LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MTE DUE MTE DUE DATE DUE 1M chlRC/DahfinpfiS—p.“ MOLTEN SALT SYNTHESIS OF QUATERNARY CHALCOANTIMONATES AND THIOPHOSPHATES By Jason A. Hanko A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1998 ABSTRACT MOLTEN SALT SYNTHESIS OF QUATERNARY CHALCOANTIMONATES AND THIOPHOSPHATES By Jason A. Hanko We used the now proven molten alkali polychalcogenide flux method to synthesize new quaternary chalcoantimonate and thiophosphate materials. The chalcoantimonate and thiophosphate fluxes are formed by the in situ fusion of AZQ/Sb/Q (A = alkali metal; Q = S, Se) or AZS/PZS5/S (A = alkali metal) forming highly reactive [beQyP' and [PxSy]"' units solubilized in the excess polychalcogenide flux. These molecular building blocks coordinate to metal ions in a multitude of ways, building extended lattices stabilized by alkali cations. By examining the coordination chemistry of these systems, we discovered that the thiophosphate system is best thought of as completely different than the corresponding selenophosphate systems and different materials are formed under similar experimental conditions. In the thiophosphate system the P5+ species appears to be stable under a variety of conditions, and increasing the binary P285 concentration stabilized higher nuclearity [PxSy]“’ units. While the Sb5+ species is readily observed in the chalcoantimonate systems, it is exclusively observed as the discrete tetrahedral unit. The enhanced stability of the Sb3+ species increases the diversity of the coordination chemistry displayed by the chalcoantimonate system due to the complicated equilibrium stabilizing higher nuclearity [beSy]“‘ units. Changing the Lewis basicity in of the chalcoantimonate fluxes resulted in the stabilization of recognizable [beSy]“' structural fragments that could be used as building blocks forming compounds with several potential applications. Several new [PxSyP' and [beQy]“' units have been synthesized and their coordination chemistry examined. In this dissertation, the synthesis, and characterization and properties of several new silver chalcoantimonates AzAngS4 (A = K, Rb, Cs), Cs3AgZSb3Q8 (Q = S, Se), a,B-RbAg28bS4, and three new silver-rich gold thioantimonate compounds (LB- szAg208b4819 and CszAgZOSb4819 will be discussed. Extending this methodology to gold, three new compounds, A2AquS4 (A = Rb, Cs) and szAu6Sb4SIO were synthesized, further proving the generality of the method at accommodate a variety of metals. Utilizing the corresponding thiophosphate flux a number of new quaternary compounds A2CuP389 (A = K, Rb, Cs), K3CuPZS7, CSZCuzPZS6, AzAuPS4 (A = K, Rb, Cs), and AAuPZS7 (A = Rb, Cs), were discovered. The synthetic work performed here provides the ground work for systematic synthesis, and further exploration into these and other systems. DEDICATION To my wife, Pauline for her patience and understanding. iv ACKNOWLEDGMENTS Iwould like to thank Prof. Mercouri G. Kanatzidis for his patience, understanding and his deep burning passion for chemistry. I could not have accomplished any of this without him. I would also like to thank Professors Dunbar, Crouch, and McCracken for serving as members of my guidance committee. I will not forget their kind words of encouragement. Iwould like to especially thank Prof. Dunbar, for withstanding the sonic barrage while Calvin and I studied for 812 during our first semester. Throughout my time here] was fortunate enough to have worked and interacted with some of the finest scientists, support staff, and personalities. I thank them all for their unique experiences and the fruitful discussions we shared. I want to especially thank Dean Lantero Carl N. Iverson, and Ms. Joy Heising for several experiences that will last a lifetime. Last, but certainly not least, I would like to thank my wife, Pauline, and my children, Courteny and Trent. I am truly blessed to have such a wonderful wife and two loving children that understood what all of this work and late nights were about. Financial support from the National Science Foundation, NSF-CRG, NSF- MRSEC, the Petroleum Research Fund administered by the American Chemical Society, and the Center for Fundamental Material Research are gratefully acknowledged for funding this work. TABLE OF CONTENTS LIST OF TABLES .................................................................................. xii LIST OF FIGURES ............................................................................... xvii CHAPTER] 1. Introduction ...................................................................................... 1 2. Nature of the Chalcophosphate Fluxes ...................................................... 11 3. Nature of the Chalcoantimonate Fluxes ..................................................... 18 CHAPTER2 Chemistry Of Silver in Molten Alkali Metal Polychalcoantimonate Fluxes. Synthesis and Characterization of the Quaternary Compounds A2AngS4 (A = K, Rb, Cs), and CS3Ag2$b3Q3 (Q = S, Se). 1. Introduction ..................................................................................... 33 2. Experimental Section .......................................................................... 34 2.1. Reagents ............................................................................. 34 2.2. Syntheses ............................................................................ 35 A28 (A = K, Rb, Cs) and C328e ............................................... 35 K2AngS4 (I) .................................................................... 35 szAngS4 (II) .................................................................. 35 CszAngS4 (III) ................................................................. 36 CS3Ag28b333 (IV) ............................................................... 36 Vi CS3Ag23b3Seg (V) .............................................................. 36 2.3. Physical Measurements ............................................................ 37 3. Results and Discussion ........................................................................ 53 3.1. Description of Structures ........................................................... 53 Structure of K2AngS4 (I) ..................................................... 53 Structure of szAngS4 (II) ................................................... 54 Structure of CszAngS4 (III) .................................................. 55 Structure of CS3AgZSb3Se3 (V) ................................................ 56 3.2. Synthesis, Spectroscopy and Thermal Analysis ................................ 72 3.3. Structural Relationships in AzAngS4 (A = K, Rb, Cs). The Counterion Effect .............................................................. 86 4. Conclusions ..................................................................................... 87 CHAPTER 3 Thioantimonate Flux Synthesis of oc,|3-Rb AgngS4: Two New Three-Dimensional Compounds With Acentric Structures. 1. Introduction ..................................................................................... 92 2. Experimental Section ........................................................................... 93 2.1. Reagents .............................................................................. 93 2.2. Synthesis ............................................................................. 95 RbZS ................................................................................ 93 a—RbAg2SbS4 (I) ................................................................ 93 B—RbAngbS4 (II) ............................................................... 94 2.3. Physical Measurements ............................................................. 94 3- Results and Discussion ....................................................................... 102 3.1. Description of Structures ......................................................... 102 vii a—RbAgZSbS4 (1) .............................................................. 102 B-RbAngbS4 (II) .............................................................. 104 3.2. Synthesis and Physicochemical Properties ...................................... 117 3.3. Thermal Analysis ................................................................... 118 4. Conclusions ................................................................................... 136 CHAPTER 4 Synthesis and Characterization of the Novel Mixed Valent Quaternary Silver-Rich Thioantimonates A2Ag208b4S19 (A = Rb, Cs) and B—szAgzoSb4819. A Thioantimonate with a Large Supercell. 1. Introduction .................................................................................... 140 2. Experimental Section .......................................................................... 141 2.1. Reagents ............................................................................. 141 2.2. Syntheses ........................................................................... 141 A28 (A = Rb, Cs) ................................................................ 141 szAgzoSb4819 (I) ............................................................. 142 CSzAg208b4819 (II) ............................................................ 142 B-szAgzoSb4819 (III) ........................................................ 142 2.3. Physical Measurements ............................................................ 143 3. Results and Discussion ....................................................................... 154 3.1. Description of Structures ......................................................... 154 CszAgzoSb4819 (II) ............................................................ 154 B-szAgzoSb4Slg (III) ........................................................ 156 3.2. Synthesis, Spectroscopy and Thermal Analysis ................................ 181 4- Conclusions ................................................................................... 188 viii CHAPTER 5 The Chemistry of Au in Molten Polythioantimonate Fluxes. Synthesis and Characterization of the New Multinary Gold Thioantimonates AzAquS4 (A = Rb, Cs) and Rb2Au68b4Sm. A Novel Compound with a Binary and Ternary Interconnected Framework. 1. Introduction .................................................................................... 192 2. Experimental Section .......................................................................... 193 2.1. Reagents ............................................................................. 193 2.2. Syntheses ........................................................................... 194 A28 (A = Rb, Cs) ............................................................... 194 szAquS4 (I) .................................................................. 194 CszAquS4 (II) ................................................................. 194 szAu6Sb4Sm (III) ............................................................ 194 2.3. Physical Measurements ............................................................ 195 3. Results and Discussion ....................................................................... 205 3.1. Description of Structures .......................................................... 205 szAquS4 (I) .................................................................. 205 Rb2Au68b4810 (III) ............................................................ 206 3.2. Synthesis, Spectroscopy and Thermal Analysis ............................... 217 4. Conclusions ................................................................................... 227 CHAPTER6 A2CuP389 (A = K, Rb), CszCu2P285, and K3CuP287: New Phases from the Dissolution of Copper in Molten Polythiophosphate Fluxes. 1- Introduction .................................................................................... 232 2- EXperimental Section .......................................................................... 234 ix 2.1. Reagents ............................................................................. 234 2.2. Syntheses ........................................................................... 234 A28 (A = K, Rb, Cs) ........................................................... 234 K2CuP389 (I) .................................................................... 234 Rb2CuP389 (II) ................................................................. 235 CszCu2P2$6 (III) ............................................................... 235 K3CuP287 (IV) .................................................................. 235 2.3. Physical Measurements ............................................................ 236 3. Results and Discussion ....................................................................... 250 3.1. Description of Structures .......................................................... 250 Structure of K2CuP389 (I) ..................................................... 250 Structure of CSZCUZP286 (III) ................................................. 251 Structure of K3CuP2S7 (IV) ................................................... 252 3 .2. Synthesis and Comparison between the Thiophosphate and Selenophosphate Fluxes ...................................................... 264 3.3. Physicochemical Properties ....................................................... 265 3.4. Band Structure Calculations ..................................................... 276 4. Conclusions ................................................................................... 277 CHAPTER7 Chemistry of Gold in Molten Alkali Metal Polychalcogeno—phosphate Fluxes. Synthesis and characterization of the low dimensional Compounds A2AuPS4 (A = K, Rb, Cs) and AAuP287 (A = K, Rb) 1. Introduction .................................................................................... 283 2- EXperimental Section .......................................................................... 284 2.1. Reagents ............................................................................. 284 2.2. Syntheses ........................................................................... 285 A28 (A = K, Rb, Cs) ............................................................. 285 K2AuPS4 (I) ..................................................................... 285 szAuPS4 (II) ................................................................... 285 CszAuPS4 (III) .................................................................. 286 KAuP2S7 (IV) ................................................................... 286 RbAuP287 (V) ................................................................... 286 2.3. Physical Measurements ............................................................ 287 3. Results and Discussion ....................................................................... 300 3.1. Description of Structures .......................................................... 300 Structure of K2AuPS4 (I) ...................................................... 300 Structure of CszAuPS4 (III) ................................................... 301 Structure of KAuP287 (IV) ..................................................... 302 3.2. Synthesis, Spectroscopy and Thermal Analysis ................................ 316 4. Conclusions .................................................................................... 329 CONCLUSIONS AND OUTLOOK ............................................................ 333 APPENDIXI Thiophosphate Flux Synthesis of CszCuP3S9: An Unusual Acentric Compound with One-Dimensional Screw Helices ............................................ 335 xi 1-1. 1-2. 2-1. 2-2. 2-4. 2-5. 2-7. 2-9. 2-11. 2-12. 2-13. 2-14. LIST OF TABLES Synthetic conditions for the different [Psz]"' and [Sbsz]“’units. (M = metal, A2Q = alkali chalcogenide) ................................................ 22 Structure and coordination example of the various [beQyP' units ................. 23 Calculated and Observed X—ray Powder Patterns for K2AngS4 (I) ............... 40 Calculated and Observed X-ray Powder Patterns for szAngS4 (H) ............. 42 Calculated and Observed X-ray Powder Patterns for CszAngS4 (HI) ............ 44 Calculated and Observed X-ray Powder Patterns for CS3Ag2Sb3Se3 (V) .......... 46 Crystallographic Data for K2AngS4, szAngS4, CszAngS4, CS3Ag23b3Seg ............................................................................. 48 Fractional Atomic Coordinates and B(eq) Values for KzAngS4 (I) with Estimated Standard Deviations in Parentheses ................................... 49 Fractional Atomic Coordinates and B(eq) Values for szAngS4 (H) with Estimated Standard Deviations in Parentheses ................................... 50 Fractional Atomic Coordinates and B(eq) Values for CszAngS4 (III) with Estimated Standard Deviations in Parentheses ................................... 51 Fractional Atomic Coordinates and B(eq) Values for CS3Ag28b3Se3 (V) with Estimated Standard Deviations in Parentheses ........... 52 Selected Distances (A) and Angles (deg) for KzAngS4 (I) with Standard Deviations in Parenthesesa .................................................... 58 Selected Distances (A) and Angles (deg) for szAngS4 (II) with Standard Deviations in Parenthesesa .................................................... 59 Selected Distances (A) and Angles (deg) for CszAngS4 (III) with Standard Deviations in Parenthesesa .................................................... 60 Selected Distances (A) and Angles (deg) for CS3Ag2Sb3Seg (V) with Standard Deviations in Parenthesesa .............................................. 61 Infrared and Raman Data (cm'l) for (I) - (V) ........................................... 77 xii 2-15. 3-1. 3-3. 3-4. 3-5. 3-6. 3-7. 3-8. 3-9. 4-1. 4-2. 4-4. 4-5. 4-6. Optical Band Gaps from Powder and Single Crystal Measurements and Melting Point Data for (I) - (V) .................................... 78 Calculated and Observed X-ray Powder Patterns for oc-RbAgzsbS4 (I) ............ 98 Calculated and Observed X-ray Powder Patterns for B-RbAgzsbS4 (II) ........... 99 Crystallographic Data for a-RbAgQSbS4 and B-RbAgZSbS4 ....................... 100 Atomic Coordinates ( x 104) and Equivalent Isotropic Displacement Parameters (A2 x 103) for Ot-RbAgZSbS4 (I) with Estimated Standard Deviations in Parentheses ..................................................... 101 Fractional Atomic Coordinates and B(eq) Values for B-RbAg2SbS4 (H) with Estimated Standard Deviations in Parentheses .................................. 101 Selected Distances (A) and Angles (deg) for a-RbAgZSbS4 (I)with Standard Deviations in Parenthesesil ................................................... 106 Selected Distances (A) and Angles (deg) for B-RbAg28b84 (II) with Standard Deviations in Parentheses‘il ................................................... 107 Summary of the thermal behavior of or-RbAgZSbSd, ................................ 123 Infrared and Raman Data for (I) and (II) ............................................... 125 Calculated and Observed X-ray Powder Patterns for CszAgzoSb4819 (II) ........ 148 Crystallographic Data for CszAg208b4S19 and B-szAgzoSb4S19 ........................................................................ 150 Fractional Atomic Coordinates and B(eq) Values for CszAgzoSb4819 (II) with Estimated Standard Deviations in Parentheses .................................. 151 Atomic coordinates ( x 104) and equivalent isotropic displacement parameters (A2 x 103) for B-szAg208b4519 (ID) with Estimated Standard Deviations in Parentheses ..................................................... 153 Selected Distances (A) for CszAgzoSb4819 (H) with Standard Deviations in Parentheseszil .............................................................. 159 Selected Bond Angles (0) for CszAg208b4819 (H) with Standard xiii 4-7. 4-8. 5-1. 5-3. 5-4. 5-5. 5-6. 5-7. 6-2. 6-3. 6-4. 6-5. 6-6. 6-7. Deviations in Parentheses a .............................................................. 163 Selected Distances (A) for B-szAgzoSb4819 (HI) with Standard Deviations in Parentheses a .............................................................. 165 Selected Bond Angles (0) for B-szAg208b4819 (H) with Standard Deviations in Parentheses a .............................................................. 167 Calculated and Observed X-ray Powder Patterns for szAquS4 (I) .............. 198 Calculated and Observed X - Ray Powder Patterns for szAu6Sb4S10 (III) ...................................................................... 200 Crystallographic Data for szAquS4 and szAu68b4SIO ......................... 202 Fractional Atomic Coordinates and B(eq) Values for szAquS4 with Estimated Standard Deviations in Parentheses .................................. 203 Fractional Atomic Coordinates and B(eq) Values for szAu6Sb4Sm with Estimated Standard Deviations in Parentheses .................................. 204 Selected Distances (A) and Angles (deg) for szAquS4 (I) with Estimated Standard Deviations in Parentheses a ....................................... 220 Selected Distances (A) and Angles (deg) for szAu6Sb4S10 ....................... 221 Calculated and Observed X-ray Powder Patterns for K2CuP3S9 (I) ............... 240 Calculated and Observed X-ray Powder Patterns for CszCu2P286 (1H) ........... 243 Calculated and Observed X-ray Powder Patterns for K3CuP2S7 (IV) ............. 244 Crystallographic data for K3CuP389, CS2C112P235, and K 3 C u P 2 S 7 ................................................................................. 246 Fractional coordinates and equivalent isotropic atomic displacement parameters of K2C u P 38 9 ................................................................ 247 Atomic coordinates (x 10‘) and equivalent isotropic displacement parameters (A2 x 103) for CszCu2P286 ................................................. 248 Atomic coordinates (x 10‘) and equivalent isotropic displacement parameters (A2 x 103) for K3CuP2S7 ................................................... 249 xiv 6-8. 6-9. 6-10. 6-11. 7-1. 7-2. 7-3. 7-4. 7-5. 7-6. 7-7. 7-8. 7-9. 7-10. 7-11. Selected band distances (A) and angles (°) for K2CuP389 (I) with Standard Deviations in Parenthesesa ................................................... 254 Selected Distances (A) and Angles (deg) for CszCu2P286 (HI) with Standard Deviations in Parenthesesa ................................................... 255 Selected Distances (A) and Angles (deg) for K3CuPZS7 (IV) with Standard Deviations in Parenthesesa ................................................... 256 IR and Raman Spectroscopic Data for K2CuP389, CS2CU2P236, and K3CuPZS7 ............................................................................ 275 Calculated and Observed X-ray Powder Patterns for K2AuPS4 (I) ................ 290 Calculated and Observed X-ray Powder Patterns for CszAuPS4 (III) ............. 292 Calculated and Observed X-ray Powder Patterns for KAuPZS7 (IV) .............. 294 Crystallographic data for K2AuPS4, CszAuPS4 and K A 11 P 2 S 7 .................................................................................. 296 Positional parameters and B(eq) for K2AuPS4 with Estimated Standard Deviations in Parentheses a ................................................... 297 Positional parameters and B(eq)a for CszAuPS4 with Estimated Standard Deviations in Parentheses .................................................... 298 Positional parameters and B(eq)3 for KAuPZS7 with Estimated Standard Deviations in Parentheses .................................................... 299 Selected Distances (A) and Angles (deg) for K2AuPS4 (I) with Standard Deviations in Parenthesesa ................................................... 304 Selected Distances (A) and Angles (deg) for CszAuPS4 (IH) with Standard Deviations in Parenthesesa ................................................... 305 Selected Distances (A) and Angles (deg) for KAuP2S7 (IV) with Standard Deviations in Parenthesesa ................................................... 306 Synthetic conditions for the different [Psz]“' units. (M : metal, A2Q = alkali chalcogenide) .............................................................. 317 7-12. Optical Band Gaps and Melting Point Data for (I) - (V) ............................. 320 7-13. Infrared and Raman Data (cm'l) for (I), (III) and (IV) .............................. 322 xvi 1-1: 1-3: 1-4: 1-5: 2-1: 2-2: 2-3: 24: LIST OF FIGURES Packing diagram of ABiP2S7 (A = K), as viewed down the a-axis. The alkali cations have been removed for clarity .................................................... 5 The anionic chains of CsSbS6, The alkali cations have been removed for clarity. Large open balls; S, shaded balls; Sb .......................................... 7 A Single anionic layer of Cszsb4Sg as viewed down the c - axis. Large open balls; S, shaded balls; Sb. The alkali cations have been removed for clarity .......................................................................... 9 (A) A view of a single [HngS3]"‘ layer, with labeling. (B) Packing diagram of KHngS3 as viewed down the a -axis, with labeling .................... 10 (A) View of a single [AgZPZSe6]2“' chain, with labeling. (B) Unit cell of K2Ag2P28e6 as viewed down the a - axis. (C) A Portion of the [Ag2P28e6]2“‘ framework, highlighting the [P28e6]2“‘ binding modes ............................... 15 ORTEP representation of K2AngS4 as viewed down the a - axis. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; Sb, boundary ellipsoids; S, boundary and axis ellipsoids; K (50% probability ellipsoids) ................................................................................... 63 Polyhedral representation of K2AngS4 (I) as viewed down the c - axis. Filled polyhedra; SbS4, pattern shaded polyhedra; AgS4, alkali cations are omitted for clarity ...................................................... 64 Polyhedral representation of szAngS4 (H) as viewed down the a - axis. Filled polyhedra; SbS4, pattern shaded polyhedra; AgS4, large open circles Rb+ cations ........................................................... 65 (A) The [Ag28b283]4‘ building block with labeling.(B) A portion of the [A128i2]2' layer adapted from Reference #23. (C) Relationship between the [Ag28b283]4' building block in (H) and the double xvii six-rings in CaAlzsiz ..................................................................... 66 2-5: A single layer of szAngS4 (II) with labeling, highlighting the 15-membered rings ........................................................................ 67 2-6: ORTEP representation of CszAngS4 (IH) as viewed down the a - axis. Small octant Shaded ellipsoids; Ag, principal axis ellipsoids; Sb, boundary ellipsoids; S, boundary and axis ellipsoids; Cs (50% probability ellipsoids) ...................................................................... 68 2-7: ORTEP representation of a single chain of CszAngS4 (IH), with labeling (50% probability ellipsoids) ............................................... 69 2-8: ORTEP representation of C53Ag25b3Se3 (V) as viewed down the b - axis. Small octant Shaded ellipsoids; Ag, principal axis ellipsoids; Sb, boundary ellipsoids; S, boundary and axis ellipsoids; Cs (50% probability ellipsoids) ...................................................................... 70 2-9: A view perpendicular to a single chain of CS3Ag28b3Seg (V) with labeling, highlighting the different Ag-Se binding modes (50% probability ellipsoids) ................................................................................... 71 2-10: Raman spectra of (A) K2AngS4 (I), (B) szAngS4 (H), and (C) CszAngS4 (III) ...................................................................... 79 2-11: Raman Spectra of (A) CS3Ag2Sb3S3 (IV) and (B) CS3Ag28b3Seg(V) .............. 80 2-12: (A) DTA diagram for CS3AgZSb383 (first cycle). Heat is absorbed at 386 0C as the material melts but no corresponding exothermic peak is observed. (B) Second DTA cycle showing the crystallization of CS3AgZSb383 at 273 °C, followed by its subsequent melting and recrystallization at 384 0C and 285 0C, respectively ................................... 81 2- 13: (A) DTA diagram for CS3Ag28b38eg (first cycle). Heat is absorbed at 336 0C as the material melts but no corresponding exothermic peak is observed. (B) Second DTA cycle showing the crystallization of xviii 2-15: 3-1: 3-2: 3-3: 3-4: 3-5: 3-6: 3-7: CS3Ag2Sb38e3 at 274 °C and 282 0C, followed by its subsequent melting and thermal decomposition at 323 0C and 368 °C, respectively ...................... 82 Single crystal optical absorption Spectra of (A) KzAngS4 (I), (B) szAngS4 (H), and (C) CszAngS4 (H1). The sharp features at high absorbance are noise and due to the very low transmission of light .................. 84 (A) Single crystal optical absorption spectrum of CS3AgZSb3S3 (IV) (B) Solid—State optical absorption spectrum of CS3Ag2Sb3Se3 (V) ................. 85 ORTEP representation of a—RbAg28bS4 as viewed down the a- axis. Small octant Shaded ellipsoids; Ag, principal axis ellipsoids; Sb, boundary ellipsoids; S, boundary and axis ellipsoids; Rb. (50% probability ellipsoids) ............................................................. 108 ORTEP representation of or-RbAg2SbS4 as viewed down the [110] direction. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; boundary ellipsoids; S, boundary and axis ellipsoids; Rb (50% probability ellipsoids) ............................................................. 109 ORTEP representation of a-RbAg2SbS4 as viewed along the c - direction. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; boundary ellipsoids; S. The cations are removed for clarity (50% probability ellipsoids) ............................................................. 110 Polyhedral representation of a—RbAg2SbS4 as viewed down the [110] direction , pattern shaded polyhedra; SbS4, ball and stick polyhedra; AgS4. The cations are omitted for clarity ....................................................... 111 Polyhedral representation of the “AgZS4” framework in or-RbAgZSbS4 ........... 1 12 Ball and Stick view of the “Ag2S4” framework in OL-RbAgZSbS4. Open circles; S, and shaded circles; Ag ............................................... 113 ORTEP representation of B - RbAgzsbS4, as viewed down the a - axis. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; xix 3-8: 3-9: 3-11: 3-13: 3-14: 3-15: 3-16: 3-17: boundary ellipsoids; S, boundary and axis ellipsoids; Rb. (50% probability ellipsoids) .................................................................................. 114 ORTEP representation of the "Ag284" layer in B - RbAgZSbS4 Small octant Shaded ellipsoids; Ag, boundary ellipsoids; S (50% probability ellipsoids) .................................................................... 115 Polyhedral representation of B-RbAgZSbS4 (H) as viewed down the b - axis, highlighting the highly distorted AgS4 units Pattern shaded polyhedra; AgS4, ball and stick polyhedra; SbS4. The cations are omitted for clarity ......................................................................... 116 Single-crystal optical transmission spectra of or-RbAgZSbS4 (I) and B-RbAgzsbS4 (II) .................................................................. 126 Absorption edge of a-RbAgzsbS4 (I) as a function of energy: (A) (abs) dependence (direct gap) and (B) x/a—b; dependence (indirect gap) ................. 127 Absorption edge of B-RbAg2SbS4 (I) as a function of energy: (A) (abs) dependence (direct gap) and (B) W dependence (indirect gap) ................. 128 The Raman spectra of polycrystalline samples of (A) a—RbAg2SbS4 (I) and (B) B-RbAgzsbS4 (II) .............................................................. 129 (A) DTA diagram for oc-RbAgZSbS4 (first cycle) (B) Second DTA cycle of a-RbAgzsbS4 , showing an exothermic peak upon heating. (C) Sixth DTA cycle of or-RbAg2SbS4 ................................................ 130 (A) DTA diagram for or-RbAgzsbS4 (first cycle) (B) Second DTA cycle of a-RbAgZSbS4, showing two phases ........................................ 131 (A) Third DTA cycle of or-RbAgzsbS4. (B) Fourth DTA cycle of a-RbAgzsbS4 ........................................................................ 132 (A) DSC diagram for oc-RbAg28b84 (first cycle). (B) Second DSC cycle of a-RbAgZSbS4 ........................................... 133 (A) DTA diagram for oc-RbAgZSbS4 (first cycle). (B) Sixth 3-19: 4-1: 4-4: DTA cycle of a-RbAgZSbS4 .......................................................... 134 (A) DTA diagram for or-RbAgzsbS4 (first cycle). (B) Second DTA cycle of a-RbAgZSbS4. (C) Third DTA cycle of or-RbAgZSbS4 ............... 135 Packing diagram for CszAgzoSb4819 as viewed down the b - axis ................ 168 (A) The coordination of the tetrahedral [SbS4]3' units to Ag+ ions, with labeling. (B) The coordination of the tetrahedral [SbS3] 3' units to Ag+ ions, with labeling ........................................................................ 169 (A) A parallel view of the (Ag3SbS4) sheet in (H). (B) A perpendicular view of the (Ag3SbS4) sheet, highlighting the Ag3S3 rings connecting the neighboring tetrahedral [SbS4]3' units. (C) A parallel view of the (Ag3SbS4) sheet, highlighting the Ag+ ions used in connecting the ternary (Ag3SbS4) sheet to the binary AggS "belt", black circles; Sb, gray circles; Ag, open circles; S .................... 170 (A) A parallel view of the (Ag3SbS3) Sheet. (B) A perpendicular view of the (Ag3SbS3) sheet, highlighting the Ag3S3 rings connecting the neighboring pyramidal [SbS3]3' units. (C) A parallel view of the (Ag3SbS3) Sheet highlighting the Ag+ ions used in connecting the ternary (Ag3SbS3) sheet to the binary Ag2S "belt", black circles; Sb, gray circles; Ag, open circles; S .............................................................. 172 A sequential decomposition of the [Ag208b4319]2“" layers to [Ag103b283] "sub-layers", and the further separation into the two ternary (AgngS4) and (Ag38bS3) sheets and the binary Ag2S “belt” ............. 174 A view of a portion of the layers in B-RbgAg208b4S19 (HI), with label in g ..................................................................................... 176 (A) A parallel view of the (Ag3SbS4) Sheet in (IH). (B) A perpendicular view of the (Ag3SbS4) sheet, highlighting the Ag3S3 rings connecting the neighboring tetrahedral [SbS4]3' units. (C) A parallel view of the xxi 4-8: 4-9: 4-10: 4-11: 4-12: 4-13: 5-1: 5-2: 5-3: 5-4: 5.5: (Ag3SbS4) sheet highlighting the Ag“ ions used in connecting the ternary (Ag3SbS4) sheet to the binary AggS "belt", black circles; Sb, gray circles; Ag, open circles; S ............................................................... 177 (A) A parallel view of the (AgngS3) sheet in (IH) (B) A perpendicular view of the (Ag3SbS3) sheet, highlighting the Ag383 rings connecting the neighboring pyramidal [SbS3]3' units. (C) A perpendicular view of the (Ag3SbS4) Sheet highlighting the Ag+ ions used in connecting the ternary (Ag3SbS4) Sheet to the binary Ag2S "belt", gray circles; Sb, light gray circles; Ag, open circles S ............................................................... 179 (A) Solid-state optical absorption spectrum for szAg20Sb4S19(I). (B) Solid-state optical absorption spectrum for CszAg208b4819 (II) ............... 183 Solid-state optical absorption spectrum for B-szAngb4S19 (III) ............... 184 (A) Raman spectrum for or-szAgZOSb4819(I). (B) Raman spectrum for B-szAgZOSb4S 19 (III) .................................... 185 (A) DTA diagram for a-szAgZOSb4819(I). (B) DTA diagram for B-RbgAg208b4S19 (III) .................................................................. 186 (A) DTA diagram for B-szAgZOSb4819(IH). (B) DSC diagram for B-szAgzoSb4S 19 (III) .............................................................. 187 View of a single [AquS4]n2“' chain in szAqu84 (1), with labeling ............ 208 ORTEP representation of szAquS4 (I) as viewed down the c-axis (50% probability ellipsoids) ............................................................. 209 ORTEP representation of Rb2AquS4 (I) as viewed down the b-axis (50% probability ellipsoids) ............................................................. 210 ORTEP representation of szAu6Sb4Sm (IH) as viewed down the c-axis (50% probability ellipsoids) ..................................................... 211 ORTEP representation of the [Au3Sb4Sg]' undulating layer with labeling (50% probability ellipsoids) ............................................. 212 xxii 5-6: 5-8: 5-10: 5-11: 5-12: 5-13: 5-14: 6-3: 6—4: 6-5: 6-6: ORTEP representation of the [Sb4S7]2' chain with labeling (50% probability ellipsoids) ............................................................. 213 ORTEP representation of the [Au382]' layer highlighting the pyramidal sulfides in the undulating layer (50% probability ellipsoids) ........................ 214 ORTEP representation of a perpendicular view of the [Au382]' layer with labeling (50% probability ellipsoids) ............................................. 215 ORTEP representation of the Au---Au interactions of the Au column. The solid lines represent Au-uAu interactions under 3.25A and the dotted lines represent Au---Au interactions between 3.25 and 3.60A ............... 216 Solid-state optical absorption spectra for Rb2AquS4 (I) and CszAquS4 (II) ...................................................................... 222 Solid-state optical absorption spectrum for szAu6Sb4S10 (III) .................... 223 Raman spectra of szAquS4 (I) and Rb2Au6Sb4810 (III) ......................... 224 DTA diagram for CszAquS4 (II) ...................................................... 225 Two cycles for DTA diagrams for szAu6Sb4S10 (III) ........................... 226 ORTEP representation of K2CuP3S9 down the a-axis (50% probability ellipsoids). Small octant shaded ellipsoids; Cu, principal axis ellipsoids; P boundary ellipsoids; S, boundary and axis ellipsoids; K ..................................................................... 257 Perspective view of a [CuP339]2' chain with the labeling scheme .................. 258 Perspective views of the tri-tetrahedral [P3Sg]3' ring observed in K2CuP3S9 (a), the di-tetrahedral [P2S6]2' ring (b) and quadri-tetrahedral [P4Slz]4' (c) observed respectively in 2D- and 3D-NbP2S3 ........................ 259 ORTEP representation of the packing diagram of CSzCUszS6 as viewed along the [101] direction. ...................................................... 260 View of a single [Cu2P286]2"' chain with labeling ................................... 261 View of a single [CuP287]2"' chain with labeling .................................... 262 xxiii 6-7: 6-8: 6—10: 6-11: 6-12: 6-13: 6-14: 6—15: 7-1: 7-3: 7-4: 7-5: 7-6: ORTEP representation of K3CuP287 as viewed down the b-axis .................. 263 Single crystal optical absorption spectra of (A) KzCuP389 (I) and (B) RbZCuP389 (H). The Sharp features at high absorbance are noise and due to the very low transmission of light .............................. 267 (A) Single crystal optical absorption spectrum of K3CuPZS7 (IH). (B) Solid-state optical absorption spectrum of CszCu2P236 (IV) .................. 268 Raman spectra of (A) K2CuP389 (I), (B) C82CU2P286 (HI) and (C) K3CuPZS7 (IV) .................................................................. 269 Typical DTA diagrams for K2CuP389 (I) and CszCu2P286 (III) ................... 270 A portion of the electronic band structure in selected directions of reciprocal space. The energy band-gap, which is indirect, if found between the B and 1" points .............................................................. 271 Total density of states and Cu, P, and S projected density of states of K2CuP3S9 .............................................................................. 272 Cu-s, Cu-p and Cu-d projected density of states of KzCuP3S9 ..................... 273 Absorption edge of K2CuP389 plotted as a function of energy: (A) (abs)2 dependence (direct gap) and (B) #515 dependence (indirect gap) ......... 274 View of a Single [AuPS4]n2“' chain in K2AuPS4 (I), with labeling ................ 307 ORTEP representation of K2AuPS4 (I) as viewed down the b -axis (50% probability ellipsoids) ............................................................. 308 ORTEP representation of K2AuPS4 (I) as viewed down the c - axis (50% probability ellipsoids) ............................................................. 309 View of a single [AuPS4]n2“' chain in CszAuPS4 (III), with labeling ............ 310 ORTEP representation of CszAuPS4 (IH) as viewed down the c - axis (50% probability ellipsoids) ............................................................ 311 ORTEP representation of CszAquS4 (III) as Viewed down the b - axis (50% probability ellipsoids) ............................................................ 312 xxiv 7-7: 7—8: 7-9: 7-10: 7-11: 7-12: 7-13: 7-14: View of a Single [AuPZS7]n"' chain in K2AuPZS7 (IV), with labeling ............ 313 ORTEP representation of K2AuP2S7 (IV) as viewed down the a - axis (50% probability ellipsoids). small octant shaded ellipsoids: Au, principal axis ellipsoids: P, boundary ellipsoids: S, boundary and axis ellipsoids: K ...... 314 ORTEP representation of K2AuP2S7 (IV) as viewed along the [101] direction (50% probability ellipsoids). small octant shaded ellipsoids: Au, principal axis ellipsoids: P, boundary ellipsoids: S, boundary and axis ellipsoids: K ........... 315 Single crystal optical absorption spectra of (A) K2AuPS4 (I), (B) szAuPS4 (II) and (C) CszAuPS4 (III) ............................................... 324 Single crystal optical absorption spectra of (A) KAuP287 (IV), and (B) RbAuPZS7 (V) ........................................................................ 325 Raman spectra of (A) K2AuPS4, (B) CS2AuPS4, and (C) KAngS7 .............. 326 (A) DTA diagram for a—CszAuPS4 (first cycle).(B) Second DTA cycle of a-CszAuPS4, showing two phases ................................................. 327 (A) DTA diagram for KZAuPS4. (B) DTA diagram for KAuP2S7 ................ 328 CHAPTER 1 1. Introduction Solid-state chemistry is the cornerstone on which numerous aspects of our present and future technologies will be built. Several commercial applications now entrenched as "essential" to modern life owe their conception to solid state chemistry. Semiconducting chalcogenide compounds are viable candidates for a variety of applications such as solar cellsl, m radiation detectionz, electroluminescent displays3, thermoelectric4 and nonlinear optical5 devices, second generation high density storage batteries6, and catalysis.7 These technologies rely on the refinement of existing solid-state materials or the synthesis of novel materials with enhanced properties to advance current and future applications. Therefore, exploratory solid-state synthesis continues to be a highly active area of research. This field accurately deserves the term "exploratory" because the methods available to solid-state chemists lack the same level of predictability that synthetic chemists in the other disciplines sometimes take for granted. The lack of predictability is largely due to the high reaction temperatures (>800 oC) required by solid-state "heat & beat" syntheses. The majority of the starting materials are typically solids; very high temperatures are necessary to ensure sufficient diffusion between the reactants. In most cases, the high reaction temperatures alone are not enough to ensure a complete reaction. For a reaction to proceed to completion, the material must be cooled to room temperature, thoroughly ground to expose fresh surface area, and the subsequently reheated to allow the reaction to proceed. The elevated temperatures drive the products to the most thermodynamically stable products, usually binary or ternary compounds. As a result, quaternary compounds are difficult to form, the preference lying with the more stable binary and ternary compounds. Motivated by the desire to move way from the high temperatures of classical solid state synthesis, new low temperature methods such as chemical vapor deposition (CVD),8 solventotherrnal9 and the relatively new supercritical solvent10 synthesis were developed to address the problems of diffusion. Of these, the solventotherrnal method proved to be conducive to the low temperature synthesis of zeolites11 and a variety of other materialslz' 13. While direct combination and solventotherrnal methods are an excellent choice for the synthesis of new materials, they left an intermediate temperature range, between 200 - 800°C, largely unexplored. To explore this regime, the use of a low-melting alkali polychalcogenide A2Qx flux (A = alkali metal; Q = S, Se, Te) was developed as a new synthetic method.14 By utilizing this method, a staggering number of AIM/Q (A = alkali metal; M = transition, main group, or rare earth metal; Q = S, Se, Te) compounds have been prepared.15 Several of these compounds display interesting structural,16 thermal,17 and optical properties.13t19 The molten polychalcogenide synthetic method proved to be highly accommodating, allowing the coordination chemistry of transition”, main group19' 20, and rare earth metals21 to be investigated without the formation of undesirable competing phases. This approach can stabilize kinetically stable phases that cannot be formed by other methods. While investigating the ternary AIM/Q systems, certain molecular fragments were found to be stable in this molten reaction medium. This observation resulted in the utilization of discrete molecular building blocks or recognizable structural fragments to synthesize new multinary compounds. These molecular units are formed by the in situ fusion of A2Q/E/Q and contain various [Eny]“' units (E = main group element) in a molten polychalcogenide solvent. By this method, the polychalcogenide flux can act as a solvent by which the previously unexplored coordination chemistry of the highly basic [EXQyP' units can be investigated. To gain access to this area of chemistry, the polychalcogenide flux was modified by adding a second main group metal or binary chalcogenide to the starting AzQ/FJQ reactants, increasing the complexity of the reaction system, in the hopes of forming quaternary AwaEsz compounds. The lower reaction temperature increases the mobility of the reactants, favoring the probability of stabilizing metastable phases while preventing the formation of the thermodynamically stable phases. Although thermodynamic influences cannot be totally avoided, it is clear that by enhancing the diffusion rate of the reactants, solid-state reactions can be performed at reduced reaction temperatures, favoring the formation of new multinary materials. While this discussion has emphasizes the use of molten polychalcogenide fluxes as a conducive solvent to investigate the coordination chemistry of the various [Eny]“‘ units, they are far from “innocent” solvents. The alkali polychalcogenide fluxes allow for the easy incorporation of alkali metal cations into the anionic [MxEyQZP' frameworks. The incorporation of alkali metal can lower the dimensionality of a compound or cause a structural transformation to a previously unknown structure. The use of these highly reactive solvents at intermediate temperatures has allowed the dismantling of stable compounds (i.e. Sb2Q3, and P2Q5) into reactive intermediates for the formation of new materials. The multicomponent polychalcogenide flux method was first utilized by the incorporation of tin19t22 into the polychalcogenide fluxes, followed by tellurium23, the P2Q5 glasses (Q = 824, Se”), and most recently germanium26, all with remarkable results. The coordination chemistry of chalcometalates such as TeS32‘, P28e64‘, GeS44' and SnS44‘ have been largely ignored in favor of silicate, aluminate, and phosphate species that readily react in aqueous or other more traditional solvents. By utilizing the low melting alkali polychalcogenide salts as a reactive solvent the more non-classical chemistry of these various chalcometalates units can be explored. These non-classical chalcogenide ligands are more suitable for binding to "softer," chalcophilic metals, in contrast to the oxide- containing ligands, which prefer "hard," electropositive ions. The molten nature of the polychalcogenide flux allows rapid diffusion of reactants, promoting crystal growth. Since the major means of characterization of new solid-state materials is determining the single crystal X-ray structure, methods that routinely provide well crystallized products are necessary. \ . -~ The growing importance of the Chalcophosphate class of compounds prompted our lab to begin investigating this and other related systems via our developing polychalcometalate method. Although the preliminary investigation involved the coordination chemistry of both the polythiophosphate and polyselenophosphate systems, the emphasis quickly focused on further exploration of the selenophosphate system. The polyselenophosphate system blossomed into a rich coordination chemistry of the various [PxSey]“' units with a variety of metal ions leading to a number of new materials whose structures vary from dense three-dimensional frameworks to molecular species“. We were interested in the continued exploration of the polythiophosphate system to develop this chemistry to a level of maturity similar to that observed in the polyselenophoshate system. During the course of our investigations, we discovered that the chemistry of the two systems was completely divergent, and they should be looked upon as two separate areas. The first compounds reported from a Ax[Pszl flux synthesis were ABiP287 (A = K, Rb)”. The Structure consists of corrugated [BiP287]"' layers separated by A+ ions. The layers are constructed from Bi3+ ions and multiply bonding (P287]4' units forming irregular eight-membered rings of alternating Bi-S-P atoms. The PS4 tetrahedrons of the pyrothiophosphate unit coordinate in a bidendate chelation mode to Bi+3 ions and acts as a bridge to a second Bi“3 ion, forming the top side of the eight-membered ring. The Bi+3 ions are connected at the bottom of the ring by the PS4 tetrahedron of a neighboring [P287]4' unit that acts as a bridge to the two Bi+3 ions, see Figure 1-1. The rings are linked in two dimensions by P-S-Bi bridges, forming the layer. The layers stack with the ei t-membered rings in registry, forming channels that run along the a-axis. The Bi“3 ion is in a distorted monocapped trigonal prismatic coordination geometry. The distortion presumably arises from the stereochemically active 632 lone pair of the Bi+3 ion. Bi b <— Figure 1-1: Packing diagram of ABiP287 (A = K), as viewed down the a-axis. The alkali cations have been removed for clarity. Under these proven conditions, we wished to explore the coordination chemistry of polythiophosphate and polychalcoantimonate units with transition metals. Because of sulfur's high affinity to form copper and silver antimony chalcogenide minerals, we focused our preliminary investigations on the group 11 transition metals in hopes of obtaining new quaternary alkali polychalcoantimonate compounds. The exploration of the polychalcoantimonate fluxes looked very appealing because the vast majority of the known multinary antimony chalcogenide compounds are minerals,28 containing Sb+3 species in [beQy]“' units neutralized primarily by silver and/or lead cations. The [beQy]“' frameworks containing the Sb+3 species exhibit a rich structural diversity due to the stereochemical effect of the inert lone pair and the tendency for the Sb+3 species to adopt three- 29 four- 30 or five-fold coordination.31 Taking advantage of the thioantimonate units high affinity for coinage metals, we felt this was a convenient entry to the largely unexplored area of [MbeszP' frameworks stabilized by alkali metal cations. A variety of synthetic routes have been reported to synthesize ternary alkali antimony (poly)chalcogenide compounds, with high temperature direct combination32 and solventothermal synthesis33 being the most successful. The compounds CS25b4Sg and CsSbS634 represent the first examples of thioantimonate compounds synthesized using the polythioantimonate fluxes. The structure of CsSbS6 is one-dimensional with Sb282 rhombic units linked into a chain by bridging (S5)2‘ ligands, see Figure 1-2. Each Sb atom is bonded to four S atoms with distances that are typical for a four-coordinate SbS4 species with a stereochemically active lone pair. Figure 1-2: The anionic chains of CsSbS6. The alkali cations have been removed for clarity. Large open balls; S, Shaded balls; Sb. The geometry about the Sb atoms is trigonal bipyramidal with the lone pair in an equatorial position exerting the expected distortions from an ideal trigonal bipyramidal coordination. The structure of CSsz4Sg is two dimensional containing of the same Sb282 rhombi observed in CSSbS6, but now they are linked by trigonal pyramidal [SbS3]3* units that connect the rhombi into chains running along the [110] direction. These chains are then linked into layers via (S2)2' ligands that bridge [SbS3]3' units from neighboring chains. The Sb atoms in both of the [beSy]"' units have a mixture of long and Short contacts to neighboring S atoms, see Figure 1-3. The Sb(l) atom has a square pyramidal geometry of S atoms with the lone pair presumably occupying the opposite axial site. The Sb(2) atom coordination is best described as distorted trigonal bipyramidal with three normal distances and one longer distance. Within the anionic layers of Cs28b4Sg are 14- membered rings that stack in registry from layer to layer, forming tunnels that run along the c - axis. The first reported quaternary alkali antimony sulfide compound was KHngS3.35 The Structure consists of discrete pyramidal [SbS3]3' units comer sharing to four- coordinate Hg+2 cations is a sea-saw geometry, forming a two - dimensional compound separated by K+ cations, see Figure 1-4. The first examples of a quaternary alkali antimony chalcogenides incorporating the tetrahedral [SbS4]4‘ unit, prepared in supercritical ammonia, were only recently reported36. Figure 1-3: A single anionic layer of Cs2$b4S3 as viewed down the c - axis. Large open balls; S, shaded balls; Sb. The alkali cations have been removed for clarity. (A) (B) Figure 1-4: (A) A View of a single [HngS3]"' layer, with labeling. (B) Packing diagram of KHngS3 as viewed down the a - axis, with labeling. 11 2. Nature of the Chalcophosphate Fluxes A typical reaction involves the combination of A2Q/M/P2Q5/Q in various ratios. The reactants are elemental metal M, alkali chalcogenide A2Q (A = alkali metal cation), the binary P2Q5 (Q = 8, Se) and elemental chalcogen Q (Q = S, Se, Te). The flux formation is the result of the self-redox reaction of A2Q reacting with n amounts of Q (Q = S, Se, Te) forming A2Qn+1 ligands that act as a reactive solvent. Reactions between the metals and the molten sz' ligands are performed in situ. The most direct way to incorporate an additional main group metal into these systems is to add the elemental main group metal into the reaction mixture. For the Chalcophosphate systems the binary P2Q5 compound was chosen to provide a preoxidized phosphorous source into the reaction mixture. Since the two systems have divergent chemical prOperties the polyselenophosphate and polythiophosphate systems; the methods will be described in general and differences between the two systems will be described as needed. A typical reaction mixture consists of A2Q/M/P2Q5/Q in various stochoimetric ratios. The powdered reagents are loaded, in a dry box under inert atmosphere, into a glass vial and thoroughly mixed to ensure a homogeneous mixture. The resulting mixture is loaded into a Pyrex or quartz tube which serves as the reaction container. The choice of reaction container depends on the desired reaction temperature, Pyrex tubes are routinely used for intermediate temperature reactions (200°C - 550°C), while quartz tubes are used for high temperature reactions (2 550°C). The tubes are evacuated to a pressure of ~1x10' 3 torr and flame sealed. The sealed tubes are loaded into a computer controlled furnace and then subjected to a preprogrammed heating profile. Upon heating, the A2Q/P2Q5/Q fuse together, forming a Ax[Psz] flux in which the coordination chemistry of the various [Pny]“' units can be explored with a variety of metal ions. Although the redox chemistry that occurs is quite complex, and the exact composition of the resulting molten fluxes 12 remains unknown, one can gain insight by applying the basic principles learned from the polychalcogenide chemistry. In molten polychalcogenide systems, the (Qx)2' ligands exist in a variety of lengths by undergoing a complicated self-redox equilibrium as shown in Scheme 1. Scheme 1 _ _2_ Q — _ - 2- / Q 72 01 Q 2- Q I Q/ (L + J. *— ‘\ +| \ Q Q Q L J _ _ Scheme 1 shows an elementary example; another example is the initial flux formation reaction, where Q2' reacts with n equivalents of Q forming the starting Q(n+1)2' flux. The actual equilibrium is far more complicated than the examples presented here and is still not completely understood. The various (Qx)2' ligands serve a dual purpose, the terminal chalcogenide atoms have a formal oxidation state of negative one, while the internal chalcogenide atoms are all formally neutral, keeping with the Zintl concept. This produces a basic, yet oxidizing media in which the metal is solubilized by a redox reaction with the internal chalcogenide atoms of the various polychalcogenide ligands in the flux, forming Mn+ ions and breaking the (Qx)2‘ ligands into smaller chalcogenide fragments by reducing a Q-Q bond. The addition of P2Q5 renders the Lewis acid-base equlibria described above even more complex. The P2Q5 reacts with the molten polychalcogenide ligands to form various Ax[PyQZ] units. The previously solvated metal cations are then coordinated by the basic 13 [PszP' units or (Qx)“' ligands forming soluble intermediates; acting as potential nucleation Sites for the growth of Single crystals. These fluxes, being liquids, allow rapid diffusion of these intermediates, via a dissolution-reprecipitation process, promoting the growth of high-quality single crystals. This process of single crystal growth, commonly called a mineralizer effect, is particularly effective since the flux can redissolve small or poorly formed crystallites and redeposit the material onto larger well-formed crystallites or form new nucleation sites. Since the major means of characterization of new solid—state materials is determining the single crystal X-ray structure, methods that routinely provide well crystallized products are necessary. The composition of the A2Q/M/P2Q5/Q mixture is a very important variable that can be easily manipulated. Changes in the AzQ/M/P2Q5/Q flux composition can alter the basicity of the flux, dramatically changing the coordination chemistry of the various [Pny]“' units with metal ions. The guiding principles for changing the Lewis basicity of the polychalcophosphate flux are: (I) The elemental chalcogenide concentration is inversely proportional to the Lewis basicity of the flux. (H) The amount of A2Q is directly proportional to the basicity of the flux. (IH) After extensive experimental observations, the amount of P2Q5 was determined to be critical in controlling the Lewis basictiy of the resulting flux, stabilizing various [Pny]“‘ units. Increasing the P2Q5 concentration (increasing the Lewis acidity) to the point where the elemental sulfur dilutant was removed, favoring the formation of nuclearity [PxSy]“‘ units, containing the P+5 Species. (IV) The Size of the cation plays a significant role in the dimensionally of the resulting [MxPySzln‘ framework. An example of the cation effect is observed in the structures of AzAg2P28e6 (A = K, Cs)25°. When the [Ag2P28e6]2“' framework is stabilized by Cs+ cations, a chain structure of alternating ethane-like [P2Se6]4' units bound to Ag?“2 dimers are observed, see Figure 1—5A. However, substitution of the larger Cs+ l4 cations by the smaller K+ cations results in a dramatic structural change, stabilizing a three- dimensional [Ag2P28e6]2"‘ framework. This novel yet complicated three-dimensional tunnel framework consists of AgSe4 tetrahedrons linked to [P28e6]4' units forming K+ filled channels that run along the crystallographic a - axis, see Figure 1-5B. The [P2Se6]4" units are assembled into layers that are connected by the AgSe4 tetrahedrons into a dense three-dimensional network. The [P28e6]4‘ units bridge four Ag+ ions in two different binding modes, see Figure 1-5C. Surprisingly, no Ag+ - - -Ag+ contacts were observed, which could not have been predicted based on the structure of CS2Ag2PZSe6. .1. In 2. T - -|/ .. ‘C‘ {xi J. . - Put 0 a..." “SQ—0“ ‘ “’ifi‘ . fi'$,t. o .. C’JK \. (O I, ' --' .. .. .0 .—.,.':Luls O 3" . . ._-:I u \I . \. . .n. -:£.-\'.~'--h':.°' o a. "an, 0 ° 0 ° Figure 1-5: (A) View of a single [AgzPZSean' chain, with labeling. (B) Unit cell of K2Ag2P2Se6 as viewed down the a - axis. (C) A Portion of the [Ag2P2S66]2"' framework, highlighting the [P2Se6]2"' binding modes. l6 Throughout our exploration of the Chalcophosphate fluxes we have observed a Significant difference between the selenophosphate and thiophosphate systems. While the thiophosphate flux appears to favor the P5+ species, the selenophosphate flux displays a tendency to favor the reduced P+4 species. The stability of the reduced P“4 species is consistent with the lower oxidizing power of the Sexz' ligands compared to 83‘ ligands. The Lewis basicity of the Chalcophosphate flux controls the nature of the [PnyP‘ units observed. In the selenophosphate system, the basicity is controlled primarily by changing the Azse concentration, while in the thiophosphate system, the Lewis basicity is controlled by varying the P285 concentration. The tetrahedral [PQ4]3‘ (Q = S, Se) unit is observed in both systems under highly basic conditions (high A2Q or low P28 5). Although the [P28e6]4' unit is repeatedly observed under Lewis acidic conditions (low A2Q), other reduced phosphorous species have also been reported.37 Further increasing the Lewis acidity, as observed in the synthesis of K2Cu2P4Se10,33 condenses two [P28e6]4‘ units together via bridging selenides and forms a [P4Se10]4' unit containing a cyclohexane—like ring. In the thiophosphate system, increasing the Lewis acidity of the flux (increasing the P285 concentration), favors the formation of higher nuclearity [PxSy]n' units. To further explore these phosphorous-rich conditions the elemental sulfur was removed from the reaction. A molten thiophosphate flux still forms by the combination of A28 and P285, but the oxidative properties are different. With the lack of the elemental sulfur, a portion of the thiophosphate flux now must be sacrificial to oxidize the coinage metal. The exact influence of these phosphorous-rich conditions is not understood. Phosphorous—rich reactions produced A2CuP389 (A = K, Rb)39 and C32CuP3894O a one-dimensional compound with an unprecedented acentric helical packing arrangement of the anionic framework. Increasing both the P285 and C823 concentrations lead to the synthesis of C82Cu2P2S5,39 a thiophosphate compound containing a reduced P"'4 species in the ethane- 17 like [P286]4' unit. This unprecedented observation suggests that slight changes in the Lewis acidity, via Specific C528/P285 ratios, may stabilize the reduced P+4 species. 18 3. Nature of the polychalcoantimonate fluxes The methodology of the polychalcoantimonate fluxes is very similar to the polychalcophosphate fluxes. A typical reaction involves the combination of A2Q/M/Sb/Q in various ratios, just as in the polychalcophosphate system, the only difference is the choice of antimony source. In the thioantimonate system, elemental Sb was experimentally observed to give the best results, while in the selenoantimonate system, the binary Sb2Se3 compound with a preoxidized Sb source is used. The reactions temperatures employed by the thioantimonate reactions (280 - 500°) were not sufficient to dismantle the Sb28e3 framework. The elevated reaction temperatures utilized by the selenoantimonate system (2 500°C ) were required to dismantle the Sb2Se3 extended framework. Upon heating, the A2Q/Sb2Q3/Q fuse together, forming a Ax[Sbsz] flux in which the coordination chemistry of the various [beQy]“' units can be explored with a variety of metal ions. The addition of Sb or the binary Snge3 renders the Lewis acid-base equlibria described above even more complex. The Sb or the binary Sb28e3 reacts with the molten polychalcogenide ligands to form various Ax[Sbsz] units, which can coordinate to the solvated metal cations forming a variety of [MbeszP' frameworks, stabilized by alkali cations. The composition of the A2Q/M/Sb2Q3/Q mixture is a Si gnificant variable that can be easily manipulated. Changes in the A2Q/M/Sb2Q3/Q flux composition can alter the basicity of the flux resulting in dramatic changes in the coordination chemistry with metal ions. A comparison of this system to the polychalcophosphate fluxes provides insight into the structural and chemical complexity of the two systems. The polychalcoantimonate system follows the experimentally observed trend of the polychalcophosphate system;1° basic conditions (increase A2Q) favor the tetrahedral [EQ4]3' unit (E = P, Sb; Q = S, Se) l9 and acidic conditions (decrease A2Q) favors a reduced species. In the basic system, the [SbQ4]3' unit is observed as the discrete tetrahedral species, while the [PQ4]3’ unit (Q = S , Se), exhibits a complex condensation equilibrium. Examples include [P287]4' 24", [1333913- 39,40, [PzSeg]4‘ 24b, and [p233914325c For the polychalcophosphate system, Lewis acidic conditions (decreasing A2Q or increasing the chalcogenide concentration) favor the stabilization of the reduced lH4 species which is readily observed in the ethane - like [P2Q6]4' unit (Q = S, Se). The stability of the reduced P+4 species is consistent with the lower oxidizing power of the Sexz' ligands compared to sz' ligands. In the case of the Sexz‘ ligands, the lesser oxidative power of the Se atoms cannot oxidize the P atoms to the highly oxidized P5+ species, resulting in the stabilization of [Pny]“' units containing P4+ species due the excess of polychalcogenide ligands in the flux. Conceptually, stabilizing a [beSy]n' unit containing the Sb4+ is possible but it has yet to be reported. Under Lewis acidic conditions, the pyramidal Sb3+ species is the only variable oxidation state species observed. Although there is not enough experimental data to predict the exact nature of these [beSy]n‘ units, Lewis acidic conditions can stabilize a variety of species, such as the tetrahedral [SbQ4]3' , pyramidal [SbS3]3' , and various oligomeric [beSyjn‘ units. In the thioantimonate fluxes increasing the Sb concentration stabilizes oligomeric [beSy]n‘ units, containing the Sb+3 Species, increasing the repertoire of [beSy]n‘ units available for the formation of new extended frameworks. The advantages to the thiophosphate and chalcoantimonate flux methods over traditional high temperature reactions are: (I) The use of intermediate reaction temperatures (280 - 550°C) allows the stabilization of metastable phases and the isolation of new [PxSy]“' and [beQy]n" units. (II) The dual nature of the polychalcogenide fluxes, as an reactive solvent and a minrealizer, provides a mechanism for the formation of high quality crystalline materials, critical for the crystal structure determination and any potential applications. 20 (IH) The tunable Lewis basicity of the fluxes allows for the stabilization of a specific [PxSyP' and [beQy]n' unit, or a completely unexpected unit. (IV) The [PxSy]“' and [beQy]n' units readily bind with a variety of metals with a ever increasing repertoire of bonding modes. (V) The ease of isolation of highly pure materials due to the residual flux's high solubility in common organic solvents. The [PxSy]n' and [beQy]n' units were thought to be the critical factor in determining the structure. They are typically thought of as molecular or ologomeric building blocks that bond to coinage metal cations in the polychalcogenide flux, building extended frameworks stabilized by alkali metal cations. The growing diversity of the phases synthesized owe their conception to the great diversity of the [PxSy]n' and [beQy]n' units observed and the staggering number of different binding modes displayed by these units. A unique feature of the thioantimonate system is the stabilization of novel [MbeySzP' frameworks that are dominated by the coinage metal chalcogen framework, not by the thioantimonate unit. This is a completely unexpected result, suggesting that an entirely new area of this chemistry can be explored. Even with this result, the thiophosphate and chalcoantimonate systems is critical for designing new materials since the [PxSy]n' and [beQy]“' units can be controlled by manipulating the Lewis basicity of the flux. To give a better indication of the great variety of binding modes of the [beQy]n‘ units we have constructed Tables 1-1 and 1-2. Table 1-1 summarizes the synthetic conditions that have stabilizes the various [PxSy]n' and [beQy]n' units discovered throughout the course of this dissertation. Table 1-2 lists the [beQy]“‘ units observed and an example of the binding mode displayed by these units. The vast majority of results shown in this table were unknown before the start of this research. Even more importantly, these two systems Should be thought of as completely different systems, even when [PxSy]"' and [beQy]n' units are similar the resulting chemistry is rarely even closely related. 21 This dissertation focuses on the synthesis, characterization, and properties of several new thiophosphates and chalcoantimonates. The chemistry revolved around the coinage metals because of their highly chalcophilic nature and because the quaternary thiophosphate and chalcoantimonate chemistry was largely unexplored. The exploratory synthetic work performed here lays the ground work for the systematic synthesis and further exploration into this chemistry. 22 Table 1-1. Synthetic conditions for the different [Psz]°' and [Sbsz]"'units. (M = metal, A2Q = alkali chalcogenide). M / P28e5 / A286 / Se P11+ / ligands References 1/1-3/ 1-2/ 10 P“ / [P2Se6]4' 16 l / 1.5-2 / 3-4/ 10 P5+ / [PSe4]3', [PSe5]3', [PzSeg]4' 16, 24b, 250 1 /2-3 /2/ 10 P5+ / [P28e7]4', [P25e9]4' 16, 25c 1 / 1 / 2/ 10 P3+ / [P3Seg]4' 37 M / P285 / A28 / 5 PM / ligands References 1 / 1.5-3 / 24 / 4-12 P5+ / [PS4]3‘, [P287l4' 24b 1 / 3 / 2 / 4 P4+ / [P286143 24a 1 /2-3/ 2 / - P5+ [P3S9l3‘ 39,40 1 l3 /3 / - 134+ / [P286143 39 M / Sb / A2S / S Sb“ / ligands References 1-2/ 1 / 2 / 8 Sb5+ / [Sbs4]3- 36,43 1 / 1.5/ l / 8 Sb5+ / Sb3+ [SbS4]3', [Sb284]2' 36,43 34/ 1 l2 / 8-16 . Sb5+ / Sb3+ [SbS4l3', [Sb8313' 36,43 2/ 1 /2 / 8 Sb3+ / [Sb4S7]2' 49 M / Sb / A2Se / Se Sbr|+ / ligands References 4/1/1/16 Sb3+ / [Sb2Q6]4' 4g 2/ 1 /2 / 8 Sb3+ / [Sb28e5]4' 50 1 /3a Sb3+ / [SbSe2]' , {[SbSe2]}22' 51 aThe Sb and A2Sex source was the ternary compound Rb3SbSe3. 23 Table 1-2. Structure and coordination examples of the various [beQy]n- units. a o I [SbS4l3' in A2AquS4 (A = Rb, Cs)40 Q/Sb\"'” O —-M [SbS4l3' in ngAgZSbgsg36A3 L Q —M [SbSe4l3- in CS3AgZSb3Seg36A3 M/:\Sb.«|\\s \ [SbS4]3' in KzAngS436,43 M [SbS4]3' in KAgZSbS436 :>S\S b...n\S\M /°’ \, M l S . 3- ' 36,43 / VS/M [SbS4I m szAngS4 [Sb34]3’ in a-RbAgZSbS436A3 (Cont.) 24 Table 1-2. (Continued) M\ /M M/S\S b..«\\S /\M MS\S/ b‘S/M [SbS413-in B-RbAgZSbS445 M/ \ M “I" S\ S \ Sb "“ M [Sbs4l3- in a,B-Cs2Angs436.43 M M M\\ / M f M . \S,Sb..,,, / [SbS4l3' 111C3821"xg203b431946 / ‘ S\ S M / \ M M M .Sb a” M/8 I\3 S / \ . M i (Sbs313- 1n NazCquS344 l (Cont.) 25 Table 1-2. (Continued) M M \ Q/ /Sb.. Sb... 0 A "” ""0 [sz841} in CS3Agzsbgsg36A3 M/ \M [SbZSe4]2' in CS3AgQSb3Seg43 Q -—-—- M M"\/ I (I2 / \ Qlllh. Sb .«\\\Q’II. ,.¢\\\Q / V f \ Q (I) Q (I) Q [szS6]6‘ in K2La28b23948 [szSe6]6' in K2Gd23b286948 5 [Sb 3 12- in szAu6Sb4S 1049 / \ / 4 7 Sb Sb 1 \ S s ‘M S l s 51 M Sb Sb 86/ I .""Ise/‘ "q’Se Se Se . Jl/ \hlfl [Sb2365]4' 1n KTthzSe650 M l . Se A "m, [SbSez]‘ 1n RbCquZSe4-H2051 Se M/ \M (Cont.) Table 1-2. (Continued) [SbS3]3' in CS3Ag28b3S752 [Sb4S7]2' in CsAng4s752 [SbSe3]3' in CU3SbSe353 [SbSe3]3‘ in Ag3SbS354 [SbSe4]3' in CU3SbSe455 [SbS4]3' in CU3SbS456 [Ssz]“' in AngSz59 (Cont.) 27 Table 1-2. (Continued) . M Q?“ M-Q ' 4 M K M ‘ 3- - 60 M M [SbS4] 1n CU3SbS4 ‘ M Sb M r Q-M AC3 MW M E M [SbS]3' in Ag3SbS361 M/Q\ [SbS]3' in Ag53b3462 Sb _ . l 1 Q/ \QrM [SbS]3 1n CU3SbS363 Ml \ Id 9M [SbS]3' in Cu128b481364 M 10. 11. 12. 13. 28 List of References Zweibel, K.; Michell, R. CuInSez and CdTe: Scale-up for Manufacturing, Solar Energy Research Institute, Dec. 1989 and references therein. (b) Mickelsen, R.A.; Chen, W.S., In Ternary and Multinary Compounds, Eds. Deb, S.K.; Zunger, Z., Proceedings of the 7th Conference, Materials Research Society, 1987, 39. Smity, R.A. Semiconductors, Cambridge University Press, 1978, 438. (b) Barlett, B.E. Infared Phys., 1969, 9, 35. Ballman, A.A.; Byer, R.L.; Eimerl, D,; Feigelson, R.S.; Feldman, B.J.; Goldberg, L.S.; Menyuk, N.; Tang, C.L. Applied Optics, 1987, 26, 224. (a) Rowe, D. M.; Bhandari, C. M. Modern Thermoelectrics, Holt, Rinehart, and Winston, London, 1983, p. 103. (b) Borkowski, K.; Pyzyluski, J. J. Mater. Res. Bull., 1987, 22, 381. Ballman, A.A.; Byer, R.L.; Eimerl, D,; Feigelson, R.S.; Feldman, B.J.; Goldberg, L.S.; Menyuk, N.; Tang, C.L. Applied Optics, 1987, 26, 224. (a) The Sodium-Sulfur Battery, Sudworth, J. L.; Tilley, A. R., Eds., Chapman and Hall, New York, 1985. (b) Fischer, W. Mater Res. Soc Symp Proc., 1989, 135, 541. (0) Powers, R.W.; Karas, B.R. J. Electrochem. Soc., 1989, 136, 2787. (a) Chianelli, R. R.; Pecoraro, T. A.; Halbert, T. R.; Pan,W.-H.; Stiefel, E. 1.]. Catal., 1984, 86, 226. (b) Pecoraro, T. A.; Chianelli, R. R. J. Catal., 1981, 67, 430. (c) Harris, 8.; Chianelli, R. R. J. Catal., 1984,86, 400. Steigerwald in Inorganometallic Chemistry, T.P. Fehlner, Ed., Plenum Press, New York, 1992, p.333. Rabenau, A. Angew. Chem, Int. Ed. Engl., 1985, 24, 1026. Wood, P.T.; Pennington, W. T.; Kolis, J. W. J. Am. Chem. Soc., 1992, 114, 9233. Barrer, R.M. Hydrothermal Chemistry of Zeolites, Academic, New York, 1982. (a) Morris, R.E.; Wikinson, A. P.; Cheetham, A. K. Inorg. Chem, 1992, 31, 4774. (b) Harrison, W. T. A.; Stucky, G. D.; Morris, R.E.; Cheetham, A. K. Acta Crystallogr., 1992, C48, 1365. (c) Morris, R.E.; Harrison, W. T. A.; Stucky, G. D.; Cheetham, A. K. J. Solid State Chem, 1991, 94, 227. ((1) Harrison, W. T. A.; Morris, R.E.; Cheetham, A. K. Acta Crystallogr., 1992, C48, 1182. (e) Morris, R.E.; Hriljac, J. A.; Cheetham, A. K. Acta Crystallogr., 1990, C46, 2013, (1990). (f) Morris, R.E.; Cheetham, A. K. Chem. Mater., 1994,6, 67. (g) Harrison, W. T. A.; Stucky, G. D.; Cheetham, A. K. Eur. J. Solid State Inorg. Chem, 1993,30, 347. (a) Liao, J.-H.; Kanatzidis, M.G. Inorg. Chem, 1992, 31, 431. (b) Kim, K.- W.; Kanatzidis, M.G. J. Am. Chem. Soc., 1992, 114, 4878. (c) Chou, J.-H.; 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 29 Kanatzidis, M.G. Chem. Mater., 1995, 7, 5. (d) Chou, J.-H.; Kanatzidis, M.G. Inorg. Chem, 1994, 33, 1001. (e) Stephan, H.-O.; Kanatzidis, M.G. J. Am. Chem. Soc., 1996, 118, 12226. (D Chou, J.-H.; Hanko, J.A.; Kanatzidis, Inorg. Chem, 1997, 36, 4. (g) Bowes, C.L.; Ozin, G.A.; Adv. Mater., 1996, 8, 13. Kanatzidis, M. G. Chem. Mater., 1990, 2, 353. Sutorik, A.C.; Kanatzidis, M.G. Prog. Inorg. Chem, 1995, 43, 151. Kanatzidis, M.G. Curr. Opinion Solid State and Mater. Sci, 1997, 2, 139. (a) Dhingra, 8.; Kanatzidis, M.G. Science, 1992, 258, 1769. (b) Marking, G.A.; Hanko, J.A.; Kanatzidis, M.G. Chem. Mater., 1998, 10, 1191. Marking, G.A.; Liao, J .-H.; Kanatzidis, M.G. manuscript in preparation. (b) Kanatzidis, M.G.; Liao, J .-H.; Marking, G.A., United States Patent 5,618,471. (c) Kanatzidis, M.G.; Liao, J .-H.; Marking, G.A., United States Patent 5,614,128. Liao, J .-H., Ph.D. Thesis, Michigan State University, East Lansing, MI, 1993. Liao, J.-H.; Varotis, C.; Kanatzidis M.G. Inorg, Chem, 1993, 32, 2453. Sutorik, A.C.; Kanatzidis, M.G. Chem. Mater., 1997, 9, 387. Liao, J.-H.; Kanatzidis M.G. Chem. Mater., 1993, 5 , 1561. (a) Zhang, X.; Kanatzidis M.G. J. Am. Chem. Soc., 1994 116, 1890. (b) Zhang, X.; Kanatzidis M.G. Inorg, Chem, 1994 33, 1238. (a) McCarthy, T.J.; Kanatzidis M.G. Chem. Mater., 1993, 5 , 1061. (b) Chondroudis, K.; Hanko, J .A.; Kanatzidis, M.G. Inorg. Chem, 1997, 36, 2632. (c) McCarthy, T.J.; Kanatzidis M.G. J. Alloy Comp., 1996, 236 , 70. (a) McCarthy, T.J.; Kanatzidis M.G. J. Chem. Soc. Chem. Commun., 1994, 1089. (b) McCarthy, T.J.; Kanatzidis M.G. Inorg. Chem, 1995, 34, 1257. (c) Chondroudis, K.; Kanatzidis, M.G. Inorg. Chem, 1995, 34, 5401. (a) Marking, G. A.; Kanatzidis, M.G.Abstract from 210th Fall ACS Meeting, Chicago I]. (b) Marking, G.A.; Hanko, J.A.; Kanatzidis, M.G. Chem. Mater., 1998, 10, 1191. McCarthy, T.J.; Kanatzidis M.G. J. Chem. Soc. Chem. Commun., 1994, 1089. Takeuchi, Y.; Sadanaga, R. Z. Kistallogr., 1969, 33, 1205. Dittmar, V.G.; Schafer, H. Z. Anorg. Allg. Chem, 1977, 437, 183. Cordier, G.; Schafer, H. Z. Anorg. Allg. Chem, 1982, 437, 519. Cordier, G.; Schwidetzky, C.; Schafer, H. Rev. Chem. Min., 1982, I 9, 179. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47 48. 49. 30 (a) Olivier—Fourcade, B.; Phillippot, J .; M. Maurin Z. Anorg. Allg. Chem, 1978, 446, 159. (b) Graf, H.A.; Schafer, H. Z. Anorg. Allg. Chem, 1975, 414, 211. (c) Kanishcheva, A.S.; Kuznetsov, V.G.; Lazasrev, V.B.; Tarasova, T.G.; Zh. Strukt. Khim., 1977, 18, 1069. (d) Kanishcheva, A.S.; Mikhailov, Y.N.; Kuznetsov, V.G.; Batog, V.N. Dolk. Adad. Nauk SSSR, 1980, 251, 603. (e) Cordier, G.; Schafer, H. Rev. Chem. Min., 1981, 18, 218. (f) Dorrscheidt, W.; Schafer, H. Z. Naturforsch., 1981, 36b, 410. (a) Ko Y.H.; Tan K.M.; Parise J .B.; Darovsky A. Chem. Mater., 1996, 8 , 493. (b) Parise J.B.; Ko Y.H. Chem. Mater., 1992, 4 , 1446. (c) Parise J.B. J. Chem. Soc. Chem. Commun., 1990, 1553. (d) Sheldrick, W.S.; Hausler, H.-J. Z Anorg. Allg. Chem, 1988, 56], 149. (e) Dittmar, V.G.; Schafer, H. Z. Anorg. Allg. Chem, 1978, 441, 93. (f) Sheldrick, W.S.; Hausler, H.—J. Z. Anorg. Allg. Chem, 1988, 557, 105. (g) Eisenmann, B.; Schafer, H. Z. Narurforsch,. 1979, 34b, 383. (h) Graf, H.A.; Schafer, H. Z. Narurforsch., 1972, 27b, 735. McCarthy, T.J.; Kanatzidis M.G. Inorg. Chem, 1995, 34, 1257. Imafuku, M.; Nakai, I.; Nagashima, K. Mat. Res. Bull., 1986, 21, 493 (a) Wood, P.T.; Schimek, G.L.; and Kolis J .W. Chem. Mater., 1996, 8,. (b) Schimek, G. L.; Jerome, J. E.; Pennington, W. T.; Kolis, J. W. J. Solid State Chem, 1996, 123, 277. Chondroudis, K.; Kanatzidis, M.G. Angew. Chem, 1997, 36, 1324. Chondroudis, K.; Kanatzidis, M.G. Submitted for publication. Hanko, J.A.; Sayettat, J.; Jobic, S.; Brec, R.; Kanatzidis, M.G. Submitted for publication. Hanko, J .A.; Kanatzidis, M.G. Submitted for publication. Hanko J .A.; Kanatzidis,M.G. J. Alloy and Comp., Accepted for publication. Drake, G.W.; Kolis, J.W. Coord. Chem. Rev. , 1994, I3 7, 131. Hanko, J. A.; Kanatzidis, M. G. Abstract #0413 from 210th Fall ACS Meeting, Chicago Il 1995. Jerome, J.B.; Schimek, G.L.; Drake, G.W.; Kolis, J.W. Eur. J. Solid State Chem, 1996, 33, 765. Hanko, J .A.; Kanatzidis, M.G. To be published. Hanko, J .A.; Kanatzidis, M.G. Submitted for publication. Lee. J.; Chen, Z.; Wang, X.; Proserpio, D. M. J. Alloy and Comp., 1997, 262- 263, 28. Choi, K.S.; Hanko, J .A.; Kanatzidis, M.G. Manuscript in preparation. Hanko, J.A.; Kanatzidis, M.G. J. Chem Soc. Chem Comm, 1998,725. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 31 Choi, K.S.; Iordanidis, L. Chondroudis, K.; Kanatzidis, M.G. Inorg. Chem, 1997, 36, 3804. Hanko, J.A.; Kanatzidis, M.G. Angew. Chem. Int. Ed., 1998, 37, 342. Schimek, G.L.; Jerome, J .E.; Pennington, W.T.; Kolis, J .W. Submitted for publication. Pfitzner, A. Z. Anorg. Allg. Chem, 1995, 621, 685. Harker, D. J Chem. Phys, 1936, 4, 381. Pfitzner, A. Z. Kistallogr., 1995, 209, 1205. Garin, J.; Parthe, B.; Oswald, H.R. Acta Crystallogr., 1972, 828, 3672. Imamov, R.M.; Pinsker, Z.G.; Ivchenko, A.I. Krista, 1964, 9, 853. Hofmann, W. Z. Kistallogr., 1932, 84, 177. Knowles, C.R.; Acta Crystallogr., 1964, 317, 847. Alieva, A.G.; Pinsker, Z.G. Krista, 1961, 6 , 204. Kutoglu, A. Neues Jahrbuch Miner., 1968, 5, 145. Petrunina, A.A.; Maksimov, B.A.; Ilyukhin, V.V.; Belov, N.V. Dolk. Adad. Nauk SSSR, 1969, 188, 324. Machatschki, F. Z. Kistallogr., 1928, 68 , 204. Wuensch, B.J. Z. Kistallogr., 1964, 119, 437. CHAPTER 2 Chemistry of Silver in Molten Alkali Metal Polychalcoantimonate Fluxes. Synthesis and Characterization of the Quaternary Compounds AzAngS4 (A = K, Rb, CS), and CS3AgQSb3Q8 (Q = S, SC). 32 33 1. Introduction Over the past several years, we have demonstrated that molten salt (poly)chalcogenide syntheses at intermediate temperatures (ZOO-500°C) can produce a wide range of new multinary compounds.l During these investigations, it was discovered that certain molecular fragments are stable in this molten reaction medium. This observation led to an interesting twist to this chemistry, namely the utilization of molecular building blocks in the synthesis of new solid—state lattices. These building blocks are formed by the in situ fusion of A2Q/E/Q to form AxEsz species. (A = K, Rb, Cs; E = main group element, and Q = S, Se). By this method, the polychalcogenide flux can act as a solvent in which the coordination chemistry of the highly basic [Eny]n' ligands can be investigated. This approach was first explored in the tin systemz, followed by tellurium3, the P2Q5 glasses (Q = 8“, Se5), and most recently germanium", all with surprisingly good results. In this context, we explored the coordination chemistry of polychalcoantimonate ligands with transition metals, particularly group 11 metals . During the past two decades, a number of solid-state ternary alkali metal antimony (poly)chalcogenide compounds have been synthesized by either high temperature direct combination reactions7 or solventothermal synthesis.8 Recently, CSZSb4Sg and CsSbS69 were synthesized by the now proven molten polysulfide flux method. The vast majority of the quaternary antimony chalcogenide compounds are minerals,10 containing [beQy]“' framework neutralized primarily by silver and/or lead cations. In most of these mineral examples, the antimony exists as Sb+3 species in pyramidal coordination. The [beQyP' frameworks containing the Sb+3 species exhibit a rich structural diversity due to the stereochemical effect of the inert lone pair and the tendency for Sb to adopt threel 1- four”- or five-fold coordination78. The first reported 34 quaternary alkali antimony sulfide compound was KHngS3.I3 The structure consists of discrete pyramidal [SbS3]3' units linked together by distorted tetrahedral Hg2+ cations forming a two - dimensional compound separated by K+ cations. The first examples of a quaternary alkali antimony chalcogenides incorporating the tetrahedral [SbS4]4' unit, prepared in supercritical ammonia, were only recently reported”. Because of the high stability of copper and silver antimony chalcogenide minerals”, we investigated the group 11 transition metals in hopes of obtaining new quaternary alkali polychalcoantimonate compounds. Here we report the synthesis, structural characterization, optical, and thermal properties of the new solid-state alkali quaternary polychalcoantimonates, A2AngS4 (A = K, Rb, Cs) and CS3Ag28b3Q3 (Q = S, Se). Complementing the small number of isostructural compounds, synthesized in supercritical ammonia, recently reported by Kolis, et. al.14 The CS3Ag2Sb3Q3 (Q = 8, Se) represent a rare example of a mixed valence antimony compound with both Sb5+ and Sb3+ centers in the same structure. 2. Experimental Section 2.1 . Reagents The reagents mentioned in this study were used as obtained unless noted otherwise: (i) antimony powder 99.999% purity, -200 mesh, Cerac Inc., Milwaukee, WI; (ii) silver powder 99.95% purity, -325 mesh Alfa AESAR Group, Seabrook, NH; (iii) cesium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (iv) rubidium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (v) potassium metal, analytical reagent, Aldrich Chemical Co., Milwaukee, WI; (vi) sulfur powder, sublimed, J.T. Baker Chemical Co., Phillipsburg, NJ; (vii) N,N- 35 dimethylformamide (DMF) reagent grade, EM Science, Inc., Gibbstown, NJ; (viii) diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. 2.2 Syntheses. A28 (A = K, Rb, Cs) and Cs28e were prepared by reacting stoichiometric amounts of the elements in liquid ammonia as described elsewhere.16 All manipulations were carried out under a dry nitrogen atmosphere in a Vacuum Atmosphere Dri-Lab glovebox. Preparation of KzAngS4 (I). An amount of 0.164g (1.50 mmole) KZS, 0.054g (0.50 mole) Ag, 0.092g (0.75 mmole) Sb, and 0.128g (4 mmole) S were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame-sealed in vacuo (~10'3 Torr). The reaction mixture was heated to 400°C over 12 hrs in a computer- controlled furnace. It was isothermed at 400°C for 4 days, followed by cooling to 100°C at a rate of 4°C/hr and then to room temperature in 1 hour. The air and moisture sensitive product was isolated by dissolving the K2Sx and any Kx[SbySz] flux with DMF under inert atmosphere to obtain orange-yellow crystals in 46% yield based on Sb. Quantitative microprobe analysis on single crystals gave KLgAngS45 (average of three acquisitions). Preparation of szAngS4 (II). An amount of 0.102g (0.50 mmole) szs, 0.027g (0.25 mmole) Ag, 0.031 g (0.25 mmole) Sb, and 0.080g (2 mmole) S were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame- sealed in vacuo (~10'3 Torr). The reaction mixture was heated to 350°C over 12 hrs in a computer-controlled furnace. It was isothermed at 350°C for 4 days, followed by cooling to 110°C at a rate of 4°C/hr and then to room temperature in 1 hour. The product, which is air and water stable, was isolated by removing the excess flux as in (I) to obtain yellow 36 to yellow-orange crystals in 52% yield based on Sb. Quantitative microprobe analysis of single crystals of the two crystals gave Rb1,9Ag1,1SbS4,7 (average of three acquisitions). Preparation of CszAngS4 (111). An amount of 0.298g (1 mmole) CS2S, 0.054g (0.50 mole) Ag, 0.062g (0.50 mmole) Sb, and 0.128g (4 mmole) S were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame-sealed in vacuo (~10'3 Torr) and heated as in (H). The product, which is air and water stable, was isolated by removing the excess flux as in (I) to obtain yellow crystals in 63% yield based on Sb. Quantitative microprobe analysis on single crystals gave Cs],5AngS4_5 (average of three acquisitions). Preparation of CS3Ag2Sb383 (IV). An amount of 0.164g (0.55 mmole) C82S, 0.054g (0.50 mmole) Ag, 0.092g (0.75 mmole) Sb, and 0.128g (4 mmole) S were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame- sealed in vacuo (~10'3 Torr). The reaction mixture was heated to 280°C over 12 hrs in a computer-controlled furnace. It was isothermed at 280°C for 4 days, followed by cooling to 100°C at a rate of 2°C/hr and then to room temperature in 1 hour. The product, which is air and water stable, was isolated by removing the excess flux as in (I) to obtain red crystals in a 72% yield based on Sb. Quantitative microprobe analysis on single crystals gave a atomic ratio of CstAngLsSM (average of three acquisitions). Reactions at 280°C were contaminated with a small amount, ~5%, of Ag3SbS3. Increasing the reaction temperature to 350°C produced (IV) in a pure form. Preparation of CS3Ag2Sb3Se3 (V). An amount of 0.104g (0.30 mmole) C8236, 0.032g (0.30 mole) Ag, 0.055g (0.45 mmole) Sb, and 0.126g (1.60 mmole) Se were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame- sealed in vacuo (~10'3 Torr). The reaction mixture was heated as in (IV). The product, 37 which is air and water stable, was isolated by dissolving the excess flux as in (I) to obtain black crystals in a 75% yield based on Sb. Quantitative microprobe analysis on single crystals gave an atomic ratio of Csl.3Ag1,2Sb1,3Se4,45 (average of three acquisitions). 2.3 Physical Measurements Powder X-ray Diffraction . Analyses were performed using a calibrated Rigaku- Denki/RW400F2 (Rotaflex) rotating anode powder diffractometer controlled by an IBM computer, operating at 45 kW 100 mA and with a 1°/min scan rate, employing Ni-filtered Cu radiation. The calculated powder patterns for (I) - (III) and (V) were prepared with the CERIUS2 software.l7 Tables of calculated and observed XRD patterns are summarized in Tables 2-1 to 2-4, respectively. Infrared Spectroscopy . Infrared spectra in the far-IR region (600-50 cm-l), were recorded in 4 cm-1 resolution on a computer controlled Nicolet 750 Magna-IR Series H spectrophotometer equipped with a TGS/PE detector and a silicon beam splitter . The samples were ground with dry CsI into a fine powder and pressed into translucent pellets. Raman Spectroscopy . Raman spectra were recorded with a BIO-RAD FT Raman spectrometer with a Spectra-Physics Topaz T10-106c 1.064 nm YAG laser running at 11 amps. The samples were ground into a fine powder and loaded into melting point capillary tubes. Solid State U V/Vis/Near IR Spectroscopy. Optical diffuse reflectance measurements were performed at room temperature using a Shimadzu UV-3101PC double beam, double monochromator spectrophotometer. The instrument is equipped with integrating sphere and controlled by a personal computer. BaSO4 was used as a 100% reflectance standard for all materials. Samples were prepared by grinding them to a fine powder and spreading them on a compacted surface of the powdered standard 38 material, preloaded into a sample holder. The reflectance versus wavelength data output can be used to estimate the band gap of the material by converting reflectance to absorption data as described earlier.18 Single crystal optical transmission spectroscopy. Room temperature single crystal optical transmission spectra were obtained on a Hitachi U-6000 Microscopic FI‘ Spectrophotometer mounted on an Olympus BH2-UMA metallurgical microscope over a range of 900 to 380 nm. Crystals lying on a glass slide were positioned over the light source and the transmitted light was detected from above. Difierential Thermal Analysis (DTA). DTA experiments were performed on a computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~ 20 mg) of ground crystalline material was sealed in quartz ampoules under vacuum. A quartz ampoule of equal mass filled with A1203 was sealed and placed on the reference side of the detector. The sample was heated to the desired temperature at 10 °C/min, isothermed for 10 minutes, and finally cooled to 50 °C at the same rate. Residue of the DTA experiment was examined by X-ray powder diffraction. To evaluate congruent melting we compared the X—ray powder diffraction patterns before and after the DTA experiments. The stability/reproducibility of the samples were monitored by running at multiple cycles. Semiquantitative microprobe analyses. The analyses were performed using a JEOL JSM-6400V scanning electron microscope (SEM) equipped with a TN 5500 EDS detector. Data acquisition was performed with an accelerating voltage of 20kV and twenty second accumulation time. Single crystal X-ray Crystallography. Intensity data for (I), (III) and (V) were collected on a Rigaku AFC6 diffractometer, using 00/20 scans. Intensity data for (H) was collected on a Nicolet p3 four circle diffractometer using 0 scans. Data for (11), (HI) and (V) were collected at -100° C while the data for (I) was collected at room temperature. In all cases, graphite-monochomatized radiation was used. The crystals showed no 39 significant intensity decay, monitored by three standard reflections measured every 150 reflections throughout the data collection. The space groups were determined from systematic absences and intensity statistics. The structures were solved by direct methods using SHELXS-8619 and refined by full-matrix least-squares techniques of the TEXSAN package of crystallographic programs.20 The secondary extinction coefficient was refined for all structures. An empirical absorption correction based on w scans was applied to each data set, followed by a DIFABSZ' correction to the isotropically refined structures. All atoms were eventually refined anisotropically. All calculations were performed on a VAX station 3100/76 computer. After full refinement the final R/Rw for (I) were 3.0/4.5%. Since the structure is non-centrosymmetric refinement was attempted on the other enantiomorph which converged with R/Rw = 3.2/4.8%. Therefore, the first solution was retained. The complete data collection parameters and details of the structure solution and refinement for (I) - (II) and (V) are shown in Table 2-5. Tables 2-6 to 2-9 contain the coordinates of all atoms, average temperature factors, and their estimated standard deviations. 40 Table 2-1. Calculated and Observed X-ray Powder Patterns for K2AngS4 (I). hk] dcalc, A d obsd, A I/Imax(0bSd) 011 101 6.32 6.29 26 1 11 5.38 5.38 54 210 4.62 4.62 12 121 4.03 4.00 14 220 3.67 3.67 100 112 3.48 3.48 34 130 3.32 3.31 4 022 3.16 020 3.13 3.13 37 131 3.06 3.06 l 1 31 1 3.01 3.01 15 231 2.72 222 2.69 2.69 34 040 2.63 2.62 35 400 2.57 2.57 27 132 2.54 2.54 16 312 2.51 5.51 16 141 2.42 2.42 4 41 1 2.38 2.38 4 232 2.33 2.33 4 123 2.29 2.92 4 241 2.24 2.24 2 042 2.19 2.19 4 402 2.15 2.16 5 412 2.1 1 2.1 1 4 332 2.08 2.08 1 l 051 2.03 2.03 6 242 2.015 2.015 6 422 1 .994 1 .997 12 323 1.941 1.946 8 114 1.907 1.913 3 251 1.892 1.891 3 521 1.862 1.868 5 440 152 512 333 224 134 260 352 620 451 044 404 1.839 1.829 1.797 1.740 1.697 1.661 1.634 1.623 1.595 1.580 1.566 41 1.841 1.798 1.744 1.697 1.660 1.634 1.623 1.592 1.582 1.567 42 Table 2-2. Calculated and Observed X-ray Powder Patterns for szAngS4 (I). hkl dcalc, A d obsd, A I/Imax(0bSd) 011 7.47 7.48 6 110 6.52 6.54 8 101 6.32 6.34 19 1117 5.58 5.59 1 1 111 5.45 5.47 8 012 4.66 4.66 10 112 4.11 4.11 9 210 3.83 3.84 19 022 3.73 3.74 100 211— 3.62 3.63 18 211 3.55 3.57 24 123 3.43 3.43 6 122 3.37 3.37 8 203 3.25 3.26 10 103— 3.21 3.21 16 217 3.11 3.12 15 212 3.03 3.04 32 023 2.904 2.912 3 13? 2.797 2.799 4 230 040 2.705 2.712 39 213— 2.597 2.586 27 213 2.522 2.516 7 114— 2.421 2.424 6 232 2.762 2.386 6 233 2.338 2.338 2 145 2.308 2.313 4 2.278 2 323 2.226 2.219 3 322 2.179 2.174 2 233 2.148 2.154 7 331 2.118 2.119 9 143— 2.086 2.073 4 242 2.054 2.054 13 134 224 115 152 125— 234 422' 252 252 135— 062 334 2.021 2.000 1.954 1.933 1.890 1.868 1.848 1.811 1.803 1.785 1.761 1.702 1 .684 43 2.021 1.994 1.961 1.939 1.891 1.872 1.844 1.819 1.803 1.789 1.764 1.743 1.706 1.688 :onwooswoxosox WQNUJ Table 2—3. Calculated and Observed X-ray Powder Patterns for CszAngS4 11k] dcalc, A d obsd, A I/Imax(0bSd) 011 8.56 8.48 20 002 6.46 6.40 8 111 4.94 4.91 12 013 4.03 4.00 21 112 3.96 3.93 40 122' 3.87 3.85 100 122 3.39 3.38 11 200 3.34 3.35 15 032 202' 3.28 3.27 54 004 3.23 3.21 44 113 3.18 3.16 22 131 3.13 3.11 24 014 3.11 3.10 37 211 2.98 2.97 20 040 2.86 2.84 34 132 2.83 2.82 40 12? 2.78 2.78 7 204 2.62 2.62 20 115’ 2.54 2.53 15 230 2.518 2.508 15 133 2.501 2.491 14 134‘ 2.447 2.453 6 231 2.399 2.392 5 215 051 2.252 2.239 8 310 2.194 2.186 6 242‘ 135 2.152 2.146 18 204 2.109 2.104 11 214 2.075 2.068 6 206' 2.021 2.015 12 331— 1.962 1.955 14 3231— 1.936 1.929 20 250 061 1.890 1.894 6 322 054 25 l 126 162 332 063 136 055 244 413' 333 263— 345— 1.875 1.866 1.841 1.830 1.792 1.760 1.743 1.723 1.712 1.697 1.673 1.649 1.605 1.586 45 1.881 1.869 1.826 1.788 1.756 1.739 1.717 1.712 1.693 1.667 1.645 1.601 1.582 04> O\\IO\\)OOO\UI\JOO 46 Table 2-4. Calculated and Observed X-ray Powder Patterns for C83Ag28b3Se3 hkl dcalc, A d obsd, A I/Imax(0bSd) 001 12.90 12.90 25 100 10.55 10.59 50 101— 9.10 9.10 13 102 6.07 6.07 7 111 200 5.29 5.29 37 102 5.07 5.08 8 012 4.89 4.90 3 112 4.72 4.72 6 202’ 4.55 4.56 5 210 4.32 4.32 5 112 4.20 4.21 12 212 3.89 3.90 24 020 3.75 3.74 42 103 013 3.73 3.74 42 301— 3.58 3.59 43 120 3.53 3.53 100 212 113 3.34 3.34 7 10? 3.27 3.27 62 022 3.24 3.24 51 ZZT 3.05 3.06 51 311 2.97 2.98 38 221 222— 2.92 2.91 44 123— 2.83 2.83 18 401— 2.69 2.70 23 30? 2.65 2.65 54 302 2.566 2.577 10 205— 2.525 2.527 11 3174— 2.502 2.501 11 115 2.483 2.486 16 123' 2.466 2.461 1 1 214 2.404 2.408 4 223 2.367 2.374 14 235 2.191 2.200 5 033 025 116 215 125 422 502 514 512 217— 126 333 610 325 127 2.162 2.126 2.106 2.072 2.022 1.953 1.898 1.889 1.840 1.805 1.786 1.765 1.713 1.700 1.681 47 2.163 2.122 2.109 2.066 2.028 1.953 1.898 1.884 1.848 1.809 1.787 1.762 1.717 1.699 1.681 10 17 16 22 20 15 21 10 20 10 48 Table 2-5. Crystallographic Data for KzAngS4, szAngS4, CszAngS4, and CS3Ag25b3Seg I 11 111 v Formula K2AngS4 szAngS4 CszAngS4 CS3Ag2Sb3Se3 fw 436.05 528.79 623.67 805.70 a,A 10.287(4) 8.186(2) 6.8718(9) 10.791(4) b,A 10.537(3) 10.821(3) 1 1.442(2) 7.539(5) c,A 7.898(3) 10.331(2) 13.234(1) 13.177(3) 13, deg 90.000 9181(2) 102.500(9) 101.40(4) 2; v, A3 4; 856.1(4) 4; 914.7(7) 4; 1015.9(2) 4;1051(1) A, A 0.71069 0.71069 0.71069 0.71069 space group Pnn2 (#34) P21/n (#14) le/C (#14) P21/m (#11) deale. 8/Cm3 3.383 3.839 4.077 5.093 ”(m Ka)’ 72.66 161.82 123.22 244.43 cm'1 T, 0C 23 -100 -100 -100 29 max, deg 50.0 50.0 50.0 50.0 no. data collected 918 1830 2060 21 10 no. unique data 918 1719 1894 2002 no. data observed, 776 l 199 1398 1401 I > 36(1) no. of variables 75 73 74 89 final R/Rw.a% 3.0/4.5 3.8/3.9 4.7/6.0 3.0/3.9 GOF 2.30 3.46 2.32 1.41 a R = 2(11“ol'll'7clyleol; Rw = [ZWUFol—1Fc1)2/2W|F012]”2 49 Table 2-6. Fractional Atomic Coordinates and B(eq) Values for KzAngS4 (I) with Estimated Standard Deviations in Parentheses. atom x y z BequZ Agl 0 1/2 0.5211(4) 2.61(7) Ag2 1/2 1/2 0.7394(3) 1.94(7) Sbl 0.228490) 0.27467(7) 0.5302 0. 82(3) S1 0.2198(3) 0.0786(3) 0.3916(5) 1.2( 1) S2 0.0483(3) 0.2963(3) 0.71 19(5) 1.2(1) S3 0.2142(3) 0.4471(3) 0.3438(5) 1.2( 1) S4 0.4227(3) 0.2781(3) 0.6823(5) 1.4( 1) K1 0.2465(3) 0.2302(3) 0.0302(9) 1.7( 1) K2 0 1/2 0.0283(8) 1.3(1) K3 1.0000 0 0.7009(7) 1. 7(2) a B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as Beq = (4/3)[aZB(1, l) + bZB(2, 2) + czB(3, 3) + ab(cosy)B(1,2) + ac(cosB)B(1,3) + bc(cosa)B(2, 3) 50 Table 2-7. Fractional Atomic Coordinates and B(eq) Values for szAng84 (11) with Estimated Standard Deviations in Parentheses. atom x y z BeSLaAZ Agl 0.1275(1) -0.03136(7) 0.28040(7) 1.54(3) Sbl 0.14243(7) 0.22640(5) 0.46172(6) 051(3) S1 0.0164(3) 0.2279(2) 0.661 1(2) 1.0( 1) S2 0.2574(3) 0.4242(2) 0.4428(2) 1 . 1(1) S3 -0.0455(3) 0.1869(2) 0.2924(2) 0.9( 1) S4 0.3379(3) 0.0715(2) 0.4499(2) 0.9( 1) Rbl 0.1271(1) 0.5095 1 (8) 0.73726(9) 1.07(4) Rb2 0.1417(1) 0.22750(9) 0.00657(9) 1.23(4) a B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as Beq = (4/3)[a2B(1, 1) + bZB(2, 2) + c2B(3, 3) + ab(cosy)B(1,2) + ac(cosB)B(1,3) + bc(cosot)B(2, 3) 51 Table 2-8. Fractional Atomic Coordinates and B(eq) Values for CszAngS4 (HI) with Estimated Standard Deviations in Parentheses. atom x y 2 Bequ2 Agl 0.7921(2) 0.0252(1) 0.4285(1) 309(5) Sbl 0.3017(1) 0.21431(8) 0.43906(6) 1.27(4) S1 0.1159(5) 0.0553(3) 0.3587(3) 1.8(1) 82 0.6134(5) 0.2143(3) 0.3918(3) 1.9(1) S3 0.3489(5) 0.1831(3) 0.6177(3) 1.8(1) S4 0.1525(6) 0.3933(3) 0.3956(3) 2.6(1) Cs 1 0.3841(2) 0.00680(9) 0.16619(7) 2.52(4) CS2 -0.1306(1) 0.2431(1) 0.15744(8) 283(4) 3‘ B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as Beq = (4/3)[azB(1, 1) + b2B(2, 2) + c23(3, 3) + ab(cosy)B(1,2) + ac(cosB)B(l,3) + bc(cosa)B(2, 3) Table 2-9. 52 with Estimated Standard Deviations in Parentheses. Fractional Atomic Coordinates and B(eq) Values for C83Ag28b3Seg (V) atom x y z Bequ2 Agl 0.8692(1) -0.0359(2) 0.89192(8) 200(5) Sbl 0.8658(1) 1/4 1.2073(1) 097(5) Sb2 0.7659(1) 1/4 0.6241(1) 093(5) Sb3 0.1806(1) 1/4 0.8882(1) 1.02(5) Sel 0.6953(1) -0.0028(2) 1.1902(1) 1.24(5) S62 0.7862(2) 1/4 0.4415(2) 200(9) 563 0.3172(2) 1/4 1.0688(1) 1.27(5) S64 0.5449(2) 1/4 0.6472(2) 1.66(8) SeS 0.9136(2) 1/4 1.0205(1) 1.25(7) Se6 0.8850(1) -0.0184(2) 0.6984(1) 1.6 1 (5) C81 0.5841(1) 1/4 0.9316(1) 1.47(5) CS2 0.1393(1) 1/4 0.5599(1) 1.97(5) CS3 0.5554(1) -l/4 0.6512(1) 1.96(5) a B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as Beq = (4/3)[aZB(1, 1) + bZB(2, 2) + czB(3, 3) + ab(cos‘y)B(1,2) + ac(cos[3)B(1,3) + bc(cosoL)B(2, 3) 53 3. Results and Discussion 3.1 Description of Structures. Structure of K2AngS4 (I). The structure features tetrahedral Ag and Sb atoms linked by bent sulfur atoms forming the [AngS4]2"' framework, see Figure 2-1. This framework is constructed from [Ang288]5' units sharing corners to four Ag+ sites forming channels filled by K7“ cations. The [Ang283]5' unit is formed from two [SbS4]3' units edge sharing to a central Ag+ cation. A polyhedral representation of (I) down the c - axis is shown in Figure 2-2. The two crystallographic silver ions are both located on a 2-fold crystallographic sites. The two Ag sites have significantly elongated tetrahedral coordination, with a single Ag-S contact of 2.669(4)A for Ag( 1) while Ag(2) has distances in the range from 2.510(3)A and 2.692(4)A, average : 2.63(5)A. The S-Ag(1)~ S angles range from of 82.1(1)° to 115.6(2)°, average: 110(7)°. The S-Ag(2)-S angles range from 93.4( 1) to 159.3(1), average: 110(11)°. Inspection of the S(2)—Ag(1)—S(3') angle of 128.1(1)° and the S(2)-Ag(1)—S(3) angle of 88.5(1)° reveal a significant deviation from an ideal tetrahedral geometry. These obtuse and acute angles are due to a strained four-membered (Sb(1)-S(3)—Ag(l)—S(2)) rings brought about by the neighboring edge sharing tetrahedrons. The second silver site is as distorted as the first with S(4)— Ag(2)-S(4') and S(1)—Ag(2)—S(1') angles of 159.3(2)° and 127.0(2)° respectively. The [SbS4]3' ligand is a regular tetrahedron, with Sb—S distances in the range of 2.332(4) to 2.340(3)A, av: 2.331(4)A. The S—Sb—S angles range from 104.8(2)° to 112.2(2)°, average : 109(1)°. The three — dimensional [AngS4]n2“‘ framework is separated by K+ ions that are located in three different sites. In K2AquS4, K(1) is coordinated by seven s atoms [range of K(1)-S distances, 3.311(7)-4.368(5)A; av: 3.37(3)A], K(2) is 8- 54 coordinate [3.187(4)-3.373(6)A; av: 3.29(2)A], and K(3) is 6 - coordinate [3.163(3) - 3.431(5)A, av: 3.26(5)A]. Selected bond distances and angles are given in Table 2-10. Structure of szAngS4 (II). The structure of szAngS4 is a layered compound consisting of tetrahedral Ag+ cations linked to [SbS4]4' units. The layers propagate along the [b101] direction and are separated by two Rb+ cations. A polyhedral representation of (II), viewed parallel to the layers, is shown in Figure 2-3. The layer can be thought of as a two-dimensional array of interconnected [Ag2Sb283]4' building blocks. The [Ag28b283]4‘ unit is isostructural to the [Ag28b283]4' unit observed in 0t - CszAngS414a and related to the [Cd4S6]4' cluster observed in K2Cd28322. The [Ag2Sb2S3]4‘ consist of two eclipsed AngSz rhombi connected by bridging sulfides, see Figure 2-4a. The final sulfide from the [SbS4]3‘ unit then comer shares to the silver atom of an adjacent [Ag28b2S3]4* unit coordinatively saturating the Ag+ and forming the layer. The Ag — S distances range from 2.511(5)A to 2.759(5)A with an average Ag—S contact of 2.63(5)A. The coordination environment for the Ag site is a slightly distorted tetrahedron. The S-Ag-S angles range from 86.2(1)°, for the S—Ag—S angle associated with the AngS2 rhombus, to 119.6(2)°, average: 109(5)°. The [SbS4]3' is a regular tetrahedron, with Sb-S distances in the range from 2.319(5) A to 2.340(5) A, average : 2.331(4)A. The S-Sb—S angles range from 104.8(2)0 to 112.2(2)0 , average: 109(l)°. The structure of the [AngS4]2“' anion is closely related to the Zintl phase CaA128i223. The CaA12Si2 compound is a layered compound constructed of [A128i2]2' rhombi linked in two dimensions with the gallery space filled by Ca2+ ions. The structure of CaA12Si2 is unique in that the silicon atoms have an unusual "umbrella" coordination, with three Si-Al bonds in a pyramidal coordination and a fourth axial bond perpendicular to the plane of the base. A portion of the [A128i2]2' layer, highlighting the "umbrella" coordination of the silicon, is shown in Figure 2-4b. Although rare, chalcogenide examples of this coordination are known, and include Cu3VS424 and Na6Cd7S1022. The 55 [AngS4]2°‘ polyanion can be thought of as a defect of the A128122- framework. In the [AngS4]2' anionic framework the two Al"‘3 sites are replaced by a monovalent and a pentavalent metal ion, (M4) and a (M45) and the two Si4' anions are replaced by four sulfides providing the necessary negative charge. With the removal of two aluminum atoms from the opposite corners of the Al6Si6 double six ring the [Ag28b283]4' building block is formed, see Figure 2-4c. The strain of the "umbrella" coordination of the silicon atoms is removed by the low coordination preference of the sulfides. The layer can alternatively be described as being composed of lS—membered rings of alternating metal and sulfur atoms. These rings are connected by bridging sulfides to form the layer. A single layer, containing the 15—membered ring, is shown in Figure 2-5. The layers stack with their 15—menbered rings in registry so that they form channels running down the c—axis. The dimensions of the ring are 13.63A from S(2) to S(2') by 4.69.31 from S(l) to 8(3). The [AngS4]n2°“ layers are separated by Rb+ ions that are located in two different sites. In RbgAngS4, Rb(1) is coordinated by seven S atoms [range of Rb(1)-S distances, 3.270(5)-3.670(5)A; av: 3.40(5)A], and Rb(2) is 8-coordinate [3.346(5)- 3.762(5)A; av: 3.46(6)A]. Selected bond distances and angles are given in Table 2-11. Structure of CszAngS4 (III). The structure found for (H1) is isostructural to CL - CszAngS4 reported earlier.143 The lAngS4] “2'“ macroanion is a chain running along the a-axis, see Figure 2-6. The chains are comprised of the same [Ag28b283]4' building blocks observed in (II) connected through the formation of a Ag282 rhombi, see Figure 2- 7. The Ag282 rhombus is characterized by a long Ag-S contact of 2.899(4)A and a normal Ag-S bond of 2.610(4)A. The AgS4 tetrahedron is fairly distorted, with Ag-S bonds in the range from 2.483(4)A to 2.899(4)A and S-Ag-S angles in the range from 127.6(1)° to 85.8(1)°. The [SbS4]3' unit has a fairly regular tetrahedral geometry with Sb- S bond distances in the range from 2.312(4)A to 2.348(4)A, average: 2.31(1)A and an 56 average S-Sb-S bond angle of 109.5(9)°. The [AngS4]n2“' chains are separated by Cs+ cations that are located on two different sites. In CszAngS4, Cs(l) is coordinated by seven S atoms [range Cs(l)-S distances, 3.497(4)A - 3.872(4)A; av: 3.65(5)A], and Cs(2) is coordinated by seven S atoms [range Cs(2)-S distances, 3.550(4)A - 3.744(4)A; av: 3.64(4)A]. Selected bond distance and angles are given in Table 2-12. Structure of Cs3AgZSb3Q3 (Q = S (IV), Se (V)). Since these two compounds are isostructural the single-crystal structure determination was performed only on the selenide compound (V), therefore the discussion will refer mainly to this compound. The structure of the [AgZSb3Se8]n3“' macroanion is a complicated one - dimensional corrugated ribbon running along the crystallographic b-axis, separated by Cs+ cations (Figure 2-8). The [AgZSb3Se8]n3°' anion is more accurately described as [Ag2 (SbSe4)(SbZSe4)]n3“', providing the first selenide example of the rare class of mixed valent quaternary antimony chalcogenides. The [SbZSe4]2“' unit is an infinite chain of pyramidal Sb3+ units sharing opposite comers, leaving the third selenide terminal. This is typical for Sb3+ species with the lone pair stereochemically expressed. The third antimony is in a tetrahedral coordination environment common for Sb5+. The [SbSe4]3' unit is a regular tetrahedron with Sb—Se distances in the range from 2.460(2)A to 2.491(2)A, average: 2.476(8)A. The Se-Sb-Se angles range from 104.22(6)° to 113.3(1)0 : average 109(1)°. The terminal selenium atoms display the shorter bond distances. The [SbZSe4]2“‘ infinite chain contains two crystallographic Sb sites with Sb-Se distances in the range of 2.613(2)A to 2.628(2)A and 2.540(3)A to 2.623(2)A for Sb(l) and Sb(2), respectively. Selected bond distances and angles for (V) are given in Table 2-13. The two terminal selenides of the [SbZSe4]2"' oligomer bind asymmetrically to silver ions in an alternating bridging and four-coordinate binding modes. The two - coordinate selenide (Se(3)) has one symmetry equivalent Sb-Se distance of 2.708(2)A. The four- coordinate selenide, Se(5), has two crystallographically distinct Ag-Se distances of 57 2.725(2)A and 2.891(2)A, respectively. Figure 2-9 shows a view perpendicular to a single chain, highlighting the different Ag-Se binding modes. The structure of (V) can also be viewed as a AgZSe ribbon surrounded by [beSey]°' units. The ribbon is built from corner-sharing AgZSe/z rhombi. The corrugated ribbon structure is a result of the alternating nature of the [Sb2Q4]2"‘ oligomer, binding to silver in an alternating four- and two-coordinate motif. The [AgZSb3Seg]n3"' chains are separated by Cs”r cations that are located in three different sites. In Cs3AgQSb3Se8 Cs(l) is coordinated by eight sulfur atoms [range of Cs(1)-Se distances, 3.637(2) - 3.917(2)A; av: 3.66(4)A] , Cs(2) is 7 - coordinate [3.821(3) - 4.1 15(2)/31; av: 3.67(5)A], and Cs(3) is 7 - coordinate [3.647(3) - 4.189(2); av: 3.74(3)A]. 58 Table 2-10. Selected Distances (A) and Angles (deg) for K2AngS4 (I) with Standard Deviations in Parenthesesa. Sb(1)—S(1) 2.340(3) S(1)-Sb(l)-S(2) 104.8(2) Sb(1)—S(2) 2.355(4) S(l)—Sb(1)-S(3) 111.1(2) Sb(l)-S(3) 2.343(3) S(1)-Sb(l)-S(4) 1 12.2(2) Sb(1)—S(4) 2.332(4) S(2)-Sb(1)-S(3) 111.1(2) Sb—S(mean) 2.331(4) S(2)~Sb(1)-S(4) 1 12.1(2) S(3)-Sb(l)-S(4) 105.6(2) Ag(1)—S(2) 2.669(4) x 2 Ag(l)—S(3) 2.669(4) x 2 Ag(1)—S(mean) 2.669(4) S(l)-Ag(l)-S(2) 104.2(2) S(l)-Ag(1)-S(3) 117.9(1) Ag(2)—8(1) 2.692(4) x 2 S(l)-Ag(l)—S(4) 119.6(2) Ag(l)—S(4) 2.510(3) x 2 S(2)-Ag(1)—S(3) 114.1(2) Ag—S(mean) 2.63(5) S(3)-Ag(1)-S(4) 86.2(1) K(1)—S(l) 3.282(7) K(l)—S(2) 3.311(7) Sb(l)-S(1)—Ag(l) 115.6(2) K(l)—S(2') 3.432(5) Sb( 1 )-S(2)—Ag( 1) 1 14.0(2) K(1)—S(3) 3.387(7) Sb(1)-S(3)-Ag(l) 82.1(1) K(1)—S(3') 3.351(5) Sb(l)-S(4)-Ag(l) 84.5(1) K(1)—S(4) 3.542(5) K(1)-S(4') 3.330(7) K(l)—S(mean) 337(3) K(2)-S(1) 3.187(4) x 2 K(3)—8(1) 3.431(5) x 2 K(2)—S(2) 3.331(6) x 2 K(2)—S(2) 3.163(3) x 2 K(2)—S(3) 3.373(6) x 2 K(2)4(3) 3.199(4) x 2 K(2)—S(4) 3.271(4) x 2 K(3)—S(mean) 326(5) K(2)—S(mean) 329(2) aThe estimated standard deviations in the mean bond lengths and the mean bond angle are calculated by the equations 61 = {Endn - [)2/n(n - 1)}1/2, where ln is the length (or angle) of the nth bond, l the mean length (or angle), and n the number of bonds (or angles). Table 2-1 1. 59 Standard Deviations in Parenthesesa. Selected Distances (A) and Angles (deg) for szAngS4 (II) with Sb(1)-S(1) 2.334(4) S(l)-Sb(1)-S(2) 104.8(2) Sb( 1 )-S(2) 2.340(5) S(1)-Sb(1)-S(3) 1 1 1.1(2) Sb(1)-S(3) 2.331(5) S( 1 )-Sb(1)-S(4) l 12.2(2) Sb(1)-S(4) 2.319(5) S(2)-Sb( l)-S(3) 111.1(2) Sb-S(mean) 2.33 1 (4) S(2)-Sb( l )-S(4) 1 12.1(2) S(3)-Sb(1)-S(4) 105.6(2) Ag(1)-S( 1) 2.51 1(5) S-Sb—S(mean) 109(1) Ag( 1 )-S(2) 2.567(5) Ag( 1 )-S(3) 2.759(5) S(1)-Ag(1)-S(2) 104.2(2) Ag(1)-S(4) 2.663(5) S(1)-Ag(l)-S(3) 1 17.9(1) Ag-S(mean) 263(5) S(1)-Ag(l)-S(4) 1 19.6(2) S(2)-Ag(1)-S(3) 1 14.1(2) Rb(1)-S(l) 3.270(5) S(3)-Ag(1)-S(4) 86.2( 1) Rb( 1)-S(2) 3.382(5) S-Ag-S(mean) 109(5) Rb(1)-S(2') 3.670(5) Rb(1)-S(3) 3.454(5) Sb(l)-S( 1 )-Ag( 1) 1 15.6(2) Rb( l )-S(3') 3.367(5) Sb( 1 )-S(2)-Ag( 1) 1 14.0(2) Rb(1)-S(4) 3.393(5) Sb(l)-S(3)-Ag(1) 82.1(1) Rb(1)-S(4') 3.308(5) Sb( 1 )-S(4)-Ag( l) 84.5(1) Rb-S(mean) 340(4) Rb(2)-S( 1) 3.687(5) Rb(2)-S(1') 3.439(5) Rb(2)-S(2) 3.596(5) Rb(2)-S(2') 3.422(5) Rb(2)-S(3) 3.402(5) Rb(2)-S(3') 3.556(5) Rb(2)-S(4) 3.762(5) Rb(2)-S(4') 3.346(5) Rb-S(mean) 346(6) aThe estimated standard deviations in the mean bond lengths and the mean bond angle are calculated by the equations 01 = {}:.,,(ln — l)2/n(n - 0}”, where In is the length (or angle) of the nth bond, 1 the mean length (or angle), and n the number of bonds (or angles). Table 2—12. 60 Standard Deviations in Parenthesesa. Selected Distances (A) and Angles (deg) for CszAngS4 (III) with Sb(1)-S(1) 2.348(4) S(l)—Sb(1)-S(2) 108.7(1) Sb(1)-S(2) 2.345(4) S(1)-Sb( 1)-S(3) 106.2(1) Sb(1)-S(3) 2.344(4) S( 1 )-Sb( 1 )-S(4) l 13.6(1) Sb( 1 )-S(4) 2.312(4) S(2)-Sb(1)-S(3) 109.8(1) S(2)-Sb( l )-S(4) 108.6(1) Ag( 1 )-S( 1) 2.610(4) S(3)-Sb(1 )-S(4) 109.9(1) Ag( l)-S(2) 2.483(4) S-Sb-S(mean) 109.5(9) Ag(1)-S(3) 2.595(4) Ag( 1 )-S(3') 2.899(4) S(l)-Ag(1)-S(2) 103.8(1) S(1)-Ag(l)-S(3) 1 10.0(1) Ag(l)-Ag(1') 3.121(3) S(2)—Ag( 1 )-S(3) 127.6(1) Cs(1)-S(1) 3.497(4) Sb( 1)-S(l)-Ag(l) 1 11.7(1) Cs( 1 )-S(2) 3.872(4) Sb(1)-S(2)-Ag(1) l 12.8(1) Cs(1)-S(2') 3.435(4) Sb(1)-S(3)-Ag(1) 86.6(1) Cs(1)-S(3) 3.604(4) Ag(1)-Ag(1')-S(1) 6000(9) Cs( 1 )-S(3') 3.741(4) Ag(l)-Ag(1')-S(2) 128.9(1) Cs(1)-S(4) 3.768(4) Ag(1)-Ag( l')-S(3) 102.7(1) Cs(1)-S(4') 3.691(4) Cs-S(mean) 365(5) Cs(2)-S( 1) 3.581(4) Cs(2)-S( 1‘) 3.554(4) Cs(2)-8(2) 3.607(4) Cs(2)-S(2') 3.896(4) Cs(2)-8(3) 3.550(4) Cs2—S(3') 3.550(4) Cs(2)-S (4) 3.744(4) Cs-S(mean) 3.64(4) aThe estimated standard deviations in the mean bond lengths and the mean bond angle are calculated by the equations cl = {2n(ln - [)2/n(n - 1)}1/2, where In is the length (or angle) of the nth bond, l the mean length (or angle), and n the number of bonds (or angles). 61 Table 2-13. Selected Distances (A) and Angles (deg) for CS3Ag28b38eg (V) with Standard Deviations in Parenthesesa. Sb(1)—Se(l) 2.628(2) x 2 Se(l)—Sb(1)-Se(1') 9298(9) Sb(1)—Se(5) 2.613(2) Se(1)—Sb(1)—Se(5) 100.62(6) Sb(1)—Se(mean) 2.626(1) Se( 1 ')—Sb(1)—Se(5) 100.62(6) Sb(3)—Se(1) 2.623(2) x 2 Se(1)—Sb(3)—Se(l') 9055(8) Sb(3)—Se(3) 2.540(3) Se( 1)—Sb(3)—Se(3) 9699(6) Sb(3)-Se(mean) 259(2) Se( 1 ')-Sb(3)—Se(3) 9699(6) Sb(2)-Se(2) 2.460(2) Se(2)-Sb(2)—Se(4) 1 13.3( 1) Sb(2)—86(4) 2.463(2) Se(2)—Sb(2)—Se(6) 104.22(6) Sb(2)—Se(6) 2.491(2) x 2 Se(2)—Sb(2)—Se(6') 104.22(6) Sb(2)-Se(mean) 2.476(8) Se(4)—Sb(2)—Se(6) 1 1285(5) Se(4)-Sb(2)-Se(6’) 1 1285(5) Ag(1)—Se(3) 2.708(2) Se(6)—Sb(2)—Se(6') 108.6(1) Ag(1)-Se(5) 2.725(2) Ag(1)-Se(5') 2.891(2) Se(3)—Ag( l )—Se(5) 1 1349(7) Ag( 1 )-Se(6) 2.593(2) Se(3)—Ag(1)—Se(5) 99.5 1 (7) Ag( 1 )-Se(mean) 272(6) Se(3)—Ag(1)—Se(6) 1 1445(7) Se(6)—Ag( l )-Se(6) 9947(6) Ag(1)—Ag(1') 3.228(3) Se(5)—Ag(1)—Se(6) l2226(7) Ag( 1 )—Ag( 1 ") 3.632(2) Se(5)—Ag( l )—Se(6) 10244(7) Cs(lHeU) 3.881(2) x 2 Sb(3)—Se(3)—Ag(l) 8214(7) Cs(1)—Se(1') 3.637(2) x 2 Sb(3)—Se(3)—Ag(l) 8214(7) Cs(1)—Se(3) 3.691(3) Sb(3)—Se(3)—Ag( 1) 7318(8) Cs(l)—Se(3') 3.917(2) x 2 Sb( l)—Se(6)—Ag(1) 126.3(2) Cs(1)—Se(4) 3.688(3) Sb( 1 )—Se(6)—Ag( 1') 126.3(2) Cs(1)—Se(mean) 3.78(4) Sb(1)—Se(6)—Ag( 1) 89.3(2) Sb(l)—Se(6)—Ag(1) 89.3(2) Cs(2)—Se(1) 3.896(2) x 2 Sb(3)-Se(6)—Ag(1) 109.8(3) Cs(1)—Se(2) 3.821(3) Cs(2)-Se(2') 3.855(2) x 2 Ag(1)-Se(3)—Ag(l) 75.2(2) Cs(2)—Se(6) 4.115(2) x 2 Ag(l)—Se(6)—Ag( 1) 99.0(3) 62 Cs(2)—Se(mean) 3.93(5) Ag( 1 )—Se(6)—Ag( 1) l35.95(7) Ag( 1 )—Se(6)—Ag( 1) 77.6( 1) Cs(3)—Se(1) 4.189(2) x 2 Ag(l)—Se(6)—Ag(l) 6788(8) Cs(3)-Se(2) 3.647(3) Ag( 1 )—Se(6)«Ag( l) l35.95(7) Cs(3)-Se(3) 3.673(3) Ag(1)—Se(6)-Ag(1) 8053(6) Cs(3)-Se(4) 3.771(3) x 2 Cs(3)—Se(6) 3.900(2) x2 Sb(1)—Se(1)—Sb(3) 9838(6) Cs(3)—-Se(mean) 388(5) aThe estimated standard deviations in the mean bond lengths and the mean bond angle are calculated by the equations cl = {}.'.,,(ln - l)2/n(n - 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 (or angles). 63 ORTEP representation of K2AngS4 as viewed down the a- axis. Small Figure 2-1: octant shaded ellipsoids; Ag, principal axis ellipsoids; Sb, boundary ellipsoids; S, boundary and axis ellipsoids; K. (50% probability ellipsoids). 65 Figure 2-3: Polyhedral representation of szAngS4 (II), as viewed down the a - axis. Filled polyhedra; SbS4, pattern shaded polyhedra; AgS4, large open circles Rb+ cations. 66 (A) Figure 2-4: (A) The [Ag28b2S314' building block with labeling. (B) A portion of the [A128i2]2' layer adapted from Reference #23. (C) Relationship between the [Ag2Sb2S3]4‘ building block in (H) and the double six—rings in CaA12S12. Figure 2-6: ORTEP representation of CszAngS4 (III) as viewed down the a- axis. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; Sb, boundary ellipsoids; S, boundary and axis ellipsoids; Cs. (50% probability ellipsoids). 69 Figure 2-7: ORTEP representation of a single chain of CszAngS4 (III), with labeling. (50% probability ellipsoids). 70 Figure 2-8: ORTEP representation of CS3Ag28b3868 (V) as viewed down the b-axis. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; Sb, boundary ellipsoids; S, boundary and axis ellipsoids; Cs (50% probability ellipsoids) /\ Figure 2—9: 71 ORTEP representation of a single chain of C83Ag28b3Se3 (V) with labeling, highlighting the different Ag-Se binding modes. (50% probability ellipsoids). 72 3.2. Synthesis, Spectroscopy and Thermal Analysis This work dealt with the reactivity of group 11 metals with Ax[Sbsz] fluxes (A = K, Rb, Cs; Q = S, Se). The syntheses were the result of a redox reaction in which the group 11 metal was oxidized by polychalcogenide ions in the Ax[Sbsz] flux. The M“ cations are then coordinated by the highly basic [beQy]n' units. The molten polychalcoantimonate flux is very effective for crystal growth in this system and is conducive to the isolation of pure materials because the residual Ax[Sbsz] flux is soluble in aqueous and polar organic solvents. In the AgS/Cu/Sb/S system (A = K, Rb, Cs) a large number of reactions, with a variety of conditions, resulted in the formation of black crystals. Microprobe analysis carried out on several randomly selected crystals gave an average composition of Cu 2,5SbS 3,6. The powder XRD experiments confirmed the black crystals were Cu3,SbS4.25 In sharp contrast, the reaction of silver metal with a Ax[Sbsz] flux (A = K, Rb, Cs; Q = S, Se) produced a variety of quaternary alkali silver polychalcoantimonate compounds. For the sake of simplicity, each system will be discussed individually. In the KZS/Ag/Sb/S system, yellow-orange polyhedral crystals of (1) form from a metal/antimony ratio of 1:1.5 in a basic 3KZS3 flux at 400° C. Reactions varying either the Sb concentration or the basicity of the flux afforded severely twinned crystals of (1). Different compounds form with varying the metal/antimony compositions. A metal/antimony ratio of 1.5:1 gave a mixture of (I) and dark red crystals that had an average composition of KAg2,2Sb1,1S4,6. The powder X-ray diffraction (XRD) experiment confirmed it was isostructural to (NH4)Ag2AsS4.26. The [AngsS4]°' macroanion has a three - dimensional structure built from layers of comer sharing AgS4 tetrahedrons, stitched together by edge sharing tetrahedral [ASS4]3' units forming tunnels 73 filled by NH4+ cations. Reactions with a metal/antimony ratio above 2:1 resulted in phase separation. In the Rng/Ag/Sb/S system, yellow crystals of (II) were synthesized from a metal/antimony ratio of 1:1 in a Lewis basic 2Rb285 flux at 3500 C. Varying the flux composition resulted in different compounds. Increasing the Angb ratio to 1.521 in a less basic Rb289 flux resulted in the formation of a new phase as evidenced by EDS and XRD. This product crystallizes as orange hexagonal plates that are air and water stable.27 Lowering the reaction temperature also afforded a different compound. Decreasing the Ag/Sb ratio to 1:15 in a Rb2S9 flux at 300 °C resulted in red-brown crystals that had an average composition of RbAg58b287. Attempts to grow single crystals suitable for single crystal diffraction studies are also underway.28 In the Cs28/Ag/Sb/S system, yellow crystals of a-CszAngS4 were synthesized from a metal/antimony ration of 1:1 in a basic 2C8285 flux at 350° C. Rational attempts to synthesize B-CszAngSal‘la from a molten alkali metal polysulfide flux were unsuccessful. Lower reaction temperatures also resulted in a different compound. Red crystals of (IV) were synthesized from an metal/antimony ratio of 1:1.5 in a less basic C8289 flux at 280 °C. At this temperature a small amount (~ 5%) of Ag3SbS329 was observed as an impurity. Red crystals of (IV) were synthesized in a pure form by increasing the reaction temperature to 350°C. In the CszslAg/Sb/Se system, black crystals of (V) were synthesized from a metal/antimony ratio of 1:1.5 in a CSZSe633 flux at 280° C. The synthesis of (V) does not appear to be dependent on the Sb concentration in the flux. However, the formation of (V) appears to be very dependent on the flux composition. Changing the basicity, by changing either the Se (decreasing basicity) or the C828e (increasing basicity), resulted a mixture of orange water-soluble crystals and black powder. Semiquantiative microprobe analysis of the orange crystals indicate they are C83SbSe4. 30 74 Investigations in the KzslAu/Sb/Q (S = S, Se) system resulted in largely amorphous black glassy chunks for both systems. Microprobe analysis revealed that both systems were quaternary, however the largely amorphous nature of the materials made further characterization difficult. Attempts to grow crystals suitable for further characterization are in progress.28 A comparison of the polychalcoantimonate to the polychalcophosophate fluxes provides insight into the structural and chemical complexity of the two systems. The polychalcoantimonate system follows the experimentally observed trend of the polychalcophosphate system.“ Basic conditions (increasing A2Q) favor the tetrahedral [EQ4]3' unit (E = P, Sb; Q = S, Se). Under these conditions, the [SbQ4]3' unit is observed as a discrete tetrahedral species, while the [PQ4]3‘ unit can participate in complex condensation equilibra, forming higher nuclearity Chalcophosphate units such as [P287]4* , [P28e8]4‘: 3‘ , and [P2Se9]4": 32. Under acidic conditions (increasing Q or decreasing A2Q) both P and Sb favor reduced species. In the polythiophosphate system the reduced P“4 species is readily observed in the ethane - like [P2Q6]4' unit (Q = S, Se). While the antimony analog of the [P2Q6]4* unit has yet to be reported, the pyramidal Sb3+ species is the only variable oxidation state species observed. The pyramidal [SbQ3]3' unit (Q = S, Se) also participates in a separate yet complex condensation equilibria, forming higher nuclearity [beQy]"‘ units. The far-IR and Raman data were in good agreement and the results are summarized in Table 2-14. The far-IR spectra of (I) - (III) display an ill resolved doublet at ~385 cm‘1 which are tentatively assigned as Sb-S and Ag-S stretching. Since both metals have very similar coordination geometries, accurate assignment is very difficult. The far-IR spectra of (IV) - (V) are more complicated due to the complex binding modes of the [Sb2Q4]2°' ( Q = S, Se) oligomer. The spectra shows an absorption assigned to the [SbQ4]3' unit at 388 cm'1 and 269 cm'1 for (IV) - (V), respectively. The spectra also shows two absorptions assigned to the [Sb2Q4]2n' oligomer at 373 cm“1 and 357 cm‘1 for 75 (IV) and 258 cm"1 and 250 cm“1 for (V), respectively. The Raman spectra of (I) - (IV) display absorptions in the 405 - 380 cm“1 region that are tentatively assigned to Sb - S vibrations and the absorptions below 380 cm'1 are assigned to Ag-S stretching vibrations. Similarly, the Raman spectrum of (V) displays absorptions in the 267 cm:1 - 256 cm'l region that are tentatively assigned to Sb - Se vibrations and the absorptions below 220 cm'1 are assigned to Ag - Se stretching vibrations. The Raman spectra of (1) -(HI) are shown in Figure 2-10 and the Raman spectra of (IV) and (V) are shown in Figure 2]]. Differential thermal analysis (DTA) data, followed by careful XRD analysis of the residue was examined for all seven compounds. The DTA of (I) displayed a single endothermic peak at ~441°C, but two exothermic peaks at ~423°C and ~391°C appear upon cooling. The two exothermic peaks did not increase in intensity with a second cycle. Examination of the residue, after the second cycle, by powder XRD revealed that the sample melted congruently, suggesting a reversible phase change. The DTA of (H) indicates that it melts congruently at ~418°C. The DTA of (III) shows that it melts incongruently at ~407°C forming a mixture of (III) and decomposition products. Although the glassy nature of the resultant powder pattern made accurate indexing of the decomposition products difficult, B - CszAngS414a was not observed. The DTA of (IV) shows a single sharp endothermic peak at ~386°C but no exothermic peaks are observed upon cooling, see Figure 2-12a. Although no exothermic peaks were observed upon cooling, suggesting a glass transition9, the powder XRD was crystalline and was indexed as a mixture of (IV) and Cs28b4S733. As shown in Figure 2- 12b, upon subsequent heating of (IV), a broad exothermic peak at ~273 °C is observed, followed by a sharp endothermic peak at ~384 °C. Upon cooling, a broad exothermic peak is observed at ~285 °C. Examination of the residue by powder XRD indicated that it was weakly crystalline and all the peaks were indexed to (IV), suggesting the bulk of the material had had undergone a glass transition. The DTA of (V) also shows a single sharp endothermic peak at ~336°C with no peaks observed upon cooling, see Figure 2- 76 13a. Examination of the residue by powder XRD showed similar behavior to (IV) revealing a mixture of (V) and ternary decomposition products. As shown in Figure 2- 13b, upon subsequent heating of (V), a two broad exothermic peaks were observed at ~247°C and ~282°C, followed by a sharp endothermic peak at ~323°C. Upon cooling, a broad exothermic peak is observed at ~368°C. Examination of the residue by powder XRD revealed a mixture of (V), CSZSbZSe434a and AngSez34b . Table 2-15 summarizes the melting point and optical data (see below) for all five compounds. The optical absorption properties of (I) - (V) were evaluated by examining the solid - state UV/vis diffuse reflectance and/or single crystal optical transmission spectra of the materials. The spectra obtained from both methods are in good agreement and confirm the semiconducting nature of the materials by revealing the presence of sharp optical gaps (see Table 15). The transparent crystals of (I) - (IV) were suitable for single crystal optical transmission measurements. The A2AngS4 (A = K, Rb, Cs) compounds exhibit steep absorption edges from which the band—gap, Eg, can be estimated at 2.35 eV (I), 2.58 eV (H), and 2.61 eV (HI), respectively. Representative spectra for (I) - (IH) are given in Figure 2-14. This is in good agreement with that reported earlier for 0:- CszAngS414a. The trend in widening band-gaps from (I) - (HI) correlates nicely with the change in the dimensionality of the compounds. As the dimensionality of the [AngS4]2°' framework decreases the extent of the Ag—S orbital overlap also decreases, causing a narrowing of the bands and increasing the energy gap. The band-gaps of CS3Ag28b3Q3 (Q = S, Se) are 2.06 eV (IV) and 1.50 eV for (V), respectively (see Figure 2-15). The change in the band-gaps of (IV) - (V) can be attributed to the substitution of the smaller 82‘ for the larger Se2'. The larger more diffuse orbitals of the selenide ions broadenthe bands, increasing the extent of Ag - Se overlap, thus lowering the band-gap. 77 Table 2-14. Infrared and Raman Data (cm‘l) for (I) - (V). KzAngS4 szAngS4 CSzAngS4 CS3Ag2Sb3Sg CS3Ag2Sb3Seg (I) (III) (IV) IR Raman IR Raman IR Raman IR Raman IR Raman 619 519 405 408 395 397 395 392 392 392 387 382 384 388 380 382 369 369 370 373 367 366 362 361 361 357 332 321 323 309 316 288 283 263 269 267 255 258 256 250 240 241 237 244 241 235 238 234 222 212 206 193 193 199 180 183 173 176 175 164 161 160 153 158 154 L 143 78 Table 215. Optical Band Gaps from Powder and Single Crystal Measurements and Melting Point Data for (I) — (V). Formula & (eV) Powder Es (eV) Single Crystal Melting Point (°C) KzAngS4 2.38 2.35 441 szAngS4 2.58 2.58 418 CszAngS4 2.66 2.61 407i C83Ag2Sb3S3 2.03 2.06 386i CS3Ag2Sb3Se3 1.50 —NA— 3361 i = incongruent melting. 79 5 l 1 l l 361 3 4 " 1 'E :3 E 3 b -l 2‘ 8 E 2 1' 392 ‘ I: 241 g 206 173 a: 1 ... l J .. 0 “ _1 LM 600 500 400 300 200 100 Wavenumber (cm‘ ') (B) Raman Intensity (arb. units) 600 500 400 300 200 100 Wavenumbcr (cm'l) (C) s t 362 I ‘3 4 r ‘ g _ 332 1 3w «2 el . a b 1 El - 4 52: 1 g t 392 153 : - 175 - “1 “3111 1 O —‘ I‘nnr A; 7 600 500 400 300 200 100 Wavenumber (cm") Figure 2-10: Raman spectra of (A) KzAngS4 (I), (B) szAngS4 (II) , and (C) CszAngS4 (IH). 80 (A) 5 l l T r 2? _. .4 'E 4 :1 E 3 - . 2.7 '5 G 2 5 2 - _ E a: 1 - 4 O l 600 500 400 300 200 100 Wavenumver cm'I (B) 16 l 1 I I 212 14 - - 8 310 t- _. 2: .g 8 h- —1 8 '5 6 l- - 183 4 h 256 " 267 2 '- —1 0 _.1 1 1 600 500 400 300 200 100 Wavenumver cm‘I Figure 2]]: Raman spectra of (A) Cs3AgZSb388 (IV) and (B) Cs3AgZSb3Se8 (V). (A) exo 8 l l l l ”l e - - 9 3 a) 4 7 ‘ (D 8 3 2 _ 386 _ a: D: 1 ° ' ‘ end°2 I l 1 L O 100 200 300 400 500 Temperature (°C) (13) em:8 T l I l I 273 l e - - 9 384 3 0 4 "' -l 0) C 8 m 2 T ‘ 03 a: 47 o _ 285 _ end°2 .4 1 1 1 L, 0 100 200 300 400 500 Temperature (°C) Figure 2-12: (A) DTA diagram for CS3Ag23b383 (first cycle). Heat is absorbed at 386 °C as the material melts but no corresponding exothermic peak is observed. (B) Second DTA cycle showing the crystallization of CS3Ag2Sb3Sg at 273 °C, followed by its subsequent melting and recrystallization at 384 °C and 285 °C, respectively. 82 Figure 2-13: (A) DTA diagram for CS3Ag2Sb3Seg (first cycle). Heat is absorbed at 336 °C as the material melts but no corresponding exothermic peak is observed. (B) Second DTA cycle showing the crystallization of CS3AgZSb3Seg at 274 °C and 282 °C, followed by its subsequent melting and thermal decomposition at 323 °C and 368 °C, respectively. 83 3 O t-‘ '-‘ N N U.) C U: C U! 0 U1 o T (D (B) exo 50 500 N W h C O O l l l <——- Response (uv) -—> 8 F 247 282 323 368 (D 3 Q. 0 O I 1 l 1 l l l 100 150 200 250 300 350 400 450 500 Temp (c) 84 (A) 1.4 r l 35 1.2 - C 3 1 - -E 3 0.8 - .,§ 0.6 ~ g 0.4 P .n < 0.2 - 0 l l 1.8 2 2.2 2.4 2.6 Energy (eV) (B) 7 r r 6 l- g 5 _ 3 a 4 - 3 8 3 " 8‘ 2 - UD 2 1 - o 1 l l I l l l 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 Energy (eV) (C) 6 I I I T I T I e 5 - ’E 3 4 _ -e' 3 3 .. CszAng S4 5 E =2.61eV 3'5 2 ,_ 3 Q 8 ‘2 1 ” o l l I l l l l 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 Energy (eV) F 18013 214: Single crystal optical absorption spectra of (A) KzAngS4 (I), (B) szAngS4 (H) , and (C) CszAngS4 (III). The sharp features at high absorbance are noise and due to the very low transmission of light. 85 (A) 2.4 1.6 - Cs Ag so 8 3 2 3 8 Eg=206eV 4 1.2 Absorption (Arb. Units) 0.8 1 l I I 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 Energy (eV) (B v .0 do w t 1 l l 1 E "a 3 90.24 - - 3 350.18 - - O 0 ‘=0.12 - — .9 § Cs3AgZSb3Se8 80.06 - E =1.50eV .- .D g < ‘L’ 5 C .5 N 3 4 5 6 7 Energy (eV) Figure 2-15: (A) Single crystal optical absorption spectrum of Cs3AgZSb388 (IV). (B) Solid-state optical absorption spectrum of Cs3AgZSb3Se8 (V). 86 3.3. Structural Relationships in AzAngS4 (A = K, Rb, C8): The Counterion Effect. The compounds (I) - (III) illustrate a beautiful example of the counterion effect. The counterion effect35 is a qualitative observation that in some systems, a change in the counterion size will lead to a change in the dimensionality and/or the coordination number of a compound. The structural change is brought about to remove repulsions caused by ineffective screening of the anions. This example provides a unique illustration of this effect, for the anionic stoichiometry is invariant, consisting of varying arrangements of [SbS4]3' units bonded to Ag+ cations. Although no change in the coordination number of the Ag cation is observed, the high mobility of the Ag“ cations allows for the dramatic changes observed in the anionic structure upon variation of the cation volume. In K2AngS4 (I) the three - dimensional structure is favored with the small K+ counterion. The larger Rb+ counterion forces the [AngS4]4‘ framework found in (I) to adopt a layered structure. The layers in (H) are built from [Ag28b283]4” clusters linked together through bridging sulfides, increasing the anionic volume to sufficiently screen the larger Rb+ cations. The larger Cs+ cations slices the [AngS4]2' layers found in (11) into [AngS4]n2“‘ chains separated by C83“ cations. The [AngS4]n2°' chains consist of the same [Ag28b283]4‘ clusters observed in (H) but are now connected through the formation of a Aggsz rhombi. If a hypothetical compound (Ph4P)2AngS4 were synthesized, its structure may result in a “zero - dimensional”, molecular compound. The larger cation should cause the further rearrangement in the [AngS4]2' framework to make it a discrete molecule, possibly some kind of [Ag28b283]4' cluster. at 51 87 4. Conclusions. The synthesis of these new members of the A/Ag/Sb/Q family emphasizes the value of the polychalcoantimonate Ax[Sbsz] (Q = S, Se) fluxes in the synthesis of new chalcoantimonate compounds. The complex structural diversity displayed is primarily due to the variety of binding modes of the [Sbsz]n‘ units, further expanding our concept of using molecular building blocks in the synthesis of extended structures. These fluxes provide a convenient entry into the unknown group 11 chemistry and preliminary results indicate that several more quaternary and possibly the unknown ternary Au/Sb/Q (Q = S, Se) compounds may also be synthesized and structurally characterized. The use of low temperatures (280 - 350°C) allows the synthesis of thermodynamically unstable compounds. Their metastable nature can be clearly seen in their thermal analysis. Although more experimental information is needed, it appears that control of flux composition-basicity results in compounds with different [Sbsz]n' units. This may enable us to achieve a level of synthetic control over the reaction outcome and ultimately allow us to construct structures with preselected ligands. The related selenoantimonate chemistry suggests that both the thio— and seleno-antimonates should be explored to fully understand the [Sbsz]n' anion chemistry. 88 List of References Sutorik A. C.; Kanatzidis M. G. Progr. Inorg. Chem, 1995, 43, 151-265. (a) Liao, J.-H.; Varotis, C.; Kanatzidis M. G. Inorg, Chem, 1993, 32, 2453. (b) Liao, J.-H.; Kanatzidis M. G. Chem. Mater., 1993, 5 , 1561. (a) Zhang, X.; Kanatzidis M. G. J. Am. Chem. Soc., 1994 116, 1890. (b) Zhang, X.; Kanatzidis M. G. Inorg, Chem, 1994 33, 1238. McCarthy, T. J.; Kanatzidis M. G. Chem. Mater., 1993, 5 , 1061. (a) McCarthy, T. J.; Kanatzidis M. G. J. Chem. Soc., Chem. Commun., 1994, 1089. (b) McCarthy, T. J.; Kanatzidis M. G. Inorg. Chem, 1995, 34, 1257. (a) Marking, G. A.; Kanatzidis, M. G. Abstract from 210th Fall ACS Meeting, Chicago II. (b) Marking, G. A.; Kanatzidis, M. G. Submitted for publication. (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.; Lazasrev, V. B.; Tarasova, T. G.; Zh. Strukt. Khim.,1977, 18, 1069. (d) Kanishcheva, A. S.; Mikhailov, Y. N.; Kuznetsov, V. G.; Batog, V. N. Dolk. Adad. Nauk SSSR, 1980, 251, 603-605. (e) Cordier, G.; Schafer, H. Rev. Chem. Min., 1981, 18, 218-223. (f) Dorrscheidt, W.; Schafer, H. Z. Naturforsch., 1981, 36b, 410-414. (g) Cordieer, G.; Schwidetzky, C.; Schafer, H. Rev. Chem. Min., 1982, 19, 179-186. (a) Ko Y. H.; Tan K. M.; Parise J. B.; Darovsky A. Chem. Mater., 1996, 8 , 493- 496. (b) Parise J.B.; Ko Y.H. Chem. Mater.,1992, 4 , 1446-1450. (c) Parise J. B. J. Chem. Soc., Chem. Commun., 1990, 1553-1554. ((1) Sheldrick, W. S.; Hausler, H.-J. Z. Anorg. Allg. Chem, 1988, 561, 149-156. (e) Dittmar, V. G.; Schafer, H. Z. Anorg. Allg. Chem, 1978, 441, 93-97. (f) Sheldrick, W. S.; Hausler, H.-J. Z. Anorg. Allg. Chem, 1988, 557, 105-111. (g) Eisenmann, B.; Schafer, H. Z. Narurforsch. 1979, 34b, 383-385. (h) Graf, H. A.; Schafer, H. Z. Narurforsch. 1972, 27b, 735-739. 10. ll. 12. l3. 14. 15. 16. 17. 18. 19. 20. 21. 89 McCarthy, T. J.; Kanatzidis M. G. Inorg. Chem, 1994, 33 , 1205. Takeuchi, Y.; Sadanaga, R. Z. Krstallogr. ,1969, 130, 346. Dittmar, V. G.; Schafer, H. Z. Anorg. Allg. Chem, 1977, 437, 183. Dittmar, V G.; Schafer, H. Z. ; Schwidetzky, C. Z. Anorg. Allg. Chem, 1978, 441, 93-97. Imafuku, M.; Nakai, 1.; Nagashima, K.; Mat. Res. Bull., 1986, 21 , 493-501. (a) Schimek, G. L.; Pennigton, T. L.; Wood, P. T.; Kolis, J. W. J. Solid State Chem, 1996, 123, 277-284. (b) Wood, P. T.; Schimek, G. L.; Kolis, J. W. Chem. Mater., 1996, 8, 721-726. (a) Wemic, J. H. Am. Miner., 1960,45, 591 . (b) Hoda, S. N.; Chang, L. L. Y. Am. Miner., 1975, 60, 621. (c) Sawada, H.; szada, I.; Hellner, B.; Toknami, M. Zeitschriftfuer Kristallographie, 1987, 180, 141. (d) Megarskaya, L.; Rykl, D.; Taborsky Z. Am. Miner., 1987, 72, 227. (e) Makovicky, B.; Munne, W.; Neuss, G.; Jahrbuchfuer Mineralogie Abhandlungen., 1983, I47, 58. (a) McCarthy, T. J.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Mater., 1994, 6, 1072. (b) Feher, F. In Handbuch der Praparativen Anorganischen Chemie; Brauer, G., Ed.; Ferdinand Enke: Stuttgart, Germany, 1954; PP 280. CERIUSZ, Version 1.6, Molecular Simulations Inc., Cambridge, England, 1994. McCarthy, T. J.; Ngeyi, S.-P.; Liao, J .-H.; DeGroot, D.; Hogan, T.; Kannewulf, C. R.; Kanatzidis, M. G. Chem. Mater., 1993, 5, 331. Sheldrick, G. M. In Crystallographic Computing 3; Sheldrick, G. M., Kruger, C., Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985; pp 175. TEXSAN: Single Crystal Structure Analysis Software, Version 5.0; Molecular Structure Corp.: The Woodlands, TX 77381, 1981. Walker, N.; Stuart, D. Acta Crystallogr. 1983, A39, 158. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 90 Enos A. Axtell, Ph.D. Dissertation, Michigan State University, East Lansing, MI, 1995. Zheng, C.; Hoffmann, R.; Nesper, R.; von Schering, H.-G. J. Am. Chem Soc., 1986, 108, 18767. (a) Kanishcheva, A. S.; Mikhailov, Y. U.; Lazarev, V.; Moshchalkova N. A. Doklady Akademii Nauk SSSR , 1980, 252 , 872. (b) Geller, 8.; Wemick, J. H. Acta. Cryst., 1959, 12, 46. Troger, F. J. Am. Mineralogist, 1966, 51, 890. Auemhammer, M.; Effenberger, H.; Irran, E,; Pertlik, F.; Rosenstingl, J. Solid State Chem. 1993, 106, 421. Hanko, J. A.; Kanatzidis, M. G. Manuscript in preparation. Hanko, J. A.; Kanatzidis, M. G. Work in progress. Indexed Ag38bS3 using "Powder Diffraction File-Inorganic Phases", vol. 21,. pp. 1173, and vol. 25, pp. 1187, 1993. Sugaki, A.; Sci. Rep. Tohoku Univ., 1993, 106, 421. Chondroudis, K.; Hanko, J. A.; Kanatzidis, M. G. Inorg. Chem. ,1997, 37 , 2323. Chondroudis, K.; McCarthy, T. J.; Kanatzidis, M. G. Inorg. Chem, 1995, 34, 1257. Indexed C82Sb4S7 using "Powder Diffraction File-Inorganic Phases", vol. 32,. pp. 214, 1987. (a) Drake, G. W.; Kolis, J. W. Coord. Chem. Rev., 1994, I37, 131. (b) Sheldrick, W.; Wackhold, M. Angew. Chem. Int. Ed. Engl., 1997, 36, 206. Kanatzidis, M. G. Phosphorous, Sulfur, and Silicon, 1994, 93-94, 159. CHAPTER 3 Thioantimonate Flux Synthesis of 0t,[3 - RbAgZSbS4: Two New Three-Dimensional Compounds With Acentric Structures. 91 92 1. Introduction The alkali polychalcogenide fluxes are conducive to the intermediate temperature synthesis of new multinary chalcometalates.l Alkali polychalcogenide fluxes allow the easy incorporation of alkali metal cations into the compound. The incorporation of alkali metal can lower the dimensionality of a compound or cause a structural transformation to a previously unknown structure. The motivation for the research reported here is to develop a methodology using molecular building blocks or recognizable structural fragments in the synthesis of new multinary compounds with potentially useful properties. To further develop this general methodology, we have explored the polythioantimonate flux method as a convenient way to synthesize new quaternary compounds with transition metals. Several unusual examples have been reported, these include: Cszsb4S3, CsSb86,2 A2AngS4 (A = K, Rb, Cs),3v4 CS3Ag2Sb3Qg (Q = S,Se),3:4 KTthZSe6,5 AzAgzoSb4819 (A = Rb, Cs),° A2AquS4 (A = Rb, Cs),7 szAu68b48108 and KHngS39. These building blocks form by the in situ fusion of A2Q/Sb/Q ( A = Alkali metal; Q = S, Se) to form various Ax[Sbsz] species in a molten polychalcogenide solvent. The metal atoms are dissolved by reaction with the polychalcogenide ligands and, along with the various [Sbsz]n‘ units, act as mineralizers in the flux to promote single crystal growth. Although ligands such as [SbS4]3' and [SbS3]3' are interesting because of the many potentially different bonding modes they can exhibit, relatively little is known about their coordination chemistry in the solid state. Continuing our investigations with the group 11 transition metals, here we report the synthesis, structural characterization, optical, and thermal properties of (1,13 - RbAgzsbS4, two new quaternary thioantimonates with an acentric structure. While 01. - RbAg2SbS4 was under investigation in our laboratory, its synthesis and 93 structure, in supercritical ammonia, was reported4b. The highly unusual thermal properties exhibited by this compound, however, were not studied and merit further investigation. 2. Experimental Section 2.1. Reagents The reagents mentioned in this study were used as obtained unless noted otherwise: (i) antimony powder 99.999% purity, ~200 mesh, Cerac Inc., Milwaukee, WI; (ii) silver powder 99.95% purity, -325 mesh Alfa AESAR Group, Seabrook, NH; (iii) rubidium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (iv) sulfur powder, sublimed, J.T. Baker Chemical Co., Phillipsburg, NJ; (v) N,N—Dimethylfonnalide (DMF) reagent grade, EM Science, Inc., Gibbstown, NJ. (vi) diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. 2.2. Syntheses. szs was prepared by reacting stoichiometric amounts of the elements in liquid ammonia as described in Chapter 2 Section 2.2. Preparation of a-RbAg2SbS4 (I). An amount of Rb28 (0.51g, 0.50 mole), Ag (0.041 g, 0.375 mmole), Sb (0.015g, 0.125 mmole), and S (0.064g, 2 mole) were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame-sealed in vacuo (~10‘3 Torr). The reaction mixture was heated 94 to 350°C over 12 hours in a computer—controlled furnace. It was isothermed at 350°C for 4 days, cooled to 150°C at a rate of 4°C/hr, and then cooled to room temperature in 10 hours. The product, which is stable in water and air, was isolated by dissolving the Ross, and any be[SbySz] flux with DMF under an inert atmosphere to give well-formed orange hexagonal and red-orange polyhedral crystals in 52% yield, based on Sb. Quantitative microprobe analysis from both crystals gave an average formula of Rb1.9Ag1.1SbS4_7. Powder X-ray diffraction experiment confirmed that the orange and red crystals were identical, having only different morphologies. Preparation of B-RbAgZSbS4 (II). An amount of Rb2S (0.110g, 0.50 mole), Ag (0.108g, 1 mole), Sb (0.03lg, 0.25 mmole), and S (0.128g, 4 mole) were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame-sealed in vacuo (~10'3 Torr). The reaction mixture was heated to 500°C over 15 hours in a computer-controlled furnace. It was isothermed at 500°C for 4 days, cooled to 110°C at a rate of 4°C/hr, and then cooled to room temperature in 1 hour. The product, which is stable in water and air, was isolated by dissolving the Rb2Sx and any be[SbySz] flux with DMF under an inert atmosphere to give bright red crystals in 60% yield, based on Sb. Quantitative microprobe analysis of single crystals gave RbAg2,1Sb1_SS6,6 2.3. Physical Measurements Powder X-ray Diffraction The compounds were examined by Powder X- ray Diffraction for the purpose of phase purity. Analysis for (I) and (II) were performed using a calibrated Rigaku-Denki/RW400F2 (Rotaflex) rotating anode powder diffractometer controlled by an IBM computer, operating at 45 kW 100 mA 95 and with a 10/min scan rate, employing Ni-filtered Cu radiation in a Bragg- Brentano geometry. Powder patterns were calculated with the CERIUS2 software.lo Tables of calculated and observed XRD patterns for a-RbAgZSbS4 and B- RbAngbS4 are given in Tables 3-1 and 3-2, respectively. Infrared Spectroscopy Infrared spectra, in the far-IR region (600-50 cm: 1), were recorded on a computer controlled Nicolet 750 Magma-R Series H spectrophotometer equipped with a TGS/PE detector and silicon beam splitter in 4 cm"1 resolution. The samples were ground with dry CsI into a fine powder and pressed into translucent pellets. Raman Spectroscopy Raman spectra were recorded with a BIO-RAD FT Raman spectrometer with a Spectra-Physics Topaz T10-106c 1,064 nm YAG laser running at 11 to 11.5 amps equipped with a Ge detector. The samples were ground into a fine powder and loaded into melting point capillary tubes. Single crystal optical transmission spectroscopy Room temperature single crystal optical transmission spectra were obtained on a Hitachi U-6000 Microscopic FT Spectrophotometer mounted on an Olympus BH2-UMA metallurgical microscope over a range of 380 to 900 nm. Crystals lying on a glass slide were positioned over the light source and the transmitted light was detected from above. Difierential Thermal Analysis (DTA) DTA experiments were performed on a computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~ 25 mg) of ground crystalline material was sealed in quartz ampoules under vacuum. A quartz ampoule of equal mass filled with A1203 was sealed and placed on the reference side of the detector. The sample was heated to the desired temperature at 10 °C/min, then isothermed for 10 minutes and finally cooled to 50 °C at the same rate. The residues of the DTA experiments were examined by X-ray powder diffraction. To evaluate congruent melting we compared the X-ray powder diffraction patterns before and after the DTA experiments. The 96 stability/reproducibility of the samples were monitored by running at multiple cycles. Differential Scanning Calorimetry (DSC) DSC experiments were performed on a computer-controlled Shimadzu DSC-50 thermal analyzer under a nitrogen atmosphere at a flow rate of 35 ml/min. The samples (~ 5 mg) of the crystalline material were crimped in an aluminum pan. The pan was placed on the sample (right) side of the DSC-50 detector and a crimped empty aluminum pan of equal mass was placed on the reference (left) side of the detector. The samples were heated to the desired temperature at 5 °C/min, isothermed for 5 minutes, and finally cooled at the same rate to 50°C. The temperatures associated with each peak in the spectrum have a standard deviation of 0.2 degrees. The adopted convention in displaying the data is as follows: exothermic peaks are associated with a positive heat flow while endothermic peaks are associated with a negative heat flow. Semiquantitative microprobe analyses The analyses were performed using a JEOL JSM-6400V scanning electron microscope (SEM) equipped with a TN 5500 EDS detector. This technique was used to confirm the presence of all elements in the compounds. Data acquisition was performed with an accelerating voltage of 20kV and thirty seconds accumulation time. Single crystal X-ray Crystallography. Intensity data for (I) was collected on a Siemens SMART-CCD diffractometer using graphite-monochromatized MoKOt radiation. The data collection covered the full sphere of reciprocal space, out to 58° 29. The individual frames were measures with a detector to crystal distance of 5 cm with an omega rotation of 0.3 deg. and an acquisition time of 30 seconds, leading to a total measurement time of about 18 hours. The SMART software11 was used for the data acquisition and SAH‘IT 11 for data reduction. The space group was determined from systematic absences and intensity statistics. No crystal decay was detected for (I). A correction for Lorentz polarization effects and an empirical 97 absorption correction (SADABS)12 were applied to the data. The structure of (I) was solved using direct methods and refined with a full-matrix least squares techniques of the SHELXTL13 package of crystallographic programs. With a full sphere of data, the correct enantiomorph could be refined to determine the correct structure. The structure as refined had a flack parameter of 0.02(2), confirming the structure as solved was the correct enantiomorph. Intensity data for (H) were collected using a Rigaku AFC6S four-circle automated diffractometer equipped with a graphite crystal monochromator. Crystal stability was monitored with three standard reflections whose intensities were checked every 150 reflections, and unless noted, no crystal decay was detected in any of the compounds. The space groups were determined from systematic absences and intensity statistics. An empirical absorption correction based on w scans was applied to all data during initial stages of refinement. The structures were solved by direct methods using SHELXS-86 software14a, and full matrix least squares refinement was performed using the TEXSAN software package14b. An empirical DIFABS correction15 was applied as recommended after full isotropic refinement, after which full anisotropic refinement was performed. Since the structure of B - RbAgZSbS4 is non-centrosymmetric, a refinement on the other enantiomorph should be performed to determine the correct structure. Only unique data was collected for this data set, preventing the determination of the correct enantiomorph. The complete data collection parameters and details of the structure solution and refinement for a-RbAgZSbS4 and B- RbAgZSbS4 are given in Table 3-3. The coordinates of all atoms, average temperature factors, and their estimated standard deviations are given in Tables 3-4 to 3-5. 98 Table 3- 1. Calculated and Observed X-ray Powder Patterns for Ot-RbAgQSbS4 (I). hkl dcalc, A d obsd, A I/Imax(0bSd) 003 5.55 5.51 100 101 5.42 5.38 12 103 3.99 3.97 25 104 3.37 3.35 100 110 3.31 3.29 60 111 3.24 3.23 24 112 3.07 3.06 9 105 2.88 2.87 31 113 2.84 201 2.82 2.77 34 203 2.54 2.56 20 204 2.36 2.35 18 115 2.34 2.26 8 107 2.19 2.19 12 205 2.17 2.17 7 211 2.14 2.14 5 116 2.12 2.12 12 212 2.09 2.08 6 213 2.01 2.01 11 206 1.99 1.99 5 108 1.95 1.95 41 214 300 1.92 1.92 12 301 1.89 1.89 13 302 1.863 1.862 11 207 1.831 1.834 10 215 1.817 1.816 11 ‘ 118 109 1.763 1.764 7 99 Table 3-2. Calculated and Observed X-ray Powder Patterns for B-RbAngbS4 (II). hkl dcalc, A d obsd, A I/Imax(0b5d) 101 5.51 5.50 30 200 3.62 3.61 33 020 3.39 3.39 26 112 3.23 3.22 66 211 2.99 2.98 6 121 2.89 2.88 8 202 2.75 2.75 6 103 2.64 2.64 6 013 2.62 2.61 6 220 2.47 2.47 25 004 213 2.12 2.12 123 2.08 2.08 5 312 2.00 2.00 22 114 1.95 1.99 13 132 1.928 1.924 100 321 1.917 1.919 56 204 1.837 1.836 8 400 1.810 1.808 32 024 1.805 1.803 38 033 1.771 1.769 9 411 1.713 1.712 4 040 1.698 1.694 8 105 1.660 1.658 5 323 224 1.616 1.615 7 420 1.597 1.596 5 g 100 Crystallographic Data for or-RbAgZSbS4 and B-RbAgZSbS4. Table 3-3. Formula or-RbAg25b54 B'RbAgZSbS4 FW 551.19 551.19 a, A 6.616(2) 7.240(3) b, A 6.616(2) 6.795(3) c, A 16.588(3) 8.528(3) 61 (deg) 90.00 90.00 B (deg) 90.00 90.00 7 (deg) 120.00 90.00 2, V(A3) 3; 628.8(5) 4, 419.5 (3) r. (Mo Kot), A 0.71069 0.71069 space group P 3221 (#154) 1222 (#23) Dcalc, 8/0m3 3.64 4.45 u,cm'1 14.99 251.93 20max, deg 57.6 49.8 Temp (°C) -73 23 Final R1/wR2a % 2.7/7.9 NA Final R/Rw,b % NA 3.0/4.8 Octants collected ih, ik, i1 h, k, 1 Total Data Measured 6604 233 Total Unique Data 1040 233 (ave) Data F02>2G(FO‘2) 1040 NA Data F02>3o(F0'2) NA 219 No. of Variables 39 21 Cgstal Dimen.,mm 0.20x0.20x0.10 0.16x0.080.40 bR = 2(1Fol - chl)/ZIF0|, Rw = {Zw(lFol - IFCI)2/ZwlFo|2}1/2. 101 Table 3-4. Atomic Coordinates ( x 104) and Equivalent Isotropic Displacement Parameters (A2 x 103) for (X-RbAngbSA, (I) with Estimate Standard Deviations in Parentheses. Atom X Y Z Uecf Sb 1 7200(1) O 6667 5(1) Rbl 4434(1) 4434(1) 5000 12(1) Agl 948(1) ~2874( l) 5954(1) 21(1) S1 5053(2) -191(2) 5529(1) 10(1) 82 7587(3) -3315(3) 6743(1) 9(1) a U(eq) is defined as one third of the trace of the orthogonalized Uij tensor. Table 3-5. Fractional Atomic Coordinates and B(eq) Values for B-RbAgZSbS4 (H) with Estimated Standard Deviations in Parentheses. Atom x Y 2 Bequ2 Sbl 1/2 0 0 1.12(6) Agl 1/2 1/2 -0.203l 4.2(1) Rbl 0 1/2 0 2.0(1) s1 0.6884(5) -0.1980(6) 0.1575(4) 1.6(1) ‘1 B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as Beq = (4/3)[a2B(1, 1) + bZB(2, 2) + czB(3, 3) + ab(cosy)B(1,2) + ac(cosB)B(l,3) + bc(cosa)B(2, 3) 102 3. Results and Discussion 3.1. Description of Structures Structure of a-RbAg2SbS4 (I). The three-dimensional structure of or- RbAgZSbS4 is isostructural to RbAgzsbS44b reported by Kolis, et. al., and similar to the structure of SrCu2SnS4l°. The acentric three-dimensional framework is formed by the complex condensation of comer-sharing AgS4 tetrahedrons and [SbS4]3‘ units, see Figure 3—1. This complex binding scheme produces Rb+ filled channels that run in a non-interconnecting criss-cross fashion along the crystallographic a - and b - axes, and in the [110] direction, see Figure 3-2 A view down the c- axis, highlighting the hexagonal arrangement, is shown in Figure 3-3. This unusual arrangement for three channels is the result of the successive application of the 32 screw axis parallel to the c - axis. A polyhedral representation showing the helical nature of the [SbS4]3' units is shown in Figure 3-4. The [SbS4]3‘ unit adopts a symmetric binding mode with all four sulfur atoms coordinating with two Ag+ ions as shown in Scheme 1. Scheme 1 103 The [SbS4]3‘ unit is a fairly regular tetrahedron with Sb-S distances in the range from 2.327(2)A to 2.335(2)A and an average S-Sb-S angle of 109.4(2)o. The Ag+ ion has a distorted tetrahedral geometry with Ag-S bond distances in the range from 2490(2)A to 2733(2)A. The S-Ag-S bond angles range from 91.82(4)° to 146.19(3)°. The closest Ag-Ag contact is 4.065(3)A which does not indicate any significant interactions. Our refinement of the structure of a-RbAg2SbS4 compares well with the results reported by Kolis and coworkers.4b Although there are slight discrepancies in the final reported residuals (3.7/3.8 vs. 2.7/7.9) and the standard deviations, a direct comparison cannot be made because of the different temperature used to collect the data. The structure of a-RbAgzsbS4 was initially solved with the space group P31. After discovering the published refinement of a-RbAg2SbS44b with the higher symmetry space group, a solution on our original data set was attempted with this new space group. Initial refinements in the higher symmetry space group (P3221) were unsuccessful and a related compound, SrCuzsnS4, was solved with the lower symmetry space group (P31), casting doubt as to the correct space group. This problem was already encountered by Kolis et. a1.4b and upon careful consideration of the possible space groups the correct space group, P3221, was detennined. The space group choice was confirmed, of the CCD data set, by utilizing the SHELXL13 package of crystallographic programs to convert the lower symmetry solution to the higher symmetry space group. The structure of Ot-RbAg2SbS4 is unique in that the binary "Ag284" portion of the structure forms a three-dimensional framework in which the Sb5+ species reside. Polyhedral and ball stick representations of the stable "Ag284" framework are shown in Figures 3-5 and 3-6, respectively. This represents a departure from other chalcoantimonate and Chalcophosphate compounds, dominated by the various [Eny]"‘ (E = P, Sb; Q = S, Se, Te) units coordination to the metal ions, not the 104 metal ions framework coordinating to the various [Eny]“' units. The [Ag2SbS4]n- framework is stabilized by an eight-coordinate Rb+ ion located in the channels [range of Rb-S distances, 3.352(2) - 3.483(2)A; mean 342(1) A] Tables of selected distances and angles for or-RbAbQSbS4 are given in Table 3-6. Structure of B-RbAgzsbS4 (II). The three-dimensional structure of B-RbAgZSbS4 is similar, but not identical, to the structure of BaAngeS417 and KAgZSbS44b. A view down the a - axis is given in Figure 3-7. The three- dimensional framework is best described as a series of tetrahedral AgS4 units that form layers via comer-sharing in the fashion of Hglzlg, see Figure 3-8. These layers are linked by [SbS4]3' tetrahedral which edge-share to AgS4 units in opposite layers, resulting in channels in which the Rb“ cations reside. A polyhedral representation (Figure 3-9) clearly shows the acentric nature of the structure by highlighting the fact that all the tetrahedral [SbS4]3' units all point along the crystallographic c - direction. This arrangement produces criss-cross channels along the a and b directions. The dimensions of the channels (5.059A (Ag - Ag) by 7.240A (Sb - Sb)) appear to be just large enough to stabilize the Rb+ ions. The small difference between the [AgZSbS4]“' framework observed in B- RbAbzsbS4 and that of KAgZSbS44b is not readily apparent, see Scheme 2. As shown in Scheme 2, the projections down the crystallographic c - axis appear to be identical. Although, when the structure is stabilized by the larger Rb+ cation, the framework distorts, breaking the 4-fold rotation and lowering the symmetry. Since both compounds are acentric, it is the length of the cell edges that is important, not the choice of the unit cell. The tetrahedral AgS4 units (Figure 3-9) are highly distorted, resulting in a compressed Ag-S contact of 2495(4)A and a severely elongated Ag-S contact of 2883(4)A. The bond angles of the tetrahedral AgS4 unit are just as distorted, ranging from 9l.l7(2)° to l620(2)°, respectively. This 105 highly anisotropic nature of the tetrahedral AgS4 units is a common feature of this structure type.13 The [SbS4]3' unit is a fairly regular tetrahedron, with a single Sb- S distance of 2.340(4)A and S-Sb—S angles in the range from 108.7(2)° to 110.2(3)°. Scheme 2 [SbS4l3' war. yu.‘9 B-RbA92SbS4 KA92SbS4 The [AgzsbS4]n- framework is stabilized by an eight-coordinate Rb+ ion located in the channels [range of Rb-S distances, 3.332(4)-3.493(4)A; mean 338(2) A] Tables of selected distances and angles for li-RbAbZSbS4 are given in Table 3-7. 106 Table 3-6. Selected Distances (A) and Angles (deg) for Ot-RbAgZSbS4 (I) with Standard Deviations in Parentheses 8. Sb(1)-S(l) 2.327(2) x2 S(l)-Sb(1)-S(1') 108.64(8) Sb(1)-S(2) 2.335(2) x2 S(1)-Sb(1)-S(2) 109.40(5) x2 Sb(l)-S(mean) 233(1) S(1)-Sb(l)-S(2') 110.10(5) x2 S(2')-Sb(1)-S(2) 109. 19(8) Ag( 1 )-S( 1) 2.490(2) S-Sb-S (mean) 109.4(2) Ag(l)-S(2) 2.720(2) Ag( 1 )-S(3) 2.469(2) Ag( 1)-S(4) 2.733(2) S( 1)-Ag(1)-S( 1') 91 .82(4) Ag(1)-S(mean) 260(7) S(2)-Ag(1)-S( l) 146.19(3) S(l)-Ag(l)-S(2') 105.87(5) Rb(1)-S(1) 3.399(2) x2 S(2)-Ag(l)-S(l') 114.34(2) Rb(1)-S(1') 3.483(2) x2 S(l')-Ag(1)-S(2) 90.66(5) Rb(1)-S(2) 3.352(2) x2 S(2)-Ag(1)-S(2') 95.33(4) Rb(1)-S(2') 3.439(2) x2 S-Ag(l)—S (mean) 107(8) Rb(1)-S (mean) 341(1) 3The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 01: {2n(l,, — [)2/n(n — 1)}1’2, where 1,, is the length (or angle) of the nth bond, l the mean length (or angle), and n the number of bonds. 107 Table 3-7. Selected Distances (A) and Angles (deg) for B-RbAgZSbS4 (H) with Standard Deviations in Parentheses 3. Sb(1)-S(l) 2.340(4) S(l)-Sb(1)-S(2) 108.7(2) Sb(1)-S( 1') 2.340(4) S( 1 )-Sb( 1 )-S(3) 110.2(3) Sb(1)—S(1") 2.340(4) S(1)-Sb(1)-S(4) 109.7(3) Sb(1)-S(l “0 2.340(4) S(2)-Sb(l)-S(3) 109.2(2) S(2)-Sb(1)-S(4) 109.8(3) S(3)-Sb(1)-S(4) 109.2(2) Ag(1)-S(l) 2.495(4) x2 S-Sb-S (mean) 109(2) Ag(l)-S( 1') 2.883(4) x2 Ag(l)—S(mean) 26(7) S(l)-Ag(1)-S( l ') 162.0(2) S( 1)-Ag(1)-S( 1 ") 9622(7) S(l')-Ag(1)-S(l"') 91.17(7) S(l)-Ag(1)—S(l") 162.0(2) Ag( 1)-Ag( 1') 3.464(5) S(l')-Ag(1)-S(1") 9622(7) S(3)-Ag( 1)-S(4) 131.3(2) Rb(1)-S(1) 3.493(4) x4 S-Ag(1)-S (mean) 123(13) Rb(1)-S(1) 3.332(4) x4 Rb(l)-S(mean) 338(2) Ag(1)-S(1)-Ag(1’) 138.4(1) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 0'1 = (2,,(ln — l)2/n(n — 1)}1/2, where 1,, is the length (or angle) of the nth bond, l the mean length (or angle), and n the number of bonds. 108 Figure 3-1: ORTEP representation of or-RbAgZSbS4 as viewed down the a- axis. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; Sb, boundary ellipsoids; S, boundary and axis ellipsoids; Rb. (50% probability ellipsoids). Figure 3-2: ORTEP representation of Ot—RbAgZSbS4 as viewed down the [110] direction. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; boundary ellipsoids; S, boundary and axis ellipsoids; Rb. (50% probability ellipsoids). 110 Figure 3-3: ORTEP representation a—RbAgZSbS4 as viewed along the c - direction. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; boundary ellipsoids; S. The cations are removed for clarity. (50% probability ellipsoids). 111 The cations are omitted for clarity. Polyhedral representation of a—RbAgZSbS4 as viewed down the [110] direction pattern shaded polyhedra SbS4 ball and stick polyhedra; AgS4. Figure 3-4: 113 Figure 3-6: Ball and Stick view of the “Ag284” framework or-RbAgZSbS4. Open circles; S, and shaded circles; Ag 114 Figure 3-7: ORTEP representation of B-RbAgZSbS4, as viewed down the a - axis. Small octant shaded ellipsoids; Ag, principal axis ellipsoids; boundary ellipsoids; S, boundary and axis ellipsoids; Rb. (50% probability ellipsoids) 115 Figure 3-8: ORTEP representation of the "Ag2S4u layer in B-RbAgzsbS4 Small octant shaded ellipsoids; Ag, boundary ellipsoids; S. (50% probability ellipsoids) 116 Figure 3-9: Polyhedral representation of B-RbAg2SbS4 (H) as viewed down the b - axis highlighting the highly distorted AgS4 units. Pattern shaded polyhedra; AgS4, ball and stick polyhedra; SbS4. The cations are omitted for clarity. 117 3.2. Synthesis and Physicochemical Properties: The syntheses were the result of redox reactions in which the silver is oxidized by polysulfide ions in the Ax[SbySz] flux. The Ag+ centers are then coordinated by the highly charged [SbySz]n- ligands. The molten polythioantimonate flux is very effective for crystal growth in this system. The isolation of pure crystalline products is facilitated by the flux solubility in aqueous and organic solvents. In a-RbAgZSbS4, the orange and red-orange crystals were found to be the same phase, but with different morphologies. The thioantimonate fluxes appear to follow the trends observed in the corresponding Chalcophosphate fluxes, where Lewis basic fluxes favors the tetrahedral [EQ4]3' units (E = P, Sb; Q S, Se). We were interested in exploring fluxes with higher Ag/Sb ratios in hopes of incorporating additional equivalents of Ag+ ions into the resulting [AngbySzP‘ framework. In the previous chapter, reactions with an Ag/Sb ratio = 2 resulted in either phase separation or the formation of a known compound. Silver rich frameworks were only stabilized when the Angb ratio .>_ 3. In addition, the Rb+ cation may provide the best combination of size and basicity to stabilize novel [AngbySzP‘ frameworks. The optical absorption properties of were evaluated by examining the single crystal optical transmission spectra of the materials. The spectra confirm the semiconducting nature of the materials by revealing the presence of sharp optical gaps. The spectrum of a-RbAgZSbS4 exhibits a steep absorption edge, from which the band-gap, Eg, can be assessed at 2.23 eV. The spectrum of B- RbAg2SbS4 also exhibits a steep absorption edge with a band-gap, Eg, of 1.96 eV. Representative spectra for (I) and (II) are given in Figure 3-10. Since the optical gap was determined by single crystal data, plots of (abs)2 vs. hv and «labs vs. hv, 118 shown in Figures 3-11 and 3-12, can be used to distinguish between a direct and an indirect band gap in semiconductors.19 The spectra confirm the indirect nature of the band gap for both (I) and (H). The infrared spectrum of Ot-RbAgZSbS4 and [S-RbAngbS4 show a strong absorption at ca. 390 cm’l, characteristic of the tetrahedral [Sbs413- unit“. The Raman spectrum of (I) and (H) show absorbencies at ca. 361 cm-l, and 174 cm.1 are characteristic of the tetrahedral [SbS4]3‘ unit by comparison with the Raman spectrum of Na3Sb84'9HzO.20 The absorptions at ca. 392 cm‘1 and 380 cm'1 are assigned to Ag-S vibrations. The far-IR and Raman data are summarized in Table 3-8. The Raman spectra for (I) and (H) are shown in Figure 3-13. 3.3. Thermal Analysis The thermal behavior of chalcogenide compounds is investigated to determine the stability of the materials. Materials which melt congruently can be mechanically processed, making large single crystals or thin films from a melt. This is property is essential for materials to be viable candidates for applications requiring well-formed single crystals. The acentric structures of (I)-(H) makes these materials candidates for several potential applications especially if they melt congruently. Materials that melt incongruently highlight the metastable nature of the compounds synthesized by polychalcometalate fluxes. Careful examination of the residue by powder x-ray diffraction offers insight into the compounds stability. In materials that melt incongruently, identification of the decomposition products often allows for the determination of the kinetically and thermodynamically stable phases when multiple phases of the same formula exist or when proposing possible mechanisms for product decomposition”. Occasionally, some materials exhibit interesting thermal properties, such as reversible glass formation”. The thermal 119 behavior of or-RbAgZSbS4 was so completely unprecedented that we are struggling to understand the results. The results that follow are extraordinary in the research of chalcogenide compounds. These results differ from batch to batch as well as from sample to sample from the same batch. The latter happens in seemingly pure single phase samples. The complex thermal behavior of a-RbAgZSbS4 is summarized in Table 3—8. The first sample from Batch I displayed reversible glass formation. Although rare, this behavior is observed in the (Ph4P) [M(Se6)2] (M = Ga, In, T0223 and CszHggMzsg (M = Sn, Ge).22b Reversible glass formation is characterized by an endothermic peak upon heating and the corresponding exothermic peak is observed upon subsequent heating. As shown in Figure 3-14A, upon heating the first sample of or-RbAgzsbS4, a single endothermic peak is observed at 396°C. Upon cooling, no corresponding exothermic peak is observed. Upon subsequent heating, a broad exothermic peak is observed at 306°C, followed by a sharp endothermic peak at 392°C. Upon cooling, again no corresponding exothermic peak is observed, see Figure 3-14B. The compounds sixth cycle is shown in Figure 3- 14C. The powder X-ray diffraction pattern of the residue at room temperature was very weak, showing only a few peaks, indexed to Sb and Ag metal. A second sample, from Batch 1, showed a completely different series of thennal events. Upon heating, three endothermic peaks are observed at 326°C, 400°C, and 430°C, respectively. Upon cooling, two exothermic peaks are observed at 364°C and 267°C, respectively, see Figure 3-15A. Upon subsequent heating, the three exothermic peaks are again observed at 326°C, 400°C, and 430°C, respectively. Upon cooling, only a single endothermic peak at 364°C was observed, see Figure 3—15B. Upon heating for a third cycle two exothermic peaks were observed at 246°C and 306°C; followed by two endothermic peaks at 400°C and 430°C, respectively. A single sharp exothennic peak at 357°C was observed 120 upon cooling. A fourth cycle showed an exothermic peak, at 246°C, upon heating folloowed by the two endothermic peaks at 400°C, and 430°C, respectively. Upon cooling, only a single sharp exothermic peak at 357°C is observed. The third and fourth DTA cycle of sample 2 from Batch I are shown in Figure 3-16. Examination of the ingot by powder X-ray diffraction revealed that or-RbAgQSbS4 had undergone a structural transformation to a new material with a large unit cell, as determined by the low angle peaks observed in the powder X—ray diffraction of the residue. A new sample from a second batch (Sample 1 Batch H) was run to confirm the results from the second sample of the first batch, see Table 3-8. Examination of the ingot by powder X-ray diffraction revealed that the material had undergone a structural transformation, forming a mixture of or—RbAgZSbS4 and the phase observed in the x-ray powder pattern observed for Sample 2 Batch 1. The thermal analysis of a—RbAgZSbS4 was also investigated by Differential Scanning Calorimetry (DSC). Although the results obtained by the two methods are very similar, the sample containers are not. The sample container for the DSC is a crimped alumina pan while the sample container for the DTA is an evacuated quartz ampoule. The sample container for the DSC allows for oxygen contamination leading to the decomposition of the material, dramatically effecting the results. Upon heating a sample from a third batch of crystals (Sample 1 Batch HI), two sharp endothermic peaks were observed at 330°C and 400°C, respectively. Upon cooling two groups of weak exothermic peaks were observed ca. 347°C and 306°C, see Figure 3-17. Subsequent heating showed a weak exothermic peak at 246°C followed by a single endothermic peak at 407°C. Upon cooling, the weak exothermic peaks had resolved into four peaks at 406 °C, 355°C, 305°C, and 279°C, respectively. A third cycle was attempted but oxygen contamination rendered the results unreliable. Due to oxidation of the sample, a powder pattern was not be run on the residue. 121 A second sample (Sample 2 Batch HI) was examined by DTA in order to evaluate the results obtained by the DSC measurement. Surprisingly, the first three cycles were fairly consistent, showing a single endothermic peak at ca. 405°C upon heating. Upon cooling 8 single exothermic peak at 314°C for the first two cycles and at 307°C for the third cycle, see Figure 3-18. The fourth cycle showed two endothermic peaks at 390°C and 405°C, indicating the formation of a second phase. Upon cooling two exothermic peaks were observed at 333°C and 310°C, respectively. If a second phase formed, the endothermic peak at 390°C was strangely absent in the fifth cycle. A single endothermic peak was observed at 405°C upon heating and two exothermic peaks were observed at 350°C and 305°C, upon cooling. Another sample from a fourth batch of crystals (Sample 1 Batch IV) was examined by DTA. This sample displayed two endothermic peaks at 332°C and 403°C upon heating and a single exothermic peak at 313°C upon cooling, see Figure 3-19. The next three cycles were fairly consistent, see Table 3-8. The spectra reported here consistently display an endothermic peak at ca. 400°C, suggesting a melting point. Variable temperature powder X-ray diffraction experiments confirm that the material was still weakly crystalline at 390°C. Rational attempts to synthesize this unknown phase above 400°C resulted in the formation of a third phase A2Ag20Sb4519 (A = Rb, and Cs). The unknown phase may be a polytype of A2Ag20Sb4819 (A = Rb, and Cs) with a narrow window of thermal stability, see Chapter 4. This could explain the above thermal results where once in the molten state, the high mobility of the Ag+ could cause areas of inhomogenity in the sample, resulting in the formation of an amorphous material from which the phase is produced. At this point, any attempts to propose a mechanism for the observed decomposition product are only speculation. What is 122 amazing is that we never observed the conversion of a—RbAgZSbSrt to B- RbAgZSbS4 in these experiments. In stark contrast to the thermal behavior of a-RbAgZSbS4, the DTA data showed that B-RbAgzsbS4 melted with decomposition at 400°C, forming a mixture of (H) and Ag3SbS3. Attempts to index the powder pattern of the residue with or- RbAgzsbS4 were inconclusive, due to the overlap of a-RbAgZSbS4 with [3- RbAgZSbS4, and Ag3SbS3, 123 Table 3-8. Summary of the thermal behavior of a-RbAngbS4 Batch I Sample cycle endothennic exothennic Figure heating (°C) cooling (°C) 1 1 396 -NA- 3-14A 2 306a,392 -NA- 3-14B 3-6 3063, 390 -NA- 3-14C 2 l 326, 400, 430 364, 267 3—15A 326, 400, 430 364 3-15B 3 246a, 3063 400, 430 3-16A , 357 4 246a 3-16B 400, 430 357 Batch 11 Sample cycle endothermic exothermic heating (0C) cooling (°C) 1 l 326, 400 ,430 364, 267 a exothermic peak upon heating. 124 Batch 111 Sample cycle endothermic exothermic Figure heating (°C) cooling (°C) 1b lb 330 404 347 306 3-17A 2b 246 a 407 406, 359, 305, 3-17B 279 2 l 405 314 3-18A 2 405 314 3 403 307 4 390, 405 333, 310 5 405 307 6 405 350, 305 3-18B Batch IV Sample cycle endothermic exothermic Figure heating (0C) cooling (°C) 1 l 332, 403 313 3-19A 2 405 334 3-19B 3 403 306 3-19C 4 405 302 a exothermic peak upon heating. b examined by DSC. 125 Table 3-9. Infrared and Raman Data for (I) and (II). oc-RbAngbS4 B-RbAngbS4 IR Raman IR Raman 547(m) 509(m) 435(m) 392(w, sh) 396(w) 386(m) 384 (m) 385(s) 380 (w) 361(s) 361(s) 332(m) 327(m) 304(w) 300(m) 260(b, sh) 279(m) 253(m) 246(m) 250(b, sh) 247(m) 227(m) 227(w) 203(m) 170(m) 174(w) 176(m) 174(b) 146(w) 150(m) 126 (A) 3.5 - — 2 - ot - RbAg2Sbs4 - 1.5 _ lag = 2.23eV _ Absorption (Arb. Units) 0.5‘ ' I 1.8 2 2.2 2.4 2.6 2.8 Energy (eV) (B) 3.5 r l 2.5 - B- RbAgZSbS4 E3 = 1.96 eV Absorption (Arb. Units) 2.5 2 Energy (eV) Figure 3-10: Single-crystal optical transmission spectra of Ot-RbAgZSbS4 (I) and B-RbAngbS4 (H) 127 (A) or - RbAgZSbS 4 5 l 1 T l l l 5 _ — -y=-160.08 +71.55x F1: 0.90735 ' _ _ 3 O .E I 3 4 l— I .4 a 40’ cu _ I _ a 3 I ' ,._. I a) I 3 2 2 - I O “ v ’1 . o 1 _ . . o .. O 9 .l.‘ 0 J l 1 1 1 1 2.23 2.24 2.25 2.26 2.27 2.28 2.29 2.3 Energy (eV) (B) a - RbAgZSbS4 1 .6 l l r l r T _ _ . y = -24.703 + 11.401x F1: 0.97968 . 1.5- . ’- 1.4 F ’/ - 1.3 — l - 4 1.2 L I, — 1,1_ I. 1h I — 0.9 — . 0’ 0.8 L4 1 l 1 l l 2.23 2.24 2.25 2.26 2.27 2.28 2.29 2.3 (Abs.)”2 arb. units Energy (eV) Figure 3-11: Absorption edge of or-RbAgZSbS4 (I) as a function of energy: (A) (abs) dependence (direct gap) and (B) \labs dependence (indirect gap). (A) (B) (Abs.)“2 arb. units Figure 3-12: (Abs.)2 arb. units 128 B - RbAgZSbS4 9 I l r — — - y = -43.605 + 22.592x h: 0.94391 8 - a 7 F q 6 r- ‘ :2 5 — ' - 4 ’1 f ” . C 3 _ I o _ I.’o ’ . 2 .. (o _ r°'>' 1 4 l I I I 1.95 2 2.05 2.1 2.15 2.2 Energy (eV) [3 - RbAgZSbS 4 -8 T I 1 l -- — - y = -3.299 + 2.2205x R: 0.98616 .7L - .6 - .; I .5 .. ’0’ - 4 ” u - . .1 I.’. .3 - I ’o - C {O .2 - ‘4' 1 1 . if I L l 1.95 2 2.05 2.1 2.15 2.2 Energy (eV) Absorption edge of B-RbAgZSbS4 (I) as a function of energy: (A) (abs) dependence (direct gap) and (B) Vabs dependence (indirect gap). (A) Raman Intensity (arb. units) (13) Raman Intensity (arb. units) Figure 3-13: 129 100 2 t r l t I 1 I— 1 t I r C 361 . 1.5 - _ C -1 1 — - . 384 260 - ; 247 Z 0-5 f 392 / j . 332 . . 174 . o l l l l I l L L l l j A- 600 500 400 300 1 200 Wavenumbers (cm’ ) 25 ”fr 1 T I I I t I I I T r I l I 1 I I [fl U T q 20 L 361 L )- .1 .5 e -: t : . 380 . 1° : 1 : 227 g 5 -_ 396 / J : Ef/J 251 174 3 o t -1 . _5 L l L J J LL L l L1 1 1 l l l l 1 l I 4 d 600 500 400 300 200 100 Wavenumbers (cm'1 ) The Raman spectra of polycrystalline samples of (A) (x-RbAgzsbS4 (I) and (B) B-RbAgzsbS4 (II) 130 5.0 I u l i 1 exo start / 0.0-)- ~ 3‘ 3 0 g -50 ~ « 8 \ 0: end l 10 >- 396 -L endo -1S .L l l l % 0 100 200 300 400 500 Temperature (°C) exo 50 41 f I l 306 I 0.01- . > a I 0 start (I) g -5.0« . Q. (I) 0 a: l .‘o-i— JI- endo -1s 1 t t 1 100 200 300 400 500 Temperature (°C) 10 1 1 . 1 T exo 3‘ 3 0.0-— -- 0 0 C O 3 0 50 db .11- n: endo 15 l l l i + 0 100 200 300 400 500 Temperature (°C) Figure 3- 14: (A) DTA diagram for Ot-RbAgZSbS4 (first cycle) (B) Second DTA cycle of Ot-RbAgZSbS4 , showing the exothermic peak upon heating. (C) Sixth DTA cycle of a-RbAgzsbS‘t 131 0 C l l r l i 6X0 400 I ~2.0-- start—> ~~ 326 > -4,o.. l .. E; 267 § -6 0 b 364 8 x a .2 \ -e.o-- end 430 -+ l -10-- 4L endo -12 l l ‘r 5 . 0 100 200 300 400 500 600 Temperature (° C) (B) exo 5.0 t t + i J. 400 000- ‘- start—> > £1 g -5.0-- -- 8 to 0 I l -10.. .. endo '15 i 1 I I 0 100 200 300 400 500 600 Temperature (° C) Figure 3- 15: (A) DTA diagram for a-RbAg28b84 (first cycle). (B) Second DTA cycle of a-RbAgZSbS4, showing two phases. 132 (A) 5.0 db db '0' -- -5.0-- Response (pV)_—"‘D’ % -15« endo l 0 100 200 300 400 500 Temperature (°C) 1 (B) 5.0’ ”Ti q. «(I- q- 246 01)" .51).”. Response (uv)—_-'> g -104- 4511- endo -20 1 1 1 1 r 0 100 200 300 400 500 Temperature (° C) Figure 3-16: (A) Third DTA cycle of or-RbAgZSbS4. (B) Fourth DTA cycle of or-RbAgZSbS4. 133 2.0 1 l l i exo E E, O m - C o O. U) 0 a: endo O 100 200 300 400 500 Temp (°C) 2.0 i : ; 4 exo 246 T 1.0"! 'stan ._._) .1. g? 0.01L § § -1.o~~ 8 (I) 0 ‘1 -2.o-b i -3.0-- endo -4.0 l I r 0 100 200 300 400 500 Figure 3-17: (A) DSC diagram for or-RbAgzsbS4 (first cycle). (B) Second DSC cycle of a-RbAngbS4, 134 (A) 20 -b 8X0 15* Response (11 v)—————-i> 51}- endo o.c ' 1 1 1 1 O 100 200 300 400 500 600 Temperature (°C) (B) 14 12.1- Response (11v)——i> g endo l l 1 l l I O 100 200 300 400 500 600 Temperature (°C) Figure 3-18: (A) DTA diagram for ut-RbAngbS4 (first cycle) (B) Sixth DTA cycle of OL-RbAgZSbS4. 135 (A) <— Response 01v) —-> g endo O. 1 41 1 1 1 0 100 200 300 400 500 600 Temperature (°C) (B) 20m- 15"? 101* Rasponse (pv) —-> g endo 0.0 1 1 100 200 300 400 500 600 Temperature (°C) (C) 20 101% 50.1.- <— Response (11v) —> g endo 0.0 1 1 1 1 100 200 300 400 500 600 Temperature (°C) Figure 3-19: (A) DTA diagram for (tt-RbAgZSbS4 (first cycle). (B) Second DTA cycle of Ot-RbAngbS4. (C) Third DTA cycle of Ot-RbAgZSbS4. 136 4. Conclusions The great structural diversity displayed in these [Angb4](3'x)“' frameworks is fascinating and is due to the large number of binding modes available to the [SbS4]3- group and the tetrahedral AgS4 units. The thermal behavior of Ot- RbAngbS4 was unanticipated, and calls for further exploration of this property. The isolation of Gt, [3 -RbAg2SbS4 and the decomposition product(s) of Ot- RbAgZSbS4 suggests that there may be several phases in the Rb/Ag/Sb/S system which may yet be discovered. 10. 11. 12. 13. 137 List of References (a) Sutorik, A.; Kanatzidis, M. G. Progr. Inorg. Chem, 1995, 43, 151. 0)) Kanatzidis, M. G. Curr. Opinion Solid State and Mater. Sci., 1997, 2, 139. McCarthy, T. M.; Kanatzidis, M. G. Inorg. Chem, 1994, 33, 1205. Hanko, J. A.; Kanatzidis, M. G. Abstract #0413 from 210th Fall ACS Meeting, Chicago IL 1995. (a) Schimek, G. L.; Pennigton, T. L.; Wood, P. T.; Kolis, J. W. J. Solid State Chem, 1996, 123, 277. (b) Wood, P. T.; Schimek, G. L.; Kolis, J. W. Chem. Mater., 1996, 8, 721. Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G. Inorg. Chem, 1997, 34, 1234. Hanko, J. A.; Kanatzidis, M. G. manuscript in preparation. Hanko, J. A.; Kanatzidis, M. G. J. Alloys Comp., Accepted for publication. Hanko, J. A.; Kanatzidis, M. G. J. Chem. Soc. Chem. Commun., 1998, 724. Imafuku, M.; Nakai, I.; Nagashima, K. Mat. Res. Bull., 1986, 21 , 493. CERIUSZ, Version 1.6, Molecular Simulations Inc., Cambridge, England, 1994. SMART and SAINT. Data collection and Processing Siftware for the Smart CCD system. Siemens Analutical X-ray Instruments Inc., 1995. Blessing , R. H. Acta Crystallogr., 1995, A51, 33. SHELXTL Version 5, Reference Manual. Siemens Industrial Automation Inc., 1994. 14. 15. l6. 17. 18. 19. 20. 21. 22. 138 (a) Sheldrick, G. M., in Crystallographic Computing 3 ; Sheldrick, G. M., Kruger, C., Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985, p. 175. (b) Gilmore G. J., Appl. Cryst. ; 1984, I7, 42. Walker, N.; Stuart, D. Acta Cryst, 1983, A39, 158. Honle, W.; Wibbelmann C.; Brockner, W. Z. Naturforsch. 1984, 39b, 1088. (a) Teske, C.L.; Z. Naturforsch, B, 1979, 34, 1979. (b) Auemhammer, M.; Effenberger, H.; Irran, E,; Pertlik, F.; Rosenstingl, J. Solid State Chem, 1993, 106, 421. Jeffery, G.A.; Vlasse, M. Inogr. Chem, 1967, 6, 396. Hanko, J. A.; Sayettat, J.; Jobic, S.; Brec, R.; Kanatzidis, M. G. Submitted for publication. Siegert, H. Z. Anorg. Allg. Chem, 1954, 275, 225. Chondroudis, K; Hanko, J. A.; Kanatzidis, M. G. Inorg. Chem, 1997 37 , 2323. (a) Dhingra, S.-D.; Kanatzidis, M. G. Science, 1992, 258, 1769. (b) Marking, G. A.; Hanko, J. A.; Kanatzidis, M. G. Chem. Mater. 1998, 10, 1191. CHAPTER 4 Synthesis and Characterization of the Novel Mixed Valent Quaternary Silver-Rich Thioantimonates AzAgZOSbAIS19 (A = Rb, Cs) and B- szAgzoSb4Slg. A Thioantimonate with a Large Supercell. 139 140 1. Introduction In previous chapters, we have demonstrated that the molten polychalcoantimonate flux method is conducive to the synthesis of new multinary chalcoantimonate compoundsl‘ 9. The polychalcoantimonate fluxes form by the in situ fusion of AZQ/Sb/Q (A = Na, K, Rb, Cs; Q = S, Se) and contain various [belen- units in a molten polychalcogenide solvent. Solid state frameworks can be generated by the self-assembly of these various [belen- units with dissolved metal ions, forming extended lattices stabilized by alkali cations. Examples include: 101131323116? A2AngS4 (A = K, Rb, Cs)3'4, KAgZSbS4,4 a- 3,4 5 6 RbAgZSbS4,4’3 Cs3AgZSb3Q8 (Q = S, Se) KHngS3 , Rb2Au68b4Slo, and AzAquS4 (A = Rb, Cs)7. More recently, the rare earth chalcoantimonates K2La2-be4+xSe12 (Ln = La, Ce, Pr, and Gd)8 and K2La28b2Q9 (Ln = La, Q = S; Ln = Gd, Q = Se)9 have been reported. Although the various chalcoantimonate [belen- units are interesting because of the many different binding modes they can exhibit, relatively little is known about their coordination chemistry in the solid state. In particular, it is necessary to elucidate the correlation between Lewis basicity and the [beQy]“‘ (Q = S, Se) units stabilized from the chalcoantimonate flux. As experimentally observed, in the chalcoantimonate system Lewis basic conditions favor the tetrahedral [SbQ4]3' units (Q = S, Se), while Lewis acidic conditions produce an even more complicated systems by the greater stability of the Sb3+ species, forming higher nuclearity [beQy]"' units in the resulting frameworks.3’619 While the Chalcophosphate system the [PQ4]3‘ units (Q = S, Se) units are stabilized under Lewis basic conditions, Lewis acidic fluxes produce chemically divergent results between the selenophosphatelo and thiophosphate systems.11 Continuing our investigations of the coordination chemistry of the [beQyP‘ units with the coinage metals, we report the synthesis, structural characterization, and physical properties of two isostructural quaternary silver thioantimonate compounds, AzAgZOSb4Slg (A = Rb, Cs) (I)—(II) and the related high-symmetry compound B-szAgZOSb4819 (III). The 141 structures of (I)-(II), and the related compound (IH), are layered, featuring discrete tetrahedral [Sva4]3- and pyramidal [SbmS3]3- units aggregated to opposite sides of [AgIOszsg] "sub-layers" for (I)—(III). In both cases, the different "sub-layers" are connected together in a centrosymmetric fashion with sulfur atoms forming a “sandwiched” double layer. Due to their high silver concentration, these compounds may display interesting Ag+ ion mobility. As previously described for several silver systems such as Ange,12 y-AggGeTe6,13 and Ag2T12P2811,14 Ag+ ion mobility may be evaluated by examining the anharmonic thermal parameters of their sites. 2. Experimental Section 2.1. Reagents Chemicals were used as obtained: (i) antimony powder 99.999% purity, -200 mesh, Cerac Inc., Milwaukee, WI; (ii) silver powder 99.95% purity, -325 mesh Alfa AESAR Group, Seabrook, NH; (iii) cesium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (iv) rubidium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (v) sulfur powder, sublimed, J.T. Baker Chemical Co., Phillipsburg, NJ; (vi) N ,N-dimethylformamide (DMF) reagent grade, EM Science, Inc., Gibbstown, NJ. (vii) diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. 2.2. Syntheses. A28 (A = Rb, Cs) were prepared by reacting stoichiometric amounts of the elements in liquid ammonia as described in Chapter 2 Section 2.2. 142 Preparation of szAgzoSb4819 (I). An amount of RbZS (0.051g, 0.25 mole), Ag (0.054g, 0.50 mmole), Sb (0.016g, 0.125 mmole), and S (0.080g, 2 mole) were thoroughly mixed and transferred to a 6—ml Pyrex tube which was subsequently flame-sealed in vacuo (~10'3 Torr). The reaction mixture was heated to 500°C over 15 hrs in a computer-controlled furnace. It was isothermed at 500°C for 4 days, followed by cooling to 100°C at a rate of 4°C/hr and then to room temperature in 1 hour. The product, which is air and water stable, was isolated by dissolving the RbZSx and any be[SbySz] flux with DMF under inert atmosphere to give black crystals in 45% yield, based on Sb. Quantitative microprobe analysis of several single crystals gave RbAg6Sb1286. Preparation of Cs 2AgzoSb4S 19 (II). An amount of C328 (0.079g, 0.25 mole), Ag (0.054g, 0.50 mmole), Sb (0.016g, 0.125 mmole), and S (0.080g, 2 mole) were thoroughly mixed and transferred to a 6-ml Pyrex tube as above and heated as in (I). The product, which is air and water stable, was isolated by removing the excess flux as in (I) to give black crystals in 58% yield, based on Sb. Quantitative microprobe analysis on single crystals gave CsAggsbmSm. Preparation of B-szAgzoSb4819 (111). An amount of Rb28 (0.051 g, 0.25 mole), Ag (0.054g, 0.50 mmole), Sb (0.016g, 0.125 mole), and S (0.080g 2 mole) were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame-sealed in vacuo (~10'3 Torr). The reaction mixture was ramped to 400°C over 15 hrs in a computer-controlled furnace. It was isothermed at 400°C for 4 days, followed by cooling to 100°C at a rate of 4°C/hr and then to room temperature in 1 hour. The product, which is air and water stable, was isolated by dissolving the RbZSx and any be[SbySz] flux with DMF under inert atmosphere to give black crystals in approximately 10% yield, based on Sb. Quantitative microprobe analysis of single crystals gave RbAg9.7Sb2,1Sg,6. 143 Crystals suitable for single crystal X-ray diffraction studies were produced from the following conditions: an amount of szS (0.102g, 0.50 mmole), Ag (0.108g, 1 mole), Sb (0.032g, 0.25 mole), and S (0.160g, 4 mole) were thoroughly mixed and transferred to a 6-ml Pyrex tube which was subsequently flame-sealed in vacuo (~10'3 Torr). The reaction mixture was heated to 350°C over 15 hrs in a computer-controlled furnace. It was isothermed at 350°C for 4 days, followed by cooling to 100°C at a rate of 4°C/hr and then to room temperature in 1 hour. The product, which is air and water stable, was isolated by dissolving the RbZSx and any be[SbySz] flux with DMF under inert atmosphere to give a mixture of black, orange, and yellow crystals. The black crystals were the major phase, with an estimated yield of 60%, based on Sb. Quantitative microprobe analysis of single crystals gave RbAggsbuSlm. The yellow and orange crystals were determined by powder X-ray diffraction experiments to be szAngS4 and Ot-RbAgZSbS4, respectively, 2.3. Physical Measurements Powder X-ray Dtfiraction. Analyses were performed using a calibrated Rigaku- Denki/RW400F2 (Rotaflex) rotating anode powder diffractometer controlled by an IBM computer, operating at 45 kW 100 mA and with a 10/min scan rate, employing Ni-filtered Cu radiation in a Bragg-Brentano geometry. The theoretical X-ray powder pattern was calculated using the CERIUS2 software package.15 Calculated and observed XRD patterns for (11) are given in Table 4-1. Infrared Spectroscopy. Infrared spectra in the far-IR region (600-50 cm'l) for (I) - (III) were recorded on by a DHK (pc) computer controlled N icolet 750 Magna-IR Series H spectrometer equipped with a TGS/PE detector and a silicon beam splitter with 4 cm'1 144 resolution. The samples were ground with dry CsI into a fine powder and pressed into translucent pellets. Raman Spectroscopy. Raman spectra (1000-100 cm'l) were recorded with a BIO- RAD FT Raman spectrometer with a Spectra-Physics Topaz T10-106c 1,064 nm YAG laser operating at 11 amps. Samples of (I) - (III) were ground with dry CsI into a fine powder and loaded into melting point capillary tubes. Solid State U V/VisflVear IR Spectroscopy. Optical diffuse reflectance measurements were performed at room temperature using a Shimadzu UV-3101PC double beam, double monochromator spectrophotometer. The instrument is equipped with integrating sphere and controlled by a AST Bravo 3/255 computer. BaSO4 was used as a 100% reflectance standard for all materials, compacted into a sample holder. Samples were prepared by grinding them to the fine powder and spreading them on the compacted surface of the powdered standard material. The reflectance versus wavelength data generated can be used to estimate a material's band gap by converting reflectance to absorption data as described elsewhere. “5 Dzflerential Thermal Analysis (DTA). DTA experiments were performed by a AST Bravo/386SX computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~ 20 mg) was ground into a microcrystalline powder and sealed in a quartz ampoule under vacuum. A quartz ampoule of equal mass filled with A1203 was sealed and placed on the reference side of the detector. The sample was heated to the desired temperature at 10 0C/min, then isothermed for 10 minutes and finally cooled to 50 0C at -10 °C/min. The residue of the DT A experiment was examined by X-ray powder diffraction. To evaluate congruent melting the X-ray powder diffraction patterns before and after the DTA experiments were compared. The stability of the sample and the reproducibility of the melting point was monitored by conducting multiple cycles with the same heating profile. Differential Scanning Calorimetry (DSC). DSC experiments were performed on a computer-controlled Shimadzu DSC-50 thermal analyzer under a nitrogen atmosphere at a 145 flow rate of 20 ml/min. The samples (~ 5 mg) of the crystalline material were crimped in an aluminum pan. The pan was placed on the sample (right) side of the DSC-50 detector and a crimped empty aluminum pan of approximately equal mass was placed on the reference (left) side of the detector. A sample of (IH) was heated to the desired temperature at 5 oC/min, isothermed for 5 minutes, and finally cooled at the same rate to 50°C. The temperatures associated with each peak in the spectrum have a standard deviation of 0.2 degrees. The adopted convention in displaying the data is as follows: exothermic peaks are associated with a positive heat flow while endothermic peaks are associated with a negative heat flow. Semiquantitative microprobe analyses. The analyses were performed using a JEOL JSM-6400V scanning electron microscope (SEM) equipped with a TN 5500 EDS detector. Data acquisition was performed with an accelerating voltage of 20kV and a twenty-second accumulation time. Single crystal X-ray Crystallography. Intensity data for CszAgzoSb4819 (II) were collected on a Rigaku AFC6 diffractometer, using (lo/20 scans at 23°C with graphite- monochromatized radiation. The crystals showed no significant intensity decay, as determined by monitoring three standard reflections measured every 150 reflections throughout the data collection. The space group was determined from systematic absences and intensity statistics. An empirical absorption correction based on viscans was applied to the data set during the initial stages of refinement. An empirical DIFABS17 correction was applied as recommended after full isotropic refinement, then full anisotropic refinement was performed. The structure was solved by direct methods using SHELXS-8618 and refined by the full-matrix least-squares techniques of the TEXSAN package of crystallographic programs.19 With the room temperature data collection for (H) and the large number of Ag+ sites in the structure increases the potential of thermal motion, as monitored by the size of the thermal ellipsoids. After full refinement of (11), three of the silver sites (Ag(10), Ag(l3), 146 and Ag(20)) had large thermal ellipsoids suggesting silver disorder on the sites. The final refined occupancies for Ag(lO), Ag(13), and Ag(20) were found to be 80/20%, 48/52%, and 50/50%, respectively. Further investigations into the anharmonic refinement of the Ag+ sites are currently underway”. A starting unit cell for (III) was found using the Rigaku AFC6 diffractometer. A primitive hexagonal cell of 7.523(4)A x 62.95(3)A was determined. To confirm the cell edges, one hour axial photographs were taken along the a and b directions. The photographs revealed a 4a x 4b supercell. The resulting cell, 30.57(2)A x 62.95(3)A, was determined to be too large for data collection on the Rigaku AFC6 diffractometer. Subsequently, the crystals were sent to Dr. Victor Young at the University of Minnesota for data collection on a Siemens CCD diffractometer. The subcell (7.4902(1)A x 62.663(1)A) was confirmed but the correct supercell was found to be 4a x 4b x 2c, producing a rhombohedral cell with dimensions of 29.9673(2)A x 125.6272(6)A. Intensity data for the subcell and the supercell was collected. Intensity data for (111) were collected on a Siemens SMART-CCD diffractometer using graphite-monochromatized MoKOt radiation. The data collections for both the subcell and supercell covered more than a hemisphere of reciprocal space, up to 52° 20. The individual frames were measured with a detector to crystal distance of 5 cm and an omega rotation of 0.3 deg with an acquisition time of 45 seconds per frame, leading to a total measurement time of approximately 24 hours. The SMART software21 was used for the data acquisition, SAINT21 for data extraction and reduction. The space group was determined from systematic absences and intensity statistics. No crystal decay was detected for (III). A correction for Lorentz polarization effects and an empirical absorption correction (SADABS)22 were applied to the data. The structure of (III) was solved using direct methods and refined by full-matrix least squares techniques of the SHELXTL23 package of crystallographic programs utilizing the subcell data. 147 The complete data collection parameters and details of the structure solution and refinement for CszAg208b4519 and B-szAg208b4S19 are given in Table 4-2. The cell parameters for the B-szAg20Sb4819 supercell are also given in Table 4-2. The coordinates of all atoms, average temperature factors, and the estimated standard deviations for (H) and (III) are given in Tables 4-3 and 4-4, respectively. 148 Table 4- 1. Calculated and Observed X-ray Powder Patterns for CszAgzoSb4819 (II). hkl dcalc, A d obsd, A I/Imax(0b8d) 200 21.283 21.267 100 1 1 0 14.142 14.025 23 4 0 0 10.640 10.532 36 6 0 0 7.093 7.084 21 2 02 6.502 6.551 11 0 2 1 6.469 6.461 13 2 21 6.021 6.015 14 4 2 1 5.293 5.306 20 2 T5 _ 5.042 5.068 18 6 0 2 6 2 2 4.337 4.336 19 8 2 1 3 3 1 5 3 0 4.310 4.300 22 1_13 4.012 3.991 18 2 2 3 6 2 2 3.767 0 4 0 _ 8 0 2 1321 3.734 3.729 59 4 2 3 623223 3.550 3.529 41 4 4 0 9 1 3— 9 3 T 3.446 3.449 27 2 0 {10 0 _2_ 3.251 3.250 24 1221242 11 13 3.134 3.147 25 113 T1 14 3.077 3.073 23 14 0 0 3.040 3.033 68 804 4_42 3.010 3.005 83 2 2 4 3 5 0 2.934 5 3 3_“ 3 1 2.845 2.840 24 1004224 1042 424 534— 2.705 2.706 22 1 3 4 16 0 0 2.657 2.650 40 1204 16021421 1132124i 751 _ 8041242 1312 425‘ 225— 10 4 2 1802244 1621 1800 11001442 206463 20021912 1513 825 2521442 1842 22001262 2002 1244 1733 2.649 2.600 2.586 2.521 2.460 2.376 2.364 2.337 2.167 2.149 2.147 2.061 2.166 2.008 1.932 1.907 1.877 1.850 149 2.644 2.60 2.584 2.518 2.459 2.375 2.357 2.332 2.160 2.150 2.146 2.063 2.156 2.008 1.929 1.903 1.874 1.847 61 21 20 33 3O 29 36 24 24 18 28 18 19 26 20 35 25 150 Table 4-2. Crystallographic Data for CSzAg203b4319 and B-szAgzoSb4819, Formula C82A8203b4319 B-Rb2A8203b4319 B-Rb2A8203b4319 subcell data supercell data FW 3424.48 3424.48 300 447.63 a, A 43.917(3) 7.4902 (1) 29.967(4) b, A 14.995(2) 7.4902 (1) 29.967(4) c, A 14.968(3) 62.663 (1) 125.62(2) 61 (deg) 90.00 90.00 90.00 [3 (deg) 91.84 (9) 90.00 90.00 7 (deg) 90.00 120.00 120.00 2; V(A3) 24; 10952(8) 3; 3044.60(9) 60; 97 701(1) 71 (Mo Ka). A 0.71069 0.71069 0.71069 Space group C2/c (#15) 113—(#148) P3711147) Deare. g/cm3 5.168 5.603 5.106 p, cm-1 152.84 15.37 13.38 Temp (0C) 23 23 23 Final R/Rw,a % 6.8/8.6 NA NA R1/wR2b % NA 133/25.8 NA TotalData Measured 1722 5976 352 001 Total Unique Data 1545 1221 64 634 Data F02>36(F02) 927 NA NA Data F02>26(F02) NA 1219 44 800 No. of Variables 126 73 Crystal Dimen., mm 0.49 x 0.42 x 0.25 0.39 x 0.28 x 0.17 0.10 x 0.10 x 0.10 3R = 201701 - wen/21m, Rw = {ZWUFOI - IFCI)2/2wlF012}1/2. b R1 = 2(IF012 - chl2)/2|Fol, W112 = {2w(lF0l2 - chl2)2/ZWIF012}1/2. 151 Table 4-3. Fractional Atomic Coordinates and B(eq) Values for CszAg208b4S19 (II) with Estimated Standard Deviations in Parentheses. Atom x y z B(ecL a Cs( 1) 0.4798(1) 0.6190(4) 0.0699(40) 2.9(3) Cs(2) 0.5201(1) 0.8716(4) -0.0722(0) 2.6(3) Sb( 1) 0.4343(1) 0.8755(4) 0.2045(30) 0.9(2) Sb(2) 0.2749(1) 0.8730(4) 0.4279(40) 1.4(2) Sb(3) 0.4323(1) 0.3737(4) 0.2024(30) 0.9(2) Sb(4) 0.2270(1) 0.8629(4) 0.0637(40) 1.4(2) Ag( 1) 0.2150(2) 1.0092(5) 0.2910(50) 3.4(4) Ag(2) 0.3609(2) 0.8827(6) 0.3442(50) 3.8(3) Ag(3) 0.4038(2) 1.1263(5) 0.2130(50) 3.6(3) Ag(4) 0.4082(2) 0.7298(5) 0.4169(50) 3.6(4) Ag(S) 0.2806(2) 0.6220(5) 0.4038(40) 2.9(3) Ag(6) 0.4149(2) 0.7618(5) -0.0619(0)) 3.7(4) Ag(7) 0.41 16(2) 0.6296(5) 0.2154(50) 3.5(3) Ag(8) 0.3087(2) 1.1276(5) 0.395 1 (40) 3.5(3) Ag(9) 0.3608(2) 0.9927(5) 0.0240(70) 4.8(4) Ag(10) 0.3513(3) 0.5366(7) 0.163(10) 4.0(4) Ag(lO') 0.361(1) 0.537(2) 0.124(30) 1(1) Ag( 1 1) 0.3603(2) 0.7674(5) 0.5439(60) 4.0(4) Ag(12) 0.3356(2) 0.0832(6) 0.1836(50) 4.4(4) Ag(13) 0.3615(4) 0.648(1) -0.154(1) 4.2(4) Ag(13') 0.3570(4) 0.700(1) -0. 179(1) 4.5(4) Ag(l4) 0.2750(2) 1.0155(5) 0.2017(50) 3.0(3) Ag( 15) 0.2823(2) 0.2285(5) 0.1872(50) 3.1(3) Ag(16) 0.4095(2) 0.4768(5) -0.0801(6) 3.9(4) Ag(17) 0.1898(2) 1.21 10(5) 0.2664(50) 4.0(4) Ag(18) 0.3580(2) 0.7224(5) 0.0575(60) 3.7(4) Ag( 19) 0.4129(2) 0.9759(5) -0.0819(6) 3.4(4) Ag(20) 0.3409(6) 0.461(1) -0.068(2) 6.6(5) Ag(20') 0.3458(6) 0.508(2) -0.082(2) 7.0(5) S(l) 0.3035(5) 0.876(2) 0.288(10) 2(1) S(2) 0.3020(6) 0.994(1) 0.536( 10) 2(1) 152 S(3) 0.4505(5) 0.499(1) 0.301 (20) 1( 1) S(4) 0.4511(4) 0.372(1) 0.048( 10) 1.1(8) S(5) 0.3761(5) 0.373(1) 0.162(10) 1.3(9) S(6) 0.4495(6) 0.245(1) 0.294(10) 2(1) S(7) 0.4002(5) 1.122(1) 0.012(10) 2(1) S(8) 0.1999(5) 0.992(1) -0.028( 1) 0.9(9) S(9) 0.4517(5) 1.007(1) 0.294(20) 2( 1) S( 10) 0.4526(6) 0.751(1) 0.303(20) 2(1) S(l 1) 0.4540(5) 0.874(1) 0.052(10) 1.6(9) S(12) 0.3793(5) 0.871(2) 0.168(10) 2(1) S(l3) 0.3624(4) 0.367(1) -0. 181(1) 1.3(8) S(14) 0.3028(5) 0.368(1) 0.297( 10) 2(1)0 S(lS) 0.3633(5) 0.131(2) 0.348(10) 2.2(9) S(l6) 0.2499(5) 1.139(1) 0.283(10) 3(1) S(l7) 0.3984(5) 0.381(1) 0.517(10) 2.0(9) S( l 8) 0.3025(6) 0.750(1) 0.524(20) 2(1) S(19) 0.1980(7) 0.745(1) -0.037(2) 3( 1) ‘1 B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as Beq = (4/3)[azB(1, 1) + b2B(2, 2) + c2B(3, 3) + ab(cosy)B(1,2) + ac(cosB)B( 1,3) + bc(cosa)B(2, 3)] 153 Table 4-4. Atomic coordinates ( x 104) and equivalent isotropic displacement parameters (A2 x 103) for B-szAg203b4S19 (111) with Estimated Standard Deviations in Parentheses. x y z U(eq) Occ. Sb(l) 3333 6667 1244(1) 25(1) 1 Sb(2) 6667 3333 162(1) 36(2) 1 Ag(l) 5263(19) 4705(19) 748(1) 49(4) 0.47(3) Ag( 1') 6160(42) 5818(47) 739(3) 70(9) 024(2) Ag(2) 1430(1 l) 2864(1 1) 206(1) 62(3) 064(2) Ag(2') -282(35) 5930(46) 738(3) 72(9) 0.23(2) Ag(3) 8648(12) 7296(15) 1062(3) 55(3) 0.64(6) Ag(3') -1456(21) 7092(24) 1 132(5) 42(6) 0.32(6) Ag(4) 1 1094( 18) 2187(19) 401(2) 51(5) 0.32(2) Rb(l) 6667 3333 1535(2) 63(3) 1 8(1) 3333 6667 864(2) 33(4) 1 S(2) 1642(16) 3282(15) 1365(2) 34(2) 1 S(3) 5040(15) 4945(15) 356(2) 36(2) 1 S(4) 6667 3333 1014(3) 35(4) 1 S(S) 0 0 753(3) 62(6) 1 8(6) 0 0 0 109(16) 1 U(eq) is defined as one third of the trace of the orthogonalized Uij tensor. 154 3. Results and Discussion 3.1 Description of Structures Structure of A2A gzoSb4S 19 (A = Rb, Cs) (1) - (11): Since A2Ag208b4819 (A = Rb, Cs) are isostructural, the single-crystal structure determination was only conducted on CszAgzoSb4819; therefore, this discussion will refer primarily to this compound. The structure of CszAgzoSb4819 is strikingly complex, with [AgzoSb4S19]2“' layers separated by Cs” cations, see Figure 4~1. The [Ag208b4819]2"' layers are actually composed of two [AgloszSg] "sub-layers" connected by sulfur atoms. The [AgloSb289] "sub-layers" are in turn composed of three different interconnecting sheets: two ternary Ag3Sb83 and Ag3SbS4 layers segregated to opposite sides of the binary Agzs sheet. As a result (CszS)(Ag28)4(Ag3SbS3)2(Ag3SbS4)2 is a more descriptive formula. A unique feature of this compound is the presence of both Sb+5 and Sb"3 species segregated to opposite sides of the [AgloSb2S9] "sub-layers". Although CS3Agzsb3Q3 (Q = S4, Se3) was the first reported alkali chalcoantimonate compound to contain both Sb3“5 and Sb+3 species, the Sbf3 species was observed as an oligomeric [Sb2Q4lnzn' (Q = S, Se) chain, not as discrete pyramidal [SbS3]3' units. The Ag3SbS4 sheet is comprised of two tetrahedral [SbS4]3' units exhibiting a complex bonding motif with Ag+ ions, see Figure 4-2A. The three basal sulfides are doubly bridging to Ag+ ions while the fourth apical sulfide binds to three Ag+ ions. The Ag+ ions are bridged by the basal sulfides of neighboring [SbS4]3' units, forming two different Ag3S3 rings linking the Sb+5 species together. The two different Ag3S3 rings resemble a cyclohexane ring in a chair configuration. Figure 4-3A shows a view parallel to the [AgloSb283] “sub-layer” and a perpendicular view of the Ag3SbS4 sheet, highlighting the extended connectivity shown in Figure 4-3B. A third view (Figure 4-3C) shows the 155 connectivity of the Ag+ atoms bonded to the apical sulfide, connecting the Ag3SbS4 to the AgZS sheet. The two tetrahedral [SbS4]3' units are regular tetrahedrons with Sb-S bond distances in the range from 230(2) to 233(2) and 232(2) A to 238(3) A for Sb(l) and Sb(3), respectively. The S-Sb-S angles range from 107.0(9)° to 111.5(8)°. and 109.2(8) to 110.2(8) for Sb(l) and Sb(3), respectively. The sheet containing the two pyramidal [SbS3]3' units displays an even more complex bonding mode with Ag+ ions, see Figure 4-2B. Two sulfides from the [SbS3]3‘ units bond to three Ag+ ions while the third sulfide bonds to four Ag+ ions in a rare distorted square pyramidal coordination. Although rare, this coordination is observed in the structure of CS3Ag28b3Sg.3 Figure 4-4A shows a view parallel to the Ag3SbS3 sheet in the [Angbzsg] “sub-layer” and a perpendicular view (Figure 4-4B), highlighting the extended connectivity. The complex Ag-S bonding arrangement forms highly distorted Ag3S3 rings, connecting the Sb+3 species together by bridging neighboring pyramidal [SbS3]3' units. As shown in Figure 4-4C, the darker shaded Ag+ ions bridge the ternary Ag3SbS3 layer to the binary Ag2S sheet. The pyramidal [SbS3]3' units are highly distorted, with Sb—S bond distances in the range from 2.41(2)A to 2.45(2)A and 2.37(2)A to 2.45(2) for Sb(2) and Sb(4), respectively. The S-Sb-S angles range from 97.7(7)0 to 99.4(7)°. and 93.3(7) to 99.8(8) for Sb(2) and Sb(4), respectively. The union of the three sheets into the [Angszg] "sub-layer" involves a very complicated bonding scheme that is best described graphically. As shown in Figure 4-5, the two ternary Ag3SbS3 and Ag3SbS4 and the binary Agzs “belt” combine together to assemble the [AgloSb289] "sub-layer" which are in turn connected by bridging monosulfide ions, forming [AgzoSb4819]2‘ double layers. The parallel views of the ternary Ag3SbS4 and Ag3Sb83 sheets (Figures 4-3C, and Figure 4-4C) highlight the Ag+ ions used to assemble the "sub-layer". As the ternary Ag3SbS4 sheet is brought next to the binary AgZS “belt” the apical sulfide (8(5) and S(l2)) of the tetrahedral [SbS4]3' unit binds to three Ag+ ions from the AgZS “belt”, connecting the layers together. The connection of 156 the ternary Ag3SbS3 sheet to the binary AgZS “belt” is slightly more complicated due to the participation of all three sulfides of the pyramidal [SbS3]3’ unit in bonding to the Agzs “belt”. As shown in the fragment of the Ag3SbS3 sheet (Figure 4-5), two of the three sulfides (3(1) and S(18) for Sb(2) and S( 14) and S(l9) for Sb(6)) from the pyramidal [SbS3]3' unit binds to one Ag+ while the third sulfide (S(2) and S(8)) binds to two Ag+ ions, and these four Ag+ ions connect the sheets together. The resulting [Ag10Sb2S9] "sub-layers" are further connected together by bridging sulfides in a centrosymmetric fashion, forming the [AgzoSb4819]2' double layers. The coordination environments around the Ag+ ions can be assembled into three groups; trigonal planar, tetrahedral, and linear. The geometries around Ag(l), Ag(2), Ag(S), Ag(9), Ag(lO), Ag(l 1), Ag(13), Ag(l4), Ag(15), and Ag(18) are distorted trigonal planar with Ag-S bond distances in the range from 2.45(3)A to 2.80(3)A. The geometries around Ag(3), Ag(4), Ag(6), Ag(7), Ag(8), Ag(16), Ag(17) , and Ag(19) are tetrahedral. The geometries around six of these tetrahedral Ag+ ions are highly distorted with two normal Ag-S distances3v4 at 2.62(3)A and 2.58(3)A, and two Ag-S bond distances that range from 2.77(2)A to 2.75(3)A. The S-Ag-S angles for these six Ag+ ions range from 91.0(8)O to l20.7(8)°. The environment about the remaining two tetrahedral Ag+ ions is even more highly distorted with very long Ag-S contacts at 2.82(3)A and three normal Ag- S distances3i4 that range from 2.71(3)A to 2.54(2)A. The S-Ag-S angles reflect this highly distorted nature in the range from 91 .5(8)° to 139.5(8)°. The geometry around Ag(12) and Ag(20) are best described as "linear" with slightly compressed Ag-S distances that range from 2.38(2)A to 2.54(2)A, and S-Ag-S angles that range from 139(1)o to l65.6(8)°. Selected bond distances and angles are given in Tables 4-5 and 4-6, respectively. Structure of B-szAgzoSb4819 (III): The structure of B'szAgZOSb4819 is related to (I) and (H), with [Ag208b4819]2"' layers separated by Rb“ cations, see Figure 4- 6. The [AgzoSb4819]2“’ layers are similar to the layers observed in the structure of (1)-(II). 157 The difference between the structures is that the layers in (1)-(II) stack in registry, while in (III) they are shifted with respect to one another, giving rise to the hexagonal packing arrangement. A more descriptive formula for B-szAg208b4819 is the same that can be written for (1)-(II), (Rb2S)(Ag28)4(Ag3SbS3)2(Ag3SbS4)2, it represents a new member of a class of compounds with the general formula (A28)w(AgZS)x(Ag3SbS3)y(Ag3SbS4)z. The [SbS4]3' unit is regular tetrahedron with Sb-S bond distances in the range from 2.321(9) A to 239(2) A and S-Sb-S angles in the range from 109.0(3)o to 110.0(2)o. The sheet containing the two pyramidal [SbS3]3' unit is fairly regular, with a single Sb-S bond distance of 2.42(1)A, and a single S-Sb-S angle of 92.2(3)°. Although the structure of (III) can also be assembled as the combination of two ternary Ag3SbS3 (Figure 4-7) and Ag3SbS4 (Figure 4-8) and a binary AgZS sheet, it is difficult to accurately describe the connections due to the highly disordered nature of the Ag+ sites. Although the refinement is still in progress, the basic [AngbySzP' layer found in (H1) is similar to (I) - (II) but with partial occupancy observed in all Ag+ sites. This partial occupancy is attributed to the folding of the 4a x 4b x 2c supercell in to the subcell. The occupancies of each Ag+ site were refined independently and are given in Table 4-4. Attempts to achieve charge neutrality for the model resulted in silver deficient layers; assuming the empirical formula for (H1) is Rb(),33Ag3,33Sbo,67S3,167. Attempts to establish charge neutrality by decreasing the sulfur concentration or locating the missing Ag+ density were unsuccessful. Although none of the sulfur sites appeared to be disordered, there was one problematic position at the origin. Extensive attempts to model the site with disorder or substitution of the site with Ag+ gave significantly higher residuals. The site is octahedrally coordinated an unusual environment for sulfur. The origin of the supercell may be attributed to slight movement of this sulfur atom from cell to cell. During the course of refinement, the Sb and 158 Rb sites do not appear to be disordered. The final R1 for the isotropically refined structure at this stage is 13.26%. The Ag+ ions adopt two different coordination environments, trigonal planar and tetrahedral. The geometry around Ag(l), Ag(l'), Ag(2), and Ag(3') is distorted trigonal planar with Ag-S bond distances in the range from 2.45(2)A to 2.70(3)A and S-Ag-S angles from 98.2(8)° to 139.8(5)°. The geometry around Ag(2'), Ag(3), and Ag(4) is best described as tetrahedral with Ag-S bond distances that range from 2.53(2)A to 2.88(3)A. The S-Ag-S angles for these 3 Ag+ ions range from 84.9(6)° to 125.3(9)°. Selected bond distances and angles are given in Tables 4-7 and 4—8, respectively. An attempt was made to solve the supercell (see Table 4-2) with little success. The immense size of the supercell makes its solution a very interesting challenge. As shown in Table 4-2 over 350,000 reflections were collected, resulting in a useable data set of well over 40,000 reflections. The current subcell refinement was sent to Dr. Michel Evain, at the Institut des Matériaux de Nantes in France, for further examination. Preliminary examination of the subcell refinement suggests that (111) may be an ion conductor and that a possible diffusion path has been observed.20 159 Selected Distances (A) for CSzAgzoSb4819 (II) with standard deviations in Table 4—5. parentheses.a Sb( 1 )-S(9) 232(3) Ag(10)-S(5) 265(3) Sb(l)-S(10) 230(2) Ag(10)-S(8) 258(3) Sb(1)-S(l 1) 232(2) Ag(10)-S(l3) 249(3) Sb(l)-S( 12) 233(2) Sb(l)-S(mean) 231(5) Ag( 10')-Ag(18) 291(4) Ag(lO')-Ag(20) 267(4) Sb(2)-8(1) 241(2) Ag( 10')-Ag(20') 267(5) Sb(2)-S(2) 245(2) Sb(2)-S(18) 241(2) Ag(lO')-8(5) 253(4) Sb(2)-S(mean) 242(3) Ag(lO')-S(8) 273(5) Sb(3)-S(3) 233(2) Ag( 10')-S(17) 260(5) Sb(3)-S(4) 233(2) Sb(3)-S(5) 238(2) Ag(l 1)-Ag(12) 3.20(1) Sb(3)-S(6) 231(2) Sb(3)-S(mean) 233(4) Ag(l 1)-S(5) 265(2) Ag( 1 1)-S(7) 248(2) Sb(4)-S(8) 2.45(2) Ag(11)-S(18) 249(3) Sb(4)-S (14) 244(2) Sb(4)-S( 19) 237(3) Ag(12)-Ag( 14) 287(1) Sb(4)-S(mean) 242(4) Ag(12)-Ag(15) 3. 19( 1) Ag(1)-Ag(5) 3 .09( 1) Ag(12)-S(2) 247(2) Ag( l)-Ag(10) 309(2) Ag(12)-S(15) 235(2) Ag(l)-Ag(l4) 3.07( 1) Ag(1)-Ag( 17) Ag(1)-S(8) Ag(1)-S(14) Ag(l)-S(16) Ag(2)-A80 1) Ag(2)-Ag(19) Ag(2)-Ag(4) Ag(2)-Ag(9) Ag(2)-8(1) Ag(2)-S(7) Ag(2)-802) Ag(3)-Ag(1 1) Ag(3)-Ag( 12) Ag(3)-Ag( 19) Ag(3)-S(6) A8(3)-S(7) AEGO-S(9) Ag(3)-S(15) Ag(4)-Ag(7) Ag(4)-Ag(1 1) Ag(4)-Ag( 16) 321(1) 258(2) 244(2) 247(3) 3.13(1) 3.11(l) 309(1) 299(1) 2.45(2) 249(2) 262(2) 305(1) 298(1) 305(1) 272(2) 261(2) 278(3) 272(2) 305(1) 297(1) 309(1) 160 Ag(13)-Ag(13') Ag(13)-Ag(18) Ag(13)-Ag(20) Ag(13)-8(5) Ag(13)-S(14) Ag(13)-S(l7) Ag(13’)-Ag(18) Ag(13')-S(5) Ag(13')-S(14) Ag(l3')-S(15) Ag( 14)-3( 1) Ag(l4)-S(2) Ag(l4)-S(l6) Ag(15)-S(14) Ag(15)-S(16) Ag(15)-S(18) Ag(16)-Ag(20) Ag(16)-Ag(20) Ag(16)-S(3) Ag(16)-S(4) Ag(16)-S(l3) 084(2) 301(2) 249(3) 261(2) 253(3) 251(3) 3.10(2) 265(3) 252(3) 255(3) 258(2) 266(2) 251(2) 260(2) 248(2) 247(2) 299(3) 283(3) 265(2) 271(2) 272(3) Ag(4)-S(4) Ag(4)-S(7) Ag(4)-S( 10) Ag(4)-S( 1 3) Ag(5)-Ag( 14) Ag(5)-Ag(15) Ag(5)-Ag(17) Ag(5)-Ag(20) Ag(5)-S(8) Ag(5)-S(16) Ag(5)-S(18) Ag(6)-Ag( 13) Ag(6)-Ag(13) Ag(6)-S(6) Ag(6)-SO 1) Ag(6)—S(15) Ag(6)-S(l7) Ag(7)-A800) Ag(7)-Ag(lO') Ag(7)-Ag( 16) Ag(7)-Ag( 18) 274(2) 260(2) 266(2) 254(3) 299(1) 3. 19(1) 308(1) 288(3) 256(2) 256(2) 255(2) 292(2) 285(2) 264(2) 262(3) 283(3) 255(2) 293(2) 268(4) 312(1) 311(1) 161 Ag(16)-S(l7) Ag(17)-8(1) Ag(17)-S(l3) Ag(17)-S(l6) Ag(17)-S(l9) Ag(18)-S(12) Ag(18)—S(l7) Ag(18)-S(l9) Ag(19)-S(7) Ag(19)-S(9) Ag( 19)-S(1 1) Ag(19)-S(15) Ag(20)-Ag(20) Ag(20)-S(8) Ag(20)-S(13) Ag(20)-S(8) Ag(20')-S(13) Ag(20)-S(14) Cs(l)-S(3) Cs(1)-S(3) 257(2) 259(2) 262(3) 278(3) 261(2) 268(2) 247(2) 248(3) 263(2) 258(2) 268(2) 270(3) 075(3) 247(3) 2.37(3) 265(3) 266(4) 281(4) 3.93(2) 3.91(2) 162 Cs( 1 )—S(3) 3.62(3) Ag(7)-S(3) 268(3) Cs(1)-S(4) 3.91 (2) Ag(7)-S( 10) 263(3) Cs(1)-S(4) 3 .65(2) Ag(7)-S( 13) 275(2) Cs( 1 )-S( 10) 402(2) Ag(7)-S( 17) 256(2) Cs(1)-S(10) 368(3) Cs(1)-S(1 1) 398(2) Ag(8)-Ag(9) 3.1 1(1) Cs(1)-S(17) 3.46(3) Ag(8)-Ag( l4) 3.15(1) Cs(1)-S (mean) 379(6) Ag(8)-Ag(15) 3.12(1) Cs(2)-S( 10) 392(2) Ag(8)-S(2) 278(2) Cs(2)-S(1 1) 357(2) Ag(8)-S(15) 2.5 8(2) Cs(2)-S(1 1) 396(3) Ag(8)-S(l6) 267(3) Cs(2)-S(4) 3.85(2) Ag(8)-S( 19) 263(2) Cs(2)-S(6) 3.65(2) Cs(2)-S(6) 3.84(2) Ag(9)-Ag(12) 290(1) Cs(2)-S(7) 3.39(2) Ag(9)-Ag(19) 2.88( 1) Cs(2)-S(9) 3.64(3) Cs(2)-S(9) 383(2) Ag(9)-S(2) 260(3) Cs(2)-S (mean) 373(6) Ag(9)—S(7) 262(3) Ag(9)—S(12) 264(2) Ag(10)-Ag(10’) 0.79(4) Ag(10)-Ag(l8) 3.15(1) Ag(10)-Ag(20) 3. 16(3) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 01 = {Ewan -— [)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 4-6. in parentheses.a 163 Selected Bond Angles (°) for CszAgzoSb4819 (II) with standard deviations S(9)-Sb(1)-S( 10) S(9)-Sb( l )-S(l 1) S(9)-Sb(l)-S(12) S(10)-Sb(1)-S(11) S(10)-Sb(l)-S(12) S(11)—Sb(l)-S(12) S( l )-Sb(2)-S(2) S(1)-Sb(2)-S(18) S(2)—Sb(2)-S( 18) S(3)-Sb(3)-S(4) S(3)-Sb(3)-S(5) S(3)-Sb(3)-S(6) S(4)-Sb(3)-S(5) S(4)-Sb(3)-S(6) S(5)-Sb(3)-S(6) S(8)-Sb(4)-S(14) S(8)-Sb(4)-S( l 9) S(14)—Sb(4)-S(l9) S(8)-Ag(l)-S(14) S(8)-Ag(1)-S(16) S(l4)-Ag(l)-S(16) S(1)-Ag(2)-S(7) S(1)-Ag(2)-S(12) S(7)-Ag(2)-S(12) S(6)-Ag(3)-S(7) S(6)-Ag(3)-S(9) S(6)-Ag(3)-S(15) S(7)-Ag(3)-S(9) S(7)-Ag(3)-S(15) S(9)-Ag(3)-S(15) S(4)-Ag(4)-S(7) S(4)-Ag(4)-S(10) S(4)-Ag(4)-S(13) S(7)-Ag(4)-S( 10) S(7)-Ag(4)-S(13) S(10)-Ag(4)-S(l3) S(8)—Ag(5)-S(16) S(8)-Ag(5)-S(18) S(16)-Ag(5)-S(18) S(6)-Ag(6)-S(1 1) S(6)-Ag(6)-S(15) 1 l 1.5(8) 107.6(8) 110.2(9) 109.8(8) 107.0(9) 110.7(8) 99.4(7) 97.7(8) 97.5(7) 1 10.2(8) 1 10.1(9) 1 10.2(8) 109.7(8) 107.4(8) 109.2(8) 93.3(7) 99.8(8) 98.4(8) 108.0(7) 108.3(7) 142.6(7) 137.6(7) 101.9(7) 119.8(8) 107.0(7) 81 .4(7) 104.6(7) 104.9(7) 137.3(8) 107.5(7) 107.9(7) 86.7(7) 110.9(8) 110.0(7) 123.6(8) l 1 1.5(7) 136.2(8) 98.4(7) 125.3(8) 922(7) 103.9(7) S(15)-Ag(8)-S(16) S(15)-Ag(8)-S(19) S(16)-Ag(8)—S(19) S(2)-Ag(9)-S(7) S(2)-Ag(9)-S(12) S(7)-Ag(9)-S(12) S(5)—Ag(10)-S(8) S(5)-Ag(10)-S(13) S(8)-Ag(10)-S(13) S(5)-Ag(10')-S(8) S(5)-Ag(10')-S(17) S(8)-Ag(10’)-Sl7 S(5)—Ag(l 1)-S(7) S(5)-Ag(l 1)-S(18) S(7)-Ag(] 1)-S(18) S(2)-Ag(12)—S( 15) S(5)—Ag(13)-S(14) S(5)-Ag(13)-S(17) S(14)-Ag(13)-S(17) S(1)-Ag(l4)-S(2) S(1)-Ag(l4)-S(16) S(2)-Ag(14)-S(16) S(l4)—Ag(15)-S(16) S(14)-Ag(15)-S(18) S(16)-Ag(15)-S(18) S(3)-Ag(16)-S(4) S(3)-Ag(16)-S(l3) S(3)-Ag(l6)-S(l7) S(4)-Ag(16)-S(13) S(4)-Ag(16)-S(17) S(13)-Ag(16)-S(17) S(1)-Ag(17)-S(13) S(1)-Ag(17)-S(16) S(1)-Ag(17)-S(19) S(13)-Ag(l7)-S(16) S(13)-Ag(17)-S(19) S(16)-Ag(l7)-S(l9) S(12)-Ag(18)-S(17) S(12)-Ag(18)-S(l9) 134.1(7) 113.7(8) 95.9(8) 127.0(8) 100.2(7) l 16.9(8) 98.5(8) 121.4(9) 131.6(9) 98(1) 1 15(2) 1 19(2) 121.5(8) 96.7(8) 140.9(9) 166.7(8) 100.1(9) 124(1) 128.6(9) 93.8(7) 128.5(7) 133.7(7) 109.2(7) 103.3(8) 147.4(8) 89.3(7) 109.7(7) 112.8(8) 106.5(7) 1 10.1(7) 123.1(8) 115.6(8) 104.2(8) 920(7) 125.8(8) 118.4(8) 94.1(8) 116.8(8) 97.6(8) 164 S(6)-Ag(6)-S(l7) 120.4(7) S(17)-Ag(18)-S(19) 145.6(9) S(l l)-Ag(6)-S(15) 103.5(7) S(l l)-Ag(6)-S(17) 121.6(7) S(7)-Ag(19)-S(9) 115.8(8) S(15)-Ag(6)-S(17) 112.1(8) S(7)-Ag(19)-S(11) 110.8(7) S(7)-Ag(19)-S(15) 115.3(8) S(3)-Ag(7)-S( 10) 91 .2(7) S(9)-Ag(19)-S(1 1) 924(8) S(3)-Ag(7)-S(l3) 107.6(7) S(9)-Ag(19)-S(15) 1 14.3(8) S(3)-Ag(7)—S(17) 111.8(7) S(11)-Ag(19)-S(15) 105.1(7) S(lO)-Ag(7)-S(13) 106.3(7) S(10)-Ag(7)-S(17) 1 18.7(8) S(8)-Ag(20)—S(13) 153(1) S(13)-Ag(7)-S(17) 117.8(8) S(8)-Ag(20')-S(13) 125(1) S(2)-Ag(8)—S(15) 1 13.4(8) S(8)-Ag(20')-S(14) 81(1) S(2)-Ag(8)-S(16) 100.8(7) S(13)-Ag(20')-S(14) 1 19(1) S(2)-Ag(8)-S(19) 88.9(6) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 0'1 = {E,;(ln — [)2/n(n — 1)}1/2, where In is the length (or angle) of the nth bond, l the mean length (or angle), and n the number of bonds. 165 Table 4—7. Selected Distances (A) for B-szAgzoSb4S19 (IH) with standard deviations in parentheses.a Sb(1)-S(2) 2.321(9) x3 Ag(2')-S(3) 253(2) Ag(2')-8(1) 260(2) Sb(1)-S(1) 239(2) Ag(2')-S(4) 275(3) Sb(2)-S(3) 2.426(11) x3 Ag(2')-8(5) 295(3) Ag(1)-Ag(1') 295(2) Ag(3)-Ag(3') 3.18(2) Ag(1)-Ag(2') 299(2) Ag(3)-Ag(3) 3.04(2) Ag(1)-Ag(3) 3.01(2) Ag(3)-Ag(3) 3.04(2) Ag(1)-Ag(3) 3.03(2) Ag(3)-Ag(1) 3.01(2) Ag( 1)-Ag( 1) 3.12(2) Ag(3)-Ag(2') 258(3) Ag(1)-Ag(1) 3.12(2) Ag(1)-Ag(3') 325(3) Ag(3)-S(4) 2.589(9) Ag(1')-Ag(3') 291(4) Ag(3)-8(5) 262(3) Ag(3)-S(2) 273(2) Ag(1)-S(4) 2.45(2) Ag(3)-S(2) 273(2) Ag(1)-S(3) 2.473(13) Ag(1)-S(1) 2.627(9) Ag(3')-Ag(3') 3.27(3) x2 Ag(3')-Ag(l') 291(4) Ag(1')-Ag(2') 1.54(4) Ag(3')—Ag( 1) 326(2) Ag(l')-Ag(3) 260(3) Ag(3')-Ag(1) 3.25(3) Ag(1')-Ag(2') 262(3) Ag(3')-Ag(3) 3.18(2) x2 Ag(1')-Ag(1) 295(2) Ag(1')-Ag(2') 3.32(3) Ag(3')-S(2) 249(2) x2 Ag(1')-Ag(4) 292(3) Ag(3')-S(4) 2.549(13) Ag(1')-S(3) 252(2) Ag(4)-Ag(2) 3.098(14) Ag(1')-S(l) 262(2) Ag(4)-Ag(2) 3.096(14) Ag(l')-S(4) 270(3) Ag(4)-Ag(1') 292(3) Ag(4)-Ag(2') 286(3) Ag(2)-Ag(2) 3.217(13) x2 Ag(4)-Ag(4) 246(2) x2 Ag(2)-Ag(2) 3.179(12) x2 Ag(2)-Ag(4) 3.098(14) Ag(4)-8(5) 262(2) Ag(2)-Ag(4) Ag(2)-S(3) Ag(2)-S(6) Ag(2')—Ag(1') Ag(2')-A80) Ag(2')-Ag(3') Ag(2')-Ag(4) Ag(2')-Ag(3) Ag(2‘)-A80) 3.096(14) 253(1) x2 2.262(7) 3.32(3) 299(2) 290(4) 286(3) 258(3) 262(3) 166 Ag(4)-S(6) Ag(4)-S(3) Ag(4)-S(3) Rb(1)-S(4) Rb( l )-S(2) Rb( l )-S(2) Rb(1)-S(2) 288(1) 265(2) 264(2) 3.27(2) 3.895(11) x2 3.893(11) x2 3.452(12) x3 21The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 0'1 = {E,;(ln — [)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 4-8. 167 deviations in parentheses. Selected Bond Angles (°) for B-szAgZOSb4S 19 (HI) with standard S(2)-Sb(l)-S(2) 110.0(2) x3 S(4)-Ag(3)-S(5) 125.3(9) S(2)-Sb(l)-S(1) 109.0(3) x3 S(5)-Ag(3)-S(2) 110.2(4) x2 S(4)-Ag(3)-S(2) 109.4(6) x2 S(3)-Sb(2)—S(3) 97.2(3)x3 S(2)-Ag(3)-S(2) 84.9(6) S(4)-Ag(l)-S(3) 139.8(5) S(2)—Ag(3')-S(2) 95.6(10) S(4)-Ag(1)-S(1) 121.1(5) S(2)-Ag(3')-S(4) 119.1(8) S(3)-Ag(1)—S(1) 99.1(5) S(2)-Ag(3')-S(4) 119.0(8) S(3)-Ag(l')-S(l) 98.2(8) S(5)-Ag(4)-S(3) 117.2(5) x2 S(3)-Ag(1')-S(4) 124.9(1) S(3)-Ag(4)-S(3) 93.3(6) S(1)-Ag(1')-S(4) 112.7(9) S(5)-Ag(4)-S(6) 117.8(5) S(3)-Ag(4)-S(6) 103.9(4) x2 S(6)-Ag(2)-S(3) 130.5(4) S(6)-Ag(2)-S(3) 130.7(4) S(3)-Ag(2)-S(3) 98.8(5) S(4)-Rb(1)-S(2) 141.9(2) x3 S(2)-Rb(1)-S(2) 64.6(3) x3 S(3) -Ag(2')-S(l) 98.4(9) S(3)-Ag(2')-S(4) 122.6(9) S(1)-Ag(2‘)-S(4) 1 1 1.6(9) S(3)-Ag(2')-S(5) 1 10.0(9) S(1)-Ag(2')-S(5) 104.0(9) S(4)-Ag(2')-S(5) 108.4(8) 2‘The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 61 = {2n(ln — [)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. 168 A9 Sb Figure 4—1: Packing diagram for CszAg20Sb4819 as viewed down the b-axis. 169 (A) A96 A919’ A94’ A918 Q S11 " . S4 ’ $12 A 1 A918 . \. .A99 9 . Sb3 A910 A 7 f‘ 89 A93 86 f\ . g ’ V o ’ e V o ”7 S10 . \ 33 C 0 A92 . .A913 85 0 A94 A919 A96 A916 (13) A95 A91 A95, A915’ . 33 0 g 818 ’ c" A A 1' A920 Q A910 A917 A911. 9 A91 0 Sb6 . .sz . 0 A914 a 0 A 13 a (. A 9 a S14 0 9 S19 81 A92 9 . 82 A913 0 O . g A 14 A915 A98 9 A98’ Figure 4-2: (A) The coordination of the tetrahedral [SbS4]3‘ units to Ag+ ions, with labeling. (B) The coordination of the pryamidal [SbS3] 3' units to Ag+ ions, with labeling. 170 Figure 4—3: (A) A parallel view of the (Ag3SbS4) sheet in (II). (B) A perpendicular view of the (Ag3SbS4) sheet, highlighting the Ag3S3 rings connecting the neighboring tetrahedral [SbS4]3' units. (C) A parallel view of the (Ag3SbS4) sheet highlighting the Ag+ ions used in connecting the ternary (Ag3SbS4) sheet to the binary Ag2S "belt", black circles; Sb, gray circles; Ag, open circles; S. 171 (A) 172 Figure 4-4: (A) A parallel view of the (Ag3SbS3) sheet. (B) A perpendicular view of the (Ag3SbS3) sheet, highlighting the Ag3S3 rings connecting the neighboring pryamidal [SbS3]3‘ units. (C) A parallel view of the (Ag3SbS3) sheet highlighting the Ag+ ions used in connecting the ternary (Ag3Sng) sheet to the binary Agzs "belt", black circles; Sb, gray circles; Ag, open circles; S. 173 174 Figure 4-5: A sequential decomposition of the [Ag20Sb4S19]2"“ layers to [Angszg] "sub-layers", and the further separation into the two ternary (Ag3SbS4) and (Ag3SbS3) sheets and the binary Ag2S “belt”. 199119 [Esqsefiv] (O . .‘ .‘ _ F‘ I‘ . a 10‘ > to M (I) l—l m 3' CD (D F. 199113 [781189590 175 163/121 elqnop dalmstqsoafivl ale/191 qnsr [9339301511] 176 dds/1! \4 .. , 1.. t .466 .2 . . 19mm»; .. _ . ul/ 2v ., ..i #6311 91.1.\\M 3),». \ .a/f /.9 44;; 107 “a {a .1 1.6 f. /. \xm‘flewrr/C.b// \.\\ , .\. \ 4%,.‘234 62./1.42. 4 _ .4. . ta (“W/x ,. .\!..,9.. \\ . 1 7.52571‘4 4 .. ,.a\. .11 N. i ‘a/ 8.20%.. // A4 \ , Cup. 2. In 6 We. 6 % \1 Q a\ ilr’lfl/ r'l b... 4‘ .2... .6 1‘ . .1911...” 706 6w \ \ A view of a portion of the layers in B-szAg208b4819 (III), with labeling. Figure 4-6: 177 Figure 4-7: (A) A parallel view of the (Ag3SbS4) sheet in (III). (B) A perpendicular view of the (Ag3SbS4) sheet, highlighting the Ag3S3 rings connecting the neighboring tetrahedral [SbS4]3' units. (C) A parallel view of the (Ag3SbS4) sheet highlighting the Ag+ ions used in connecting the ternary (Ag3SbS4) sheet to the binary Agzs "belt", Sb, gray circles; Ag, light gray circles; S, open circles. 178 (C) Figure 4-8: 179 (A) A parallel view of the (Ag3SbS3) sheet in (111) (B) A perpendicular view of the (Ag3SbS3) sheet, highlighting the Ag3S3 rings connecting the neighboring pyramidal [SbS3]3' units. (C) A parallel view of the (Ag3Sb83) sheet highlighting the Ag+ ions used in connecting the ternary (Ag3SbS3) sheet to the binary Agzs "belt", gray circles; Sb, light gray circles; Ag, open circles; S. 181 3.2. Synthesis, Spectroscopy and Thermal Analysis. The syntheses of these compounds was the result of a redox reaction in which the silver is oxidized by the polysulfide ions in the Ax[SbySz] flux. The Ag+ centers are then coordinated by the highly charged [beQy]"‘ units. Although a number of different [beQyF' units have been isolated from the flux, the ability to predict which [beQy]“' units will be formed is not completely understood. In the present conditions the flux was silver rich and antimony poor. Similar conditions3 favored the formation of oliogmeric [beQy]“' units containing the Sbi‘3 species. Instead, the compounds contain both discrete tetrahedral [SbS4]3' and pyramidal [SbS3]3' units containing Sb+3 and Sb"5 species, respectively. Although the stability of the lower oxidation sates, Sb“3 species, increases as one goes down the row, this unprecedented result suggests the presence of a complicated Lewis acid-base equilibrium. Continuing our exploration of thioantimonate fluxes with higher Ag/Sb rations we have been able to incorporate additional equivalents of Ag+ ions forming silver-rich materials. The compounds described here represent a departure from other chalcometalate compounds, whose structural descriptions are dominated by the various [Eny]"' (E = P, Sb; Q = S, Se, Te) units coordinating to the metal ions, not the resulting framework coordinating to the various [Eny]"' units. The structures of (1)-(III) are almost completely dominated by the “Ag28” portion of the [Angbsz] 2‘" framework, to which the various [beQy]"' units coordinate. The structure of (III) represents a new member of a class of compounds with the general formula (AgS)w(AgZS)x(Ag3SbS3)y(Ag3SbS4)z, further validating the use of the polychalcoantimonate fluxes for synthesis of new materials. The optical absorption properties of (I) — (IH) were obtained in the diffuse reflectance mode using polycrystalline samples of the compound. The spectra show that these materials are wide band-gap semiconductors. The AzAngbrtslg (A = Rb, Cs) 182 compounds exhibit steep absorption edges from which the band-gap, E 8, can be assessed at 1.21eV (I), and 1.37eV (H), respectively, see Figure 4-9. The band-gap of B— szAggosmslg is 1.63eV, see Figure 4-10. The far-IR spectra of (1)-(H) display only a few absorptions in the 400 — 100 cm'1 range. The sharp absorption at ca. 378 cm'1 is in between the values expected for the Sb-S vibrations of the [SbS4]3‘ (ca. 385 cm'l) and the oligomeric [Sb284]2- units (ca. 373, and 357 cm'l) found in the far-IR spectrum of CS3Ag2Sb383, making an assignment difficult. Absorptions observed below 350 cm'1 are probably due to Ag-S vibrations. Attempts to obtain a far—H2 spectrum of (HI) were unsuccessful. The Raman spectra of (1)—(II) display absorptions in the 400—350 cm‘1 range that are assigned to Sb—S vibrations and the absorptions below 350 cm’1 are tentatively assigned to Ag-S vibrations. The sharp absorption at ca. 360 cm'1 is characteristic of the [SbS4]4" unit by analogy to A2AquS4.24 The Raman spectrum for (HI) was extremely weak, showing only a single peak at 358 cm' 1, confirming the presence of the [SbS4]3' unit. The Raman spectra of (I) and (IH) are shown in Figure 4-11. Unlike the thermal analysis described in the previous chapter, increasing the Ag+ ion concentration appears to have a stabilizing effect on the thermal properties of (1)-(H). Differential thermal analysis (DTA) data, followed by careful XRD analysis of the residues, show that or-szAgzosb4819 melts incongruently at 436°C, producing a mixture of et- szAgZOSb4519 and Ag3SbS3, see Figure 4-12A. The DTA data for CszAgZOSb4S19 reveals that the compound melts congruently at 416 °C, see Figure 4-12B. The DTA data, followed by careful XRD analysis of the residue revealed that B- szAg208b4819 melts incongruently at 450 °C, forming a mixture of a-szAg298b4S19 and B- szAgzoSb4819, see Figure 4-13A. The DSC data revealed that the there were two endothermic transitions at 179 °C and 210 0C, respectively. Upon cooling two exothermic peaks are observed at 206 °C and 167°C, suggesting a Ag+ ion transition. 183 (A) 0'4 l l l l l I 0.35 - 0.3 r 0.25 — 0.2 h 0.15 — Rb Ag Sb S 2 20 4 19 0.1 — Eg = 1.21 eV 0.05 — I - 0 1 1 1 1 1 2 5 6 a/S Absorption Coeff. (Arb. Units) (B) 0’35 l j I I l 0.3 — _ 0.25 - _ Cs Ag Sb S 2 20 419 0.1 — Eg = 1.37 eV " 0.05 - - 0 l l l l a/S Absorption Coetf. (Arb. Units) 2 3 4 Energy (eV) Figure 4-9: (A) Solid-state optical absorption spectrum for CszAgzoSb4819. (B) Solid-state optical absorption spectrum for szAg208b4819. 184 E 1-2 1 I 1 1 1 I C I) . 1 - - e < V 0.8 - -1 ::' 8 0 6 O B -Rb2Ag 20 Sb4S 19 .§ 0,4 _ 13g: 1.63 eV _ S 8 .o 0.2 - - < ‘J <0 1 1 1 1 1 \d 0 O 1 2 3 4 5 6 7 Energy (eV) Figure 4—10: Solid-state optical absorption spectrum for B-szAg208b4819. 185 (A) 6 T I I I g 360——> c 5 — - 3 .o' E, 4 " ‘ > .1: 3 — _ (I) 381 g 318 222 s 2 - 258 / 7 C 8 1 - 1 4 (U a: 0 l l l l 600 500 400 300 200 100 Wavenumbers (cm'1) (B) O N .1 .1 _o —L l I O .2 I J Raman Intensity (arb. units) 600 500 400 300 _1 200 100 Wavenumber (cm ) Figure 4-11: (A) Raman spectrum for Rb2Ag20$b4S 19 (I). (B) Raman spectrum for b-Rb2Ag208b4S 19 (HI). 186 (A) U1 - - . - - C CD sponse (11v) —> 5 6r .3 C <§—;—Re .2 U1 CD 3 O. O l l l J l r N o 0 100 200 300 400 500 600 Temperature (deg C) (B) (D X 0 our: It at on o O O 4:. —L a: O 101 I r 0) -§ 1 r I 1 Response (11v) —> [(3 r l r d I O (I) l 7 l l (D 3 O. O _1 2 1 1 1 I 1 0 100 200 300 400 500 Temperature (deg C) Figure 4-12: (A) DTA diagram for Rb2Ag208b4S 19 (I). (B) DTA diagram for CszAgzoSb4S 19 (H). 187 (A) 6X025 I I l I l g - Q) (I) c -1 o O. (D 0) . I endo_5 1 1 1 1 L O 100 200 300 400 500 600 Temperature (°C) (B) exo1c5 I I I l I i 1 - 167 206 ~ E 0.5 - 4/ end - E 0 - 4 805 . ' . D __ -1 t . Start __) W -1.5 1- 210 7 endo -2 J 1 1 I r r 0 50 100 150 200 250 300 350 Temperature (°C) Figure 4—13: (A) DTA diagram for B-szAggoSb4S19(HI). (B) DSC diagram for B-Rb2A8203b4519 (III)- 188 4. Conclusions. The exploration of the silver-rich fluxes produced three new thioantimonate compounds with a complicated layered structure. These compounds represent the first members of a new class of compounds with the general formula (AZS)W(Ag2S)x(Ag3SbS3)y(Ag3SbS4)z. The structure of (1)-(HI) represent a new direction in the chalcometalate chemistry in which the structural description is dominated by the Agzs framework coordination to the various [beSyln‘ units. While the complicated layers observed in (1)-(HI) are very intriguing from a structural standpoint, these materials sill require extensive characterization to determine if they exhibit properties which make them potenital candidated for ion conductors or other applications. The large supercell observed in (IH) is unusual and further work to is needed to completely determine the structure. \OOO\10\M 11. 12. l3. 14. 15. l6. 17. 18. 19. 20. 21 189 List of References McCarthy, T. J.; Kanatzidis, M. G. Inorg. Chem, 1994, 33, 1205. Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G. Inorg. Chem, 1997, 36, 3804. (b) Chen, J. H.; Dorhout, P. K. J. Alloys and Comp., 1997, 249, 199. Hanko, J. A.; Kanatzidis, M. G. Abstract #0413 from 210th Fall ACS Meeting, Chicago Il 1995. (a) Wood, P. T.; Schimek, G. L.; Kolis, J. W. Chem. Mater., 1996, 8, 721. (b) Schimek, G. L.; Pennigton, T. L.; Wood, P. T.; Kolis, J. W. J. Solid State Chem.., 1996, 123, 272. Imafuku, M.; Nakai, I.; Nagashima, K. Mat. Res. Bull., 1986, 21 , 493. Hanko, J. A.; Kanatzidis M. G. J. Chem. Soc: Chem. Comm, 1998, 726. Hanko J. A.; Kanatzidis,M. G. J. Alloys and Comp., In press. Chen, J. H.; Dorhout, P. K. J. Alloys and Comp., 1997, 249, 199. Choi, K.-S.; Hanko, J. A; Kanatzidis, M. G. Inorg. Chem. Submitted (a) Kanatzidis, M. G. Curr. Opinion Solid State and Mater. Sci, 1997, 2, 139. (b) Sutorik, A. C.; Kanatzidis, M. G. Prog. Inorg Chem, 1995, 43, 151 and references therein. Chondroudis, K.; Hanko, J. A.; Kanatzidis, M. G. Inorg. Chem, 1997, 36, 2623. Rom, I; Sitte, W. Solid State Ionics , 1997, 10],: 381. Boucher, F.; Evain, M.; Brec, R. J Solid State Chem, 1993, 107, 332. Gaudin, E.; Fischer, L.; Boucher, F.; Evain, M.; Petricek, V. Acta. Crystallogr., 1997, B53, 6775. CERIUS2, Version 1.6, Molecular Simulations Inc., Cambridge, England, 1994. McCarthy, T. J .; Ngeyi, S.-P.; lLiao, J .-H.; DeGroot, D.; Hogan, T.; Kannewulf, C. R.; Kanatzidis, M. G. Chem. Mater., 1993, 5, 331. Walker, N.; Stuart, D. Acta Crystallogr., 1983, A39, 158. Sheldrick, G. M., "Crystallographic Computing 3"; Sheldrick, G. M., Kruger, C., Doddard, R., Eds; Oxford University Press: Oxford, England, 1985; pp 175. TEXSAN: Single Crystal Structure Analysis Software, Version 5.0; Molecular Structure Corp.: The Woodlands, TX 77381, 1981. Hanko, J. A.; Bree, R. Evain, M. Kanatzidis, M. G. Work in progress. SMART and SAM. Data collection and Processing Software for the Smart system. Siemens Analytical X-ray Instruments Inc.,l995. 22. 23. 24. 190 Blessing , R. H. Acta Crystallogr., 1995, A51, 33-38. SHELXT L Version 5, Reference Manual. Siemens Industrial Automation Inc., 1 9 94 . Hanko, J. A.; Kanatzidis, M. G. J Alloys Comp., accepted for publication. CHAPTER 5. The Chemistry of Au in Molten Polythioantimonate Fluxes. Synthesis and Characterization of the New Multinary Gold Thioantimonates A2AquS4 (A = Rb, Cs) and szAusSbrtSm. A Novel Compound with a Binary and Ternary Interconnected Framework. 191 192 1. Introduction As shown in previous chapters, the polychalcoantimonate fluxes can be used for the synthesis of new multinary thioantimonate and selenoantimonate compoundsl'G. This method is complementary to conventional direct combination of the binary sulfides7 or hydro(solvento)thermal synthesis.8 The polychalcoantimonate fluxes are formed by the in situ fusion of A2Q/Sb/Q and contain [beQy]“' ligands (A = Na, K, Rb, Cs ; Q = S, Se) as well as polychalcogenide ligands. The key feature of this method is that the polychalcoantimonate units preferentially coordinate to coinage metal ions, building extended lattices. Examples include KTthzseg,2 A2AngS4 (A = K, Rb, Cs)3'4, CS3Ag28b3Q3 (Q = S,Se)3'4 , AAgz SbS4 (A = K, Rb)3~4, A2AgzoSb4519 (A = Rb, Cs)5 and KHngS36. The binding modes displayed by the [Sbsz]n' anions in the above materials are different from the binding modes observed in the coinage metal sulfosalt minerals, i.e. Cu3SbS4,9 CuPbSbS3,10 AngSbS3,“ and Cu12Sb4SI3.” While a variety of ternary alkali gold chalcogenide compoundsl3‘l4 are known, no well characterized ternary Au/Sb/S phase has been reported. Attempts to prepare a ternary Au/Sb/S phase via direct combination reactions were fruitless, leading us to address the question of whether a quaternary A/Au/Sb/S compound could be stabilized. Continuing our investigations of the coinage metals, particularly Au, we report here the synthesis, structural characterization, and physical properties of the new quaternary gold thioantimonate compounds, A2AquS4 (A = Rb, Cs) (1)-(H), and szAu65b4S19(HI) . The novelty of RbgAu68b4Sm is that its layers are comprised of two different, independent interwoven frameworks. The only other structurally characterized example of two interpenetrating frameworks is K2PdSeto‘5 193 2. Experimental Section 2.1 Reagents The reagents used in ths study were used as obtained unless noted otherwide: (i) gold metal (99.99%) was acquired from Liberty Coins, Lansing MI; (ii) antimony powder 99.999% purity, -200 mesh, Cerac Inc., Milwaukee, WI; (iii) cesium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (iv) rubidium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (v) sulfur powder, sublimed, J .T. Baker Chemical Co., Phillipsburg, NJ; (vi) N,N-Dimethylformamide (DMF) reagent grade, EM Science, Inc., Gibbstown, NJ ; (vii) diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. Finely divided Au metal. A Canadian Maple Leaf gold coin, (99.99%, 31.1 g) was dissolved in 400 ml of aqua regia (300 ml concentrated HCl and 100 ml of concentrated HNO3). The solution was boiled in an acid resistant fume hood to a volume of ~100 ml. The concentrate was neutralized with ammonium hydroxide and the gold was reduced with excess hydrazine hydrochloride, dissolved in 100 ml of distilled water. The resulting black suspension was heated gently, with stirring, for one hour to allow particle aggregation. The gold powder was filtered and washed with copious amounts of distilled water and acetone, and then heated in air at 130°C for 4 hours to drive off any remaining volative impurities, yielding 30.9 g of Au powder. Note: heating too long or at too high a temperature results in large intractable grain sizes. 194 2.2 Syntheses. A28 (A = Rb and Cs) were prepared by reacting stoichiometric amounts of the elements in liquid ammonia16 as described in Chapter 2, Section 2.2. szAqu84 (1). A mixture of 0.102g (0.50 mole) szS, 0.098g (0.50 mmole) Au, 0.031g (0.25 mmole) Sb, and 0.080g (2.5 mmole) S were thoroughly mixed and transferred into a 6—ml Pyrex tube which was subsequently flame-sealed in vacuo (~10‘3 Torr). The reaction mixture was ramped to 350°C over a 12 hour period in a computer-controlled furnace. The reaction was heated at 350°C for 3 days, followed by cooling to 150°C at a rate of 4°C/hr and then to room temperature in 10 hours. The product, which is slightly unstable in water and air, was isolated by dissolving the Rb2Sx and any residual be[SbySz] flux with DMF under inert atmosphere to give pale yellow crystals and a small amount of unreacted gold metal (1-5%). A yield of 75%, based on Sb was estimated. Quantitative microprobe analysis of single crystals gave Rb2,3Au1.4Sb86. CszAquS4 (II). A mixture of 0.149g (0.50 mole) Cszs, 0.098g (0.50 mole) Ag, 0.031 g (0.25 mole) Sb, and 0.080g (2.5 mmole) S was sealed under vacuum in a Pyrex tube and heated as in (I). The product, which is stable in air and water, was isolated as in (I) to give yellow crystals in 77% yield, based on Sb. Quantitative microprobe analysis on single crystals gave Csl,4AngS4.3, szAu68b4Sm (III) was synthesized from a mixture of Rb28 (0.102g ; 0.5 mmol) Au (0.098g ;0.5 mmol) Sb (0.031 g ;0.25 mmol) and S (0.064g ; 2 mmol) was sealed under vacuum in a Pyrex tube and ramped to 350°C for 4d followed by cooling to 150°C at 4°C h '1. The product, which is stable in water and air, was isolated by 195 dissolving the szSx and any residual be[SbySZ] flux with DMF under inert atmosphere to give black needles in low yield (20% yield based on Sb). Microprobe analysis carried out on several randomly selected crystals gave an average composition of RbAu4ng2,789. 2.3. Physical Measurements. Powder X-ray Difl'raction. The compounds were examined by powder X-ray diffraction for the purpose of phase purity. Accurate d spacings obtained from the powder pattern for the three compounds were performed using a calibrated Rigaku- Denki/RW400F2 (Rotaflex) rotating anode powder diffractometer controlled by an IBM computer, operating at 45 kW 100 mA and with a 10/min scan rate, employing graphite crystal filtered Cu radiation in a Bragg-Brentano geometry. The calculated powder patterns for (I) and (HI) were prepared with the CERIUS2 software.17 Listings of calculated and observed powder patterns for (I) and (HI) are given in Tables 5-1 and 5-2, respectively. Infrared Spectroscopy. Infrared spectra, in the far-IR region (600-50 cm'l), were recorded on a computer controlled Nicolet 750 Magna-IR Series II spectrophotometer equipped with a TGS/PE detector and silicon beam splitterwith a resolution of 4 cm'l. The samples were ground with dry CsI into a fine powder and pressed into translucent pellets. Raman Spectroscopy. Raman spectra (600-100 cm'l) were recorded on a BIO- RAD Fl‘ Raman spectrometer equipped with a Spectra-Physics Topaz T10-106c 1.064 nm YAG laser and a Ge detector. The samples for (I) and (II) were ground to a fine powder and loaded without modification into glass capillary tubes. The crystals for (HI) were ground with C81 into a fine powder and loaded into a glass capillary tube. Solid State U V/VisflVear IR Spectroscopy. Optical diffuse reflectance measurements were performed at room temperature using a Shimadzu UV—3101PC double beam, double monochromator spectrophotometer. The instrument is equipped with integrating sphere and controlled by a personal computer. BaSO4 was used as a 100% 196 reflectance standard for all materials. Samples are prepared by grinding them to a fine powder and spreading them on a compacted surface of the powdered standard material, preloaded into a sample holder. The reflectance versus wavelength data generated can be used to estimate a material’s band gap by converting reflectance to absorption data as described elsewhere. 18 Differential Thermal Analysis (DTA). DTA experiments were performed on a computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~25 mg) of ground crystalline material was sealed in quartz ampoules under vacuum. A quartz ampoule of equal mass filled with A1203 was sealed and placed on the reference side of the detector. The sample was heated to the desired temperature at 10°C/min, then isothermed for 10 minutes and finally cooled to 50°C at the same rate. The stability of the sample and the reproducibility of the melting point for each sample was monitored by running multiple cycles. The residues of the DTA experiment were examined by X-ray powder diffraction. To evaluate congruent melting the X-ray powder diffraction patterns before and after the DTA experiments were compared. Semiquantitative Microprobe Analyses. The analyses were performed using a JEOL JSM—6400V scanning electron microscope (SEM) equipped with a TN 5500 EDS detector. This technique was used to confirm the presence of all elements in the compounds. Data acquisition was performed with an accelerating voltage of 20kV and twenty second accumulation time. Single Crystal X-Ray Crystallography . Intensity data for (I) was collected using a Rigaku AFC6S four-circle automated diffractometer equipped with a graphite crystal monochromator. Crystal stability was monitored with three standard reflections whose intensities were checked every 150 reflections. No crystal decay was detected for (I). The space group was determined from systematic absences and intensity statistics. An empirical absorption correction, based on 111 scans, was applied to the data set of (1) during initial stages of refinement. The structure of (I) was solved by direct methods using 197 SHELXS-86 software19a, and full matrix least squares refinement was performed using the TEXSAN software package19b. An empirical DIFABS correction20 was applied as recommended, after full isotropic refinement, after which full anisotropic refinement was performed. Intensity data for (111) was collected on a Siemens SMART-CCD diffractometer using graphite-monochromatized MoKa radiation. The data collection covered over a hemisphere of reciprocal space, up to 52° 2(-). The individual frames were measured with a detector to crystal distance of 5 cm and an omega rotation of 0.3 deg with an acquisition time of 45 seconds, leading to a total measurement time of about 24 hours. The SMART software” was used for the data acquisition, SAH‘JTZI for data extraction and reduction. The space group was determined from systematic absences and intensity statistics. A correction for Lorentz polarization effects and an empirical absorption correction (SADABS)22 were applied to the data. The structure of (III) was solvend using direct methods and refined by full-matrix least squares techniques of the SHELXTL23 package of crystallographic programs. The complete set of data collection parameters and details of the structure solution and refinement for (I) and (III) are given in Table 5-3. The coordinates of all atoms, average temperature factors, anisopropic displacement parameters and their estimated standard deviations are given in Tables 5-4 to 5-5. 198 Table 5-1. Calculated and Observed X-ray Powder Patterns for szAquS4 (I). hkl dcalc, A d obsd, A I/Imax(0135d) 020 10.34 10.20 19 100 7.33 7.00 14 110 6.62 6.91 57 120 5.98 5.93 3 021 5.83 5.78 6 040 5.17 5.13 6 111 4.93 4.90 2 121 4.56 4.53 2 140 4.22 4.19 8 041 4.17 4.14 20 131 4.09 4.07 17 200 3.66 3.67 4 141 3.62 3.64 4 210150 3.13 3.64 4 002 3.53 3.51 17 220 3.45 3.46 100 022 3.34 3.32 30 230 3.23 3.27 8 211 3.21 3.23 12 151 3.20 3.19 11 102 3.18 112 3.14 3.12 29 221 3.10 3.08 20 122 3.04 3.02 5 042 2.91 2.92 5 132 2.88 170 2.74 2.71 20 080 2.58 2.57 20 212 2.52 2.50 9 222 062 2.46 2.45 3 180 2.43 2.41 7 330 2.30 2.27 2 242 2.28 2.26 2 240 271 252 133 280 350 324 321 241 322 153 223 2.20 2.187 2.166 2.131 2.112 2.104 2.068 2.000 1.984 1.972 1.970 1.945 199 2.21 2.181 2.156 2.134 2.112 2.077 2.039 1.984 1.975 1.966 1.939 mom-boo h-P-QN 200 Table 5-2. Calculated and Observed X - Ray Powder Patterns for szAu6Sb4S10(III). h kl dCfllCn A dobs” A I”max. (obs) 0 2 0 13.03 13.01 100 0 4 0 6.51 6.51 4 2 1 0 6.05 6.03 7 2 3 0 5.05 5.05 9 0 6 0 4.34 4.34 7 0 5 1 4.17 4.16 6 1 5 1 3.95 3.95 6 1 7 0 2 60 3 01 3.562 3.563 30 0 7 1 3.284 3.283 7 12 2 3 5 0 3.245 3.244 15 2 7 0 3.196 3.193 13 4 1 0 3.088 3.089 16 4 2 0 3.025 3.025 12 l 4 2 2.980 2.980 6 2 8 0 2.887 2.886 23 1 9 0 2.822 2.820 9 4 2 1 2.774 2.770 8 1 6 2 3 1 2 2.652 2.649 6 0 100 4 4 1 2.607 2.607 13 3 3 2 2.548 2.551 7 2 100 3 8 1 2.405 2.404 12 3 5 2 2.373 2.376 7 2 7 2 2.354 2.354 8 5 4 0 2.324 2.325 5 3 10 0 2.207 2.207 5 3 7 2 2.168 2.170 6 481 1120 2.140 2.141 5 4 5 2 2.119 2.119 4 1 10 2 2.058 2.059 4 2 120 620 2.047 2.042 5 1 13 0 2102 1.978 1.974 4 4 7 2 2121 1.967 1.968 4 1112 1131 651 3121 423 670 661 622 5101 4121 690 3122 214 1142 11611123 5130 1.935 1.904 1.856 1.855 1.841 1.811 1.807 1.764 1.742 1.725 1.686 1.672 1.628 1.573 1.561 201 1.931 1.904 1.859 1.850 1.842 1.813 1.807 1.762 1.742 1.721 1.686 1.672 1.629 1.574 1.561 AWNWUIQ-b-hw-bhw-bhh 202 Table 5-3. Crystallographic Data for szAquS4 and szAu6Sb4Sw. I III Formula fw a,/31 b}. c}. 01, deg 13. deg 7. deg z; V, A3 7L, A space group dcalC9 g/cm3 11(Mo K01), cm'1 Crystal dim (m1) secondary extension T, 0C 20 max, deg no. data collected no. unique data no. data observed, I > 36(1) no. of variables final R/Rw,a% GOF szAquS4 308.95 6.804(3) 20.127(4) 7.195(4) 90.000 90.000 90.000 8; 985(1) 0.71069 P bcm (#57) 4.165 278.81 0.04 x 0.04 x 0.12 3.3234x10'S 23 50.0 1016 955 548 46 4.5/6.6 3.06 szAu6Sb4S 10 2160.33 12.4402(2) 26.0790(4) 6.9614(1) 90.000 90.000 90.000 4; 2258.3(7) 0.71069 P nnm (#63) 4.043 48.83 0.400 x 0.010 x ~0.005 NA -120 45 15609 2340 1014 159 8.0/9.6 1.56 a R = 2(lFoI-IFcI)/£|Fol; Rw = [2w(lFol—|Fcl)2/ZwlFol2]1/2 203 Table 5-4. Fractional Atomic Coordinates and B(eq) Values for szAquS4 with Estimated Standard Deviations in Parentheses. atom x y 2 Bequ2 Au] 0.3929(2) 3/4 0 234(7) Sbl 0.2086(3) 0.8896(1) 1/4 1.9(1) S 1 0.200(1) 1.0045(4) 1/4 2.4(4) S2 0.393(1) 0.8638(3) -0.014( 1) 3.3(3) S3 -0.092(1) 0.8377(5) 1/4 3.9(5) Rbl 0.2981(5) 1.0118(2) -1/4 2.7(2) Rb2 -0. 1064(7) 0.8297(2) -1/4 3 .8(2) a B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as ch = (4/3)[aZB( l , 1) + sz(2, 2) + czB(3, 3) + ab(cosg)B(1,2) + ac(cosb)B(1,3) + bc(cosa)B(2, 3) 204 Table 5-5. Fractional Atomic Coordinates and B(eq) Values for szAu6Sb4S 10 with Estimated Standard Deviations in Parentheses. atom x y z B(eq) Au(l) 0.6904(2) 0.19466(9) -0.2422(3) 1.87(5) Au(2) 0.6925(3) 0.3201(1) 0.5000 2.09(8) Au(3) 0.4852(2) 0.26812(9) -0.2514(3) 206(5) Au(4) 0.7297(2) 0.2907(1) 0.0000 1.90(7) Sb( 1) 0.3175(4) 0.3976(2) 0.0000 1.6( 1) Sb(2) 0.9352(4) 0.3454(2) 0.5000 1.8(1) Sb(3) 1.0332(4) 0.3762(2) 0.0000 1.6( 1) Sb(4) 0.4516(4) 0.3842(2) 0.5000 1.7( 1) Rb(l) 0.6643(6) 0.4329(3) 0.0000 2.7(2) Rb(2) 0.1742(7) 0.4851(3) 0.5000 2.9(2) S(l) 0.911(1) 0.4104(6) 0.251(2) 2.2(3) S(2) 0.763(2) 0.2393(8) -0.5000 2.2(5) S(3) 0.913(2) 0.2803(8) 0.0000 2.3(5) S(4) 0.640(2) 0.4046(8) 0.5000 2.0(5) S(S) 0.637(2) 0.1405(8) 0.0000 1.8(5) S(6) 0.161(2) 0.4538(9) 0.0000 2.7(5) S(7) 0.553(2) 0.3163(9) 0.0000 2.5(5) S(8) 0.408(1) 0.4486(5) -0.251(2) 2.3(4) 205 3. Results and Discussion 3.1 Description of Structures Structure of AzAquS4 (A = Rb, Cs) (1) - (11): Since A2AquS4 (A = Rb, Cs) are isostructural, the single-crystal structure determination was only done on szAquS4, therefore this discussion will refer mainly to this compound. The structure of (I) and (II) belong to the same family as CszAuPS4.l6 This is the first example that We are aware of where a thioantimonate and a thiophosphate compound are isostructural. The infinite [AquS4]n2"' chains are comprised of alternating tetrahedral [SbS4]3' units, linked together by sharing adjacent comers with monovalent gold cations, see Figure 5-1. The chains run parallel to the crystallographic c-axis and are separated by alkali cations, see Figure 5-2. The [AquS 4L12” chains propagate by a glide plane forming a sinusoidal wave that runs along the crystallographic c direction, see Figure 5-3. The chains form a projection, down the chain axis, resembling the letter "C". The Au+ ion resides on a 2-fold crystallographic axis, with a Au-S distance of 2.293(7) A. The S-Au-S angle is essentially linear at 179.6(5)°. The [SbS4]3- unit is a regular tetrahedron with Sb—S bond distances in the range of 2.30(1)A to 2.334(7)A (mean 2.32(2)A) and S-Sb-S angles in the range of 103.6(2)° to 115.6(4)°. The Au-Au' distance is 3.598(2)A, too long to represent any significant bonding interaction. The [AquS4]n2n' chains are separated by Rb+ ions which are located in two different sites. In szAquS4, Rb(l) is coordinated by nine S atoms [range of Rb(l)-S distances, 3.51(1)-3.94(1)A; Ave 3.67(6)A], and Rb(2) is coordinated by eight S atoms [3.58(1)—4.01(1)A; Ave 3.80(6)A]. Selected bond distances and angles are given in Table 5-6. 206 The structure of the [AquS4]n2"' anion is not subject to a cation effect“ as one moves from the larger Cs+ to the smaller Rb+ cation. Presumably, the difference in cation volume between Rb+ and Cs‘ is not large enough to overcome the strong preference of the Au+ cations for linear coordination. As observed prevoiusly in this dissertation, and in other systemszS, structural changes may yet occur in the case of [AquS4]n2“' anion as one moves to the smaller K+, Na+ or Li+ cations. Structure of szAu65b4Sw. The strikingly complex structure of szAu6Sb4S10 is composed of two different interpenetrating layered frameworks, [Au3Sb453]' and [Au382]', see Figure 5-4. As a result, Rb2[Au3Sb4Sg][Au382] is a more descriptive formula, and to the best of our knowledge, represents the first reported example of a compound in which a binary framework is interpenetrating with a ternary one. As shown in Figure 5-5, the [Au3Sb4S3]' layer is strongly undulating and consists of alternating [Sb4S7]n2“' one—dimensional chains bound to [Au3S]+ units. The chains alternate above and below the layer in a staggered fashion. The [Sb4S7]n2"' chain is comprised of four condensed SbS3 pyramids forming a 12 membered Sb - S ring, see Figure 5-6. Two of the SbS3 units share two comers leaving one sulfide terminal, while the other two SbS3 units share all three comers. The dimensions of the ring are 6.49(3)A (Sb(4) - Sb(2)) by 7.83(3)A (Sb(l) - Sb(3)). The Sb atoms are in pyramidal coordination with Sb-S distances in the range from 2.21(6)A to 2.65(6)A; (mean 2.46(3)A) and compare well with those reported for C82Sb4S3l and CS3Ag2Sb3883'4. The discrete [Au3S]+ unit has a pyramidal sulfide linked to three linear Au+ cations. The Au - S distances range from 2.28(4)A to 2.46(5)A and compare well with those found in KAuS513, CsAu3S214a, and AAuS14b (A = Na, K, Rb). The S - Au - S angles range from 170° to 174°. The second framework, which is interwoven with the one described above is a [Au382]' layer. The [Au382]' layer is puckered with 12 membered Au-S rings in an anti- 207 B203 motif: (ring dimensions: 6.91(2)A Au(4) to Au(4) by 7.19(3)A Au(3) to Au(3)). Figure 5-7 highlights the pyramidal sulfide and the puckered nature of the layer and Figure 5-8 shows a perpendicular view. The Au-S distances are in the range from 2.25(4)A to 2.46(5)A and the S - Au - S angles are in the range from 165° to 178°. The [Au352]' layered structure is similar but not identical to that observed in CsAu3Szl4a. Upon further inspection, it was observed that the Au+ centers in szAu68b4S10 aggregate to form a column that runs along the c - axis. There are two types of Au---Au interactions: those at 3.25.131 and below and those between 3.25 and 3.6A. Figure 5-9 shows a view perpendicular to these Au-based columns. In Figure 5-8 the interactions below 3.25A are represented as solid lines where the Au---Au interactions between and 3.25 and 3.60A are represented as dashed lines. The layers are separated by ten - coordinate Rb(l)+ [Rb( 1) - S mean = 3.59(3)A] and eight - coordinate Rb(2) [Rb(2)+ [Rb(2) - S mean = 3.50(3)A]. Selected bond distances and angles are given in Table 5-7. 208 - . . . '2 . . 2: . . . z ‘1 ’1 ‘1 Aul »: : ; S3 1; 3 S]. ;; Sbl - . - 52 Figure 5-1: View ofa single [Aqusi],2"' chain in szAquS4 (I), with labeling. 209 0 8? l”. o 8 .. JJE_§’+—_(éb & by El). > 839 0—0 ll ”‘03 @Xdc ” *8 ‘9.) 5% 1933 L L ‘L__ b” l) « ”>81 : {'— 188-6192938 .6) ,7! Eva—— 30%!) (B .) gatfl—O O— 0..- $9 (3.; W/ “’7 61) 0o 62% (9(1) 6 @419 Q Q L. Figure 5-2: ORTEP representation of szAquS4 (I) as viewed down the c-axis (50% probability ellipsoids). 210 ORTEP representation of szAquS4 (I) as viewed down the b-axis (50% probability ellipsoids). Figure 5-3: 211 Figure 5-4: ORTEP representation of szAu6Sb4S10 (III) as viewed down the c-axis (50% probability elliposids).small octant shaded ellipsoids: Ag, principal axis ellipsoids: Sb, boundary ellipsoids: S, boundary and axis ellipsoids: Rb. 212 Figure 5-5: ORTEP representation of the [Au3Sb4Sg]‘ undulating layer with labeling (50% probability elliposids). 213 51 0 5122 0 gm 0 o 0 g o o 55 o o 55 58 Sbl-l Sbl 511 Figure 5-6: ORTEP representation of the [Sb487]2' chain with labeling (50% probability ellipsoids). 214 Figure 5-7: ORTEP representation of the [Au352]' layer highlighting the pyramidal sulfides in the undulating layer (50% probability elliposids). 215 . . ‘ O . . Figure 5-8: ORTEP representation of a perpendicular view of the [Au3S2]' layer with labeling (50% probability elliposids). 216 Figure 5-9: ORTEP representation of the Au---Au interactions of the Au column. The solid lines represent Au---Au interactions under 3.25A and the dotted lines represent Au-t-Au interactions between 3.25 and 3.60A. 217 3.2. Synthesis, Spectroscopy and Thermal Analysis The syntheses were the result of redox reactions in which the gold is oxidized by polychalcogenide ions in the Ax[SbySz] flux. The Au+ centers are then coordinated by the highly charged [SbySz]n- ligands. The molten polythioantimonate flux method is very effective for crystal growth in this system. The isolation of pure crystalline products is facilitated by the residual fluxes solubility in aqueous and organic solvents. Unlike other systems of this type, the thioantimnate system lacks the same control of the Lewis basicity as in the Chalcophosphate systems. A reduced Sb3+ species can be stabilized under Lewis acidic conditions, but thedifficulity arises from the ability to predict the exact nature of the Sbn+ species in that the Sb3+ species are typically observed in compounds containing ologomeric [SbySz]n- units. Investigations in the K2S/Au/Sb/Q (S = S, Se) system resulted in largely amorphous black glassy chunks for both systems. Microprobe analysis gave an average composition of K2Au1,7SbS6 for the thioantimonate system and K1,5Aqu2,1Se2_2 for the selenoantimonate system. Attempts to grow crystals suitable for further characterization are in progress.26 In the A2S/Au/Sb/S system (A = Rb, Cs), (1) - (II) were synthesized from a metal/antimony ratio of 2:1 in a 2A286 flux at 350° C. Although an excess of gold is required for the synthesis of (I) - (II), only a small amount of unincorporated Au metal is observed as an impunity in (I), while (11) is obtained in pure form. Unlike the silver system, increasing the metal concentration did not increase the incorporation of the coinage metal into the resulting compound. The systhesis of (III) was synthesized from a metal/antimony ratio of 2:1 in a 2A2S5 flux at 350° C. Suggesting that the thioantimonate method is sensitive to even slight changes in the flux composition resulting is the stabilization of a totally different [beSy] “' units. 218 The optical absorption properties of the compounds were evaluated by examining the solid-state UV/vis diffuse reflectance spectra these materials. The spectra confirm the semiconducting nature of the materials by revealing the presence of sharp optical gaps. The A2AquS4 (A: Rb, Cs) compounds exhibit steep absorption edges from which the band gaps (Eg) can be assessed at 2.43 eV (1) and 2.61 eV (II), respectively. The optical spectrum of szAu6Sb4Sm reveals the presence of a sharp optical gap of 1.37eV, suggesting the material is a semiconductor. Representative spectra for (I) and (II) are given in Figure 5-10 and the spectrum for (111) is given in Figure 5-11. The infrared spectrum of szAquS4 displayed absorptions at 610(m), 410(s), 401(m), 390(m) 360(8) 345(w), 266(w, broad), 173(w), and 153(w) cm'l. The absorption at ca. 390 cm], is characteristic of the tetrahedral [SbS4]3‘ unit3v4. The weak absorptions below 345 cm'1 are assigned to Au-S vibrations. The far-IR spectrum of szAu5Sb4Sm displays absorptions at ca. 377 and 350 cm’1 which can be tentatively assigned to Sb - S stretching modes in the "Sb284" - like backbone of the [AU3Sb4S3]' framework”. Absorptions in the 381-347 cm'1 range are tentatively assigned to the Sb-S vibrational stretching modes by analogy with the C82Sb483l and CS3Agzsb3533. Absorptions below 347 cm“1 are assigned to Au-S vibrations. By comparison with KAuS 5 the absorption at ~323 cm'1 is assigned as an Au-S stretching vibration.13 The Raman spectrum of szAquS4 (ground crystals) showed absorbencies at 453(w) 407 (w), 391(m), 362(8), 332(5), 234(w), 175(w), and 153(w) cm'l. The vibrations at ca. 362 ch, 175 cm'l, and 153 cm.1 are characteristic of the tetrahedral [SbS4]3- unit by comparison with the Rarnan spectrum of Na3SbS4-9H4O.27 By comparison with KAuS5 the absorption at ca. 330 cm.1 can be assigned to a Au-S stretching vibration”. This stretching vibration is not observed in the infrared spectrum because the S-Au-S linkage resides on a center of symmetry. The Raman spectrum of szAu6Sb4S10 displays absorptions in the range 377-350 cm'1 which are assigned to Sb-S 219 modes and the absorptions below 350 cm'1 are assigned to Au-S stretching vibrations. The Raman spectra for (I) and (IH) are shown in Figure 5-12. The DT A of (I) shows that it melts incongruently at ca. 400°C, converting partially to Au metal and residual amorphous Ax[SbySz] flux. Compound (11) melts congruently at ca. 400°C, suggesting that large crystals may be grown from a melt. Representative spectrum for CsZAquS4 is shown in Figure 5-13. Differential thermal analysis (DTA) data, followed by careful XRD analysis of the residues, shows that RbgAung4Sm melts incongruently at ca. 442°C. Examination of the residue by powder XRD revealed that the compound decomposes to an amorphous material and Au metal respectively. The DTA spectra for two cycles for (IH) is shown in Figure 5-14. Table 5-6. Selected Distances (A) and Angles (deg) for szAquS4 with (1) Estimated Standard Deviations in Parentheses 3. Sb(1)-S(l) 2.314(9) S(2)-Au(1)-S(2') 179.9(4) Sb(1)-S(2) 2.334(7) x2 Sb(1)-S(3) 2.30(1) Au(l)-S(3)-Sb(l) 100.7(3) Sb(1)-S(mean) 2.32(1) S(l)-Sb(1)-S(2) 103.6(2) Au(l)-S(2) 2.293(7) S(l)-Sb(l)-S(2’) 103.6(2) Au(l)-S(2’) 2.293(7) S(l)-Sb(l)-S(3) 1 15.6(4) S(2)-Sb(1)-S(2') 109.0(4) Au(1)-Au(l ’) 3.598(2) S(2)-Sb(l)—S(3) l 12.1(2) S(2)-Sb( l )-S(3’) l 12.1(2) S-Sb-S (mean) 109(2) Rb(1)-S(1) 3.662(3) x2 Rb(2)-S(l) 3.40(1) Rb(1)-S(1') 343(1) Rb(2)-S(2) 3.869(8) x2 Rb(1)-S(1") 340(1) Rb(2)-S(2') 3.858(8) x2 Rb(l)-S(2) 3.488(7) x2 Rb(2)-S(3) 3.602(2) x2 Rb(l)-S(2') 3.782(8) x2 Rb(2)-S(3) 3.37(l) Rb(l)-S(3) 334(1) Rb(2)—S (mean) 3.67(7) Rb(l)-S (mean) 355(5) alThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 61 = {XnUn - [)2/n(n - 1)}1’2, where 1,, is the length (or angle) of the nth bond, I the mean mength (or angle), and n the number of bonds. 221 Table 5-7. Selected Distances (A) and Angles (deg) for szAu6Sb4Sm (III) with Standard Deviations in Parentheses. Au(1)- S(3) 2.32(1) Sb(l) - S(6) 243(2) Au(l) - S(5) 2.30(1) Sb(l) - S(8) 2.47(1) x2 Au(2) - S(2) 2.28(2) Sb(2) - S(5) 2.54(2) Au(2) - S(4) 230(2) Sb(2) - S(l) 2.48(1) x2 Au(3) - S(3) 2.32(1) Sb(3) - S(6) 258(2) Au(3) - S(7) 2.31(1) Sb(3) - S(l) 2.48(1) x2 Au(4) - S(3) 229(2) Sb(4) - S(4) 2.40(2) Au(4) - S(7) 230(2) Sb(4) - S(8) 2.47(1) x2 Au(1)- Au(3) 3.192(3) Au(l) - Sb(l) 3.393(5) Au(l) - Au(4) 3.060(4) Au(l) - Sb(3) 3.233(5) Au(2) - Au(4) 3.09(1) Au(2) - Sb(2) 3.090(6) Au(1)-Au(l') 3.589(4) Au(l')-Au(3') 3.461(5) Au(3)-Au(3') 3.461(5) S(2)-Au( 1 )-S(5) 170.9(7) S( l )-Sb(3)-S( l ’) 89.4(7) S(2)—Au(2)-S(4) 173.9(8) S( l )-Sb(3)-S(6) 95.6(5) S(3)-Au(3)-S(7) 178.4(7) S( l ')-Sb(3)-S(6) 95.6(5) S(3)-Au(4)-S(7) 170.0(8) 222 (A) :19: 1-6 l l l l l l C 3 1.4 - - .o' 5 1.2 — — 3:- 1.0 — ~ 8 o 0.8 - " E o_5 — szAquS4 - 55‘ 0.4 — Eg=2.43 ev — (I) 2 0.2 - - (g 0.0 l I I l I 0 1 2 3 4 5 6 7 Energy (eV) (B) 2710-0 l l l l l l c 3 9' 8.0 — — 55 g“ 6.0 - - o 0 g 4.0 - Cs AquS " .5 2 4 9 E =2.61eV .9, 2.0 - g - n (D I l J l l B 0.0 0 1 2 3 4 5 6 7 Energy (eV) Figure 5-10: Solid-state optical absorption spectra for szAquS4 (I) and CszAquS4 (II). 223 E 0'3 T I T l T l C =fo.zs . .Q a ‘7 0.2 a: 8 0015 (3 Rb 2A1! 68b 481 0 g 0.] E8: 1.33 CV O U) .0 < (I) a O O O (n J L l- )- 0 1 2 3 4 S 6 7 Energy (eV) Figure 5-11: Solid-state optical absorption spectrum for szAu68b4Sw (IH). 224 (A) 7 I I l j A 362 a s - - E 5 L 332 7 3 4 .. - E .2 391 a, 3 '- 153 '- E 407 — 453 175 t: 2 L 234 - E a 1 ' i O l J l I 600 500 400 300 200 100 Wavenumbers (cm" ) (B) 1.4 7 V 1' I' 250 Raman Intensity (arb. units) 0.4 . - 4 600 500 400 300 200 100 Wavenumbers (cm'1 ) Figure 5-12: Raman spectra of szAquS4 (I) and Rb2Au6Sb4S10 (HI). endo <1“— DTA (11V) ———1> exo 225 4 I L l Figure 5-13: 200 300 400 500 600 Temperature ("C) DTA diagram for CszAquS4 (II). (B) 226 (D X 0 I 01 O end ‘1 5 l- -l endo -20 . . . . . 0 100 200 300 400 500 600 Temperature (°C) (A) 5 ‘I Y Y Y 8X0 1‘ .. 1 start ; \ 3 -5 -l a) 8 g -10 _ \ 393 , c: end -15 . endo -20 a 1 i 1 g 0 100 200 300 400 500 600 Temperature (°C) Figure 5-14: Two cycles for DTA diagrams for szAu6Sb4Sm (IH). 227 4. Conclusions In summary, these new members of the A/Au/Sb/S family emphasize the value of the thioantimonate variation in the flux method. These fluxes provided a convenient entry into the previously unknown gold chemistry, and the potential exists that more quaternary and possibly the unknown ternary Au/Sb/Q (Q = S, Se) compounds26 may yet be discovered. szAu68b4S10 represents the first example of a sulfosalt with two different interpenetrating anionic frameworks. 1) 2) 228 List of References McCarthy, T. J.; Kanatzidis, M. G. Inorg. Chem. 1994, 33, 1205. Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G. Inorg. Chem., 1997, 34, 1234. (a) Hanko, J. A.; Kanatzidis, M. G. Abstract #0413 from 210th Fall ACS Meeting, Chicago IL 1995. (a) Schimek, G. L.; Pennigton, T. L.; Wood, P. T.; Kolis, J. W. J. Solid State Chem. 1996, 123, 272. (b) Wood, P. T.; Schimek, G. L.; Kolis, J. W. Chem. Mater., 1996, 8, 721. Hanko, J. A.; Kanatzidis, M.G. in preparation. Imafuku, M.; Nakai, I.; Nagashima, K. Mat. Res. Bull. 1986, 21 , 493. For example see: (a) Cordier, G.; Schafer, H. Rev. Chem. Min., 1981, I8, 218. (b) Dorrscheidt, W.; Schafer, H. Z Naturforsch., 1981, 36b, 410. (c) Cordier, G.; Schwidetzky, C.; Schafer, H. Rev. Chem Min., 1982, 19, 179. (a) Sheldrick, W.; Wachhold, M. Submitted for publication. (b) Drake, G. W.; Kolis, J. W. Coord. Chem. Rev., 1994, I37, 131. Garin, J.; Parthe, B.; Oswald, H.R.; Acta Crystallogr., 1972, 328, 3672. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 229 Edenharter, A.; Nowacki, W.; Takeuchi, Y.; Z Kristallogr., 1970, 131, 397. Ito, T.; Nowacki, W.; Z Kristallogr., 1974, 139, 85. Wuensch, B. J. Z Kristallogr., 1964, 119, 437. Park, Y.; Kanatzidis, M. G. J. Alloys Compd., 1997, 25 7, 137. (a) Klepp, K. O.; Weithhaler, C. J. Alloys Comp., 1996, 243, 1. (b) Klepp, K. O.; Weithhaler, C. J. Alloys Comp., 1996, 243, 12. (c) Bronger, W.; Kathage, H. U. J. Alloys Comp., 1992, 184, 87. (d) K. O. Klepp, W. Bronger, J. Less- Common Met, 1987, 127, 65. (e) Klepp, K. O.; Bronger, W. J. Less-Common Met, 1987, I32, 173. (f) Klepp, K. O.; Bronger, W. J. Less-Common Met, 1988, I37, 13. (g) Klepp, K. O.; Brunnbauer, G. J. Alloys Comp., 1992, 183, 252. Kim, K.- W.; Kanatzidis, M. G. J. Am. Chem. Soc., 1992, 114, 4878. Chondroudis, K.; Hanko, J.A.; Kanatzidis, M.G. Inorg. Chem, 1997, 36, 2623. CERIUS2, Version 1.6, Molecular Simulations Inc., Cambridge, England, 1994. McCarthy, T. J.; N geyi, S.-P.; Liao J.-H.; DeGroot, D.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem Mater., 1993, 5, 331. (a) Sheldrick, G. M. In Crystallographic Computing 3; Sheldrick, G.M., Kruger, C., Doddard, R., Eds.; chford University Press: Oxford. England, 1985; pp 175- 20. 21 22. 23. 24. 25. 26. 27. 230 189. (b) TEXSAN: Single Crystal Structure Analysis Software, Version 5.0; Molecular Structure Copr.; The Woodlands, TX 77381, 1981. Walker, N., Stuart, D. Acta Crystallogr. 1983,A39, 158. SMART and SAINT. Data collection and Processing Software for the Smart system. Siemens Analytical X-ray Instruments Inc.,l995. Blessing , R. H. Acta Crystallogr. 1995,A51, 33. SHELXTL Version 5, Reference Manual. Siemens Industrial Automation Inc., 1994. Kanatzidis, M. G. Phosphorous, Sulfur, and Silicon, 1994, 93-94, 159. McCarthy, T. J.; Kanatzidis M.G. Inorg. Chem, 1995, 34, 1257. Hanko, J. A.; Kanatzidis, M.G. Work in Progress. Siebert, H. Z Anorg. Allg. Chem, 1954, 275, 225. CHAPTER 6 A2CuP3S9 (A = K, Rb), CszCu2P2S6, and K3CuP2S7: New Phases from the Dissolution of Copper in Molten Polythiophosphate Fluxes. 231 232 1. Introduction The recent activity in the application of Chalcophosphate fluxes has uncovered a multitude of new materials with elaborate structure types arising from a variety new binding modes displayed by the [Pny]n"' (Q = S, Se) units. An interesting trend, which has become evident from this work, is that the [PxSey]n“' species behave very differently than the corresponding [PxSy]n"‘ species. This trend is manifested in not only in how the [Pny]n°' (Q = S, Se) species bind to metal ions but also with respect to their relative stability. For example the tetrahedral [PQ4]3' (Q = S, Se) unit, typically does not lead to the same structure types when they form compounds with the same formula. In fact, they do not give compounds with the same formula. This phenomenon is reminiscent of the similar differences previously observed between the polychalcogenide 8,} and Senz' species.2 Furthermore, the analogous synthetic conditions which may give rise to a certain [PxSey]n2' unit tend to give rise to a different [PxSy]n2‘ unit, often the anions differing in the oxidation state of the phosphorous atom.3 The latter is the consequence of the difference in the electronegativity between the S and Se atoms. These facts underscore the need for the simultaneous exploration of both the thiophosphate and selenophosphate fluxes for the same metal system, and the continued search for fundamental differences in the chemistry and reactivity. The polychalcophosphate fluxes form by the in situ fusion of AzQ/PzQle (A = K, Rb, Cs; Q = S, Se) and contain various [Pnyln' units available for reaction with metal ions. Typically, this type of coordination chemistry leads anionic frameworks stabilized by alkali metal cations. The generality of this method has been demonstrated by the ability to incorporate main group“, transition5, lanthanide, and actinide5°v6 metals into solid state chalcophosptate compounds. The highly basic nature of the fluxes permits the suppression of the highly stable M2P2Q5 (Q = 8, Se) phases. The M2P2Q5 (Q = S, Se) structure type . 7.8 . . . 15 related to that of CdIz, and 1t pr0jects altematlng, octahedrally coordinated, metal ions 233 and the ethane-like [P2Q6]4’ (Q = S, Se) units. Some compounds possess interesting magnetic9 and intercalation chemistry,10 while others have been studied intensively as potential candidates for cathode materials in secondary lithium batteries.11 Substitution on the divalent metal cations by monovalent coinage metal cations results in M’4P2Q6 (e. g. M’ = Ag, Q = S; M’ 2 Cu, Q = Se),12 whose structure retains the same M2P2Q6 structural motif, but now each octahedral metal site is replaced by two M'+ ions with trigonal planar coordination. The two divalent metal ions can also be replaced by a monovalent coinage and a trivalent main group metal ion to give M’M"P2Q6, again retaining the overall M2P2Q6 structure type”. Based on the limited body of work in the multinary copper Chalcophosphate systems, a rare example is the ternary Cu3PQ4 (Q = S143, SeMb), we became interested in applying the alkali thiophosphate flux approach to these systems. The reaction of the coinage metals with polychalcophosphate fluxes has already resulted in a number of unusual compounds; examples include CszMszSeg (M = Cu, Ag)5a, A3AuP2Seg (A = K, Rb, Cs),3 AzAuszSeg (M = Rb, Cs),3 K3Ag3P3Se9,Sa K3Cu3P3Se9,5° K2Cu2P4Sero,5f AzAuPSa (A = K, Rb, Cs),3 AAuPzS7 (A = K, Rb)3 and reciently KAu5P2S3.15 An interesting counterion effect exists in the A3Ag3P3Se9 system. The large Cs+ counterion, gives [M2P2Se6]n2“' (M = Cu, Ag) consisting of infinite chains of alternating [P2Se6]4' units and Mg“ dimers. Upon moving to the smaller K+ cation, the K3Ag3P3Se9 (M = Cu5°, Ag5a) adopts a three-dimensional structure comprised of ethane-like [P2Se6]4' units and tetrahedral Cu+ or Ag+ ions in a complex but elegant bonding scheme. Although one might think that the thiophosphate and selenophosphate systems would produce isostructural compounds, the gold chalcophosphates3 provide an excellent example of how chemically divergent the chemistry of two systems can be. The chemistry of copper presents similar differences between the sulfur and selenium systems. Here, we report the synthesis, structure spectroscopic and thermal properties of the quaternary copper thiophosphate compounds, A2CuP3S9 (A = K, Rb), CuzCuzPZSG and K3CuP2S7. While 234 the selenide analog of CuzCu2P2S6 has already been reported, the selenide analogs for A2CuP389 (A = K, Rb) and K3CuP2S7 are still unknown and it seems unlikely that they would be stable. The K2CuP3S9 features the unprecedented cyclic [P3S9]3' anion. The results of the band structure calculation for this compound will also be discussed. 2. Experimental Section 2.1. Reagents The reagents mentioned in this study were used as obtained unless noted otherwise: (i) Cu metal powder electrochemical dust, Fisher Chemical Co., Pittsburgh, PA; (ii) phosphorous pentasulfide (P2S5) 99.999% purity, Aldrich Chemical Co., Milwaukee, WI.; (iii) Cesium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (iv) Rubidium metal, analytical reagent, Johnson Matthey/AESAR Group, Seabrook, NH; (v) Potassium metal, analytical reagent, Aldrich Chemical Co., Milwaukee, WI.; (vi) Sulfur powder, sublimed, J.T. Baker Chemical Co., Phillipsburg, NJ; (vii) Methanol (MeOH) ACS anhydrous, EM Science, Inc., Gibbstown, NJ. (viii) Diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. 2.2. Syntheses. A 28 (A = K, Rb, Cs) were prepared by reacting stoichiometric amounts of the elements in liquid ammonia as described in Chapter 2, Section 2.2. Preparation of KzCuP389 (I). An amount of 0.032g (0.50 mole) Cu, 0.333g (1.50 mole) P2S5 and 0.110g (1 mole) K28 was thoroughly mixed and sealed 235 under vacuum a Pyrex tube. The reaction was heated to 550°C in 15 hours in a computer controlled furnace. It was isothermed at 550°C for 4 days, cooled to 150°C at a rate of 4°C/hr, and then cooled to room temperature in 10 hours. The excess Kx[PySz] flux was removed by washing with degassed methanol under inert atmosphere revealing yellow crystals (75% yield based on Cu). The yellow crystals are stable in air for several months. but decompose in the presence of water. The product is occasionally contaminated with brown-orange crystals of Cu3PS4 and/or an intractable black powder. Preparation of szCuP389 (II): An amount of 0.032g (0.50 mole) Cu, 0.220g (1 mole) P285 and 0.203g(1 mmole) szs was sealed under vacuum in a Pyrex tube and heated as in (I). The product, which is stable in air but decompose in the presence of water, was isolated as in (I) to give yellow crystals (52% yield based on Cu). Preparation of CszCu2P285 (111): An amount of 0.032g Cu (0.50 mole), 0.330g P255 (1.50 mole) and 0.298g C82S (1 mole) was sealed under vacuum in a Pyrex tube and heated as in (I). The product was isolated as in (I) to give a mixture of dark violet and yellow crystals. The mixture was quickly washed with degassed water to give dark violet crystals in approximately 50% yield based on Cu. The resulting dark violet crystals are stable in air and water. Preparation of K3CuP287 (IV): An amount of 0.016g Cu (0.25 mmole), 0.110g P2S5 (0.50 mole) , 0.550g K28 (0.50 mmole), and 0.048g (1.5 mmole) S that was sealed under vacuum in a Pyrex tube. The reaction was heated to 500°C in 15 hours in a computer controlled furnace. It was isothermed at 500°C for 4 days, cooled to 110°C at a rate of 4°C/hr, and then cooled to room temperature in 10 hours. The product, is stable in air but decompose in water. The compound was isolated as in (I) to give transparent colorless crystals (40% yield based on Cu). 236 2.3. Physical Measurements. Powder X-ray Difi‘raction. The compounds were examined by Powder X-ray diffraction to check for phase purity. Accurate d spacings obtained from the powder pattern for (I) were recorded on a CPS 120 INEL X-ray powder diffractometer using monochromatized radiation Cu K-Lm (3:1.540598A). Accurate d spacings obtained from the powder patterns for (111) - (IV) were recorded on a Phillips XRG-3000 computer- controlled powder diffractometer with Ni filtered CuKa radiation. Both diffractometers are equipped with a curved position-sensitive detector calibrated with NazCa3A112F14. The calculated powder pattern for (I) was prepared with the Lazy Pulverix Programma. While the calculated powder patterns for (111) - (IV) were prepared with the CERIUS2 software.16b Tables of calculated and observed XRD patterns for (I), (III) and (IV) are given in Tables 6-1, 6-2, and 6-3, respectively. After crystal collection and structure refinement (vide infra) for (I), the final cell parameters were determined by least squares refinements made on the powder X-ray data and are given in Table 6-4. Infrared Spectroscopy. Infrared spectra, in the far-IR region (600-50 cm'l), were recorded on a computer controlled N icolet 750 Magna-IR Series II spectrophotometer equipped with a TGS/PE detector and silicon beam splitter in 4 cm'1 resolution. The samples were ground with dry Csl into a fine powder and pressed into translucent pellets. Raman Spectroscopy. Raman spectra (600-100 cm") were recorded on a BIO- RAD FT Raman spectrometer equipped with a Spectra-Physics Topaz T10-106c 1.064 nm YAG laser and a Ge detector. Crystals of (I), - (HI) were loaded without modification into glass capillary tubes. Crystals of (IV) were ground with Csl into a fine powder and loaded into a glass capillary tube. 237 Single Crystal Optical Transmission. Room temperature single crystal optical transmission spectra were obtained on a Hitachi U-6000 Microscopic FT Spectrophotometer mounted on an Olympus BH2-UMA metallurgical microscope over a range of 380 to 900 nm. Crystals lying on a glass slide were positioned over the light source and the transmitted light was detected from above. Solid State U V/Vis/Near IR Spectroscopy. Optical diffuse reflectance measurements were performed at room temperature using a Shimadzu UV-3101PC double beam, double monochromator spectrophotometer. The instrument is equipped with integrating sphere and controlled by a personal computer. BaSO4 was used as a 100% reflectance standard for all materials. Samples are prepared by grinding them to a fine powder and spreading them on a compacted surface of the powdered standard material, preloaded into a sample holder. The reflectance versus wavelength data generated can be used to estimate a material’s band gap by converting reflectance to absorption data as described earlier. ‘7 Differential Thermal Analysis (DTA). DT A experiments were performed on a computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~25 mg) of ground crystalline material was sealed in quartz ampoules under vacuum. A quartz ampoule of equal mass, filled with A1203, was sealed and placed on the reference side of the detector. The sample was heated to the desired temperature at 10°C/min., then held there for 10 min. and finally cooled to 50°C at the same rate. The stability/ reproducibility of the samples were monitored by running multiple cycles. Residues of the DTA experiment were examined by X-ray powder diffraction. To evaluate congruent melting the X-ray powder diffraction patterns before and after the DTA experiments were compared. Semiquantitative Microprobe Analyses. The analyses were performed using a JEOL JSM-6400V scanning electron microscope (SEM) equipped with a TN 5500 EDS detector. This technique was used to confirm the presence of all elements in the 238 compounds. Data acquisition was performed with an accelerating voltage of 20kV and twenty second accumulation time. Electronic structure calculations. The band structure was calculated using density functional theory with the local density approximation. The one-electron Shrédinger equation was solved self consistently using the tight-binding (TB) linear muffin-tin orbital (LMTO) method in the atomic spheres approximation (ASA), including the combined correction.18 This method splits the crystal space into overlapping atomic spheres (Wigner-Seitz spheres), whose radii are chosen to completely fill the crystal volume. In the present calculation, 29 additional empty spheres (E) had to be included to model the interstitial space. We allowed an overlap of 15% for the atoms centered spheres, 20% for the interstitial spheres with atomic ones, and 25% between empty spheres. A summary of the sphere radii used in our calculation is in reference (19). All reciprocal space integrations were performed with the tetrahedron method20 using 128 k-points within the irreducible Brillouin zone. The basis sets consisted of 4s, 4p and 3d orbitals for Cu, 38 and 3p orbitals for P and S, and 4s orbitals for K. The 3d orbitals for P and S, the 4p and 3d orbitals for K and E (empty spheres) p-d states were downfolded using the Lowdin’s technique.20 Single crystal X—Ray Crystallography. Intensity data for (I) were collected using a STOE Image Plate X-Ray diffractometer for room temperature data collection. The data was collected as a series of diffraction patterns recorded by rotating a randomly oriented crystal in the X-ray beam. Images were recorded over the range 4) = 00° to 200.2° with a 07° increment angle and a 4 minute irradiation per image. Cell parameters were determined from a least square analysis of the setting angles of 5000 reflections in the 38° = 29: 56.3° domain and led to the parameters a = 6.8307(3)A, b = 23.565(1)A, c = 9.5398(5)A, B = 100.213(5)°. The reflection intensities were recorded in the -18 S h S 18, -8 S k S 8, and -11 S l S 11 space. After the Lorentz polarization correction of the 10692 raw data, a set of 4032 reflections with I S 30(1) was used. A data analysis indicated a 21m Laue 239 symmetry with limiting conditions consistent with the P21/n space group. The structure was then solved from direct methods using the SHELXT L program213 and refined using the JANA’9621b software package. Conventional atomic and anomalous scattering factors were taken from the usual sources. The atomic parameters and the scale factor were refined in a full matrix mode minimizing the function wR=[2w(|Fo|-|Fc|)2/£wF02]“2. A, weighting scheme based on 0(F0) corrected with an instability coefficient was used. No anharmonic refinement was needed to account for the behavior of the Cu atomzz. Intensity data for (111), and (IV) were collected using a Rigaku AFC6S four-circle automated diffractometer equipped with a graphite crystal monochromator. Unique data for (111) - (IV) were collected out to 500 in 20. Crystal stability was monitored with three standard reflections whose intensities were checked every 150 reflections. No crystal decay was detected for either (HI) - (IV). The space groups were determined from systematic absences and intensity statistics. An empirical absorption correction based on u! scans was applied to both data sets. An empirical DIFABS correction23 was applied after full isotropic refinement, following which full anisotropic refinement was performed. The structures of (III) - (IV) were solved by direct methods using SHELXS-86 software24a, and refined with full matrix least squares techniques using the TEXSAN software package24b. The complete data collection parameters and details of the structure solution and refinement for (I), (III), and (IV) are given in Table 6-4. The coordinates of all atoms, average temperature factors, and their estimated standard deviations are given in Tables 6-5 to 6-7. Table 6-1. Calculated and Observed X-ray Powder Patterns for K2CuP3S9 (I). 240 h k l dcalc dobs Iobs (‘70) 0 2 1 7.344 7.335 99 1 0 -1 5.995 6.000 17 0 4 0 5.891 5.887 19 1 2 0 5.840 5.837 65 l 0 1 5.058 5.054 3 1 1 1 4.945 4.944 8 0 0 2 4.695 4.693 14 1 4 0 4.431 4.432 7 0 2 2 4.361 4.363 9 1 3 1 4.252 4.254 9 0 5 1 4.212 4.207 6 1 4 -1 4.202 4.207 6 0 3 2 4.030 4.029 27 -1 2 2 3.971 3.971 11 0 6 0 3.927 3.929 50 1 4 1 3.837 3.840 7 0 4 2 3.671 3.675 4 0 6 1 3.623 3.624 4 1 5 1 3.445 3.445 2 1 2 2 3.41 1 3.41 1 2 2 1 -1 3.327 3.325 16 1 3 2 3.245 3.246 5 2 2 -1 3.232 3.232 25 -1 5 2 3.143 3.146 9 0 1 3 3.103 3.101 11 0 2 3 3.025 3.024 15 2 0 -2 2.997 2.996 1 1 2 1 1 2.976 2.978 9 0 8 0 2.945 2.946 50 0 3 3 2.907 2.906 10 1 5 2 2.842 2.843 3 2 3 1 2.803 2.802 72 NNOHN—‘OO b9 c>.b \c \D to h: —-.> 4> O\2s7,33 and AAuP287 (A = K, Rb).3 The Cu+ cation is in a trigonal planar coordination with Cu-S bond distances ranging from 2.226(4)A to 2.278(4)A and S—Cu-S angles in the range from 115.3(1)° to 125.7(2)°. The coordination environment of the coinage metal site on (III) and AAuP287 (A = K, Rb) is worth discussing here. The square planar coordination of the Au3+ ions is formed by bonding to two of the three terminal sulfides from each PS4 tetrahedron. The binding scheme for the coordination of the of Cu+ ions in (IV) is completely different. One [P287]4' unit chelates in a bidendtate fashion to the Cu+ ion and the third coordination site is filled by a sulfide from a neighboring [P287]4‘ unit. The structure of (IV) is related to (I) where the chains are constructed by two of the terminal sulfides of the polythiophosphate units binding to one Cu+ ion and then bridging 253 the neighboring Cu+ ion. In (I) the [P389]3' unit chelates in a tridentate fashion forming the basal plane of the coordination sphere. The removal of a second [P82] vertex results in the [P287]4‘ unit that chelates in bidentate fashion to the Cu+ ion forming six membered rings. This coordination twists the [P287]4' unit preventing the Cu+ ion from adopting a tetrahedral coordination environment. Each PS4 tetrahedron is slightly distorted with P-S distances ranging from 1.987(4)A to 2.167(4)A. and S-P-S angles ranging from 99.2(2)o to 115.5(2)°, respectively. Selected bond distances and angles are given in Table 6-10. The potassium cations located in three different sites. In K3CuP287 K(l) is coordinated by seven sulfur atoms [range of K(l)-S distances, 3.178(5)A-3.527(5)A; average 3.32(4)A]. K(2) is also seven-coordinate [3.325(5)A-3.531(5)A. average 3.38(1)A], and K(3) is seven-coordinate [3.137(4)A-3.642(5)A, average 3.31(4)A]. 254 Table 6-8. Selected band distances (A) and angles (°) for K2CuP389 (I) with Standard Deviations in Parenthesesa. Cu-S(1) 2.305(2) S(l)-Cu-S(2) 1 1400(9) Cu-S(2) 2.299(2) S(l)-Cu-S(3) 1 10.25(8) Cu-S(3) 2.332(2) S(l)-Cu-S(4) 109.43(8) Cu-S(4) 2.303(2) S(2)-Cu-S(3) 99.49(8) S(2)-Cu-S(4) l 1 l.23(8) Cu-S(mean) 2.310(2) S(3)-Cu-8(4) 1 12. 18(8) P( l )-S(2) 1.987(3) S(2)-P(1)-S(3) 1 12.7( 1) P(1)-S(3) 1.986(3) S(2)-P( 1)-S(6) 104.1(1) P(1)-S(6) 2.1 17(3) S(2)-P(1)-S(7) 105.0(1) P(1)-S(7) 2.124(3) S(3)-P(1)-S(6) 1 14.0( 1) S(3)-P(l)-S(7) 1 12.7(1) P(1)-S(mean) 2.054(3) S(6)-P(1)-S(7) 107.5(1) P(2)-8(1) 1.989(3) S(1)-P(2)-S(7) 112.5(1) P(2)—8(7) 2.129(3) S(1)-P(2)-S(8) 1 14.3( 1) P(2)-S(8) 2.109(3) S(1)-P(2)-S(9) 115.8(1) P(2)-8(9) 1.951(3) S(7)-P(2)-S(8) 104.7(1) S(7)-P(2)—S(9) 105.0(1) P(2)-S(mean) 2.045(3) S(8)-P(2)-S(9) 103.4(1) P(3)-S(4) 1.980(3) S(4)-P(3)-S(5) l 16.1(1) P(3)-S(5) 1.975(3) S(4)-P(3)-S(6) l 12.4(1) P(3)-S(6) 2.131(3) S(4)-P(3)-8(8) 1 14.6(1) P(3)-S(8) 2.108(3) S(5)-P(3)-S(6) 103.1(1) S(5)-P(3)-S(8) 103.9(1) P(3)-S(mean) 2.049(3) S(6)-P(3)-S(8) 105.4(1) 255 Table 6-9. Selected Distances (A) and Angles (deg) for C82Cu2P256 (HI) with Standard Deviations in Parenthesesa. P(1)-S(1) 2.025(8) S(1)-P(l)-S(2) 1 113.8(2) P(1)-S(2) 2.020(7) S(1)-P(l)-S(3) 1 10.0(2) P(1)-S(3) 2.032(7) S(2)-P(1)-S(3) 1 13.5(2) S(1)-P(1)—P(2) 106.0(2) P(2)-S(4) 2.032(7) S(2)-P(1)-P(2) 105.0(2) P(2)-8(5) 2.023(7) S(4)-P(1)-P(2) 108.0(2) P(2)-S(6) 2.014(8) P(1)-P(2) 2.245(5) S(4)-P(2)-S(5) 1 13.3(2) S(4)-P(2)-S(6) l 1 1.5(2) Cu(l)—S(1) 2.293(4) S(5)-P(2)-S(6) 1 12.9(2) Cu(1)-S(2) 2.307(4) S(4)-P(2)-P(l) 107.8(2) Cu(1)-S(3) 2.264(4) S(5)-P(2)-P(1) 104.7(2) S(6)-P(2)-P( 1) 106.6(2) Cu(2)-S(4) 2.314(4) Cu(2)-8(5) 2.262(4) Cu(2)-S(6) 2.293(4) Cu(1)-Cu(1')-S(2) 81 . 1(1) Cu( 1 )-Cu( 1 ')-S(6) 105.1(1) Cu(1)-Cu(1') 2.716(3) Cu(1)-Cu(l')-S(3) 96.8(1) Cu(2)-Cu(2') 2.724(3) S(2)-Cu(1)-S(6) 109.6(1) S(2)-Cu(1)-S(3) 124.0(1) S(6)-Cu(1)-S(3) 124.4(1) Cu(2)-Cu(2‘)-S( 1) 101.4(1) Cu(2)-Cu(2')-S(4) 96.6(1) Cu(2)-Cu(2')-S(6) 83.4( 1) S(1)-Cu(2)-S(4) 109.2(1) S(1)-Cu(2)-S(5) 122.2(1) S(4)-Cu(2)-S(5) 127.2(1) 21The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 0'] = (2,10,, — [)2/n(n — 1)}1/2, where In is the length (or angle) of the nth bond, 1 the mean mength (or angle), and n the number of bonds. 256 Table 6- 10. Selected Distances (A) and Angles (deg) for K3CuP287 (IV) with Standard Deviations in Parenthesesa. P( l )-S( 1) 2.039(5) S( l )-P( l )-S(2) 100.2(2) P(1)-S(2) 2.167(4) S(1)~P(1)-S(4) 108.2(2) P(1)—S(4) 2.025(5) S(1)-P(l)-S(7) 11 1.0(2) P( 1 )-S(7) 1.987(5) S(2)-P(1)-S(4) l 12.3(2) . S(2)-P(1)-S(7) 1 l 1.6( 1) P(2)-S(3) 2.043(5) S(4)-P(1)-S(7) l 12.7(2) P(2)—S(2) 2.138(5) S-P(l)-S (mean) 109(1) P(2)-S(8) 2.017(5) P(2)-8(5) 1.994(5) S(2)-P(2)-S(3) 108.5(2) S(2)-P(2)-S(5) 1 1 1.1(2) Cu(1)-S(1) 2.278(4) S(2)-P(2)-S(8) 99.2(2) Cu(1)-S(3) 2.226(4) S(3)-P(2)-S(5) 1 12.5(2) Cu(1)-S(4) 2.226(4) S(3)-P(2)-S(8) 109.0(2) S(5)-P(2)-S(8) 1 15.5(2) S-P(2)-S (mean) 109(1) P( l )-S(2)-P(2) l 1 1.1(2) S(1)-Cu-S(3) 1 18.7(2) S(1)-Cu-S(4) 1 15.3(1) S(3)-Cu-(4) 125.7(2) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 0'1 = {2n(ln — [)2/n(n — 1)}1/2, where 1,, is the length (or angle) of the nth bond, 1 the mean mength (or angle), and n the number of bonds. 257 Figure 6-1: ORTEP representaion of K2CuP389 down the a-axis (50% probability ellipsoids). Small open cycles; Cu, small solid circles; P atoms, stripped cycles; 8, and large cycles; K. 258 Figure 6-2: Perspective view of a [CuP389]2' chain with the labeling scheme. 259 Figure 6-3: Perspective views of the tri-teuahedral [P38913' ring observed in K2CuP389 (a), the di-tetrahedral [st612' ring (b) and quadri-tetrahedral [P481214'(c) observed respectively in 2D- and 3D-NbP283. 260 Figure 6-4: ORTEP representation of the packing diagram of C82Cu2P286 as viewed along the [101] direction. 261 Figure 6-5: View of a single [Cu2P286]29' chain, with labeling. Figure 6-6: 262 57 53 P2 52 55 P1 3 3 3 C01 81 511 86 View of a single [CuP287]2"' chain with 1abe1ing. 263 Figure 6-7: ORTEP representaion of K3CuP287 as viewed down the b-axis. Small open cycles; Cu, small solid circles; P atoms, stripped cycles; 8, and large cycles; K. 264 3.2. Synthesis and Comparison between the Thiophosphate and Selenophosphate Fluxes: Throughout our exploration of the Chalcophosphate fluxes, we have observed a significant difference between the thiophosphate and selenophosphate systems. While the thiophosphate flux appears to favor the P5‘*' species, the selenophosphate flux displays a tendency to favor the reduced P4 species. The stability of the reduced P+4 species is consistent with the lower oxidizing power of the Sexz‘ ligands compared to 8x2“ ligands. The Lewis basicity of Chalcophosphate flux controls the nature of the [Pny]“' units observed. In the selenophosphate system, the basicity is controlled by changing the A28e concentration, and in the thiophosphate system varying the P285 concentration changes the Lewis basicity. The tetrahedral [PQ4]3' (Q = 8, Se) unit is observed in both systems under highly basic conditions (high A2Q or low P285). While the [P28e6]4' unit is a repeatedly observed under Lewis acidic conditions (low A2Q), other reduced phosphorous species have been reported.5d Further increasing the Lewis acidity, as observed in the synthesis of K2Cu2P48e10,5f condenses two [P28e6]4' units together via bridging selenides forming a [P4Se10]4' unit, containing a cyclohexane-like ring. In contrast, in the thiophosphate system, decreasing the Lewis basicity of the flux (i.e. higher P285 concentration), favors the formation of higher nuclearity [PxSy]“' units with phosphorous centers mainly in the +5 oxidation state. To further explore these phosphorous-rich conditions the elemental sulfur was removed from the reaction conditions. A molten thi0phosphate flux is still formed by the combination of A28 and P285, but the oxidative properties would be different because of the absence of S-S bonds. Here a portion of the thiophosphate flu becomes sacrificial, oxidizing the coinage metal. The formation of the P+4 centers in C82Cu2P286 is rationalized this way. The exact influence of these phosphorous-rich conditions on product outcome is not well understood. 265 3.3. Physicochemical Properties The optical properties of (I), (H) and (IV) were evaluated by examining their single crystal optical transmission spectra. The optical absorption properties of (III) was examined by UV/vis diffuse reflectance spectroscopy. The spectra show that these materials are wide band-gap semiconductors. The A2CuP3S9 (A = K, Rb,) compounds exhibit steep absorption edges and band-gaps, Eg, of 2.57 eV (1), and 2.60 eV (11), respectively, see Figure 6-8. The band-gap of K3CuP287 is 2.95 eV. The compound Cs2Cu2P286 exhibits a steep absorption edge and a band-gap, Eg, of 2.43 eV, see Figure 6-9. The far-IR and Raman data of (D, (111) - (IV ) are summarized in Table 6-11. The far-IR spectra of I, H and IV display several absorptions in the 600-400 cm-1 1 range. The sharp absorption at 400 cm' and 463 cm.1 represents the characteristic P-S-P stretching vibrations for I and IV, respectively. The doubly bridging bonding motif lowers the energy of the S-P-S vibration for I and II. The remaining absorptions are tentatively assigned to the -P82 stretching vibrations by analogy to AAuP287 (A = K, Rb)5. Absorptions below 400 cm"1 are assigned to Cu—S stretching vibrations. The far-IR spectrum of C82Cu2P286 displays an absorption at 441 cm'], which may be assigned to the P-P stretching vibration by analogy to N a4P286'6H20,34 at 440 cm-1. The Raman spectra of MI display absorptions in the 250-600 cm"1 range that are tentatively assigned to P-8 and absorptions below 250 cm-1 are assigned to Cu-S stretching vibrations. The sharp absorption at ca. 391 cm'1 is assigned to the -PS3 stretching vibration by analogy to AAuPS4 and AAuP287 (A = K, Rb)5. The spectrum of III displays several absorptions below 200cm'1 are assigned to Cu-S stretching vibrations. The absorptions above 300 cm'1 are tentatively assigned to P-S or P-P vibrations. The spectrum of IV exhibited a high background problem and the spectrum was only obtained in a C81 matrix, therefore not all the absorptions may not be present. The absorption at ca. 266 217 cm‘1 tentatively assigned to a Cs-S stretching vibration. The Raman spectra of I, III and IV is shown in Figure 6-10. Differential thermal analysis (DTA) data followed by careful XRD analysis of the residues, show that I and II melt congruently, at 472°C, and 470°C, respectively. Crystals of III melt congruently at 716°C. The DTA of IV shows that it melts incongruently at ~500 °C to form a mixture of K3CuP287 and an as of yet unknown phase. A typical therrnogram for I and III are shown in Figure 6-11. 267 (A) 2 I I I I a =’ 1.5- .D 3; as t- o D .§ 0.5.. E‘ 8 “fr? 0 b l l l l 2 2.2 2.4 2.6 2.8 3 Energy (eV) (B) 2 I a D .D 5 1.5— a d) O U .5. 1- i? 8 .0 <1: 0.5 J l l d 2 2.2 2.4 2.6 2.8 3 Energy (eV) Figure 6-8: Single crystal optical absorption spectra of (A) K2CuP389 (I) and (B) szCuP389 (II). The sharp features at high absorbance are noise and due to the very low transmission of light. 268 (A) U" _( .1 .J _l I K3CuPZS7 Eg = 3.04eV Absorption Coeff. (arb units) 0 in 0 1 L l l 1 2.8 2.85 2.9 2.95 3 3.05 3.1 Energy (eV) (B) 1.2 0.8 P 0.6 - CszCuzP286 ot/S absorption Coeff. (arb. units) 0.4 Eg = 2.43 eV 1 0.2 . O l l l l 1 2 3 5 6 7 4 Energy (eV) Figure 6-9: (A) Single crystal optical absorption spectrum of K3CuP287 (III). (B) Solid-state optical absorption spectrum of CszCuszSg (IV). 269 (A) Raman Intensity (arb. units) 0 I I I I I I I El 300 700 600 500 400 300 200 100 o Wavenumber (cm“) 3'9 A 10 2". E . a 355 '9 S s a E g 4 306 224 E l i 179 5 2 / 555 147 E / 1: o , 800 700 600 500 400 300 200 100 Wavenumber (cmd ) (C) Raman Intensity (arb. units) 300 700 500 500 400 300 200 100 0 Wavenumber (cm") Figure 6-10: Raman spectra of (A) K2CuP389 (I), (B) CSZCu2P2S6 (III) and (C) K3CUP2$7 (IV) 270 401) . r r 1 4r 301) 201) 101) (100 -1OJ) -201) en I I I I I 100 200 300 400 500 600 Temperature (° C) endo exo -301) O 10.0 I I I I I I I (LOO -1QD -200 -300 -40.0 -50.0 b 18 endo exo I I I ‘60.0 I I I l O 100 200 300 400 500 600 700 800 Temperature (° C) Figure 6-11: Typical DTA diagrams for KZCuP389 (I) and CszCuszSG (III). 271 Q “2 1- 0 (A9) Afiieua Figure 6—12: A portion of the electronic band structure in selected directions of reciprocal space. The energy band-gap, which is indirect, if found between the B and F points. 272 120 Total III.IJII.I 80 d O O I 60 i. ‘ II II 40 .‘ I“ - A II‘ 2 ‘1 l— ] “NJ/kl K.M‘X, \ Vii: 11 1‘1 7“” 5° S CU total (x2) 40?- ET. (#3.... 53::- 20 Ff'I—T'I‘Y'T'I'I'V I‘T‘ I I 'I ‘ I I 1 I I .II 30*- LILLLLLI 4.1.1 421.444.: .I..LJ .1.) 20: 10': P (x 6) DOS (states/eV cell) A M IIIII’YITrr G I 50 E- S (x2) 404 .2 IILIIIII 30 rYUITIII VI! 20 1O II'IYTrIY -4 -a -2 - 1 Energy (eV) Figure 6-13: Total density of states and Cu, P, and S projected density of states of K2CUP3Sg. 273 Cu s (x 24) O 'YT'T'T‘ I IYTTTT—Y’TTVT’ITTI‘V'Y‘TY‘ 1 £14] III I I I I I I I 'fi 4 ° C I - » Cup (x24) . > 3 ~_ 1 3 4 In . .9 . 1 g 2 - .. a) : I V F q . J 8 't 4, 17 Y ‘7 TI so _ CU d (X 2) J l 30 .1 20 L J - -t " -1 P 4 10 E .2 I 3 o h a - . . AL. I ‘47 A A .4 1 . . I AA -1 -4 -3 -2 O 1 2 3 4 Energy (eV) Figure 6-14: Cu-s, Cu-p and Cu-d projected density of states of K2CuP389. 274 3 I I I — — -y = -28.554 4» 11.045x R: 0.96657 ‘ A 25 b A o g 0’ ,r :1 2 - 0’ . -€ {( 3 a." N 1.5 )- ’ I g .. r: , 0 .2 I o. H I 0 e. l - ’ ." -I O ’ o. w I .. .o 0' (U .0”; 05 j... I ‘ I I I 0 l l J 2.6 2.65 2.7 2.75 2.8 Energy (eV) 1.3 - — -y = 53305 + 2.3562x R: 0.99296 ' I 0’ ° I o A .9 ‘2 r v" I: r .‘ V :3 > .D V I— l I ~ I 53 I A O 8 , ,, - ‘é- ' . , r" ' 8 * l.’ .o 4 o N " ’0 7 0-9 ’ I . '0’ L O a" 0.8 ’1 I I 1 PI I 1 4 I I I 1 I I I 1 1 l 2.6 2.65 2.7 2.75 2.8 Energy (eV) Figure 6-15: Absorption edge of K2CuP389 plotted as a function of energy: (A) (abs)2 dependence (direct gap) and (B) x/abs dependence (indirect gap). 275 Table 6-11. IR and Raman Spectroscopic Data for K2CUP389, CszCu2P286, and K3CuP287 K2CuP389 CSzCUszS6 K3CuP287 IR Raman IR Raman IR Raman 665(8) 660(w) 574(m) 555(m) 578(m) 647(sh) 582(m) 441(m) 565(m) 470(m) 400(m) 391(8) 463(8) 389(8) 368(m) 367(8) 410(w) 350(m) 355(W) 327(w) 32 1 (m) 303(m) 314(m) 305(m) 289(w) 289(w) 287(w) 290(w) 278(m) 272(w) 217(m) 268(w) 263(w) 253(w) 255(w) 246(m) 248(w) 250(w) 244(w,b) 227(8) 224(m) 225(w) 221(w) 208(w) 179(w) 202(m) 147(w) 196(w) 169(w) 174(m) 276 3.4. Band Structure Calculations Band structure calculations were performed in order to get some insight into the type of band-gap and origin of the yellow color in K2CuP389. Let us remember that a compound appears yellow because it may be absorbing light in the blue. If color originates from an absorption phenomenon, the energy gap between the two energy levels involved in the absorption process is expected to be about 2.6eV, as it indeed observed experimentally. The portion of the electronic band structure near the Fermi level is shown in Figure 6-12. The total and partial density of states of K2CuP3S9 are depicted in Figures 6-13 and 6-14 in the [-4,4] eV-energy range with the zero energy taken at the last occupied level. From the calculated electronic structure, a semiconductor behavior is predicted with an energy gap of 1.48eV. Such a value is too small compared to the experimentally determined value to explain the yellow color. Nevertheless, this discrepancy indeed originates from the limitation of the TB-LMTO-ASA method (or more exactly the density functional method), which is known to underestimate the energy gap by about 50 - 70%”. The indirect nature of the optical band gap is confirmed by the experimental data obtained from single crystals of this compound. Data plots of (abs)2 vs. hv and W vs. hv, shown in Figure 6—15, can be used to distinguish between a direct and an indirect gap in semiconductors. This is because in solids the energy dependence of the absorption coefficient is quadratic in materials with direct energy gap while in those with indirect gaps the dependence scales to the square root.35 As mentioned above, copper and phosphorus atoms are in a tetrahedral coordination of sulfur. In the corresponding M14 tetrahedral isolated molecule, the cation atomic orbitals, engaged in antibonding interactions with chalcogens, are expected to split into d-e and d-t2 molecular orbitals sets, while 8, p orbitals are located at higher energy and are unoccupied for Cu+ and P5+ Hence, the uppermost antibonding levels of the valence band (Figure 6-12a), built on copper d-orbitals as shown on Figure 6-12b, could be associated to Cu-d E and Cu-d T2 blocks. Moreover, the bottom of the conduction band (Figures 6- 277 12c), essentially constructed on 8, p phosphorus orbitals highly hybridized with sulfur ones, are associated to antibonding sigma P-S interactions. Let us notice that the charge balance K+2Cu+ P5+38'29 is checked from band structure calculation if we assign the conduction band completely to 8p orbitals of P5+, and the top of the valence band to the copper d levels. The low contribution of the Cu-s, -p orbitals (Figures 6-l3a,b) to the bottom of the conduction band make improbable the occurrence of an intra-site inter-band transfer. The yellow color would rather originate from a charge transfer from 8210 either + . . . . Cu or P5+. Hence, 1n the absence of any color center Impurrtres, the color process at work in K2CuP3S9 would be a light induced heteronuclear intervalence charge transfer. 4. Conclusions The difference in electronegativity between sulfur and selenium results in a spectacular difference in the reactivity of the thiophosphate and selenophosphate fluxes. Changing flux conditions (e.g. chalcogenide basicity) results in the formation of various thiophosphate [PxSyP' units. The four compounds reported here support the fact the thiophosphate and selenophosphte systems need to be explored separately. The experimentally observed trend of stabilizing higher nuclearity thiophosphate units, with increased P285 concentrations endows the system a tremendous amount of flexibility due to the numerous binding modes of the [PxSy]"' units. At this stage we have only a sketchy phenomenological understanding of the complex equilibria between the various P5+ containing species and any [PxSy]"' units based on P4+ species. We expect NMR spectroscopy to be of considerable help in elucidating the equilibrium issues in these systems. 278 List of References (a) Kanatzidis, M.G. Curr.0pinion Solid State and Mater. Sci., 1997, 2, 139. (a) Kanatzidis,M. G. Chem. Mater., 2, 353, (1990). (b) Sutorik. A.; Kanatzidis, M.G. Prog. Inorg. Chem, 1995, 43, 151. Chondroudis, K.; Hanko, J.A.; Kanatzidis, M.G. Inorg. Chem, 1997, 36, 2623. (a) McCarthy, T. J.; Kanatzidis, M.G. Chem. Mater., 1993, 5, 1061. (b) McCarthy, T. J.; Kanatzidis, M. G. J. Chem. Soc. Chem. Commun.., 1994, 1089. (c) McCarthy, T. J.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Mater., 1994, 6, 1072. (d) McCarthy, T. J.; Kanatzidis, M. G. J. Alloys Comp., 1996, 236, 70. (e) Chondroudis, K.; McCarthy, T. J.; Kanatzidis, M.G. Inorg. Chem, 1996, 35 840. (f) Chondroudis, K.; Kanatzidis, M.G. J. Chem. Soc. Chem. Commun.,. 1996, 1371. (a) McCarthy, T. J.; Kanatzidis, M.G. Inorg. Chem, 1995, 34, 1257. (b) Chondroudis, K.; Kanatzidis, M.G. Inorg. Chem, 1995, 34, 5401. (c) Chondroudis, K.; McCarthy, T. J.; Kanatzidis, M.G. Inorg. Chem, 1995, 35, 3451. (d) Chondroudis, K.; Kanatzidis, M.G., Angew. Chem, 1997, 36, 1324. (e) Dorhout, P.K.; Malo, T.M. Z. Anorg. Allg. Chem, 1996. 622, 385. (f) Chondroudis, K.; Kanatzidis, M.G. Sumitted for publication. (a) Chondroudis, K.; Kanatzidis, M.G. C. R. Acad. Sci. Paris, Series B, 1996, 322, 887. (b) Chondroudis, K.; Kanatzidis, M.G. J. Chem. Soc. Chem. 1997, 119, 2574. (c) Chen, J.H.; Dorhout, P.K. Inorg. Chem, 1995, 34, 5705. (d) Chen, J.H.; Dorhout, P.K.; Ostenson, J.B. Inorg. Chem, 1996, 35, 5627. (a) Klingen, W.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem, 197 3, 401, 97. (b) Toffoli, P.; Khodadad, P.; Rodier, N. Acta Cryst, Sect. B, 1978, 34, 1779. (c) Klimgen, W.; Ott, R.; Hahn, H. Z. Anorg. Allg. Chem, 1973, 396, 10. 11. 12. 13. 14. 15. 279 271. (d) Jandali, M.Z.; Eulenberger, G.; Hahn, H. Z. Anorg. Allg. Chem, 1978, 447, 105. (a) Ouvrard, G.; Brec, R.; Rouxel, J. Mater. Res. Bull., 1985, 20, 1181. (b) Lee, 8.; Colombet, P.; Ouvrard, G.; Brec, R. Inorg. Chem, 1988, 27, 1291. (c) Lee, S.; Colombet, P.; Ouvrard, G.; Brec, R. Mater. Res. Bull., 1986, 21, 917. (d) Durand, E.; Ouvrard, G.; Evain, M.; Brec, R. Inorg. Chem, 1990, 29, 4916. (a) Odile, J.-P.; Steger, J. J.; Wold, A. Inorg. Chem, 1975, 14, 2400. (b) Taylor, B.E.; Steger, J. J.; Wold, A.; Kostiner, E. Inorg. Chem, 1974, I3, 2719. (c) Taylor, B.E.; Steger, J. J.; Wold, A. J. Solid State Chem, 1973, 7, 461. (a) Lagadic, 1.; Leaustic, A.; Clement, R. J. Chem. Soc.Chem Commun., 1992, 1396. (b) Clement, R.; Audiere, J.-P.; Renard, J.-P. Rev. Chim. Miner., 1982, 19, 560-571. (c) Michalowicz, A.; Clement, R. Inorg. Chem, 1982, 21, 3872. (d) Clement, R. J. Chem. Soc., Chem. Commun., 1980, 647. (e) Joy, P.A.; Vasudevan, S. . J. Chem. Soc. Chem, 1981, 104, 7792. Thompson, A.H.; Whittingham, M.S. US. Patent 4,049,879 1977. (b) Brec, R.; Le Mehaute’. A. Fr. Patents 7,704,519 1977. Toffoli, P.; Michelet, A.; Khodadad, P.; Rodier, N. Acta Crystallogr., 1982, B38, 706. (a) Pfeiff, R.; Kniep, R. . J. Alloys Comp., 1992, 186, 111. (b) Evain, M.; Boucher, F.; Brec, R.; Mathey, Y. J. Solid State Chem, 1991, 90, 8. (c) Lee, 8.; Colombet, P.; Ouvrard, G.; Brec, R. Mater. Res. Bull., 1986, 21 , 917. (d) Leblanc, A.; Ouili, Z.; Colombet, P. Mater. Res. Bull., 1985, 20, 947. (a) Ferrari, A.; Cavalca L. Gazzetta Chemica Italiana, 1948, 78, 283. (b) Garin, J.; Parthe, E. Acta Crystallogr., 1972, 828, 3672. Loken, 8.; Tremel, W., Eur. J. Inorg. Chem, 1998, 42., 283. 16. 17. 18. 19. 20. 21. 22. 23 24. 25. 280 (a) Yvon, K.; Jeitschko, W.; Parthé, E. J. Appl. Cryst., 1977, 10, 73. (b) CERIUSZ, Version 1.6, Molecular Simulations Inc., Cambridge, England, 1994. McCarthy, T. J.; Ngeyi, S.-P.; Liao J.-H.; DeGroot, D.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M.G. Chem. Mater., 1993, 5, 331. (a) Andersen, O.K. Phys. Rev. B, 1975, 12, 3060. (b) Andersen, O.K.; Jepsen, 0. Phys. Rev. Letter, 1984, 53, 2571. (c) Jepsen, 0.; Z. Phys., 1995, 97, 35. Cu, 1.50.4, P(1), 1.15.4, P(2), 1.13.4, P(3), 1.15.4, K(l), 2.434, K(l), 2.54.4, 8(1), 1.174, S(2), 1.15.4, 8(3), 1.154, S(4), 1.15.4, 8(5), 1.14.4, 8(6), 1.304, S(7), 1.31.4, S(8), 1.30.4, 8(9), 1.134. (a) Jepsen, 0.; Andersen, O.K. Solid State Comm, 1971, 9, 1763. (b) Lambrecht, W.R.L.; Andersen, O.K. Phys. Rev. B, 1986, 34, 2439. (a) Sheldrick,G. M. “SHELXTLTM version 5”, Siemens Analytical X-Ray Instruments, Inc. Madison, WI. (b) V. Petricek, M. Dusek, “JANA’96 Cristallographic Computing System”, Institute of Physics, Academy of Sciences of the Czech Republi c, Praha. van der Lee, A.; Evain, M.; Boucher, F.; Bree, R. Z. fur Krist.., 1993, 203, 247. Walker, N.; Stuart. D. Acta. Crystallogr., 1983. A39, 158. (a) Sheldrick, GM. In: Crystallographic Computing 3; Sheldrick, G.M., Kruger, C., Goddard R.,Eds.; Oxford University Press: Oxford, England, 1985; p 175. (b) TEXSAN: Single-Crystal Structure Analysis Software, Version 1.7, 1995 Molecular Structure Corp.: 3200 Research Forest Drive, The Woodlands, TX 77381. (a) Liao, J.-H.; Kanatzidis, M.G. Chem. Mater., 1993, 5, 1561. (b) Keane, P.M.; Lu, Y.-J.; Ibers, J.A, Ace. of Chem. Res., 1991, 24, 223. (c) Lu, Y.-J.; Ibers, J.A, J. Solid State Chem, 1992, 98, 312. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 281 (A) Grenouilleau, P.; Brec, R.; Evain, M.; Rouxel, J.Rev. Chim. Miner., 1983, 20, 628 (b) Bouchetiére, M.; Toffoli, P.; Khodadad, P.; Rodier, N. Acta. Crystallogr., 1978, B34, 384. Evain, M.; Brec, R.; Ouvrard, G.; Rouxel, J. Mat. Res. Bull., 1984, I9, 41. Queignec, M.; Evain, M.; Brec, R.; Sourisseau, C., J. Solid State Chem, 1986, 63, 89. Ondik, H.; Acta. Crystallogr., 1965, 18, 226. (a) Menzel, F.; Brockner, W.; Carrillo-Cabrera, W.; von Schnering H.G. Z. Anorg. Allg. Chem, 1994, 620, 1081. (b) Carrillo-Cabrera, W.; Sabmannshausen, J.; von Schnering H.G.; Menzel, F.; Brockner, W. Z. Anorg. Allg. Chem, 1994, 620, 489. McCarthy, T.J.; Kanatzidis, M.G. J. Alloys Comp., 1996, 236, 1089. McCarthy, T.J.; Hogan, T.; Kannewurf, C.R;. Kanatzidis, M. G. Chem. Mater., 1994, 6, 1072. Durand, E.; Evain, M.; Brec, R. J. Solid State Chem, 1992, 102, 146. Mathey, Y.; Clement, R,; Sourisseau, C.; Lucazeau, G. Inorg. Chem, 1980, I9, 2773. Pankove, J.I. in "Optical Processes in Semiconductors." Dover Publications, New York, 1975. CHAPTER 7 Chemistry of Gold in Molten Alkali Metal Polychalcogeno-phosphate Fluxes. Synthesis and characterization of the low dimensional Compounds A2AuPS4 (A = K, Rb, Cs) and AAuP287 (A = K, Rb). 282 283 1. Introduction The polychalcophosphate fluxes provide the best set of experimental conditions for the synthesis of new multinary thiophosphate and selenophosphate compounds.1 These fluxes form by the in situ fusion of A2Q/P2Q5/Q and contain [Psz]“' (Q=S, Se) units in a molten polychalcogenide solvewnt. These units can coordinate to metal ions utilizing a great variety of binding modes, giving rise to interesting new materials. These materials tend to be structurally and compositionally complex and often cannot be made by standard solid-state methods. After the initial report on ABiP287 (A=K,Rb),2 the following unusual compounds were reported: A3M(PS4)2 (A=Rb, Cs; M=Sb, Bi),3 C83Bi2(PS4)3,3 Na0_16B11_28P286,3 A2MP28e6 (A=K, Rb; M=Mn, Fe),1 A2M2P28e5 (A=K, Cs; M=Cu, Ag),l KMP28e6 (M=Sb, Bi),4 C83M4(P28e6)5 (M=Sb, Bi),5 APbPSC4,6 A4M(P8e4)2 (A=Rb, C8; M=Pb, Eu),6 Rb4Ti2(P28e9)2(P28e7),7 KTiPSe5,7 A58n(PSe5)3 (A=K, Rb),8 and A6Sn28e4(PSe5)2 (A=Rb, Cs).8 More recently, KAu5P28eg9, KzUP3Se9,1°a Rb4U4P4Se261°b, and the K(RE)P28e6 (RE=Y, La, Ce, Pr, Gd)ll series of compounds were also reported. Extension of this chemistry to Au looked appealing because no structurally characterized compounds have been reported. The only compound reported is the ternary AuPS4,12 whose structure remains elusive. Om the basis of vibrational spectroscopy, a polymeric chain consisting of alternating edge-sharing tetrahedral [PS4]3' and square AuS4 planes was proposed. Our studies with Au yielded the first structurally characterized gold selenophosphate compound, the mixed-valent A6Au11,5Aum1,5(P28e6)3 (A=K, Rb)13 which features three different coordination geometries for the Au centers. As an extension of the selenophosphate flux method, the thiophosphate chemisz was 284 examined in order to investigate the existence of phases isostructural to the expanding list of selenophosphate compounds and to synthesize new structural types. Here we report the synthesis, structural characterization, optical and thermal properties of the new quaternary gold thiophosphate compounds, A2AuPS4 (A=K, Rb, and Cs), and AAuP287 (A=K, and Rb), which represent the first structurally characterized examples of quaternary gold thiophosphates. Access to each phase was achieved by modifying the flux basicity, by means of varying the amount of A28 and P285 in the starting composition (see Syntheses).1» 6'3, 1° 2. Experimental Section 2.1. Reagents The reagents mentioned in this study were used as obtained unless noted otherwise: (1) Au metal (99.99%) was acquired from Liberty Coins, Lansing MI. (ii) phorphous pentasulfide (P285) 99.999% purity, Aldrich Chemical Co., Milwaukee, Wi.; (iii) cesium metal, analytical reagent, Johnson Matthey/AESAR Gruop, Seabrook, NH; (iv) rubidum metal, analytical reagent, Johnson Matthey/AESAR Gruop, Seabrook, NH; (v) potassium metal, analytical reagent, Aldrich Chemical Co., Milwaukee, Wi.; (vi) sulfur powder, sublimed, J.T. Baker Chemical Co., Phillipsburg, NJ; (vii) Methanol (MeOH) ACS anhydrous, EM Science, Inc., Gibbstown, NJ .; (viii) diethyl ether, ACS anhydrous, EM Science, Inc., Gibbstown, NJ. Finely divided Au metal : A Canadian Maple Leaf gold coin, (99.99%, 31.1g) was dissolved in 400 ml of aqua regia (300 ml concentrated HCl and 100 ml concentrated HNO3). The solution was boiled in an acid-resistant fume hood to a volume of approximately 100 ml. The solution was neutralized with ammonium hydroxide and the gold was reduced with excess hydrazine hydrochloride, dissolved in 100 ml of distilled 285 water. The resulting black suspension was gently heated, with stirring, for one hour to allow particle aggregation. After filtering the suspension, and washing it with copious amounts of distilled water and acetone, the resulting gold powder was heated in air for 2 hours at 200°C to drive off any remaining impurities, yielding 30.9g of Au powder. Note: heating too long or at too high a temperature results in impratical grain sizes. 2.2. Syntheses. A28 (A = K, Rb, Cs) were prepared by reacting stoichiometric amounts of the elements in liquid ammonia as described elsewhere.14 Preparation of KzAuPS4 (I). A mixture of Au (0.25 mole), P285 (0.50 mmole) , K28 (0.50 mole), and (1.5 mmole) S that was sealed under vacuum in a Pyrex tube. The reaction mixture was heated to 500°C for 4 days, followed by cooling to 100°C at a rate of 4°C/hr. The product, which is stable in water and air, was isolated by dissolving the K28x flux with degassed methanol under inert atmosphere to give yellow crystals and chunky yellow crystalline Kx[PySz] flux. The yellow chunks of Kx[PySz] flux were removed by washing with degassed distilled water, followed by methanol and ether to give yellow crystals (75% yield based on Au). The crystals are stable in air and water (signs of decomposition are apparent only after several weeks). Microprobe analysis on single crystals gave an average composition of K1,3AuP1,485.6. Preparation of szAuPS4 (II): A mixture of Au (0.25 mole), P285 (0.50 mmole), Rb28 (0.50 mole), and (1.5 mmole) S was sealed under vacuum in a Pyrex tube and heated as in (I). The product, which is stable in air and water, was isolated as in (I) to give yellow crystals (52% yield based on Au). These crystals also show signs of 286 decomposition after a several week exposure. Microprobe analysis gave an average composition of Rb1,5AuP1,185_1. Preparation of CszAuPS4 (III): A mixture of Au (0.25 mole), P285 (0.50 mole) , C828 (0.50 mole), and (1.5 mmole) S was sealed under vacuum in a Pyrex tube and heated as in (I). The product, which is stable in water and air, was isolated as in (I) to give colorless transparent crystals (63% yield based on Au). Microprobe analysis gave an average composition of CS2_25AUP3.386.1. Preparation of KAuP287 (IV): KAuP287 was synthesized from a more acidic Lewis mixture of Au (0.50 mole), P285 (0.75 mole) , K28 (0.50 mole), and (1.5 mmole) S that was sealed under vacuum in a Pyrex tube. The reaction mixture was heated to 400°C for 4 days, followed by cooling to 110°C at a rate of 4°C/hr. The product which disintegrates in the presence of water, was isolated with degassed methanol to give dark red crystals (50% yield based on Au). The crystals are soluble in DMF forming an orange-red solution. Microprobe analysis gave an average composition of K1,0Au1,2P2,6810. Preparation of RbAuP287 (V): RbAuP287 was also synthesized from a more acidic Lewis mixture of Au (0.50 mole), P285 (0.75 mole) , K28 (0.50 mole), and (1.5 mmole) S that was sealed under vacuum in a Pyrex tube and heated as in (IV). The product, which disintegrates in the presence of water, was isolated as in (IV) to give dark red crystals (50% yield based on Au). These crystals are also soluble in DMF forming an orange-red solution. Microprobe analysis gave an average composition of Rbl.0AU1.2P2.6S9.2- 287 2.3. Physical Measurements Powder X-ray Difi‘raction. Analyses were performed using a calibrated Rigaku- Denki/RW400F2 (Rotaflex) rotating anode powder diffractometer controlled by an IBM computer, operating at 45 kV/ 100 mA and with a 10/min scan rate, employing Ni-filtered Cu radiation in a Bragg-Brentano geometry. Powder patterns were calculated with the CERIUS2 software.15 Calculated and observed XRD patterns for (I), (III), and (IV) are given in Tables 7-1 — 7-3, respectively. Infrared Spectroscopy. Infrared spectra, in the far-IR region (600-50 cm'l), were recorded on a computer controlled N icolet 750 Magna-IR Series H spectrophotometer equipped with a TGS/PE detector and silicon beam splitter in 4 cm:1 resolution. The samples were ground with dry CsI into a fine powder and pressed into translucent pellets. Raman Spectroscopy. Raman spectra (700 - 100 cm '1 ) were recorded with a BIO-RAD FT Raman spectrometer with a Spectra-Physics Topaz T10—106c 1.064 nm YAG laser running at 11 to 11.5 amps equipped with a Ge detector. The samples were ground into a fine powder and loaded into glass tubes. Solid State U V/Vis/Near IR Spectroscopy. Optical diffuse reflectance measurements were performed at room temperature using a Shimadzu UV-3101PC double beam, double monochromator spectrophotometer. The instrument is equipped with integrating sphere and controlled by a personal computer. BaSO4 was used as a 100% reflectance standard for all materials. Samples are prepared by grinding them to a fine powder and spreading them on a compacted surface of the powdered standard material, preloaded into a sample holder. The reflectance versus wavelength data generated can be used to estimate a material's band gap by converting reflectance to absorption data as described earlier. 16 Single crystal optical transmission spectroscopy. Room temperature single crystal optical transmission spectra were obtained on a Hitachi U-6000 Microscopic FI‘ 288 Spectrophotometer mounted on an Olympus BH2-UMA metallurgical microscope over a range of 380 to 900 nm. Crystals lying on a glass slide were positioned over the light source and the transmitted light was detected from above. Differential Thermal Analysis (DTA). DTA experiments were performed on a computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~ 25 mg) of ground crystalline material was sealed in quartz ampoules under vacuum. A quartz ampoule of equal mass filled with A1203 was sealed and placed on the reference side of the detector. The sample was heated to the desired temperature at 10 oC/min, then isothermed for 10 minutes and finally cooled to 50 0C at the same rate. Residue of the UTA experiment was examined by X-ray powder diffraction. To evaluate congruent melting we compared the X-ray powder diffraction patterns before and after the DT A experiments. The stability/reproducibility of the samples were monitored by running at multiple cycles. Semiquantitative microprobe analyses The analyses were performed using a JEOL J SM-6400V scanning electron microscope (SEM) equipped with a TN 5500 EDS detector. This technique was used to confirm the presence of all elements in the compounds, for the compositions are unreliable do to the overestimation or underestimation of certain elements. Data acquisition was performed with an accelerating voltage of 20kV and thirty seconds accumulation time. Single crystal X-ray Crystallography. Intensity data for (I) and (IV) were collected using a Rigaku AFC6S four-circle automated diffractometer equipped with a graphite crystal monochromator. Intensity data for (III) were collected using a Nicolet (Siemens) four-circle automated diffractometer equipped with a graphite crystal monochromator. Crystal stability was monitored with three standard reflections whose intensities were checked every 150 reflections, and unless noted, no crystal decay was detected in any of the compounds. The space groups were determined from systematic absences and intensity statistics. An empirical absorption correction based on I]! scans was applied to all data during initial stages of refinement. An empirical DIFABS correction17 289 was applied as recommended after full isotropic refinement, after which full anisotropic refinement was performed. The structures were solved by direct methods using SHELXS- 86 software18a (for all compounds), and full matrix least squares refinement was performed using the TEXSAN software packagelgb. The complete data collection parameters and details of the structure solution and refinement are given in Table 7-4. The coordinates of all atoms, average temperature factors, and their estimated standard deviations are given in Tables 7-5 - 7-7. 290 Table 7- 1. Calculated and Observed X-ray Powder Patterns for KzAuPS4 (I). hkl dcalcgzt ‘1 obsd, A I/Imax(°bSd) 001 9.45 9.49 7 100 6.50 6.52 20 101—011 5.49 5.50 12 101 5.23 5.23 3 002 4.72 4.74 4 112— 3.92 3.93 6 012 3.87 3.86 8 102 3.73 3.73 71 020 3.37 3.34 2 112 200 3.25 3.25 6 021 3.17 3.17 3 201‘ 3.12 3.12 100 201 3.02 3.02 4 120 2.99 2.99 5 103— 2.89 2.89 3 121— 2.87 2.87 5 211121 2.83 2.83 3 103 2.78 2.77 12 202 2.748 2.748 3 004 2.363 2.365 5 220 2.342 2.341 4 023 2.303 2.301 2 104‘ 2.259 2.260 5 104 2.185 2.185 2 222 2.068 2.068 3 302 2.012 2.014 2 024 1.935 1.934 2 223 1.914 1.907 2 223 1.847 1.842 2 124 1.834 1.834 3 321— 1.806 1.804 2 322 224 026 224 1 25 1.728 1.695 1.649 1.633 1.582 291 1.724 1.694 1.649 1.625 1.576 #Nwmw 292 Table 7-2. Calculated and Observed X—ray Powder Patterns for CszAuPS4 (HI). hkl dcalc, A d obsd, A I/Imax(ObSd) 020 10.04 10.20 12 100 6.90 6.97 9 110 6.52 6.59 25 021 120 5.69 5.73 10 040 5.02 5.06 16 130 4.80 4.84 17 041 4.08 4.09 32 131 3.96 3.98 27 150 3.47 3.48 21 210 3.40 3.41 100 060 3.34 3.36 15 022 3.31 3.32 15 220 3.26 3.28 31 102 3.13 3.14 16 112 3.09 3.08 26 121— 2.87 2.87 5 2117121 2.83 2.83 3 103 2.78 2.77 12 202 2.748 2.748 3 004 2.363 2.365 5 220 2.342 2.341 4 023 2.303 2.301 2 10? 2.259 2.260 5 104 2.185 2.185 2 222 2.068 2.068 3 302' 2.012 2.014 2 024 1.935 1.934 2 223' 1.914 1.907 2 223 1.847 1.842 2 124 1.834 1.834 3 321— 1.806 1.804 2 322 224— 026 224 125 1.728 1.695 1.649 1.633 1.582 293 1.724 1.694 1.649 1.625 1.576 #NWNW 294 Table 7-3. Calculated and Observed X-ray Powder Patterns for KAuP287 (IV). hkl dcalc, A d obsd, A I/Imax(°bSd) 002 7.47 7.54 5 110 6.01 6.05 2 111‘ 5.62 111 5.53 5.56 100 020 4.62 4.64 27 021 4.41 4.44 4 022 3.93 3.94 8 113 3.79 3.80 15 202 3.54 3.55 3 202 3.45 3.46 6 114 3.14 3.15 3 221— 2.96 2.95 2 024 2.90 2.91 8 131’ 222' 2.82 2.82 8 222 2.76 2.77 16 115 2.65 2.66 2 311’ 025 2.51 2.51 6 133— 006 311 2.49 2.49 7 224— 2.37 2.36 10 224 040 2.31 2.31 4 134' 313— 2.290 2.291 3 313 2.232 2.234 3 042 2.207 2.208 2 135 2.059 2.061 3 117 1.998 1.998 4 400 1.978 1.979 4 044 1.965 1.966 4 226 1.942 1.945 2 245 1.936 1.938 2 242 1.920 1.923 3 315 333— 008 333 420 151— 422' 118 244 422 404 13b7 208 137 153— 153 317 245 424 1 19 246 511 246 1.906 1.876 1.843 1.818 1.788 1.773 1.755 1.725 1.723 1.71 1 1.704 1.697 1.689 1.659 1.647 1.616 1.591 1.570 1.547 295 1.907 1.877 1.846 1.820 1.789 1.775 1.751 1.728 1.723 1.708 1.703 1.697 1.690 1.659 1.648 1.617 1.592 1.573 1.547 NNNNWNNNNNNN-RNNNNNN 296 Table 7—4. Crystallographic data for K2AuPS4, CszAuPS4 and KAuP287. (I) (III) (IV) Formula K2AuPS4 CszAuPS4 KAuP287 FW 432.25 621.99 522.43 a, 4 6.518(2) 6.904(3) 7.917(3) b, 4 6.747(2) 20.093(6) 9.247(2) c, 4 9.468(3) 7.025(5) 14.968(3) 61 (deg) 90.00 90.00 90.00 B (deg) 9298(2) 90.00 9184(9) 7 (deg) 90.00 90.00 90.00 2, V(A3) 4; 415.8(2) 4, 974.6(8) 4;1095.2(8) 7. (Mo KOL). .4 0.71069 0.71069 0.71069 space group P21/m (#11) Pbcm (#57) C2/c (#15) Dcalc. g/cm3 3.469 4.32 3.168 11. cm-1 196.98 232.71 152.84 20max, deg 50 50 50 Temp (°C) 23 -120 23 Final R/Rw,a % 2.3/3.0 4.7/3.9 1.6/2.2 Octants collected h, k, at] h, k, 1 h, k, :1 Total Data Measured 873 1066 1 1 14 Total Unique Data (ave) 802 1066 103 1 DataF02>3G(F0'2) 609 769 780 No. of Variables 47 46 54 Abs. ratio (min/max) 0.8679 0.6198 0.7567 Crystal Dimen., mm 0.08x0.08x0.20 0.04x0.040.28 0.08x0.06x0.38 are = 2018,) - IFc|)/21F0l, Rw = (2,,(11701 - IFCI)2/2wlF0|2}1/2. 297 Table 7-5. Positional parameters and B(eq) for KzAuPS4 with Estimated Standard Deviations in Parentheses. ‘1 Atom X Y Z Bequ2 Au] 1 .0000 0 0 1 .46(2) K1 1.4551(5) 1/4 —0. 1657(3) 2.7(1) K2 0.7958(4) 1/4 0.4641(3) 1.8(1) S 1 1.5245(5) 1/4 0.1742(3) 20( 1) 82 1.3050(5) 1/4 0.4926(3) 1.6( 1) S3 1.0927(3) -0.0034(3) 0.2370(2) 1.78(8) P1 1.2650(4) 1/4 0.2804(3) 1.0(1) ‘1 B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as ch = (4/3)[azB(1, 1) + bZB(2, 2) + czB(3, 3) + ab(cosy)B(l,2) + ac(cosB)B(l,3) + bc(cosa)B(2, 3) 298 Table 7-6. Positional parameters and B(eq) for CszAuPS4 with Estimated Standard Deviations in Parentheses. 3 Atom X Y Z Bequ2 An] 0.1130(2) 1/4 0 1.l9(5) Csl 0.7921(3) 0.4850(1) 1/4 l.26(8) C82 0.3957(3) 0.6646(1) 1/4 2.0(1) S 1 0.529(1) 0.3390(4) 1/4 1.4(3) 82 0.] 137(7) 0.3637(2) 0.0047(9) 1.2(2) S3 0.297(1) 0.4860(4) 1/4 1.2(3) P1 0.276(1) 0.3871(4) 1/4 1.0(3) 3‘ B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as ch = (4/3)[aZB(1, 1) + bZB(2, 2) + cZB(3, 3) + ab(cosy)B(l,2) + ac(cosB)B(l,3) + bc(cosa)B(2, 3) 299 Table 7—7. Positional parameters and B(eq) for KAuP287 with Estimated Standard Deviations in Parentheses. a Atom X Y Z Beq‘lA2 Aul 1/4 1/4 0 1.48(1) K1 1/2 0.2534(3) 1/4 5.1(1) Sl 0.2212(2) 0.4202(2) 0.1153(1) 2.01(6) $2 -0.0310(2) 0.1920(2) 0.0324(1) 2.06(6) S3 0 0.1872(2) 1/4 1.83(8) P1 -0.0133(2) 0.3328(2) 0.1388(1) 156(5) 3 B values for anisotropically refined atoms are given in the form of the isotropic equivalent displacement parameter defined as Beq = (4/3)[a2B(1, 1) + bZB(2, 2) + c2B(3, 3) + ab(cosy)B(1,2) + ac(cosB)B(l,3) + bc(cosa)B(2, 3) 300 3. Results and Discussion 3.1 Description of structures. Structure of KzAuPS4 (I) and szAuPS4 (II) These two compounds are isostructural. The single-crystal structure determination was performed on the K+ salt (I), and so the discussion will refer mainly to this compound. The structure of the [AuPS4]n2“' anion is a one-dimensional chain built of alternating [PS4]3' tetrahedra linked in a comer- sharing fashion to monovalent Au atoms. The chains (Figure 7-1) run along the crystallographic b-axis and are separated by alkali metal cations, see Figure 7-2. A view highlighting the semisodial wave nature of a single chain is shown in Figure 7-3. The chains are propagated by a crystallographic screw axis resulting in the terminal sulfides of the [PS4]3' ligand to appear on opposite sides of the S-Au-S linkage. This structural motif is not observed in the selenophosphates A3AuP28e319 (A = K, Rb, Cs) compounds. Remarkably, (1)-(II) are isostructural to the recently reported thioarsenate K2AUASS4 and thiophosphate T12AuPS49- The S-Au-S angle is constrained by symmetry to be linear. The Au resides on an inversion center with an average Au-S distance of 2.293(2) A. This compares well with the Au-S distance found in KAu85,2° K2Au28n84,21 K2Au28n286,22 BaAu28nS4,23 and K4Au685,24 all featuring linearly coordinated monovalent Au+. The [PS4]3' ligand is a regular tetrahedron with an average P-S bond distance of 2.04(1)A and mean S-P-S angle of 109(2)°, respectively. The [PS4]3' tetrahedra are in a staggered arrangement above and below the Au-S linear bond (see Figure 7-3). The Au-Au' distance is 3.598(2), suggesting no significant bonding interactions. Selected bond distances and angles for K2AuPS4 are given in Table 7-8. The A+ ions are located in two different sites. In K2AuPS4, K(l) is coordinated by six 8 atoms [range of K(l)-S distances, 3.227(4)- 3.481(4)A; av 3.37(3)A], and K(2) is coordinated by eight 8 atoms [range of K(2)-S distances, 3.186(4)-3.424(3)A; av 3.35(4)4]. 301 Structure of CszAuPS4 (III): The structure of the [AuPS4]n2“‘ anion is closely related to the selenophosphate compounds, A3AuP28e3 (A: K, Rb,Cs).19 The [AuPS4]n2"‘ chains run along the crystallographic c -axis and are well separated by cesium cations (Figure 7-4). As view of a fragment of a single [AuPS4]n2"' chain is shown in Figure 7.5. The overall anionic structure of the [AuPS4]n2“‘ chains is related to the [AuP28e3]n3“‘ macroanion of A3AuP28eg in that its projection also resembles the letter ""C, see Figure 7-6. The difference between the two macroanions is shown Scheme 1. Scheme 1 Au M —> O lAuP2$eeln3”' lAuPS4ln2"' The difference is the reduction of the diselenide bridge of the [P28e3]4‘ unit is reduced and replaced by a second S-Au-S linkage, creating two [PS4]3' units. A second difference between the two structures is the packing of the chains. The chains in (111) stack in a centrosymmetric parallel fashion while the [AuP28e3]n3n' chains pack in an unusual criss- cross packing motiff. These chains run in the [110] and [220] directions and are almost mutually perpendicular to each other, in a criss-cross fashion, an unusual packing for 1-D compounds The ribbons of (III) are isostructural to A2AquS4 (A = Rb, Cs)25 and are propagated by a crystallographic glide plane resulting in the temimal sulfides of the [PS4]3' 302 ligand to appear on the same side of the essentially linear S-Au-S linkage forming a projection resembling the letter "C". The S-Au-S angle is essentially linear, at 178.9(2)°. The Au resides on a 2-fold crystallographic site with Au-S distance of 2.285(2)A. The [PS4]3‘ unit binds to two monovalent Au cations leaving the other sulfides terminal. The [PS4]3' unit is a regular tetrahedron with an average P-S bond distance of 205(3) and a mean S-P—S angles of 109(2)°. The [PS4]3‘ tetrahedra are in a staggered arrangement above and below the S-Au-S linear bond. The Au-Au' distance is 3.667(2)A indicating no significant bondng interaction. Selected bond distances and angles for C82AuPS4 are given in Table 7-9. The [AuPS4]n2“' chains are separated by C8+ ions that are located in two different sites. In CszAuPS4, Cs(l) is coordinated by nine 8 atoms [range of Cs(l)—S distances, 3.394(9)-3.51(1)A; av 3.57(3)A], and Cs(2) is coordinted by eight 8 atoms [range of Cs(2)—S distances, 3.45(1)-3.54(1).4; av 3.75(7)/4]. Structure of KAuP287 (IV) and RbAuP287 (V): These two compounds are isostructural. Therefore the single—crystal structure determination was preformed on the K+ salt (IV) and so the discussion will refer mainly to this compound. The structure of [AuP287]“' is a unique one dimensional chain consisting of alternating [P287]4' units edge sharing fashion to square planar Au3+ cations. Having a d8 electron configuration, the Au3+ atom adopts a square planar coordination environment. The chains run along the (101) direction (Figure 7-6) and are separated by alkali cations (Figure 7-7). The Au center resides on an inversion center with Au-S bond distances ranging from 2.352(2) to 2.354(2)A and P-S distances ranging from 1.952(2) to 2.140(2)A. The Au-Au' distance between chains is 6.068(1)A. Inspection of the S(l)-Au(l)-S(2) angles of 8349(5) and 96.51(5)°, reveals a slight deviation from an ideal square planar geometry, which is due to a strained four-menbered (Au(l)-S(l)-P(1)-S(2)) ring. The pyrothiophsophate [P287]4' unit consists of two corner sharing [PS4]3‘ tetrahedra. Each PS4 tetrahedron is slightly distorted with S-P-S angles ranging from 9890(8) to 118.3(1)°, respectively. The 303 smallest angle is associated with the constrained S-P-S—Au four-membered rings and the largest associated with the terminal sulfide. The chains are separated by alkali metal cations residing in a pocket between the terminal sulfide (8(4)) and a bridging sulfide (S(2)) In KAuP287 the potassium caion is coordinated by six sulfur atoms [range of K(1)-S distances 3.324(2) - 3.485(2)A; av. 3.37(3)/4]. Selected bond distances and angles for KAuP287 are given in Table 7-10. 304 Table 7-8. Selected Distances (A) and Angles (deg) for KzAuPS4 (I) with Standard Deviations in Parenthesesa. P(1)—S(1) 2.012(4) S(1)-P(l)-S(2) 115.5(2) P(1)-S(2) 2.013(4) S(1)-P(1)-S(3) 111.2(1) P(l)-S(3) 2.075(3) x2 S(l)-P(l)-S(3’) 111.2(1) S(2)—P(1)-S(3) 103.8(1) Au(1)-S(3) 2.293(2) S(2)-P(l)-S(3’) 103.8(1) Au(1)-S(3’) 2.293(2) S(3)-P(l)-S(3’) 110.9(2) S-P-S (mean) 109(1) Au(l)-Au(1’) 3.373(1) S(3)-Au( l )-S(3) 180.00 Au(1)-S(3)-P(1) 107.3(1) K(2)-S( 1) 3.186(4) K(1)-S(1) 3.227(4) K(2)-S(2) 3.317(4) K(1)-S(1’) 3.377(1) x2 K(2)-S(2’) 3.224(4) K(1)-S(2) 3.331(4) K(2)-S(2”) 3.465(1) x2 K(1)-S(3) 3.481(4) x2 K(2)-S(3) 3.331(3) x2 K(1)-S (mean) 3.37(3) K(2)-S (mean) 333(4) 8The estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 61 = {Zn(l,, — [)2/n(n — 1)}1/2, where 1,, is the length (or angle) of the nth bond, 1 the mean mength (or angle), and n the number of bonds. 305 Table 79. Selected Distances (A) and Angles (deg) for C82AuPS4 (111) with Standard Deviations in Parenthesesa. P(1)-S(1) 2.00( 1) Au(1)-S(2)-P(1) 103.6(3) P(l)-S(2) 2.108(7) x2 P(l)-S(3) 199(1) S(2)-Au(l)-S(2) 179.8(3) P(1)-S(mean) 205(3) S(1)-P(1)-S(2) 110.9(3) x2 Au(1)-S(2) 2.285(5) x2 S(1)-P(1)-S(3) 114.8(5) S(2)-P(l)-S(2') 109.7(4) Au(l)-Au(1’) 3.667(2) S(2)-P(1)-S(3) 105.1(3) x2 Cs( 1)-S(1) 3.45( 1) Cs(2)-8(1) 3.542(9) CS(1)-S(3) 3.720(6) x2 Cs(2)-S(1’) 3.552(3) x2 Cs(1)-S(3) 3.586(6) Cs(2)-S(3) 3.986(5) x2 Cs(1)-S(5) 3.420(9) Cs(2)-S(3’) 3.873(6) x2 Cs(1)-S(5) 3.484(9) Cs(2)-S (mean) 3.75(6) CS(1)-S(5) 3.613(3) x2 Cs(1)-S (mean) 3.57(4) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 61 = {2n(l,, — [)2/n(n — 1)}1/2, where 1,, is the length (or angle) of the nth bond, 1 the mean mength (or angle), and n the number of bonds. 306 Table 7-10. Selected Distances (A) and Angles (deg) for KAuP287 (IV) with Standard Deviations in Parenthesesa. P(l)-S(l) 2.065(2) S(1)-Au(l)-S(1’) 180.00 P(1)-S(2) 2.058(2) S(1)-Au(1)-S(2) 83.49(5) x2 P(1)-S(3) 2.140(2) S(1’)-Au(1)-8(2’) 9651(5) x2 P(1)-S(4) 1.952(2) S(2)-Au(l)-S(2’) 180.00 P(1’)-S(3) 2.140(2) Au(l)-S(1)-P(1) 8840(7) x2 Au(1)-Au(1’) 6.086(1) Au(l)-S(2)-P(l) 8852(7) x2 Au(l)-S(1) 2.352(2) x2 S(l)—P(l)-S(2) 9890(8) Au(1)-S(2) 2.354(2) x2 S(l)-P(l)-S(3) 110.90(8) S(1)-P(l)-S(4) 115.8(1) K(l)-S(1) 3.321(2) x2 S(2)-P(l)-S(3) 101.79(9) K(l)-S4 3.485(2) x2 S(2)-P(1)-8(4) 118.3(1) K(1)-8(4’) 3.333(3) x2 S(3)-P(1)-S(4) 109.87(8) K( 1 )-8 (mean) 3.37(3) S-P-S (mean) 109(3) P(l)-S(3)-P(1’) 102.1(1) aThe estimated standard deviations in the mean bond lengths and the mean bond angles are calculated by the equation 0'1 = {Z,,(l,, — [)2/n(n — 1)}1/2, where 1,, is the length (or angle) of the nth bond, 1 the mean mength (or angle), and n the number of bonds. 307 901 51 O O O O 0 P1 53 52 Figure 7-1: View of a single [AuPS4]n2"' chain in K2AuPS4 (I), with labeling. 308 Figure 7-2: ORTEP representation of K2AuPS4 (I) as viewed down the b -axis (50% probability elliposids). 309 .111 I1 ' 11. 111 \ Figure 7-3: ORTEP representation of K2AuPS4 (I) as viewed down the c - axis (50% probability elliposids). 310 901 51 0 O O O 0 Pl ‘53 S2 Figure 7-4: View of a single [AuP84]n2"' chain in CszAuPS4 (III), with labeling. gar/K 111 r" .. 11 11.. i ‘6 I 6 13;. L42. «1. as, 11 ’9 ’9 e e 6 311 \ ’1‘ 6'2? 49 ego $ Ff, ‘A/ "3 27.3 (1).) \, O f 'e I ‘ 4 § L 69 1828 5 e9 (77': T e 6 ea 49 9 ea {:1 (53¢. \‘\\\_E_'_—— 3 O..— .31 (AV/1% (‘1 \E)“ ‘4" $ E'Lf/QF; he: e 994 a e 99 L... a Figure 7-5: ORTEP representation of Cs2AuPS4 (III) as viewed down the c - axis (50% probability elliposids). 312 ORTEP representation of Cs2AquS4 (III) as viewed down the b - axis (50% probability elliposids). Figure 7-6: 313 Figure 7-7: View of a Single [AuP287]n“' chain in K2AuP287 (IV), with labeling. 314 Figure 7-8: ORTEP representation of KzAuP287 (IV) as viewed down the a - axis (50% probability elliposids). Small octant shaded ellipsoids; Au, principal axis ellipsoids; P, boundary ellipsoids; S, boundary and axis ellipsoids; K. 315 Figure 7-9: ORTEP representation of K2AuP287 (IV) as viewed along the [101] direction (50% probability elliposids). small octant shaded ellipsoids; Au, principal axis ellipsoids; P, boundary ellipsoids; S, boundary and axis ellipsoids; K. A /— ‘1 316 3.2. Synthesis, Spectroscopy and Thermal Analysis The synthesis were the result of a redox reaction in which the metal is oxidized by polychalcogenide ions in the Ax[Psz] flux. The Au cation centers are then coordinated by the highly charged [PyQZ]“' ligands. The molten polychalcophosphate flux method is very effective for crystal growth in this system. The isolation of pure crystalline products is facilitated by the residual polychalcophosphate fluxes solubility in aqueous and/or organic solvents. The flexibility of this method is demonstrated by the easy manipulation of the fluxes Lewis basicity by varying its starting composition. This results in compounds containing not only different [Pny]“' units. An additional advantage of this method is its ability to stabilize either P5+ or P4+ species which results in compounds which form different [Pny]“‘ units. Our studies with quaternary chalcophosphates provide enough examples for the construction of Table 7-1 1 in which we summarize the conditions under which each species is stabilized. It is evident that there are Significant differences between the thiophosphates and the selenophosphates systems and each case we be discuss separately. For the thiophosphates, a wide range of conditions result in the formation of the P5+ Species, exclusively. Low P285 concentrations favors the tetrahedral [PS4]3' unit while high concentrations of P285 favors the formation of a higher nuclearity [P287]4' unit. A review of the literature revealed only a few structurally characterized quaternary alkali compounds featuring the ethane-like [P286]4' unit. Of these compounds, KMP286 (M = Mn, Fe)26 was synthesized by direct combination and only Na0,16Bi1.23P2863 was produced in a molten thiophosphate flux. In the latter case the use of the less basic Na+ counterion may be important for the stabilization of a P4+ containing [PxSyP' unit. Attempts to increase the amount of gold incorporated into a thiophosphate framework or the breakdown of the 317 Table 7-11. Synthetic conditions for the different [Psz]“' units. (M : metal, A2Q = alkali chalcogenide). M / P28e5 / A286 / 86 PM / ligands References l / 1-3/ 1-2/ 10 P417 [P256614' 1, 4, 5, 9, 12, present work 1 / 1.5-2 / 34/ 10 P5+ / [PSe413-. [P865133 [P236814’ 6,8, present work 1 /2-3 /2/ 10 P5+ / [P2867143 [P289914‘ 7 M / P285 / A28 / S PM / ligands References 1 / 1.5-3 / 2-4 / 4-12 P5+ / [P3413‘r [P257143 2, 3. present work 1 /3 / 2 /4 1"“ / [P286143 3 318 [AuPS4]n2"' chains into a previously unknown molecular [Au2(PS4)2]4‘ anion resulted in a mixture of A2AuPS4 and unreacted Au metal. According to the Table 7-11, the most decisive factor to produce P5+ or P” species in the selenophosphates flux is the A28e content. Increasing the A28e concentrations (raising the Lewis basicity), stabilizes [PxSey]n- units containing P5+ species. Decreasing the A28e concentration (lowering the basicity), favors the stabilization of a reduced P” species, and in particular the [P28e6]4- ligand. The only exception is Ti that is found coordinated with [PXSey]n- units containing P5+ species,even under low basicity conditions.7 The highly acidic Ti4+ cation may be responsible for this anomaly. Attempts to synthesize the hypothetical AAuPzSe6 compound, containing a highly oxidized Au+3 ions and a reduced P+4 species were not succesfull. Instead the mixed valent A2AuP28e612 was stabilized, containing both Au+ and Au"3 ions or AzAu2P28e6 , with only Au+ ions were obtained. Structurally the thiophosphate and selenophosphate systems possess staggering differences due mainly to the presence of different [Pny]“' units. Nevertheless, the [AuPS4]n2“' macroanion in (IH) is similar to the [AuP28e3]n3“' macroanion in A3AuP28e3l6 in that two oxidatively coupled bridged [PQ4]3‘ units give rise to a [P2Q3]4' unit. Even in this case, though there are other essential differences such as the metal/phosphorus ratio . When the same ligands and the same countercation are present it is possible to synthesize isostructural compounds. One such example are the isostructural compounds KTiPSe5 7 and KTiP85.27 The optical absorption properties were evaluated by examining single crystal optical transmission spectra of the materials (see Table 7-12). The Spectra confirm the semiconducting nature of the materials by revealing the presence of sharp optical gaps. The band gaps of A2AuPS4 (A = K, Rb, and C8) compounds exhibit steep absorption edges from which the band-gap, Eg, can be assesed at 2.51 (I), 2.59, (11) and 3.04 eV (III), respectively. The band-gaps of AAuP287 are 2.03 (IV), and 2.23 eV (V), respectively . 319 The energy-gaps for A2AuPS4 and AAuP287 (A = K, Rb) increase slightly with a corresponding increase in the cation size. The single crystal absorption specta for (1)-(III) and (IV)-(V) are given in Figures 7-9 and 7-10, respectively. 320 Table 7-12. Optical Band Gaps and Melting Point Data for (1)-(V). Formula Eg, eV mp, oC K2AuPS4 2.50 498 szAuPS4 2.59 401 CszAuPS4 3.04 375* KAuP287 2.09 398 RbAuP287 2.15 404i * = converts to B - C82AuPS4 at T = 375°C. i = incongruent melting. 321 The far-IR and Raman data were in good agreement and the results are summarized in Table 7-13. The far-IR spectra of A2AuPS4 (1)—(III) and AAuP287 (IV)-(V) are quite complex. Absorptions in the 620-400 cm'1 range for (1)-(III) are tentatively assigned to the P-S vibrational stretching modes by analogy with the AuPS4,11 MPS4 (M = In, Ga, and Ba)28 and KNiPS4.29 Absorptions below 400 cm“1 are assigned to S-P-S bending modes and Au-S vibrations.24 By comparison with KAu85 the absorption at ca. 320 cm'1 can be tentatively assigned as Au-S stretching vibration.30 The Far-IR of (IV)-(V) displays a number of absorptions in the 600-400 cm'1 range. The sharp absorption at 458 cm'1 represents the characteristic P-S-P stretching vibration while the remaning absorptions are due to the -P83 stretching vibrations by analogy to Ag4P287.31~32 Absorptions below 400 cm'1 are assigned to Au-S stretching vibrations and P—S deformation modes. The Raman spectra of (1)—(III) displayes absorbencies at ca. 395 cm], 217 cm], and 181 ch, characteristic of the tetrahedral [PS4]3' unit by comparison with the Raman spectrum of Na3PS4'9H4O.27 The characteristic Au-S vibration12 at ca. 327 ch is observed for (I) but not for (ID). The Raman Spectrum of (IV) display absorptions in the 217-670 cm'1 range that are tentatively assigned to P-8 and the absorptions below 200cm‘l are assigned to Au-S stretching vibrations. The Raman Spectra for (I), (III), and (IV) are given in Figure 7-11. Table 7-13. Infrared and Raman Data (cm‘l) for (I), (III) and (IV). 322 KzAuPS4 (I) CszAuPS4 (III) KAuPzS7 (IV) IR Raman IR Raman IR Raman 619 669 603 576 552 551 543 536 519 519 470 471 458 419 410 418 400 395 384 398 387 368 370 363 367 350 355 327 327 325 303 304 289 288 286 278 271 280 268 253 257 255 256 246 241 241 244 240 217 217 217 228 208 209 210 202 203 196 181 183 169 161 170 152 150 323 Differential thermal analysis (DTA) data followed by careful XRD analysis of the residues, Show that (1)-(II) melt congruently at 498°C and 401°C, respectively. Typical thermograrn for (H) is shown in Figure 7-12A. The DTA of (III) indicates that the compound undergoes a structural transformation above 375°C, forming a mixture of Cl- CszAuPS4 and B-CszAuPS4. The x-ray powder pattern of B-CszAuPS4 indexed with the powder pattern of (1)-(H), suggesting that a-CszAuPS4 was a kinetically stable phase. The DTA of (IV) Shows that the material melts congruently at 398°C. The DTA of (V) shows that it melts incongruently at 404°C to form a mixture of RbAuP287 and an as yet unknown phase. Typical therrnograrn for (IV) is shown in Figure 7-123. Table 7-12 Shows the optical and melting point data for all compounds. 324 A 3 I I I I 8 5 2.5 _ .o' s: 2 - Fl) AuPS s is _ 2 4 g ' E9 = 2.596V 8 1 )— .0 < 0.5 'MI 2 2.2 2.4 2.6 2.8 3 Energy (eV) 2 I T I I ’0? g 1.6 b D .9 1.2 _ K2AuPS4 3 E =2.SOeV : 9 .9. 0.8 - - '9 8 .0 0.4 :1: .. < l l l l 2 2.2 2.4 2.6 2.8 3 Energy (eV) 3.5 I I I T I 2? 3 — 'E D , 2.5 - Cs AuPS .e 2 4 g 2 __ Eg =2.956V - S 3: 1.5 - 9 < l l .° 01 2 2.2 2.4 2.6 2.8 3 3.2 Energy (eV) Figure 7- 10: Single crystal optical absorption spectra of (A) K2AuPS4 (I), (B) Rb2AuPS4 (H) and (C) C82AuPS4 (IH). 325 (A) l I I l A 3.5 — :9. 'E 3 3 _ £5 < v 2-5 - KAuPS c 2 7 .g E =2.09eV e. 2 9 _ O U) .D < 1.5 a l l l l 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 Energy (eV) (13) l I I T j A 120 - - £2 5 100 —- . 9' 80 - :5 g 60 ' RbAuPS 3: 2 7 e 40 ' E =2.23 eV ‘ O 9 g 20 - q 0 l L I A l 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Energy (eV) Figure 7-11: Single crystal optical absorption spectra of (A) KAuP287 (IV), and (B) RbAuPzS7 (V). 326 12 T r I I I I 217.5 10 _ 395.4 _ 3 419.2 355.9 g a b 241 2 I ‘ 470.6 ' 1313 3‘ 257.0 E a r J 1522 W 5:; 551.7 / a 4 I- mg .1 2 h _ o J 1 l l I l 700 600 500 400 300 200 100 0 Wavenumbers (enr‘) (B) 14 I I I I I I 12 " g.) 10 _ T: 3 a .— E" 2 s a - u. a! ‘ .1 2 - 0 700 600 500 400 300 200 100 O Wavenumbers (cm") 8 7 - 6 an arbitrary units 5 300 700 600 500 400 300 200 100 o Wavenumbers (cm" ) Figure 7—12: Raman spectra of (A) KzAuPS4, (B) CszAuPS4, and (C) KAuP287. 327 (A) (D X 0 -20- <{-—- Response (uv) —> F -25- endo 30 1 1 I I I I I O 100 200 300 400 500 600 700 Temperature (°C) (B) fill l l Stark; 326 414 ‘Fen q— u:— - (D X 0 -5- -25- I I I I 100 200 300 400 500 600 700 Temperature (°C) Figure 7-13: (A) DTA diagram for a-CszAuPS4 (first cycle).(B) Second DTA cycle of a-CszAuPS4, showing two phases. 328 (A) 15 exo i 10 ~ 0 (I) C O Q. s 5 _ D‘.’ endo 0 I I I I 1 00 200 300 400 500 600 Temp (° C) (B) 10 1 . . T . 6X0 T 5 L . 280 A 0 ' ‘ > 3 o -5 l. - U) 5 L 9, -1o 4 a? -15 L 4 ‘L -20 l- 398 " endo _25 l 1 I 1 l o 100 200 300 400 500 600 Temperature (°C) Figure 7-14: (A) DTA diagram for szAuPS4. (B) DTA diagram for KAuPZST 329 4. Conclusion In summary, the compounds reported here are the first gold thiophosphate compounds perpared in a molten polythiophosphate flux. The structural diversity displayed is astonishing and is due mainly to the variety of binding modes of the [Psz]“' units. The fluxes provided a convenient entry into the unknown Au chemistry and preliminary results indicate that the elusive ternary Au/P/Q (Q = S, Se) compounds can also be synthesized and structurally characterized.34 The use of low temperatures allows the synthesis of thermodynamically unstable compounds. Their metastable nature can be clearly seen in their thermal analysis. Control of the flux composition and basicity is paramount in controling the nature of the [Psz]“' units stabilized in the resulting materials. This knowledge provides the ability to synthesize compounds with preselected [PyQZP' units, although the final structure will be almost imposible to predict. This, in addition to, the significant chemical differences between the thiophosphate and selenophosphate systems, suggests that both systems should be explored to fully comprehend the Chalcophosphate anion chemistry. 10. ll. 12. 14. 15. 330 List of References (a) Sutorik, A.; Kanatzidis, M. G. Progr. Inorg. Chem, 1995, 43, 151. (b) Kanatzidis, M.G. Curr. Opinion Solid State and Mater. Sci., 1997, 2, 139. T. J. McCarthy, T. Hogan, C. R. Kannewurf, M. G. Kanatzidis, Chem. Mater., 1994, 6, 1072. McCarthy, T. J.; Kanatzidis, M. G. J. Alloys Comp., 1996, 236, 70. McCarthy, T. J.; Kanatzidis, M. G. J. Chem. Soc. Chem. Commun., 1994, 1089. McCarthy, T. J.; Kanatzidis,M. G. Chem. Mater., 1993, 5, 1061. Chondroudis, K.; McCarthy, T. J.; Kanatzidis, M.G. Inorg. Chem, 1996, 35 840. Chondroudis, K.; Kanatzidis, M.G. Inorg. Chem, 1995, 34, 5401. Chondroudis, K.; Kanatzidis, M.G. J. Chem. Soc. Chem. Commun.,. 1996, 1371. Loken, S; Tremel. W., Eur. J. Inorg. Chem. , 1998, 283-. (a) Chondroudis, K.; Kanatzidis, M.G. C. R. Acad. Sci. Paris, Series B, 1996, 322, 887. (b) Chondroudis, K.; Kanatzidis, M.G. J. Am. Chem. Soc., 1997, 119, 2574. (a) Chen, J. H; Dorhout, P. K. Inorg. Chem, 1995, 34, 5705. (b) Chen, J. H; Dorhout, P. K. Ostenson, J. E. Inorg. Chem. 1996, 35, 5627. Patzmann, U.; Brockner, W.; Cyvin, B. N.; Cyvin, S. J. Raman Spectroscopy, 1986, I7, 257. Chondroudis, K.; McCarthy, T. J.; Kanatzidis, M.G. Inorg. Chem, 1996, 35, 3451. McCarthy, T.J. Kanatzidis, M.G. Inorg. Chem, 1995, 34, 1257. CERIUSZ, Version 1.6, Molecular Simulations Inc., Cambridge, England, 1994. 16. l7. l8. 19. 18. 20. 21. 22. 23. 24. 25. 26 27. 28. 29. 30. 31. 331 McCarthy, T. J.; Ngeyi, S.-P.; Liao J.-H.; DeGroot, D.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M.G. Chem. Mater., 1993, 5, 331. Walker, N.; Stuart, D. Acta Cryst., 1983, A39, 158. (a) Sheldrick, G. M., in Crystallographic Computing 3 ; Sheldrick, G. M., Kruger, C., Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985, p. 175. (b) Gilmore G. J., Appl. Cryst. ; 1984, I 7, 42. Chondroudis, K.; Hanko, J. A.; Kanatzidis, M.G. Inorg. Chem, 1997, 36, 2623 Brec, R.; Ouvrard, G.; Evain, M.; Grenouilleau, P.; Rouxel, J. J. Solid State Chem, 1983, 47, 174. Y. Park, M. G. Kanatzidis, Angew. Chem. Int. Ed. Engl. 1990, 29, 914. Y. Park, Ph.D. Dissertation, Michigan State University, E. Lansing, MI, 1991. Liao, J.-H.; Kanatzidis, M.G. Chem. Mater., 1993, 5, 1561. Teske, Chr. L. Z. Anorg. Allg Chem, 1978, 445, 193-201. Klepp, K.O.; Bronger, W. J. Less Common Metals, 1988, I37, 13. Hanko, J .A.; Kanatzidis, M.G. Manuscript in preparation . (a) Menzel, F.; Brockner, W; Carrillo-Cabrera, W.; von Schnering, H.G. Z. Anorg. Allg Chem. 1994, 620 , 1081. (b) Carrillo-Cabrera, W.; Sabmannshausen, J .; von Schnering, H.G.; Menzel, F.; Brockner, W. Z. Anorg. Allg Chem, 1994, 620 , 489. Do, J.; Lee, K.; Yun, H. J. Solid State Chem, 1996 125 , 30. (a) D'ordyai, V.S.; Galagovets, I.V.; Peresh, B. Yu.; Voroshilov, Yu. V.; Gerasimenko, V.S.; Slivka, V. Yu.; Russ. J. Inorg. Chem, 1979, 24, 1603. Sourisseau, C.; Cavagant, R.; Fouassier, M.; Brec, R.; Elder, S.H.; Chemical Physics, 1995, 195, 351. Y. Park, Ph.D. Dissertation, Michigan State University, E. Lansing, MI, 1991. Menzel, F.; Ohse, L.; Brockner, W.; Heteroatom Chem, 1990, I, 357. 32. 33. 34. 332 (a) Queigrec, M.; Evain, M.; Brec, R.; Sourisseau, C.; J. Solid State Chem, 1992, 189, 209. Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements; Pergamon Press: New York, 1993, p. 1372. Chondroudis, K.; Hanko, J .A.; Kanatzidis, M.G. Work in progress. 333 CONCLUSIONS AND OUTLOOK The chalcophosphate chemistry was the center of a significant amount of research in the 1970’s through the 1980’s. By using primarily a direct combination approach a number of new ternary chalcophosphate compounds were discovered, many with new structure types. This approach proceeded until almost every element in the periodic table was studied, then the exploratory nature of the research gradually faded and the emphasis shifted to improving the properties of known phases with potential technological interest. The corresponding chalcoantimonide chemistry was in a similar state of development. Using direct combination and hydrothermal synthesis, a number of new ternary chalcoantimonate compounds covering the spectrum of structural possibilities, from molecular species to three-dimensional frameworks, were observed. The lower reaction temperatures utilized by hydrothermal synthesis allowed the replacement of the alkali cations with larger organic counterions, leading to the formation of more open frameworks that may serve as forerunners to a new class of microporous solids. Even with the extensive research in the chalcophosphate and chalcoantimonate systems, there are still metal systems which have not been studied. ’ The molten polychalcogenide flux as a new synthetic methodology, particularly the chalcophosphate and chalcoantimonate fluxes, offer a great potential to expand and characterize new classes of compounds in a systematic way. This experimental methodology was first applied to the selenophosphate system, reviving this area of research which had been dormant for several years. The same approach has now been applied to the thiophosphate system to synthesize new quaternary thiophosphate compounds and probe whether the coordination chemistry paralleled that of the selenophosphate system. This recent activity has uncovered a multitude of new materials with elaborate structure types arising from a variety of new binding modes displayed by the [Pny]n"‘ (Q = S, Se) units. An interesting trend, which became evident in this work, is that the 334 [PxSey]n"‘ species behave very differently than the corresponding [PxSy]n“' species. This trend is manifested in not only in how the [Pny]n“' (Q = S, Se) species bind to metal ions but also with respect to their relative stability. For example the tetrahedral [PQ4]3' (Q = S , Se) unit, typically does not lead to the same structure types when compounds are formed; in fact, they do not give compounds with the same formula. Furthermore, the analogous synthetic conditions which may give rise to a certain [PxSey]n2' unit tend to give rise to a different [PxSy]n2' unit, the anions often containing the phosphorous atom in a different oxidation state. The latter is a consequence of the difference in the electronegativity between the S and Se atoms. These facts underscore the need for the simultaneous exploration of both the thiophosphate and selenophosphate fluxes for the same metal system, and the continued search for fundamental differences in the chemistry and reactivity. To further develop the molten polychalcogenide flux as a general synthetic methodology, it was extended to the chalcoantimonate systems, where there are few reported quaternary alkali chalcoantimonate compounds. A comparison of the chalcoantimonate to the chalcophosphate fluxes provides insight into the structural and chemical complexity of the two systems. The chalcoantimonate system follows the experimentally observed trend of the polychalcophosphate system. Basic conditions (increasing A2Q) favor the tetrahedral [EQ4]3‘ unit (E = P, Sb; Q = 8, Se). Under these conditions, the [SbQ4]3' unit is observed as a discrete tetrahedral species, while the [PSe4]3‘ unit can participate in complex condensation equilibra, forming higher nuclearity chalcophosphate units such as [P28e3]4', and [PZSe9]4'. Under acidic conditions (increasing Q or decreasing A2Q) both P and Sb favor reduced species. In the polythiophosphate system the reduced P+4 species is readily observed in the ethane - like [P2Q6]4' unit (Q = S, Se). While the antimony analog of the [P2Q6]4‘ unit has yet to be reported, the pyramidal Sb3+ species is the only other oxidation state species observed. The pyramidal [SbQ3]3' unit (Q = S, Se) also participates in a separate yet complex 335 condensation equilibria, forming higher nuclearity [beQy]"' units. The thiophosphate system does not appear to stabilize a reduced species under acidic conditions, although increasing the P285 concentration favors a complex condensation equilibra, forming higher nuclearity chalcophosphate units such as [P257]4‘, and [P389]3'. Although the ability to synthesize new materials with preselected ligands imparts a large degree of synthetic control, a number of important parameters are still unknown. In particular, the nature of the thiophosphate and chalcoantimonate systems in the molten state can only be speculated, based on the [Pny]“' and [beQy]“' units found in the resulting compounds. Further elucidation of the kind of [PnyP' and [beQy]“‘ units, existence of intermediates, equlibria resulting in the formation of higher nuclearity units, and the interaction with the metal need to be examined. The need for in situ experiments, such as Raman spectroscopy and X-ray or neutron diffraction experiments, may be the best approach to gain this information. In addition, NMR spectroscopy should be of considerable help in elucidating the equilibrium issues in these systems. APPENDIX 1 Thiophosphate Flux Synthesis of CszCuP389: An Unusual Chrial Compound with One-Dimensional Screw Helices. 336 337 Thiophosphate Flux Synthesis of CszCngng An Unusual Chrial Compound with One-Dimensional Screw Helices. Jason A. Hanko and Mercouri G. Kanatzidis* Department of Chemistry, and Center for Fundamental Materials Research, Michigan State University, East Lansing, Michigan 48824. Abstract The reaction of Cu with a molten mixture of Cs28/P285 produces the quaternary compound CszCuP389 (I). The bright yellow crystals of CszCuP3S9 are stable in air and water for over a month. The compound crystallizes in the acentric, hexagonal space group P65 (no. 170) and at 23°C: a = 15.874(2)A; c = 35.100(7)A; v = 7660(3)A3; z = 22. The [CuP389]n2“' chains are comprised of an alternating arrangement of tetrahedral Cu+ ions and cyclic [P389]3‘ units. The structure contains two unique [CuP389]n2"' chains that pack in a helical fashion along the c-axis. The [P3Sg]3' unit is related to the molecular adamantine [P4810] unit with one [PS]3+ vertices removed and replaced by a Cu+ ion. The title compound was characterized by Differential Thermal Analysis (DTA), Far-IR and Raman, and single crystal optical transmission spectroscopy. The preliminary investigations into the non-linear properties of (I) will be reported. 338 Introduction During the past several years, we have demonstrated that the chalcophosphate fluxes are an excellent choice for the synthesis of new multinary chalcophosphate compounds.l These fluxes are commonly formed by the in situ fusion of A2Q/P2Q5/Q (A = alkali metal; Q = S, Se), forming a variety of [Psz]“' units in a molten polychalcogenide solvent. Under these conditions the coinage metals (Cu, Ag, Au) are readily oxidized by the polychalcogenide ligands from the flux and then coordinated by the various [Psz]“' (Q = S, Se) units, building up extended frameworks, stabilized by alkali cations. The exploration of the coordination chemistry of the coinage metals with these fluxes produced several unusual compounds, examples include: CszM2P28e6 (M = Cu, Ag),2 K3M3P3Se9 (M = Cu,33 Ag,3b), ALa(P28e6),4 the mixed-valent A6Au11.5Aum1.5(P28e6)3 (A: K, Rb),5 A3AuP28e3 (A = K, Rb, Cs),6 A2Au2P28e6 (A = K, Rb),6 A2AuPS4 (A = K, Rb, Cs),6 and AAuPZS7 (A = K, Rb),6 K3CuP287,7 and K2Cu2P4Se10.8 An unexplored area of the chalcophosphate chemistry is the phosphorous-rich thiophosphate fluxes, which can be achieved by the removal of the elemental sulfur. As previously reported,6 a [P28 5] rich flux favors the formation of the higher nuclearity [P257]4' unit. With the removal of the elemental sulfur delutant, the phosphorous concentration is further increased, stabilizing even higher nuclearity [PySz]n' units. Preliminary investigations with these phosphorous- rich fluxes produced the quaternary compounds A2CuP389 (A = K, Rb) containing the cyclic [P389]3' unit and C32Cu2PZS67, containing the ethane-like [P286]4' unit. Continuing our investigation of the coordination chemistry of these higher nuclearity [PxSy]"’ units with coinage metals, we report on the synthesis, and physicochemical characterization of C52CuP389,9 a novel one—dimensional compound with an acentric structure. The [CuP389]n2“‘ macroanion in (I) are very similar to the [CuP389]n2"' chains previously reported in A2CuP389 (A = K, Rb),7 but with a completely different packing arrangement. In the previous structure, the [CuP389]n2“‘ chains pack in an alternating 339 centrosymmetric fashion, but in CszCuP389 the [CuP3S9]n2“' macroanions pack in a helical screw-like arrangement. The structure of CS2CUP3SQIO contains two different [CuP3S9]n2“‘ chains of alternating cyclic [P3S9]3' units chelating in a tridendate fashion to Cu+ ions, see Figure 1. Figure 2a shows a [CuP389]n2“' chain, highlighting the helical packing along the polar c- axis. Figure 2c shows a view of a [CuP389]n2“' chain down the polar c-axis. A view of the [CuP389lnzn' chain containing Cu(1) is shown in Figure 2c. As viewed in Figure 2c, the basal plane of the [CuS4] tetrahedrons is defined by the tridendate chelation of three of the sulfurs from the cyclic [P389]3' units to the Cu+ ions while the apical site of the copper tetrahedra is filled by a sulfur atom from a second neighboring [P3Sg]3' unit. The average Cu-S distance of the three [CuS4] tetrahedrons is 2.306(9)A, 2.307(9)A, and 2.316(7)A for Cu(l), Cu(2), and Cu(3), respectively. The s- Cu-S angles display a slight distortion from the tetrahedral geometry, with an average S- Cu-S angle of 109(2)°, 109(1)°, and 109(2)° for Cu(l), Cu(2), and Cu(3), respectively. The two [CuP389]n2“' chains contain three different [P3S9]3' units, with P-S bond distance and S-P-S angles that are well within the expected range reported for A2CuP3S9 (A = K, Rb).7 The P-S distances and S-P—S angles of [P3S9]3’ unit in the [CuP389]n7-“' chain containing the Cu(l) ion are representative of the other two [P3S9]3‘ units. The P-S distances are in the range from 1.954(4)A to 2.139(3)A with an average P—S distances of 2.04(3)A, 2.05(3)A and 204(3) for P(1), P(2) and P(3), respectively. The S-P-S angles reveal only a slight distortions from the tetrahedral coordination geometry, with average S- P—S angle of 109(2)° for P(1), P(2), and P(3), respectively. The [P3S9]3‘ unit contains a [P383] ring chair-shaped a almost-equilateral triangle with a mean P-P-P angle of 60.00(5)° and a mean P-P distance of 3.483(3)A. Selected bond distances and angles are given in Table 1. As described in the structure of A2CuP3Sg (A = K, Rb),7 there are two types of P- S distances in the three different cyclic [P3Sg]3‘ units, the longer P-S distances within the 340 P383 ring and shorter P-S distances for the terminal sulfides. The average ring P-S distances were 2.11(5)A, 2.11(4)A and 2.12(4) for P(1), P(2), and P(3), respectively. The average terminal P-S distances were much shorter at 1.97(1)A, for P(1), P(2), and P(3), respectively. Other [PxSy]n' units such as the bi-tetrahedral [P286]2' (2.111(1)A versus 1.993(1)A) and quadri-tetrahedral [P4812]4‘ units (2.113(3)A versus 1.989(2)A), display a similar discrepancies in the [P383] ring P-S bond distances and the terminal P-S bond distances. In oxides, this segregation is more pronounced. Examination of the corresponding phosphates Na3P309 and the monohydrate Na3P3Og,H20,12 show that the average P-O bond distance in the P303 ring is 1.615A, while the average terminal P-O bind distance is 1.484A. The [CuP389]2' chains are separated by Cs+ ions that are located in six different sites. In CszCuP389 Cs(l) is in ninefold coordination to sulfur atoms [range of Cs(1)-S distances, 3.501(3)A to 4.107(3)A ; average 3.73(6)A], Cs(2) is in eightfold coordination to sulfur atoms [range of Cs(2)-S distances, 3.515(3)A to 3.911(3)A; average 3.70(4)A], Cs(3) is in tenfold coordination to sulfur atoms [range of Cs(3)-S distances, 3.512(3)A to 4.082(3)A; average 3.83(6)A], Cs(4) is in tenfold coordination to sulfur atoms [range of Cs(4)-8 distances, 3.575(3)A to 4.185(3)A; average 3.83(7)A], Cs(5) is in ninefold coordination to sulfur atoms [range of Cs(5)-S distances, 3.530(3)A to 4.143(3)A ; average 3.74(8)A], and Cs(6) is in ninefold coordination to sulfur atoms [range of Cs(6)-S distances, 3.552(3)A to 4.085(3)A ; average 3.73(5)A]. The optical absorption properties were evaluated by examining the single crystal optical transmission spectrum of CszCuP389, see Figure 3. The sharp optical gap in the spectrum suggest that the compound is a wide bandgap semiconductor with a band gap (Eg) of 2.61 eV. The far-IRlza spectrum of CszCuP389 displays several absorptions in the 600-400 cm'1 range. The sharp absorption at 400 cm'1 and 463 cm'1 are characteristic of the P-S-P stretching vibrations in the cyclic [P389]4‘ unit. The doubly bridging bonding motif of the 341 tetrahedral [P84]3' units lowers the energy of the S-P-S vibrations. The remaining absorptions are tentatively assigned to the -P82 stretching vibrations by analogy to AAuP287 (A = K, Rb)3. Absorptions below 400 cm'1 are assigned to Cu-S stretching vibrations. The Raman spectrumlzb of C82CUP359 displays absorptions in the 250-600 cm" range that are tentatively assigned to P-8 and absorptions below 250 cm'1 are assigned to Cu-S stretching vibrations. The stretching vibrations at ca. 386 cm'1 and 294 cm"1 are assigned to the [PS4]3' by analogy to AzAuPS4, AAuP287 (A = K, Rb)3, and A2P286 (A = K, CS)”. The Raman spectrum of Cs2CuP389 is shown in Figure 4. Differential thermal analysis (DT A) data followed by careful XRD analysis of the residues, revealed that the title compound melts incongruently at ~477°C forming a mixture of CszCuP389 and C82P23614. The non - linear optical properties were examined using a preliminary screening test developed in our lab.15 The sample displayed a weak second harmonic generation (SHG) signal. Although a weak SHG signal is observed, the test is not conclusive because the sample cannot be examined in all possible orientations. The synthesis of CszCuP389 further validates the chemical diversity of the thiophosphate system and further underscores the chemical difference between the selenophosphate system. The acentric structure of CszCuP389 is very unusual and could not have been predicted before the compound was synthesized. The highly acentric arrangement makes it a potential candidate for a number of applications. While the preliminary investigations of the non-linear optical properties of the title compound were encouraging, further testing is necessary. The dramatic structural change introduced by the slight modifications of the thiophosphate flux shows that we are still at the mercy of serendipity. For at this stage, we only have a rough idea of the complex equlibra between stabilization of these higher nuclearity [PxSyln' units. \OOOQG 10. 342 List of References (a) Kanatzidis, M.G. Curr. Opinion Solid State and Mater. Sci. 1997, 2, 139. (b) Sutorik, A.; Kanatzidis, M.G. Prog. Inorg. Chem., 1995, 43, 151. McCarthy, T. J.; M. G. Kanatzidis, Inorg. Chem. 1995, 34, 1257. Dorhout, P.K.; Malo, T.M. Z. Anorg. Allg. Chem. 1996. 622, 385. Chen, J.H.; Dorhout, P.K. Inorg. Chem. 1995, 34, 5705. (b) Chen, J.H.; Dorhout, P.K.; Ostenson, J.E. Inorg. Chem. 1996, 35, 5627. Chondroudis, K.; McCarthy, T. J.; Kanatzidis, M.G. Inorg. Chem. 1996, 35, 3451. Chondroudis, K.; Hanko, J .A.; Kanatzidis, M.G. Inorg. Chem. 1997, 36, 2632. Hanko, J .H.; Kanatzidis, M.G. Submitted for Publication. Chondroudis, K.; Kanatzidis, M.G. Submitted for Publication. (a) CszCuP3S9 was synthesized from a mixture of 0.032g (0.50 mole) Cu, 0.220g (1 mole) P285 and 0.149g (0.50 mole) C528 which was sealed under vacuum in a Pyrex tube and heated to 500°C for 4 d followed by cooling to 100°C at 4°C h'l. The excess Csx[PySz] flux was removed by washing with degassed methanol under inert atmosphere to reveal irregular bright yellow crystals (75% yield based on Cu). Microprobe analysis gave an average composition of €52.25CUP3.856.1- A Siemens SMART Platform CCD diffractometer from a crystal of 0.200 x 0.200 x 0.120 mm dimensions and Mo Ka (l = 0.71073 A) radiation. An empirical absorption correction.was applied to the data during data processing. Crystal data at 23°C: a = 15.874(2)A; b = 15.874(2)A; c = 35.100(7)A; a = 90.00; b = 90.00; g = 120.00; V = 7660(3)A3; Z = 22; Dc = 2.767 g/cm'3; space group P6s (no. 141); m = 6.804 cm -1 index ranges -19 = h = 21, -20: k= 21, -22 =1: 47 ; total data 48681; unique data 9745; (Rim = 0.0944); data with F02 > 28(F02) 4056; no. of variables, 406; final R/Rw 0030/0069; GOF 0.335; largest diff. peak and hole 11. 12. 13. 14. 15. 343 0.882/-O.767e A‘3. (b) SHELXL: Version 5.03, 1994. Sheldrick, GM. Siemens Analytical X-ray Instruments, Inc., Madison, WI 53719. Ondik, H.; Acta. Crystallogr, 1965, 18, 226. (a) Far-IR (CsI matrix) gave absorptions at 619(w), 519(w), 384(8), 370(s,sh), 241(w), and 16l(w)cm'1. (B) The Raman spectrum, in the same region, gave absorptions at 653(w,b), 586(m), 555(w), 486(w,b), 386(s), 325(m), 294(w), 254(w), 221(8), 175(m), and 150(w) Brockner, W.; Becker, R.; Eisenmann, B.; Schaefer, H. Z. Anorg. Allge. Chem. 1985, 520, 51. Sala, O.; Temperini, M.L.A. Chem. Phys. Lett. 1975, 36, 625. Marking, G.A.; Liao, J.-H.; Kanatzidis, M.G. manuscript in preparation. (b) Liao, J .-H.; Ph. D. dissertation, Michigan State University, 1993. (c) Kanatzidis, M.G.; Liao, J .-H.; Marking, G.A., United States Patent 5,618,471. (d) Kanatzidis, M.G.; Liao, J .-H.; Marking, G.A., United States Patent 5,614,128. Table 1. Selected bond distances (A) and angles (0) for Cs2CuP3S9. 344 Cu( 1 )-S(12) Cu( 1 )-S(23) Cu(l)-S(14) Cu( 1 )-S(01) Cu(l)—S (mean) Cu(2)-S(l 1) Cu(2)-S(09) Cu(2)-S(25) Cu(2)-S(O7) Cu(2)-S (mean) Cu(3)—S(05) Cu(3)-S(O6) Cu(3)-S(l9) Cu(3)-S(24) Cu(3)—S (mean) S(l 1)-Cu(2)—S(09) S(l 1)-Cu(2)-S(25) S(09)-Cu(2)-S(25) S(l 1)-Cu(2)-S(O7) S(O9)-Cu(2)-S(07) S(25)-Cu(2)—S(O7) 2.286(3) 2.300(3) 2.305(3) 2.333(3) 2.306(9) 2.281(3) 2.301(3) 2.323(3) 2.325(3) 2.307(9) 2.293(3) 2.315(3) 2.322(3) 2.337(3) 2.316(7) 116.34(10) 112.60(12) 105.2101) lO8.83(12) 104.64(11) 108.75(11) P( 1)-S( 1) P( 1 )-S(17) P(1)-S(6) P(1)-S(18) P(2)-S( 10) P(2)-S(15) P(2)-S(24) P(2)-S(26) P(3)-S(2) P(3)-S(3) P(3)-S(14) P(3)-S(18) P(4)—S(26) P(4)-S(04) P(4)-S(5) P(4)-S(l 3) P(5)-S(3) P(5)-S(8) P(5)-S(12) P(5)-S(17) 1.977(3) 2.121(3) 1.991(3) 2.124(3) 2.139(3) 1.958(4) 1.983(4) 2.121(4) 1 954(4) 2. 131(4) 1.981(4) 2.116(4) 2.135(4) 2.117(4) 1.985(4) 1 955(4) 2.122(4) 1.955(4) 1.988(4) 2.134(4) 345 S(21)-P(7)-S(07) 1 15.6(2) P(6)-S(7) 1.977(4) S(21)-P(7)-S(22) 103.9(2) P(6)-S( l 6) 2.120(4) S(O7)-P(7)-S(22) l 13.3(2) P(6)-S(21) 1.963(4) S(21)-P(7)-S(l6) 102.7(2) P(6)-S(22) 2.1 13(4) S(O7)-P(7)-S(16) 1 15.3(2) S(22)—P(7)-S( 16) 104.6(2) P(7)-S(4) 2.125(3) P(7)-S( 10) 2.101(3) S(25)-P(8)-S(09) 1 12.4(2) P(7)-S( l9) 1.977(3) S(25)-P(8)-S(27) 114.3(2) P(7)-S(23) 1.985(3) S(O9)-P(8)-S(27) 105.7(2) S(25)-P(8)-S(22) 1 14.5(2) P(8)-S(22) 2.1 19(3) S(O9)-P(8)-S(22) 103.0(2) P(8)-S(25) 1.976(3) S(27)-P(8)-S(22) 105.98(14) P(8)-S(27) 2.1 17(4) P(8)-S(9) 1 983(3) S(20)-P(9)-S(1 l) l 14.7(2) S(20)-P(9)«S( 16) 105.1(2) P(9)-S(l 1) 1.988(4) S(l 1)-P(9)-S(16) 1 13.6(2) P(9)-S(16) 2.120(4) S(20)-P(9)-S(27) 103.4(2) P(9)-S(27) 2.122(4) S(l l)-P(9)-S(27) l 13.6(2) P(9)-S(20) 1.959(4) S(16)-P(9)-S(27) 105.3(2) 346 Figure Captions Figure l: ORTEP representation of CszCuP389 as viewed down perpendicular to the c - axis (50% probability ellipsoids). ellipsoids with octant shading; Cu, boundary and axis ellipsoids; P atoms, boundary ellipsoids; S, and principal and axis ellipsoids; Cs. Figure 2: (A) View of the [CuP389]2', chain containing Cu( 1), prepindicular to the c - axis, highlighting the helical arrangemt of the chain. (B) View of the same chain, viewed down the c - axis. (C) View of a fragment of a [CuP389]2' chain containing Cu(l), with a labeling scheme. Figure 3: Single-crystal optical absorption spectrum of Cs2CuP389. The sharp features at high absorbance are noise and due to the very low transmission of light at those energies. Figure 4. Raman spectra of Cs2CuP389, 347 a 41 0’91) ob $1 4,: 1‘ s. a - O/J’Zik/ wt/ :5 (A) 348 ‘ as , L';.. .\ It 1 use. \\‘;‘ 7'” ‘- -O‘ '- — \\"‘-'-'~ new ’. ,‘ .ttk't‘a.‘_‘ " \ 95;, a, (C) Cul 81 P3 . P1 A» “ C’ C 54 . $5 $6 8:8 S9 Raman Intensity (arb. units) 10 - 700 586 600 349 386 175 221 325 254 if" 500 400 300 200 100 O Wavenumber 1cm“) Absorption Coeff. (arb units) 350 0.5 - Cs2CuP389 Eg = 2.61eV 2.2 2.4 2.6 2.8 3 Energy (eV)