0.228 g (0.5 mmol) P25e5, 0.172 g (0.5 mmol) CszSe, and 0.197 g (2.5 mmol) Se were heated as in (IV). The product was isolated by dissolving the flux with degassed DMF under inert atmosphere and then washed with anhydrous ether to give dark green rectagular block-like crystals and some water soluble flux. The residual flux was removed by washing the product with distilled water, MeOH, and ether to give 0.061 g (53% yield based on Cu) of dark green crystals. Semi-quantitative microprobe analysis on single crystals gave C51,30113P1,08e32 (average of three data acquisitions). Far-IR (CsI matrix) gave absorptions at 467(m), 454(5), 447(msh), 302(m), 213(vw) cm-l. KzAgzPZSe6 (VII). An amount of 0.027 g (0.25 mmol) Ag, 0.228 g (0.5 mmol) PZSes, 0.079 g (0.5 mmol) KZSe, and 0.197 g (2.5 mmol) Se were heated as in (IV). The product was isolated with DMF and washed with ether. Further washing with 0.1 ml PEt3 in ether (20 ml) gave 0.052g (50% yield, based on Ag) of orange crystals. Semi-quantitative microprobe analysis on single crystals gave K1,0Ag1,1P1,oSe2,9 (average of three data acquisitions). Far-IR (CsI matrix) gave absorbances at 460(m), 443(5), and 302(w) cm-l. CszAgzPZSe5 (VIII). An amount of 0.027 g (0.25 mmol) Ag, 0.228 g (0.5 mmol) P2Se5, 0.172 g (0.5 mmol) C528e, and 0.197 g (2.5 329 mmol) Se were heated as in (IV). The product was isolated with DMF and washed with ether. Further washing with 0.1 ml P( Et) 3 in ether (20 ml) gave 0.084 g (66% yield, based on Ag) of yellow-orange needle-like crystals. Semi-quantitative micr0probe analysis on single crystals gave C51,0Ag13P1,1Se3,2 (average of three data acquisitions). Far-IR (CsI matrix) shows absorptions at 452(w, broad) and 302(w) cm'l. K4Au2( P2Se6)2 (IX). An amount of 0.049 g (0.25 mmol) Au, 0.343 g (0.75 mmol) P2Se5, 0.079 g (0.5 mol) KZSe, and 0.197 g (2.5 mmol) Se were heated as in (IV) above. The product was isolated with DMF and washed with ether to give 0.163 g (80% yield, based on Au) of black rectangular block-like crystals. Semi-quantitative microprobe analysis on single crystals gave K2,3Au1,oP3,oSe3,9 (average of three data acquisitions). Far-IR (CsI matrix) shows absorptions 499 (s), 470(m), 429(5), 303(m), 240(w), and 205(w) cm- 1. KzMn0,5Feo,5P2Se6. An amount of 0.014 g (0.25 mol) Mn, 0.014 g (0.25 mmol) Fe, 0.457g (1.0 mmol) P2Se5, 0.157 g (1.0 mmol) K2Se, and 0.395 g (5.0 mmol) Se were heated and isolated as in (IV) to give orange-brown needle-like crystals and elemental selenium. The product was washed with triethylphosphine PEt3/ ether as above in (I V) to give orange-red crystals of K2Mfl05F€05PZS€6 and black FezP 28e6 as a minor impurity phase (~5%). The presence of FezP 25% was detected by X-ray powder diffraction. Semi-quantitative microprobe analysis on red-orange single crystals gave K2,1Mno,5Feo,5P23Se6,1 (average of six data acquisitions). Pure 330 sample for magnetic susceptibility measurements was obtained by manual separation of KzMn05 Fe05P2Se6 crystals. The homogeneity of (I ), (I I I), (V I), (V I I), (V I I I), and (I X) was confirmed by comparing the observed and calculated X-ray powder diffraction patterns. The C1th spacings observed for the bulk materials were compared, and found to be in good agreement with the dhkl spacings calculated from the single crystal data. 15 The results are summarized in Tables 6-1 to 6-6. 331 2.3. Physical Measurements The instruments and experimental setups for optical diffuse reflectance measurements, far-IR spectroscopy, and quantitative microprobe analysis on SEM/ EDS are the same as those in Chapter 2 (Part 1) (Section 2.3). The experimental setup for differential thermal analysis (DTA) is the same as that in Chapter 2 (Part 2) (Section 2.3). Magnetic susceptibility measurements were carried out using a MPMS Quantum Design SQUID magnetometer over the range of 2-300 K. Single crystals were ground to a fine powder to rrrininrize possible anisotropic effects. In the case of KzFePZSee, single crystals were carefully hand-picked from the product mixture because of the presence of the minor Fe2P28e6 impurity phase. Corrections for the diamagnetism of the PVC sample containers were applied. Field dependence measurements were performed (at 5 K and 300 K) to determine if the samples experienced saturation of their magnetic signal. For all compounds, magnetization increased linearly with increasing field over the range investigated (0-10,000 Gauss). A suitable magnetic field was chosen from the linear region for temperature dependence measurements (2000-3000 Gauss). The ueff and 0 values were calculated by least-squares fits for the plots of 1/XM vs. temperature in the paramagnetic (Curie-Weiss) region from 30-300 K. 332 2.4. X-ray Crystallography All compounds were examined by X-ray powder diffraction (XRD) for the purpose of phase purity and identification. Accurate dhkl spacings (A) were obtained from the powder patterns recorded on a Rigaku rotating anode (Cu Kn) X-ray powder diffractometer, Rigaku-Denki/RW400F2 (Rotaflex), at 45 kV and 100 mA with a 1 °/min scan rate. Single crystal X-ray diffraction data were collected on a Rigaku AFC6 diffractometer and the w/ 26 scan technique was used. None of the crystals showed any significant intensity decay as judged by three check reflections measured every 150 reflections throughout the data collections. The space groups were determined from systematic absences and intensity statistics. The structures were solved by direct methods of SHELXS-8616 and refined by full-matrix least-squares techniques of the TEXSAN package of crystallographic programs. 17 An empirical absorption correction based on xii-scans was applied to each data set, followed by a DIFABS 13 correction to I the isotropically refined structure. All atoms were eventually refined anisotropically. All calculations were performed on a VAXstation 3 100/ 76 computer. In the case of CSZOrzPZSee, inspection of the residual electron density map revealed two peaks of 2.3 and 2.7 e-/ A3 that were ~1.4 A away from Cu(l) and Cu(2). These sites are located near the centers of symmetry in between the Cu---Cu pairs. Refinement of these pseudo-octahedral sites as Cu positions resulted in no change in the R/Rw and in less than 5% occupancy. The presence of these small 333 peaks were judged insignificant. No residual electron density peaks were found near the Ag atoms in CszAgzPZSe6. In the case of K4Au2(PZSe6)2, refinement of the structure gave KzAu050111Au0501P2Se6 with Au( 1) in a square planar site and Au(2) found in a linear coordination site. The temperature factor of Au( 2) was rather high at 10.0 A2 so the occupancy was refined and found to be 0.25 to give a formula of KzAu050mAu0251P25e6 which does not balance the charge (R/Rw=13.4/ 15.5). The temperature factors of K(1) and K(2) were also rather high at ~6-7 A2. A small peak of 4.0 e-/ A3 was observed in the residual electron density map and can be assigned to a very small amount of Au+ but does not represent enough electron density for charge balance. The final R/Rw after final DIFABS and anisotropic refinement gave values of 7.6/ 10.3. The elemental analysis also suggests a Au deficiency but this measurement is complicated by the partial overlap of the Au Ma and the P Ka X-ray energy lines. More work is needed to obtain a higher quality crystals with improved X-ray diffraction characteristics. Better crystals must be grown with K+ or perhaps Rb+ and the data must be recollected. The complete data collection parameters and details of the structure solution and refinement for (I), (I I I), (IV), (V I), (V I 1), (VIII), and (IX) are given in Table 6-7. The coordinates of all atoms, average temperature factors, and their estimated standard deviations are given in Tables 6-8 to 6-14. 334 Table 6-1. Calculated and Observed X-ray Powder Diffraction Patterns for KZMnPZSE6. h k l dcalcg) A doing,A I/ IMObSd) 0 1 l, 0 2 0 6.40, 6.35 6.36 100 0 3 1 0 l 2 1 1-2 3.67 3.55 3.43 3.69 3.55 3.42 02 2, 200 3.20,3.19 3.19 1 2-2 l 4 0 2 1 l l 4 1 2 2-2 3.10 2.84 2.66 2.57 2.51 3.10 2.83 2.65 2.56 2.50 0 5 1 2.402 2.403 2 3-2 2.294 2.295 2 4 1 0 6 1 2.065 2.063 2.035 2.029 0 0 4 1.851 1.848 0 6 2 1.837 1.831 1 2-4 1.813 1.809 2 0-4 1.779 1.774 3 3 1 1.759 1.756 4 1-1 1.620 1.623 20 42 9 94 61 48 24 7 11 10 9 5 6 24 12 6 7 7 9 335 Table 6-2. Calculated and Observed X-ray Powder Diffraction Patterns for CszMnP28e6. h kl Ed, A dmd, A 1/ 1111a); (obsd) 0 1 1 0 2 0 1 1-1 6.79 6.51 4.79 1 2 0, 1 1 1 459,458 1 2-1 0 0 2 0 3 1 0 2 2 1 1-2 0 4 0 2 0 0 1 2-2 0 4 1 1 2 2 1 4 0 2 l 1 1 3 2 4.04 3.98 3.81 3.40 3.36 3.25 3.23 3.06 3.01 2.95 2.91 2.87 2.63 6.78 6.49 4.78 4.58 4.03 3.97 3.80 3.39 3.35 3.25 3.23 3.06 3.01 2.95 2.90 2.86 2.58 2 2-2 2.397 2.400 2 2 2 1 2 3 2.290 2.286 2.260 2.260 0 6 0 2.169 2.165 O 0 4 1.990 1.985 0 1 4 0 6 2 3 4 0 3 2 2 1.967 1.905 1.797 1.783 1.964 1.900 1.799 1.783 22 14 7 9 24 17 66 100 10 24 38 60 11 56 62 7 5 5 14 6 S 27 8 42 7 6 336 Table 6-3. Calculated and Observed X-ray Powder Diffraction Patterns for KzAgzP 25e6. h k l dcglcd, A dobsd. A I/ Imggobsd) O O 2 O 1 1 O l 2 1 0-2 1 1 1 1 1-2 O 2 1 O O 4 1 2 l 0 2 3 2 0-2 2 l O 2 l-l 2 O 2 O 2 4 2 1-2 1 1-5 1 2-4 l 1 5 1 3-1 2 2-1 2 2-2 2 0 4 1 0 6 2 1 4 1 1 6 2 2-4 3 O 0 2 l 5 1 1-7 2 3-2 3 O 2 1 4-1 2 O 6 3 1-3 l 4 2 2 1 6 3 2-1 1 4—3 2 2-6 1 O 8 10.47 10.47 9.90 7.66 6.80 6.37 5.82 5.43 5.23 4.55 4.38 4.02 3.98 3.95 3.87 3.83 3.79 3.64 3.55 3.49 3.40 3.37 3.27 3.21 3.17 3.09 3.05 2.91 2.84 2.81 2.78 2.74 2.70 2.65 2.63 2.61 2.58 2.56 2.53 2.51 9.91 7.66 6.78 6.37 5.80 5.42 5.23 4.54 4.37 4.02 3.97 3.94 3.86 3.82 3.78 3.64 3.54 3.48 3.39 3.36 3.26 3.21 3.16 3.09 3.04 2.90 2.83 2.80 2.77 2.74 2.69 2.65 2.62 2.60 2.57 2.55 2.53 2.50 2.4% 2.462 2.483 2.458 28 44 52 66 70 9 14 14 13 11 16 11 16 10 10 13 15 38 35 71 27 46 10 64 43 23 14 1(1) 52 31 41 49 36 17 11 15 14 38 23 26 15 337 Table 6-3. (cont'd) h kl dcalcd. A dobsd. A I/ Infix (obsd) 2 2 6 3 15 0 3 7 1 2-8 3 3 0 1 19 1 3-8 2 4 5 4 2 o 4 22 3 3 6 3 3-7 1 6-1 1 6 1 4 2 5 2 1-11 2 2 10 4 35 3 1-10 2 5 6 2 6-2 3 54 4 28 4 5 0 2.379 2.360 2.338 2.317 2.262 2.232 2.104 2.019 1.990 1.975 1.860 1.845 1.825 1.821 1761 1.754 1.746 1.725 1.712 1.708 1.699 1.688 1.627 1.546 2.374 2.357 2.335 2.312 2.257 2.231 2.099 2.016 1.987 1.973 1.856 1.848 1.823 1.817 1.760 1.752 1.746 1.723 1.711 1.706 1.696 1.688 1.623 1.545 11 8 39 17 8 43 16 12 10 24 8 39 10 9 11 10 8 8 11 14 8 10 8 10 338 Table 6-4. Calculated and Observed X-ray Powder Diffraction Patterns for CszAg2P28e6. h kl dcgjcd, A dew, A I/ ng (obsd) O l 1 0 2 0 O 3 1 1 3 0 0 2 2 2 0 0 1 3-1 1 2-2 2 1-1 0 4 0 1 2 2 2 2 0 2 1 1 1 4 0 0 1 3 2 2-2 1 1 3 1 4 2 6.99 6.26 3.74 3.55 3.49 3.39 3.34 3.22 3.15 3.13 3.0) 2.98 2.95 2.84 2.74 2.54 6.98 6.26 3.74 3.56 3.49 3.38 3.34 3.21 3.16 3.12 3.11) 2.96 2.93 2.84 2.73 2.55 2.457 2.447 2.310 2.308 2 4 1 2.180 2.178 0 0 4 2.105 2.102 0 4 3 2.089 2.085 2 3-3 1.998 1.9% 0 6 2 1.869 1.865 0 4 4 1.747 1.744 1 4-4 1.726 1.722 49 17 32 4 1m 9 3 34 18 16 7 10 10 12 17 3 4 4 3 9 4 3 20 7 7 339 Table 6-5. Calculated and Observed X-ray Powder Diffraction Patterns for C52CuszSe6. h k 1 dgalcd. A (10ng A I/ Img(0b8d) 1 1-1 0 2 0 2 1 o 2 0—2 1 3-1 2 2 1 2 2-2 040 2 o 2 3 1 o 1 2-3 1 3 2 3 2-1 1 4 1 3 1 1 3 2 1 0 2 4 4 2 o 3 5-1 40—4 4 1.4 4 4.1 2 62 0 3 5 1 2 5 o 5 4 443 3 4 3 2 0-6 444 5 0 2 3 45 o 64 1 1 6 6 1-2 5 4.3 460 404 6.85 6.53 4.56 4.02 3.83 3.46 3.43 3.27 3.23 3.15 3.11 3.04 2.95 2.92 2.85 2.67 2.431 2.278 2050 2.011 1.988 1.980 1.915 1.888 1.880 1.850 1.844 1.799 1.769 1.713 1.707 1.683 1.675 1.644 1.639 1.630 1.622 1.616 6.89 6.52 4.55 4.02 2.83 3.46 3.42 3.26 3.22 3.14 3.11 3.01 2.95 2.91 2.85 2.69 2.424 2.247 2.060 2.006 1.985 1.980 1.910 1.898 1.884 1.849 1.845 1.795 1.768 1.710 1.705 1.681 1.673 1.642 1.639 1.631 1.626 1.616 18 11 7 10 18 53 100 32 20 29 29 4 23 27 18 4 6 8 6 10 4 3 23 8 15 7 5 4 5 10 9 5 5 5 5 7 7 4 Table 6-6. Calculated and Observed X-ray Powder Diffraction Patterns for K4Au2(PzSe6)2. 340 , h k l dgadcd, A dob§d, A I/Imgx (ObSd) 0 1 1 0 2 0 1 1 0 0 2 1 l 0 1 1 1-1 l 1 1 l 2 0 0 0 2 0 1 2 0 3 1 0 2 2 0 4 0 1 3-1 l 3 1 0 3 2 1 2 2 2 0 0 1 4 0 2 2 0 0 4 2 2 2-1 1 1-3 0 5 1 2 3 0 2 1-2 2 0 2 2 3-1 2 3 1 2 2 2 2 4 0 0 0 4 0 1 4 1 5-2 0 2 4 0 5 3 0 7 1 2 6 O 2 4 4 4 2 0 4 3-1 8.89 8.15 6.64 6.46 5.91 5.70 5.56 5.43 5.30 5.04 4.84 4.45 4.08 4.05 4.0) 3.80 3.75 3.63 3.56 3.32 3.23 3.19 3.16 3.12 3.02 2.99 2 95 2 93 2 89 2 78 2 71 2.65 2.62 2.61 2.52 2.397 2.275 2.176 1.874 1.774 1.709 8.87 8.17 6.62 6.45 5.92 5.68 5.54 5.43 5.37 5.02 4.85 4.44 4.09 4.03 3.99 3.78 3.74 3.63 3 57 3 32 3.23 3.20 3.15 3.13 3.02 2.98 2.95 2.92 2.89 2.77 2.72 2.67 2.63 2.60 2.51 2.395 2.287 2.215 1.870 1.772 1.708 1m 23 10 7 8 4 3 10 7 4 15 23 21 6 5 19 4 7 3 25 26 12 12 21 8 19 12 12 13 7 21 26 444 15 18 7 11 11 14 11 6 341 Table 6-7. Crystallographic Data for (I ), (I I I), ( I V), (V I), (V I I), (V I I I), and (I X). I III IV Formula K2MnP28e6 CszMnP28e6 K2FeP28e6 FW a, A b, A c, A 11, deg 8, deg y, deg 668.84 856.46 669.75 6.5349(9) 6.4761(9) 6.421(2) 12.696(3) 13.006(2) 12.720(5) 7.589(2) 7.974(1) 7.535(3) 90.0 90.0 90.0 102.67(2) 93.09(1) 102.58(3) 90.0 90.0 90.0 2; v, A3 2;614.3(4) 2;670.6(2) Z;600.7(7) 1. (Mo [(61) 0.71073 0.71073 0.71073 space group P21/n (No. 14) P21/n (No. 14) P21/n (No. 14) Dealt, g/cm3 3.616 .1, cm-1 195.19 Zemax, deg Temp., °C 50 23 4.239 225.67 50 23 3.703 201.59 50 23 Final R/Rw, 96 2.9/3.4 2.7/3.1 3.9/4.5 Total Data 1236 Total Unique (Ave.) 1134 DataF02>30( F62) 755 No. of variables 53 1361 1245 676 53 1213 1113 683 53 *R=2(|Fol-|Fcl)/2|Fol Rw={2w(lFol-IFCI)2/2wlFol2}1/2 Table 6-7. (cont'd) 342 VI VII VIII IX Formula CszCu2P28e6 K3A83P3Se9 CszA82P2866 K4Auz( P2866)2 FW a, A b, A c, A 01, deg 8, deg 7, deg 928.61 1244.46 1017.25 1621.74 9.958(3) 8.528(6) 6.807(3) 7.272(5) 13.067(3) 11.251(6) 12.517(3) 1630(1) 10.730(2) 20.975(4) 8.462(3) 10.615(9) 90.0 90.0 90.0 90.0 102.46(2) 9324(3) 9575(3) 9183(8) 90.0 90.0 90.0 90.0 2; v, A3 4; 1363(1) 4; 2009(3) 2;717.3(8) 2; 1257(2) 1 (Mo K01) 0.71073 0.71073 0.71073 0.71073 space group P2 1/ C (No. 14) P2 1/ C (No. 14) P21/n (No. 14) P21/n (No. 14) beam, g/cm3 4.524 4.114 4.709 4.281 1., cm-1 244.14 198.57 229.30 297.16 Zemax, deg Temp., °C 60 23 50 23 50 23 45 23 Final mm. 96 5.6/6.7 4.5/6.4 3.4/3.9 7.6/10.3 Total Data 4163 Total Unique(Avc.) 3943 Data 102536062) 2029 4009 3745 1866 No. of variables 110 164 1433 1322 733 56 1978 1793 989 104 *R=2(|Fol-|Fcl)/2|Fol Rw={2w( lFol-IFCI)2/ZWIF0|2}1/2 Table 6-8. Fractional Atomic Coordinates and Beq Values for KzMnPZSee, with Estimated Standard Deviations in Parentheses. 343 atom x y z Beq.a A2 Se( 1) Se(2) Se(3) 0.2128(1) 0.32791(7) 0.9293(2) 2.79(4) 0.1797(1) O.58455(8) 0.7431(1) 2.53(4) -0.3123( 1) 0.44005(8) 0.7311(1) 2.60(4) Mn(1) 1/2 1/2 1.0000 2.07(7) P( 1) K(1) 0.0130(3) 0.4666(2) 0.8665(3) 1.49(8) 0.2245(3) 0.8097(2) 1.0394(4) 4.0( 1) Table 6-9. Fractional Atomic Coordinates and Beq Values for CszMnP28e6 with Estimated Standard Deviations in Parentheses. atom x y z Beq,a A2 CS Se(l) Se(2) Se(3) Mn P 0.2453(1) 0.82963(7) 1.02388(8) 2.79(4) 0.2378(2) 0.3319(1) 0.9664(1) 2.05(5) 0.2283(2) 0.5736(1) 0.7540(1) 1.92(5) -0.2492(2) 0.4195(1) 0.7536(1) 2.01(5) 1/2 1/2 1.0000 2.0(1) 0.0437(4) 0.4609(2) 0.8816(3) 1.4(1) Table 6-10. Fractional Atomic Coordinates and Beq Values for KzFeP28e6 with Estimated Standard Deviations in Parentheses. atom x y z Beq,a A2 Se( 1) 56(2) Se(3) Fe P K 0.2218(2) 0.3298(1) 0.9321(2) 250(6) 0.1876(2) 0.5841(1) 0.7458(2) 2.29(6) -0.3185(2) 0.4410(1) 0.7334(2) 2.35(6) 1/2 1/2 1.0000 1.8(1) 0.0144(5) 0.4667(2) 0.8667(5) 1.4(1) 0.2246(6) 0.8077(3) 1.0432(5) 3.6(2) aBeq = (4/3)[a28(1,1) + 6213(2,2) + c213(3,3) + ab(cosY)B(1,2) + ac(cosB)B(1,3) + bC(COS(1)B(2,3). 344 Table 6-11. Fractional Atomic Coordinates and Beq Values for C52012P2$e6 with Estimated Standard Deviations in Parentheses. atom x y z Beq,a A2 Cs(l) 0.6126(1) 0.3221(1) 0.1187(1) 2.45(5) Cs(2) 1.1343(1) 0.1549(1) 0.1595(1) 2.69(5) Se( 1) 0.3607(2) 0.1657(1) -0.1123(2) 2.23(7) Se(2) 0.8462(2) 0.1767(1) 0.3762(2) 1.88(7) Se(3) 1.0054(2) 0.4060(1) 0.2603(2) 2.13(7) Se(4) 0.7775(2) 0.0768(1) 0.0181(2) 1.64(6) Se(S) 0.7503(2) 0.5739(2) -0.0142(2) 2.25(8) Se(6) 0.4961(2) 0.5810(1) 0.2393(2) 2.03(7) Cu( 1) 0.5486(2) 0.45 20(2) 0.4040(2) 2.5( 1) Cu(2) 0.0172(3) 0.4002(2) 0.0378(2) 3.0(1) P( 1) 0.2146(4) 0.0403(3) -0.1671(4) 1.1(1) P(2) 0.7197(4) 0.4607(3) -0.1688(4) 1.2(1) aBeq = (4/3)[aZB(1,1) + b2B(2,2) + cZB(3,3) + ab(cosY)B(1,2) + ac(cosB)B(1,3) + bc(c05a)B(2,3). 345 , Table 6-12. Fractional Atomic Coordinates and Beq Values for KzAgzP 28e6 with Estimated Standard Deviations in Parentheses. atom x y z Beq,a A2 Ag( 1) 0.5399(3) 0.1570(2) 0.2781(1) 3.3(1) Ag(2) 0.1814(4) 0.0387(3) 0.3638(1) 4.9(1) Ag(3) 0.8686(3) 0.2194(2) 0.3972(1) 3.7(1) Se( 1) 0.7647(3) 0.4246(3) 0.4406(1) 2.5( 1) Se(2) 0.8274(3) 0.2472(3) 0.2727(1) 1.9(1) Se(3) 0.4483(3) -0.0552(2) 0.3107(1) 1.6(1) Se(4) 0.5165(3) 0.6416(3) 0.3538(1) 2.3(1) Se(S) 1.1559(3) 0.1958(3) 0.4522(1) 2.0(1) Se(6) 1.2695(3) -0.0106(3) 0.5771(1) 1.9(1) Se(7) 0.4093(3) 0.3242(3) 0.3541(1) 2.1(1) Se(8) 0.9135(3) 0.1670(3) 0.5870(1) 2.1(1) Se(9) 0.9125(3) 0.5601(3) 0.2649(1) 2.1(1) K(1) K(2) K(3) P( 1) P(2) P(3) 0.1877(7) 0.3442(7) 0.2096(3) 3.0(3) 0.5316(7) 0.2078(7) 0.5095(3) 3.4(3) -0.1365(8) 0.0254(7) 0.0999(3) 3.8(3) 0.5964(7) 0.4579(6) 0.3600(3) 1.4( 3) 0.7285(7) 0.4266(6) 0.2700(3) 1.4(3) 1.0699(7) 0.0758(7) 0.5256(3) 1.5(3) aBeq = (4/3)[a213( 1,1) + bZB(z,2) + cZB(3,3) + ab(cosY)B(1,2) + ac(cosB)B(1,3) + bc(COSa)B(2,3). 346 Table 6-13. Fractional Atomic Coordinates and Beq Values for CszAgzPZSe6 with Estimated Standard Deviations in Parentheses. atom X y 2 Beq,a A2 CS Ag 0.2465(2) 0.33282(9) 0.9891(1) 294(5) 0.0405(2) 0.4404(1) 1.3586(2) 3.59(7) Se(l) -0.2213(2) 0.1785(1) 1.0414(2) 250(7) Se(2) 0.2367(2) 0.0923(1) 1.2465(2) 256(7) Se(3) -0.1806( 2) -0.0830( 1) 1.2496(2) 2.05(6) P -0.0383(6) 0.0403(3) 1.1141(5) 1.6(2) Table 6-14. Fractional Atomic Coordinates and Beq Values for K4Au2(P7_Se6)2 with Estimated Standard Deviations in Parentheses. atom x y z Beg.a A2 Au(l) Au(2) Se(l) Se(2) Se(3) Se(4) Se(S) Se(6) K(1) K(2) P(1) P(2) 0 -1/2 1/2 1/2 0 0 1.4(1) 3.3(3) 0.1539(7) 0.6352(3) 0.0257(5) 2.4(3) -0.0886(8) 0.7352(4) -0.2368(6) 3.7(3) 0.0168(7) 0.5085(4) 0.2347(4) 2.8(3) -0.4876(8) 0.4860(5) 0.2163(6) 5.1(4) 0.7142(8) 0.9043(4) 0.0237(5) 3.1(3) 0.3973(8) 0.7644(5) -0.1868(6) 4.5(4) 0.295(2) 0.388(1) 0.471(2) 0.176(2) 0.821(1) -0.468(2) 7(1) 7(1) -0.225(2) 0.433(1) 0.279(1) 1.9(6) 0.168(2) 0.6828(9) 0.173(1) 1.8(6) 8113..., = (4/3)[a28(1,1) + b2B(2,2) + cZB(3,3) + ab(cosY)B(1,2) + ac(cosB)B(1,3) + bC(COSa)B(2,3). ' 347 3. Results and Discussion 3.1. Synthesis, Spectroscopy and Thermal Analysis The syntheses were a result of a redox reaction in which the metal is oxidized by polyselenide ions in the Ax[PySeZ] flux. The metal cation centers are then coordinated by the highly charged [P28e6]4' ligands. The molten polyselenophosphate flux method is very effective for crystal growth in this system and quite conducive for isolation of pure products. Triethylphosphine has been shown to be very useful for the removal of insoluble byproducts such as Ax[PySez] species and elemental Se. Due to the stability of these phases, direct combination reactions M/P/AZSe/ Se at 500 °C were implemented for the preparation of large amounts of pure material. This was particularly useful in the Mn system. The presence of colored DMF solutions during isolation of the products from these reactions suggests that some excess Ax[PySez] flux was present resulting from incomplete ' mixing during these reactions. Direct combination reactions in the Fe system resulted in impure material with the known ternary phase, FezP 2Se5, as the impurity. In the Cu/PZSes/KZSe/Se and Ag/PZSes/RbZSe/Se systems, pure material could not be synthesized. Efforts are continuing in order to obtain single phase products. Powder XRD of the inhomogeneous mixtures suggest that these phases are closely related to KzAgzP 2Se6. 348 Varying the flux composition resulted in the formation of different compounds. In the Mn/ PZSes/KZSe/ Se system, increasing the Lewis basicity of the flux, by doubling the amount of KZSe, resulted in the formation of red-orange water-soluble crystals that analyzed for K4,0Mn1,on,oSe3,4. The solubility and elemental analysis suggest a molecular compound, perhaps K4Mn(PSe4)2. Increased amounts of PZSes resulted in the formaton of the known Mn2P2Se6 compound. In the Fe/PZSes/AZSe/Se (A=K, Cs) systems, doubling the amount of AzSe afforded FeSez. All reactions with Co and Ni resulted in the formation of the stable binary pyrites (CoSez and NiSez). Similar investigations in the Mn/ P2Se5/Na28e/ Se system resulted in the formation of a new phase as evidenced by EDS and XRD. This product crystallizes as air-sensitive, golden-brown, thin plate-like material. Attempts to grow crystals suitable for single crystal X-ray analysis are in progress. 19 The Far-IR spectra of AzMPZSee (M=Mn,Fe), shown in Table 6- 15, display two strong absorptions at ~480 and ~465 cm-l. These vibrations can be assigned to PSe3 stretching modes by analogy with MnPZSe6 (444 cm'l).20 A description of the medium absorption band at ~305 cm-1 in A2MP28e6 is not as straightforward. In anPZSe6, a medium absorbance at 316 cm'1 was ascribed to an out-of-phase PSe3 mode, corresponding to a P-P vibration. The P-P vibration is expected to be IR inactive because it resides on a center of symmetry, however, it was not observed in the Raman spectrum. Similar results were obtained for Na4PZS6-6HZO. The spectra of Na4P286-6H20 were described in terms of internal modes of PS 3 349 groups and combinations of in-phase and out-of-phase translational and rotational motions.20 By analogy to these compounds, the medium intensity absorption band at ~305 cm'1 for A2MP28e6 can tenatively be ascribed to the out-of-phase translational PSe3 mode. The weak absorptions below 200 cm-1 in AzMnPZSee, have been assigned to translational motions of the metal ions.20 A typical spectrum is shown in Figure 6-1a for KzMnPZSe6. The A2M'2P28e6 (M'=Cu,Ag) compounds possess absorptions'in the 440-470 cm-1 range which have been assigned to PSe 3 stretching modes by analogy to the A2MP2Se6 compounds. The absorption at 302 cm-1 can be assigned to the out-of-phase PSe3 translational mode (see Table 6- 15). The far-IR spectra of CszOuzPZSee and K2Ag2P28e6 are shown in Figures 6-1b and 6-1c. The optical properties of (H X) were assessed by studying the UV/near-IR reflectance spectra of the materials. The spectra confirm the semiconducting nature of the materials by revealing the presence of sharp optical gaps. The AzMnPZSee, (A=K, Rb, Cs) compounds exhibit steep absorption edges from which the band-gap, Eg, can be assessed at 2.33 (I), 2.41 (I I), and 2.19 (I I 1) eV respectively (see Figure 6-2). The band-gaps of AzFePZSee can be assessed at 1.72 (IV), and 2.02 (V) eV respectively (see Figure 6-3). The diffuse reflectance spectrum of CszmzPZSee suggests the presence of two band-gaps at ~O.8 and 2.44 eV (see Figure 6-4a). The band gaps of (V I I) and (VIII) are 2.39 and 2.55 eV and are shown in Figures 6- 4b and 6-4c. Higher energy absorptions are readily resolved in these spectra and are assigned to electronic Se-->M charge transfer transitions. The difference between the band gaps of K2Ag2P28e6 350 , (2.39 eV) and CszAgzPZSe6 (2.55 eV) can be attributed to the change in the structural architectures of the two compounds. The Ag-Se orbital overlap in the well-separated one-dimensional chains of CszAgzPZSee is expected to be lower than that of the three- dimensional KzAgzPZSe6 compound. This is facilitated by the increased coordination number of Ag in the K+ compound. This effect gives rise to broader bands and thus a smaller band gap for K2Ag2P28e6. The band gap of (IX) is 1.55 eV and is shown in Figure 6-5. Differential thermal analysis (DTA) shows that all compounds melt congruently, suggesting that large single crystals or microcrystalline thin films can be grown from the melt. Table 6-16 shows optical and melting point data for all compounds. A typical thermogram for RbZMnPZSe6 is shown in Figure 6-6. Table 6-15. Far-IR Data (cm‘l) for AzMP25e6, AzM‘ 2P 2596, and K4Au2( P25e6) 2 Compounds. 351 KzMnP25e6 szManSee, CszMansef, KzFestef, CstePzSeé 478(5) 480(5) 481(5) 475(5) 478(5) 463(5) 465(5) 464(5) 466(5) 469(m) 303(m) 304(m) 305(m) 303(m) 2 10-143(w) 201-148(w) 189-152(w) C5201;P 25% KzAgzP zsefL CslAgzP zSe6 K4Au;( PzSe6) z 467(m) 460(m) 452(w,broad) 499 (s) 454(5) 447(msh) 443(5) 470(m) 429(5) 302(m) 302(w) 302(w) 303(m) 213 (vw) 240(w) 205(w) Abbreviations: s=strong, m=medium, w=weak, v=very. 352 4 .J 4 . .J J 8 3 g i E E J °“ J (A) (B) (C) 1‘76 123 all? 1151 “T 369 35? 252 229 unveuurlaen Figure 6-1. Far-IR spectra of: (A) KzMnPZSee (B) C82012P25e6 and (C) Kz'AgzPZSea 352 '1 l J l 3 . .5 E l "‘ J (A) (B) (C) INT: 3603352§zéa 133123 567 1:51 unveuunaen Figure 6-1. Far—IR spectra of: (A) KzMnPZSee (B) CszalzPZSeG and (C) KiAgszSee Table 6-16. Optical Band Gaps and Melting Point Data for AzM P2Se6, A zM' 2P 28e6, and K4Auz( P25e6)2 3 5 3 Formula E; (eV) M. P. (°C) KzMnP25e6 szManSe6 CszManSe6 KzFePZSe6 CstePzSe6 CszCuszSe6 K2Ag2P 2596 CszAgszSe6 K4Au;(PZSe6)2 2.33 2.41 2. 19 1.72 2.02 2.44 2.39 2.55 1.55 7 1 7 78 1 83 1 662 769 670 542 594 456 3 E 3 Kzanzs€6 13,-2.33ev _ ~ if; 3~ u 8 s 2 7 < 1— 1’ a 0 I— t. I I I I I 0 l 2 3 4 5 6 7 Erml'xymv) 0 J szMnPZSeg E.-2.4leV ” ) s t i n u . b r a ( . f f c o C n o i t p r o s b A S / u A 2 i l l l 1 3 '5 g 1.6“ 8 1.2"1 u (C) .3 0.84 S- CSzanzses 13,-2.19ev L. .— )4 < 0.4“ 1’ :5 o .— I I I I ‘7— I I 0 1 2 3 4 5 6 7 EncrchV) Figure 6-2. Optical absorption spectra of (I ), (I I), and (I I I). 355 " m I N l I — 1 I KzFePzSe6 0 ‘ 7 E,-l.72eV __ ) s t i n u . b r a ( . f f e o C n o i t p r o s b A S / u 0 . ! l U O l — . 0 1 2 3 4567 Energy (eV) 5 6 i l g S {a .o’ I- a 7.7 4_.. 1:: £5 g 3 '3 2“ 3.11 (B) CstePZSeg E1-2.02ev L- 8 .o < Q 5 Energy (eV) Figure 6-3. Optical absorption spectra of (I V) and (V). 356 3'06 5.40 (A) .. 4 A ) 0 Z '1 1 a; CSIClhpzsefi 15,-2.44ev — - '— 0 ' r ) s t i n u . b r a ( . ' l ' i c o C n o i t p r o s b A S / a 0 1 2 3 4 S 6 7 Ell-w (eV) 0.008 . 330 4 i 1 1 0.0049 0 ' o o ( ) a L KzABszseos E3-2.39eV 0.0027 0.001 n f i I I 0 t” n _ ,_ T—‘fi-—'—+ 0 1 2 3 4 5 6 7 Enamhv) A ‘ l 1 l 1 1 1 ) s t i n u . b r a ( . h c o C n o i t p r o s b A S / u 3 - S 3.5-1 g 3“ 3 21 g 15‘ .3 . 5 “ < 0.59 1’ 5 0 C ( )_ )— _. CszAszzses E‘s-2.55ev r _ I fl f I r I 0 1 2 3 4 5 6 7 Enerng) Figure 6-4. Optical absorption spectra of (V I), (V I I), and (V I I I). 357 0.05 1 l l l i l 7%? 0.044 E 0.03“ 8:3 8 K2A02P2866 (g 0.02“ Ef155¢V 2 0.019 < m B O I I I I I I 0 1 2 3 4 5 6 Encrgy(cV) Figure 6-5. Optical absorption spectrum of (I X). 358 40 1 1 1 1 20'- exo 0.. 1"” 751 U..- 11V -40 ‘ H -60 " 731 - .. - _ ”‘ - endo 80 -I 00 — L l T r r 0 200 400 600 800 1 000 Temperature (°C) Figure 6-6. Thermogram of szMnPZSee. Melting is observed at 781 °C, followed by recrystallization at 751 °C. 359 3.2. Description of Structure A2MP25e6. The structure of the [MP2Se6]n2n' (M=Mn, Fe) anion is closely related to the TiI3 structure type. Doubling of the TiI3 formula gives T1216. The [MPzSe6]n 211' anion can be viewed as an ordered substitution of two octahedrally coordinated Ti atoms by M and P2 respectively and a replacement of six I atoms by Se. The transition metal ion and the P-P pairs of K2MnP2Se6 reside in Se octahedra that share triangular faces in the a-direction (Figure 6-7a). These chains are well separated by alkali metal ions (Figure 6-7b). Mn resides on an inversion center with Mn-Se bond distances ranging from 2.737(2) A to 2.818(2) A with a normal P—P' bond of 2.243(5) A and P-Se distances ranging from 2.173(3) A to 2.183(3) A. The Mn-Mn' distance is 6.5249(9) A. Inspection of the Se(1)-Mn- Se(2) angle of 78.38(3)° for K2MnP2Se6 reveals a deviation from an ideal octahedral geometry. This small angle is due to a strained four- membered (Mn-Se(1)-P-Se(2)) ring whereas the less strained five- membered envelope-shaped rings give rise to Se(1)-Mn-Se(3) and Se(2)-Mn-Se(3) angles of 88.23(3)° and 88.60(3)° respectively. Se(1) and Se(2) are above (1.809 A) and below (-1.730 A) the calculated Mn-Se(3)-P-P' least squares plane. There are small differences between the crystal structures of CszMnP2$e6 and K2MnP2Se6 as evidenced by unit cell 3 angles of 93.09(1)° and 102.67(2)° respectively. In addition, there is a reduction of the a-axis upon moving from K to C5 which corresponds to a shortening of the Mn-Mn distance from 6.5 249(9) A to 6.476(2) A. A slight reorientation of the [P28e6]4’ ligand results in a Mn- 360 Se(3)-P angle of 99.07(7)° in K2MnP2Se6 and 9S.82(8)° in the C5 analog. This conformational change is probably due to differences in packing forces caused by the size difference of the alkali ions. Selected bond distances and angles for (I ), ( I I I), and (IV) are given in Table 6-17. Comparison of AzMPZSe6 with the known layered MszQ, structure type reveals some interesting structural relationships. It is useful to view that the MszQ, lattice is broken up into chains as a consequence of the substitution of one M2+ by two K+ cations. A single layer of MszQ, is shown in Figure 6-8a. The M atoms form a hexagonal network and the [P2C5]4' ligand coordinates to three different metal centers as shown in scheme (A). Q\ /Q ~‘P/Q\ M M \ “51K / Q Q Scheme (A) Removal of one-half of the M2+ cations gives the [MPZQJZ' anion as shown in Figure 6-8b. In order to maintain electroneutrality, two alkali metal ions must be introduced. Although this metal-deficient structure (related to that of AlCl3) may very well be stable with the proper counterion, apparently, the lack of a low energy packing arrangement for the alkali atoms within the layer causes a structural change to a 1-D structure. This generates more space for efficient packing of the alkali ions. It is interesting to speculate that with alkaline earth cations, such as Ba2+, it may be possible to stabilize the 361 [MPZQP' layer. 362 (A) I'I'l'l‘ “,‘\\\\‘ Figure 6-7: (A) ORTEP representation of a single [Manseehzn' chain. (B) Packing diagram of KzMnPZSe6 viewed down the [100] direction with labeling. 363 (A) (B) h. ' 1‘s- 0 o fl fl ’ \ hib‘N a I iov‘fl no: to! ’ ' ‘0‘ I l ’ | Figure 6-8: (A) A single layer of the anPzQ framework. The dashed lines indicate the hexagonal arrangement of an+ ions. (B) Removal of one-half of the Mn2+ ions results in a single layer of the hypothetical [MnPZQInZD' anion. Dashed lines highlight the possibility of chain formation. 364 Table 6-17. Selected Distances (A) and Angles (°) for (I ), (III), and (I V) with Standard Deviations in Parentheses.a Distances K 2MnP28e6 CsMn P28e6 K2FeP2Se6 M-M' 6.5349(9) 6.476(2) 6.421 (2) M-Se( 1) M-Se( 2) M-Se( 3) 2.853(1) 2.774(1) 2.782(2) 2.746(1) - 2.737(2) 2.676(2) 2.711(1) 2.818(2) 2.614(2) M-Se (ave.) 277(7) 278(4) 269(8) P-P' 2.236(5) 2.243(5) 2.225(7) P-Se(1) P-Se(2) P-Se(3) 2.181(2) 2.183(3) 2.182(4) 2.181(2) 2.178(3) 2.176(3) 2.176(2) 2.173(3) 2.180(4) P-Se (ave.) 2.179(3) 2.178(5) 2.179(3) A—Se( 1) A-Se(1') A-Se(1") A-Se( 2) A-Se( 2') A-Se( 2") A-Se(3) A-Se(3 ') A-Se( 3") 3.677(4) 3.916(2) 3.683(5) 3.399(3) 3.773(2) 3.398(4) 3.950(2) 3.607(3) 3.965(2) 3.597(4) 3.322(3) 3.880(2) 3.330(4) 3.757(2) 3.433(3) 3.978(2) 3.414(4) 3.525(3) 3.742(2) 3.538(4) 3.602(3) 3-695(2) 3.573(4) A-Se (ave.) 3.5(1) 3.9(1) 3.5( 1) Angles Se(1)-M-Se(2) 78.38(3) 80.80(4) 80.11(5) Se( 2)-M-Se( 3) 88.60(3) 8983(5) 87.43(5) Se( 1 )-M-Se( 3) 8822(3) 8930(4) 8936(5) M-Se(1)-P 7615(6) 7625(8) 75.9( 1) M-Se( 2)-P 7855(6) 77.14(7) 78.4(1) M-Se( 3 )-P 99.07(7) 9S.82( 8) 99.1(1) Se(1)-P-Se(2) 108.45(9) 110.0(1) 107.4(1) Se(1)-P-Se(3) 117.0(1) 115.3(1) 118.3(2) Se(2)-P-Se(3) 115.6(1) 115.8(1) 116.7(2) Se( ltP-P' 105.6(1) 104.8(2) 105.4(2) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01=12n(1n- DZ/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 365 CszCu2P2Se6. The structure of the [CuszSe6]n2n' anion is closely related to the A2MP2$e6 structure type and is shown in Figure 6-9a. The M2+ ions are replaced by weakly interacting Cu+~01+ dimers that alternate with the P-P pairs along the [101] direction. The Cu(1)-Cu(1') and Cu(2)~Cu(2') distances are 2.761(5) A and 2.73 1(5) A respectively. For comparison, the Cu-Cu distance in copper metal is 2.56 A. This attractive interaction also leads to the displacement of Cu( 1) and Cu( 2) from their respective trigonal planes by 0.2 A. These classical d10-d10 interactions have been observed in either molecular21 or solid state structures such as KCu3Sz,22 190138633 C53013Se6,7-4 and KCuS.ZS Normally, the interaction of two filled metal orbitals results in a nonbonding situation due to formation of a bonding and an equally antibonding state. However, extended Hiickel calculations by Hoffmann and coworkers have shown that mixing of empty 4s and 4p orbitals with occupied 3d orbitals in these dimers lowers the energy of the o and 0* orbitals (see Figure 6-10).26 This hybridization gives rise simultaneously to a more bonding and a less bonding M-M interaction and therefore an overall stabilization. It has also been proposed for several d10-d1O complexes that the bridging ligands force the metals closer in order to maximize M-L interactions.27 Since CszOrszSee features two [P2Se6]4- bridging ligands for each dimer, it is possible that the multidentate [P25e6]4' coordination is playing an important role in forcing the metals to close proximity. It is not clear whether the interactions contribute significantly in determining the structure or 366 the metals are simply forced to tolerate the short contact due to packing forces. It is probable that the AzMPzSe6 structure type is quite stable and since Cu+ does not prefer an octahedral coordination, and by itself is not enough to balance the negative charge on that octahedral site, it forms pairs of CuSe3 trigonal planes to conform to the imposed geometry. The Cu-Se distances range from 2.373(3) A to 2.417(4) A which compare well with other Cu-Se trigonal planar species such as (Ph4P)4[Cuz(S<24)(Ses)2]28 and (Ph4P)2[C114(S€4)2.4(S€5)0.6I-29 The Se(1)-Cu(1)-Se(6), Se(3)-Cu(2)-Se(5) angles are less than 120° at 110.6(1)° because of strained (Cu-Se-P-P'-Se) five-membered rings whereas the remaining Se-Cu-Se angles range from 121.3(1)° to 125.7(1)°. The shortest Cu-Se distances for Cu(1) (2.373(3)A) and Cu(2) (2.375(3)A) are trans to the small 110.6(1)° angle as more space is allowed for the Se(2) and Se(4) to get closer to the metal center. Selected bond distances and angles for CszouszSee are given in Table 6-18. CszAgszSee. The structure of the [AgzPZSedn2n' anion is closely related to that of CszcuszSe6 and is shown in Figure 6-9b. The major difference is that the volume is roughly half that of CszCuzP2Se6 so there is only one Ag+---Ag+ dimer per unit cell. The Ag-«Ag distance of 2.919(3) A falls in the 2.75-3.00 A range for many solid state Ag compounds with close Ag+---Ag+ contacts.30 The Ag+ ions are not displaced from their distorted trigonal planes as is the case for Cu in (V I I I). 367 The coordination geometry of Ag is distorted trigonal planar with Ag-Se distances ranging from 2.545(2) A to 2.641(3) A. These distances compare well with other Ag-Se trigonal planar species.31 The Se(1)-Ag-Se(2) angle is less than 120° at 105.55(7)° because of a strained (Ag-Se-P-P'—Se) five-membered envelope-shaped ring whereas the remaining Se-Ag-Se angles are 121.77(8)° and 132.04(9)°. The Se(1)-Ag-Se(3) angle of 132.04(9)° is quite large and compares to the 138.8° angle found in the [(Ph4P)Ag(Se4)]n one- dimensional polymer.31a The shortest Ag-Se distance (Ag-Se(3) = 2.545(2) A) is trans to the small 105.55(7)° angle as more space is allowed for the Se(3) to get closer to the metal center. Selected bond distances and angles for CszAg2P28e6 are given in Table 6-19. The existence of Cu+~O.1+ or Ag+---Ag+ dimers located in octahedral sites in these compounds can hopefully shed some light on the question concerning the nature of Cu(Ag) coordination sites in several cation substitution compounds such as CuCrngS(,,32 CuVPZS 6,54 010,52Mn1,74P2S633 and AgzMnP28(,.9b These compounds adopt the I stable M2P2S6 structure with cation substitution occurring in the octahedral sites. The crystal structure of CuCrP2S6 reveals a Cu electron density cloud which exhibits three extrema.32 One of them is in the middle on a pseudo-octahedral site and the other two are near the trigonal sulfur planes which suggests that CuSe octahedral sites, Cu---Cu pairs and vacancies all exist at room temperature and can be fit by a disorder model. Complimentary EXAFS34 studies have confirmed this disorder model but recent low-temperature neutron powder diffraction studies have disputed the existence of bimetallic 368 entities.3S Cu+---Ou+ pairs were also assumed to exist in CuVP286,5d which features the same Cu positions as found in CuCrPZS6 plus a third tetrahedral site located in the van der Waals gap. The location of copper atoms in Cu(),szMnL74P2S6 remains uncertain due to some inconsistencies between X-ray diffraction and EXAFS data.33 It appears that the occurrence of copper pairs in these copper substitution phases is still an open question. In the case of Ag, the structure determination of AgzMnP2S69b confirmed the presence of two silver atoms per octahedron with 81% of the Ag( 1) atoms found in trigonal planar coordination (Ag-Ag = 3.38 A). On each side of Ag( 1) are two sites of minor contribution (Ag(2) 13%; Ag(3) 8%). We think that the unequivocal presence of M+---M+ pairs in our compounds serves to support the presence of M+---M+ pairs in octahedral sites within the layers of CuM111P286 compounds. K2Ag2P28e6. The substitution of K for C5 resulted in a dramatic change in the structure of the [Ag2P28e6]2' anion. K2Ag2P25e6 possesses a complicated three-dimensional tunnel framework made up of AgSe4 tetrahedra linked by P28e6 units (see Figure 6-1 1). The channels run along the crystallographic a-axis and are filled with K+ ions. There are two types of [PZSe6J4' ligands in this structure, see scheme (B). Both ligands contain four Se atoms that each bond to a Ag atom and two Se atoms that each bridge two other Ag atoms. The P-P bond in Type (2) lies on a center of symmetry while that of Type (1) does not. In Type (2), the two Se atoms that bridge Ag atoms are located on opposite sides of the P-P bond. In Type (1), the two Se atoms are situated on the same side of the P-P bond. 369 Scheme (B) These are new bonding modes for the [P28e6]4' ligand and differ from those found in KMPZSee2 and C53M4(P2Se5)5.3 The [P28e6]4- units connect AgSe4 tetrahedra as shown in Figure 6-12. For comparison, the [P25e6]4‘ ligand in the known ternary compound, Ag4P 28e6, bonds to ten Ag atoms with two Se atoms each bonding to two Ag atoms and the remaining four Se atoms each bonding to three Ag atoms.36 In this structure, P28e6 building blocks are assembled into layers and are linked via AgSe4 tetrahedra to form a dense three-dimensional network. As in the case of the one- dimensional A2M2P28e6 compounds, K2Ag2P25e6 can also be viewed as a derivative of Ag4P28e5 generated by breaking down the Ag4P2Se6 framework by incorporation of K4P25e6. The addition of KzSe gives rise to a more open channel structure. Selected distances and angles for K2Ag2P2Se6 are found in Table 6-20. Ag(1,2) are located in distorted tetrahedral coordination 370 , environments as evidenced by Se-Ag-Se angles that range from 95.1(1) to l30.5(1)° for Ag(l) and 80.9(1) to 130.4(1)° for Ag(Z). Ag(3) resides in a fairly regular tetrahedral site with normal angles (103.2(1)-122.1(1)°). The Ag(1,2,3)-Se bond distances fall in the 2.578(4) to 2.797(4) A range which compare well with those observed in Ag4P25e6.36 No Agu-Ag contacts are observed as the Ag- Ag distances are long at 3.715(4) and 3.869(4) A. Normal P—P bonds in the [P28e6]4' ligands are observed at 2.278(8) and 2.31(1) A and the P-Se bonds are typical for Se atoms in bridging positions (2.19( 1) A (mean)). The higher coordination number of Ag+ most likely results from the effect the relatively small K+ cation exerts on the [Ag 2P 28e6]2- structure. A correlation between the metal coordination number (CN) and the size of the counterion was observed and discussed in the Ag+/Sex2' system earlier.313 The change of the CN for Ag+ from three, in CszAgzPZSee, to four, in KzAgzPZSee, agrees with the general trend identified in Group 10 polychalcogenides that larger counterions favor a small CN for the Group 10 metal. While a high CN will produce compact structures, a small CN tends to result in expanded or low dimensional structures. K4Au2( P28 e6)2. The preference of Au for a linear or square planar coordination results in a different structure type. The structure of the [Au2(P2Se6)2]n4n' anion consists of linear Au+ and square planar Au3+ units that are linked by bridging [PZSe6]4- ligands (Figure 6-13). The Au(l)-Au(2) distance is 3.636(3) A so no metal 37 1 interactions are present. This compound represents one of the few examples of a mixed valent Au+/Au3+ compound with the only other examples being a-(B-)AuSe.37 Selected distances and angles for K4Au2(P25e6)2 are found in Table 6-21. Au( 1) resides on a center of symmetry and is surrounded by four Se atoms in a square planar geometry with distances of 2.484(6) and 2.493(5) A. This compares well with the square planar distances of 2.474(8) and 2.489(8) A for a-AuSe and 2.496(9) A for B-AuSe.37 The linear Au(2) also resides on a center of symmetry and possesses a Au-Se distance of 2.302(7) A. This distance is shorter than those reported for linear Au+ in AuSe (2.42( 1) to 2.433(9)) A. This short distance could be a result of the partial occupancy on the Au( 2) site due to averaging. A close contact between Au(2) and K(2) was observed at 3.21(2) A. It is known that Au+ has a remarkable affinity for alkali metal cations. Such interactions have been observed in other Au+ chalcogenide complexes including [KAugTe7]4-,38 [KzAu4Te4(en)4]2',39 and the inorganic cryptand, [NaAu128e3]3'.40 Due to the inconsistencies in the charge balance and the rather high temperature factors and R/Rw values, a structural redetermination of K4Au2( P25e6)2 must be performed. 372 (A) , Sel 81 t'; Se3 v '1 .. Cu1\ se4 8; I i.) a Cul P2 Se5 Cu2' . ,0 V \0 .0 “ $82 "’ Se6 Figure 6-9: (A) A single [CuzPZSeGLZD' chain. (B) A single [AgszSealnzn' Chain. 373 d orbitals alone (1 + s + p O. I ’I 1 “ I I I 11:11: ‘\ \\ f—I- ——-—-\ \\‘\ '1’ I, 0* I I \ ‘ [I I“! \\ \ “\ \ 5“ \‘\\ 4” 6* 2“ . _._-—--7——r~ ,- \ \ \ \ \ I ’ I I I \————I O I , , \ \ 0 N \ \ \ \ \ \ \ \ ‘_——d 0 ’ 4’ I I I I I I I I I I A g I: I” m fl- .5 ~~ 1 \ Figure 6-10: Energy levels of the (1 block of Cu22+ at a separation of 2.58 A. Mixing of filled d orbitals with empty 5 and p orbitals results in a lowering of the antibonding (0*) and bonding (0) energy levels which gives rise to a net bonding interaction. Adapted from reference 26. 374 Figure 6-1 1: Packing diagram of the three-dimensional structure of KzAgzPZSee viewed down the [100] direction with labeling. 375 FigureG-IZ: ORTEP representation of a fragment of the [AgszSeGLZH' anion showing the local AgSe4 coordination environment with labeling. 376 SeS .. .. Se2 .. « Se6 g 3" P1 :2 P2 4 ’ . ‘9 o SE4 ‘. 0 Se3 Q. 0 Au2 , Aul , E. 0 ”~. 0 h” 3 CD 0 C. 4! 0 0. Figure 6-13: A single [Au2(PzSe5)z]nzn' chain. 377 Table 6-18. Selected Distances (A) and Angles (°) for CszCuszSe6 with Standard Deviations in Parenthesesfil Cu(1)-Cu(l') Cu(2)-Cu(2') Cu(1)-Se(1) Cu(1)-Se(4) Cu(1)-Se(6) Cu(2)-Se(2) Cu(2)-Se(3) Cu(2)-Se(5) 2.761(5) 2.731(5) 2.400(3) 2.373(3) 2.416(3) 2.375(3) 2.417(4) 2.407(4) Se(1)-Cu(l)-Se(4) 125.7(1) Se(l)-Cu(l)—Se(6) 110.6(1) Se(4)-Cu(l)-Se(6) 121.3(1) Se(2)-Cu(2)-Se(3) 123.7(1) Se(2)-Cu(2)-Se(5) 122.5(1) Se(3)-Cu(2)-Se(5) 110.6(1) Se(1)-Cu(l)-Cu(l') 87.3(1) Se(S)-Cu(2)-Cu(2') 84.0(1) Cu-Se (ave.) 240(2) Se(1)-P( 1)-Se(3) P( 1 )-P( 2) 2.260(6) Se(3)-P( l)-Se(4) Se(1)-P(l)-Se(4) 112.6(2) 114.1(2) 111.4(2) 114.1(2) 110.4(2) Se(2)-P(2)-Se(5) Se(2)-P( 2)-Se(6) Se(S)-P(2)-Se(6) 113.3(3) Se-P-Se (ave.) 113(2) P(1)-Se(1) P(1)-Se(3) P(1)-Se(4) P(2)-Se(2) P(2)-Se(5) P(2)-Se(6) P—Se (ave.) 2.186(4) 2.177(4) 2.202(5) 2.186(4) 2.194(5) 2.187(4) 2.189(8) Cs(l)-Se(l) 3.733(2) Cs(2)-Se(l) Cs(l)-Se(2) 3.725(2) Cs(2)-Se(3) Cs(l)-Se(2') 3.847(3) Cs(2)-Se(3') Cs(1)-Se(4) 3.861(2) Cs(2)-Se(4) Cs(l)-Se(5) 3.948(2) Cs(Z)-Se(4') Cs(l)-Se(5') 3.797(3) Cs(2)-Se(5) . Cs(l)-Se(6) 3.887(2) Cs(2)-Se(6) Cs(1)-Se(6') 3.759(2) Cs-Se (ave.) Cs(1)-Se(6") 3.967(2) 3.766(2) 3.766(2) 3.711(3) 3.694(2) 3.780(2) 3.877(2) 3.733(2) 3.80(8) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01={2n(1n- DZ/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 378 Table 6-19. Selected Distances (A) and Angles (°) for CszAgzPZSe6 with Standard Deviations in Parenthesesa Ag-Ag' 2.919(3) Se(1)-Ag-Se(2) 105.55(7) Ag-Se( 1) Ag-Se( 2) Ag-Se( 3) 2.595(2) Se( 2)-Ag-Se( 3) 121.77(8) 2.641 (3) Ag-Ag'-Se( 2) 2.545(2) Ag-Ag'-Se( 3) 9405(9) 9572(8) Se( 1 )-Ag-Se( 3) 132.04(9) Ag-Se (ave.) 2.S9(S) P—Se( 1) P-Se(2) P-Se( 3) 2.183(5) Se(1)-P-Se(2) 2.182(4) Se(1)-P—Se(3) 2.205(4) Se( 2 )-P-Se( 3) P-Se (ave.) 2.19(1) Se(1)-P-P' P-P' 2.285(8) Cs-Se( 1) Cs-Se(1') 3.788(2) Cs-Se(2") 3.819(3) Cs-Se( 3) Cs-Se( 1") 4.091(3) Cs-Se(3') 3.720(2) Cs-Se(3") 110.1(2) 115.7(2) 109.9(2) 106.5(3) 3.939(2) 3.790(2) 3.663(3) 3.725(2) Cs-Se(2) Cs-Se(2') 3.963(3) Cs-Se (ave.) 3.8(1) 21The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 61={zn(1n- DZ/n(n-1))1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 379 Table 6-20. Selected Distances (A) and Angles (°) for KzAgzPZSe6 with Standard Deviations in Parentheses.al Ag( 1 )-Ag(2) Ag( 1 )-Ag(3) Ag(l)-Se(2) Ag(l)-Se(3) Ag(1)-Se(4) Ag(1)-Se(7) Ag(2)-Se(3) Ag(2)-Se(3) Ag(2)-Se(3) Ag(2)-Se(3) Ag(3)-Se(1) Ag(3)-Se(1) Ag(3)-Se(1) Ag(3)-Se(1) Ag-Se (ave.) P(1)-P(2) P(3)-P(3') P(1)-Se(1) P(1)—Se(4) P(1)-Se(7) P(2)-Se(2) P(2)-Se(3) . P(2)-Se(9) P(3)-Se(5) P(3)-Se(6) P(3)-Se(8) P-Se (ave.) 3.869(4) 3.715(4) 2.662(4) 2.615(4) 2.786(4) 2.743(4) 2.797(4) 2.578(4) 2.678(4) 2.783(4) 2.653(4) 2.633(4) 2.662(4) 2.695(4) 2.69(7) 2.278(8) 231(1) 2.188(7) 2.178(8) 2.191(7) 2.186(8) 2.213(7) 2.180(7) 2.205(7) 2.192(7) 2.165(7) 2.19(1) K(1)-Se(4) K(1)-Se(6) K(1)-Se(7) K(1)-Se(8) K(1)-Se(9) K(1)-Se(9') K(2)-Se(4) K(2)-Se(5) K(2)-Se(6) K(2)-Se(6') K(2)-Se(7) K(2)-Se(8) K(3)-Se(4) K(3)-Se(5) K(3)-Se(7) K(3)-Se(8) 1((3)-Se(9) K-Se (ave.) ‘ Se(1)-P(1)-Se(4) Se(1)-P(1)-Se(7) Se(4)-P(1)-Se(7) Se(2)-Ag(l)-Se(3) 130.5(1) Se(2)-P(2)-Se(3) Se(2)-Ag(l)-Se(4) 95.1(1) Se(2)-P(2)-Se(9) Se(3)-Ag(2)-Se(5) 130.4(1) Se(3)-P(2)-Se(9) Se(3)-Ag(2)-Se(9) 80.9(1) Se(S)-P(3)-Se(6) Se(1)-Ag(3)-Se(6) 122.1(1) Se(5)-P(3)-Se(8) Se(1)-Ag(3)-Se(2) 102.0(2) Se(6)-P(3)-Se(8) 3.702(8) 3.454(7) 3.487(7) 3.379(7) 3.614(8) 3.360(8) 3.376(7) 3.361(7) 3.661(8) 3.384(7) 3.613(8) 3.586(7) 3.676(8) 3.867(9) 3.423(8) 3.499(9) 3.351(8) 3.5(2) 113.4(3) 111.7(3) 115.0(3) 110.5(3) 111.1(3) 111.1(3) 109.7(3) 111.5(3) 113.9(3) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 61={2n(1n- 1)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds. 380 Table 6-2 1. Selected Distances (A) and Angles (°) for K4Au2(PZSe6)2 with Standard Deviations in Parentheses.a Au(1)-Au(2) 3.636(3) Se(1)-Au(1)-Se(1') 180(1) Au(1)-Se(1) x 2 2.484(6) Se(1)-Au(1)-Se(3') 99.6(2) Au(1)-Se(3) x 2 2.493(5) Se(3)-Au(l)-Se(3') 180.CI) Se(1)-Au(1)-Se(3) 80.4(2) Au(2)-Se(4) x 2 2.302(7) Se(4)-Au(2)-Se(4') 1800) P(1)-Se(3) P(1)-Se(4) P(1)-Se(5) P(2)-Se(1) P(2)-Se(2) P(2)-Se(6) 2.21(1) 2.18( 1) 2.14(1) 2.25(1) 2.14(1) 2.14(1) Se(3)-P(l)-Se(4) 114.0(7) Se(3)-P(1)-Se(5) 109.0(6) Se(4)-P( l )-Se(5) 112.9(6) Se(1)-P(2)-Se(2) 111.7(6) Se(1)-P(2)-Se(6) 109.7(6) Se(2)-P(2)-Se(6) 113.7(7) P-Se (ave.) 2.18(5) Se-P-Se (ave.) 112(2) P(1)-P(2) 2.25(2) K(1)-Se(2) K(1)-Se(2') K(1)-Se(2") K(1)-Se(4) K(1)-Se(5) K( 1 )-Se( 5') K(1)-Se(5") K(1)-Se(6) 3.51(2) 346(2) 3.49(2) 3.S6(2) 3.72(2) 3.49(2) 3.57(2) 3.38(2) K(2)-Se(1) K(2)-Se(1') K(2)-Se(2) K(2)-Se(2') K(2)-Se( 5) K(2)-Se(6) K( 2)-Se(6') K(2)-Se (ave.) 3.55(2) 3.87(2) 3.46(2) 3.49(2) 368(2) 3.46(2) 335(2) 3.6(2) K(1)-Se (ave.) 3.5(1) KL2)-Au(2) 3.21(2) x2 aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equations 01={2n(1n- I)2/n(n-1)}1/2, where In is the length (or angle) of the nth bond, I the mean length (or angle), and n the number of bonds. 381 3.3 Magnetic Susceptibility. The temperature dependent magnetic susceptibility was collected for KzMnPZSe6 (I ), szMnPZSeg, (I I), CszManSe6 (I I I), and KzFeP28e6 (IV). All compounds are paramagnetic above ~30 K and obey Curie-Weiss law. Typical susceptibility plots at 2000 G are shown in Figure 6-14 for (I I I) and (IV). Table 6-22 lists the magnetic parameters for (I), (I I), (I I I) and (IV). The calculated moments are 5.8, 5.6, 6.0 and 5.4 BM for (I -I V), respectively. These values indicate that the M2+ ions in these compounds are in high-spin configuratons and agree well with the calculated spin-only moments (ueff = [4S(S + 1)]1/2). For comparison, the magnetic moments for anPZSe6 and FezP28e6 are 6.1 and 5.1 BM, respectively.41 Below 30 K, the susceptibility show maxima indicating antiferromagnetic coupling below this temperature. Since it is well-known that exchange interactions are very sensitive to small changes in the bond length and angles of the bridging system,42 we hoped to observe such effects in the AzMnPZSe6 family of compounds. The effect of the small reduction in the M---M distance and slight reorientation of the [P28e6]4' ligand upon moving from KzMnPZSee (Mn-Mn=6.5349(9) A) to CszMnPZSen (Mn-Mn=6.476(2) A) is not reflected in the magnetic data. The Neél temperatures (TN) for AzMnPZSee vary between 10 and 13 K with Weiss constants (6) in the -14 to -25 K range. Since the M-M distances are large (> 6.4 A), it is assumed that the weak antiferromagnetic interactions proceed via a superexchange pathway through the bridging [P28e6]4- ligand. A stronger antiferromagnetic interaction is observed in KzFePzSe6 with higher TN and 0 values of 20 K and -43 K. A similar increase in the 382 interaction upon moving from Mn to Fe is also observed for the strongly antiferromagnetic, two-dimensional Msztx (M=Mn, Fe, Ni) compounds.41 The temperature dependent magnetic susceptibility of the solid solution compound, KzMn05 Feo,5P28e6, was performed (at 2000 G) to investigate the possibility of unusual magnetic behavior due to incomplete cancellation of magnetic moments. The magnetic parameters (TN=10 K, 0=-32 K) are comparable to those observed for the AzM PZSee compounds shown above. Table 6—22. Magnetic parameters for (I ), (I I), (11 I), and (I V). Formula neff (BM) 0 (K) TN (K) K2MnP28e6 5 .8 szMnP28e6 5.6 CszManSee KZFegSee 6.0 5.4 - 2 5 - 1 8 - 14 -43 10 1 2 13 20 383 m x l l 3'. u 7“.) r '0 To u'o I'D T K I I l O 50 100 150 200 250 300 350 Temperature(K) E 1. H I. u 3. run—m 0 II I l l l O l l . I 0 50 100 150 200 250 300 350 Temper-WOO Figure 5.14; Plots of l/xM vs. T taken at 2000 G of applied field over 2-300 K for: (A) CszanPzSee and (B) KzFestea. The inset graphs show expanded views of the region 2-30 K. ' 384 3.4. Conclusions The general family of one-dimensional A2M(P28e6) compounds derives from the well known two-dimensional M2P28e6 family by incorporation of A4P28e6. The synthesis of new quaternary selenophosphates in molten Ax[PySeZ] salts provides a useful and broad synthetic methodology that can be applied to main group elements as well as transition metals. The AXIPySeZ] fluxes provide reliably [P28e6]4- units that display remarkable versatility in terms of ligand binding to metals. The high negative charge of these units makes them hard to stabilize in conventional aqueous or organic solvents. It would be interesting to see whether changes in the nominal stoichiometry of these fluxes can result in other [PxSeyPl- units as well. For example, by increasing the basicity of the flux, the [PSe4]3' ligand can be stabilized in CstPSe4.19 385 List of References - : 9 1 9 9 $ McCarthy, T. J.; Kanatzidis, M. G., Chem. Mater. 1993, 5,1061-1063. McCarthy, T. J.; Kanatzidis, M. G. accepted by J. Chem. Soc., Chem. Commun.. McCarthy, T. J.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. submitted for publication. (a) Klingen, W.; Eulenberger, G.; Hahn, H. Z. Anorg. AIIg. Chem., 1973, 401, 97-112. (b) Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst, Sect. B, 1978, 34, 1779-1781. (c) Klingen, W.; Ott, R.; Hahn, H. Z. Anorg. Allg. Chem., 1973, 396, 271-278. (d)Jandali, M. Z.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem., 1978,447,105-118. (a) Ouvrard, G.; Brec, R.; Rouxel, J. Mater. Res. BuII., 1985, 20,1181-1189. (b) Lee, S.; Colombet, P.; Ouvrard, G.; Brec, R. Inorg. Chem., 1988, 27, 1291-1294. (c) Lee, S.; Colombet, P.; Ouvrard, G.; Brec, R. Mater. Res. BuII., 1986, 21, 917-928. ((1) Durand, E.; Ouvrard, G.; Evain, M.; Brec, R. Inorg. Chem., 1990, 29,4916-4920. (a) Odile, J.-P.; Steger, J. J.; Wold, A. Inorg. Chem. 1975, 14,2400-2402. (b) Taylor, B. E.; Steger. J. J.; Wold, A.; Kostiner, E. Inorg. Chem. 1974, 13, 2719-2421. (c) Taylor, B. E.; Steger. J. J.; Wold, A. J. Solid State Chem. 1973, 7,461-467. (a) Lagadic, 1.; Leaustic, A.; Clement, R. J. Chem. Soc., Chem. Commun. 1992, 1396-1397. (b) Clement, R.; Audiere, J.-P.; Renard, J.-P. Rev. Chim. Miner. 1982, 19, 560—571. (b) Michalowicz, A.; Clement R. Inorg. Chem. 1982, 21,3872-3877. (c) Clement R. J. Chem. Soc., Chem. Commun. 1980, 647-648. (d) Joy, P. A.; Vasudevan, S. J. Am. Chem. Soc., 1981, 114, 7792- 7801. Thompson, A. H.; Whittingham, M. S. U. S. Patent 4,049,879 1977. (b) Brec, R.; Le Mehaute', A. Fr. Patents 7, 704,519 1977. (a) Pfeiff, R.; Kniep, R. J. AIons and Compounds, 1992, 186, Ill-133. (b) Evain, M.; Boucher, F.; Brec, R.; Mathey, Y. J. SoIid State Chem. 1991, 90, 8-16. (c) Lee, S.; Colombet, P.; Ouvrard, G.; Brec, R. Mat. Res. BuII., 1986, 21, 917-928. (d) Leblanc, A.; Ouili, Z.; Colombet, P. Mat. Res. BuII., 1985, 20, 947-954. 10. 11. 12. Lacroix, P. G.; Clement, R.; Nakatani, K.; Zyss, J.; Ledoux, 1. Science, 1994, 263, 658-660. Garin, J.; Parthe, E. Acta Crystallogn, 1972, 828,3672-3674. Fritz, 1. J.; Isaacs, T. J.; Gottlieb, M.; Morosin, B. SoIid State Commun., 1978, 27, 535. 13. (a) O'Neal, S. C.; Pennington, W. T.; Kolis, J. W. Angew. Chem. Int. Ed. EngI. 1990, 29, 1486-1488. (b) Zhao, J.; Pennington, W. T.; Kolis, J. W. J. Chem. Soc., Chem. Commun. 1992, 265-266. 14. Feher, F. Handbuch der Praparativen Anorganischen Chemie: Brauer, G., Ed.; Ferdinand Enke: Stuttgart, Germany, 1954; pp. 280—281. 15. CERIUS: Molecular Simulation Software, Version 3.0, (1992), Cambridge Molecular Design, Waltham, MA 02154. 16. G. M. Sheldrick, In CrystaIIographic Computing 3; Sheldrick, G. M., Kruger, C., Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985; PP175-189. 17. TEXSAN: Single Crystal Structure Analysis Software, Version 5.0, (1981). Molecular Structure Corportion, The Woodlands, TX 77381. 386 18. 19. 20. 21. Walker, N.; Stuart, D. Acta Cryst, 1983, A39,158—166. McCarthy, T. J.; Chondroudis, K.; Kanatzidis, M. G. work in progress. Mathey, Y.; Clement, R.; Sourisseau, C.; Lucazeau, G. Inorg. Chem., 1980, 19,2773-2779. (a) Dance, 1. G. Polyhedron 1983, 2, 1031-1043. (b) Hollander, F. J.; Coucouvanis, D. J. Am. Chem. Soc. 1977, 99, 6268-6279. (c) Chadha, R.; Kumar, R.; Tuck, D. G. J. Chem. Soc., Chem. Commun. 1986, 188-189. ((1) Coucouvanis, D.; Murphy, C. N.; Kanodia, S K. Inorg. Chem. 1980, 19, 2993-2998. Burschka, C.; Bronger, W. Z. Naturforsch. 1977, 323, 11-14. Burschka, C. Z. Naturforsch. 1979, 348, 675-677. Schils, H.; Bronger, W. Z Anorg. Allg. Chem. 1979, 456, 187-193. Savelberg, G.; Schafer, H. Z. Naturforsch. 1978, 338,711-713. (a) Merz, K. M.; Hoffmann R. Inorg. Chem. 1988, 27, 2120-2127. (b) Mehrotra, P. K.; Hoffmann, R. Inorg. Chem. 1978, 17,2187-2189. (a) Cotton, F. A.; Feng, X.; Matusz, M.; Poli, R. J. Am. Chem. Soc. 1988, 110, 7077-7083. (b) Lee, S. W.; Trogler, W. C. Inorg. Chem. 1990, 29,1659-1662. Muller, U.; Ha-Eierdanz, M.-L.; Krauter, G.; Dehnicke, K. Z. Naturforsch. 1990, 45b,1128-1132. Cusick, J.; Scudder, M. L; Craig, D. C.; Dance; 1. G. Polyhedron, 1989, 8, 1139-1141. Janzen, M. Angew. Chem. Int. Ed. Engl. 1987, 26,1098-1110. (a) Huang, S.-P.; Kanatzidis, M. G. Inorg. Chem. 1991, 30, 1455-1466. (b) Kanatzidis, M. G.; Huang, S.-P. J. Am. Chem. Soc. 1989, 111,760-761. Colombet, P.; Leblanc, A; Danot, M.; Rouxel, J. J Solid State Chem. 1982, 41, 174-184. Maisonneuve, V.; Cajipe, V. B.; Payen, C. Chem. Mater. 1993, 5, 758-760. Mathey, Y.; Michalowicz, A.; Toffoli, P.; Vlaic, G. Inorg. Chem. 1984, 23, 897-902. Mathey, Y.; Mercier, H.; Michalowicz, A.; Leblanc, A. J. Phys. Chem. Solids, 1985, 46,1025-1029. Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst. 1978, 834,1779-1781. Rabenau, A.; Schulz, H. J. Less-Common Metals, 1976, 48, 89-101. Kanatzidis, M. G.; Huang, S.-P. Phosphorous and Sulfur, 1992, 64,153- 160. Haushalter, R. C. Angew. Chem. Int. Ed. Engl., 1985, 24, 432-433. Kanatzidis, M. G.; Huang, S.-P. Angew. Chem. Int. Ed. Engl., 1992, 31 ,787- 789. (a) Brec, R.; Schleich, D. M.; Ouvrard, G.; Louisy, A.; Rouxel, J. Inorg. Chem., 1979, 18, 1814-1818. (b) Odile, J.-P.; Steger, J. J.; Wold, A. Inorg. Chem. 1975, 14, 2400-2402. (c) Taylor, B. E.; Steger, J. J.; Wold, A.; Kostiner, E. Inorg. Chem. 1974, 13, 2719-2421. (d) Taylor, B. E.; Steger. J. J.; Wold, A. J. Solid State Chem. 1973, 7,461-467. 42. Crawford, V. H.; Richardson, H. W.; Wasson, J. R.; Hodgson, D. J.; Hatfield, W. E, Inorg. Chem., 1976, 15,2107-2110. "'1 "'1'; A“ “71111111111111!(Jrilfylfllllgijllml'ES