SYNTHESIS AND CHARACTERIZATION OF NEW LANTHANIDE CHALCOGENIDES Except where reference is made to the work of others, the work described in this dissertation is my own or was done in collaboration with my advisory committee. This dissertation does not include proprietary or classified information. ????????????????????????????????? Geng Bang Jin Certificate of Approval: ?????????????????????????????? ?????????????????????????????? Thomas R. Webb Thomas E. Albrecht-Schmitt, Chair Associate Professor Professor Chemistry and Biochemistry Chemistry and Biochemistry ?????????????????????????????? ?????????????????????????????? Andreas J. Illies Peter D. Livant Professor Associate Professor Chemistry and Biochemistry Chemistry and Biochemistry ?????????????????????????????? Joe F. Pittman Interim Dean Graduate School SYNTHESIS AND CHARACTERIZATION OF NEW LANTHANIDE CHALCOGENIDES Geng Bang Jin A Dissertation Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Auburn, Alabama August 4, 2007 iii SYNTHESIS AND CHARACTERIZATION OF NEW LANTHANIDE CHALCOGENIDES Geng Bang Jin Permission is granted to Auburn University to make copies of this dissertation at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. ______________________________ Signature of Author ______________________________ Date of Graduation iv VITA Geng Bang Jin, son of Biao Sheng Jin and Xian Xiang Jin, was born October 8, 1981, in Donyang, Zhejiang, China. He graduated from Donyang Second High School in 1999. He attended East China University of Science and Technology in Shanghai for four years, where he graduated in July of 2003, with a Bachelor of Science degree in Applied Chemistry. In August of 2003, he entered Graduate School at Auburn University pursing a doctoral degree in Chemistry under the tutelage of Dr. Thomas E. Albrecht-Schmitt. He married Pan Chen, daughter of Wang Heng Chen and Xiao Xiang Li, on January 26, 2007. v DISSERTATION ABSTRACT SYNTHESIS AND CHARACTERIZATION OF NEW LANTHANIDE CHALCOGENIDES Geng Bang Jin Doctor of Philosophy, August 4, 2007 (B.S., East China University of Science and Technology, 2003) 299 Typed Pages Directed by Thomas E. Albrecht-Schmitt A large number of new ternary and quaternary lanthanide chalcogenides have been synthesized through solid-state reactions of corresponding elements with the aid of alkali metal halides and Sb2Q3 (Q = S, Se) fluxes. Molten Sb2Q3 (Q = S, Se) fluxes have been valuable media for accessing interlanthanide chalcogenides. The structures of these compounds were determined by single crystal X-ray diffraction. Their optical and magnetic properties were characterized by UV-vis-NIR diffuse reflectance spectroscopy, magnetic susceptibility measurements, and M?ssbauer spectroscopy. Their structures and physical properties depend highly on the choices of lanthanides and chalcogenides. They have shown a variety of structures including ordered and disordered types under different reaction conditions. Most compounds are semiconductors with wide tunable vi band gaps. Different magnetic behaviors have been found in these systems, namely Curie-Weiss type paramagnetism, van Vleck paramagnetism, antiferromagnetism, ferromagnetism, and spin glass performance. Possible spin-frustrations in some of these interlanthanide compounds were also discussed. vii ACKNOWLEGMENTS There are a lot of people to whom I would like to express my sincere gratitude. I feel so lucky to have met you all during my four years in Auburn. First of all, I would like to thank my advisor, Professor Thomas E. Albrecht-Schmitt, without whom none of this would have been possible. Thank you for your support, your guidance, and most importantly, your friendship. You have been so much more than an advisor to me. Also, I want to thank my committee members, Dr. Illies, Dr. Webb, and Dr. Livant for the valuable suggestions they have provided. I am also indebted to all my group members and collaborators. You have all been so nice to me and have shown great patience. I know I could be quite trying at times with my long list of questions. Tyler, Shehee, Daniel, and Phil, thank you for helping me get my research started when I first came to the lab. Ling Jie, Tanya, Noel, Yaqin, Anna, and Andrea, thank you for all of your support. Travis, I want to thank you most for what you have done outside of work. You have been so helpful for fixing my car, teaching me English, and helping me assimilate American culture...I would like to thank Dr. Choi and Dr. Booth for their substantial help in magnetic property measurements all along the way. Finally, I would like to thank my family. Dad and Mom, thank you for being supportive of every choice I?ve made. Pan, your companionship and encouragement have helped me to keep doing my best in my studies and work. viii Style manual or journal used: American Chemical Society Style Computer software used: Microsoft Word 2000, Microsoft Excel, Atoms v.5.0, & 6.0, CorelDRAW 10, Microcal Origin 6.0 ix TABLE OF CONTENTS LIST OF FIGURES ........................................................................................................ xvii LIST OF TABLES.......................................................................................................... xxv CHAPTER 1. INTRODUCTION .......................................................................................1 SOLID STATE SYNTHESIS..........................................................................3 STRUCTURES................................................................................................7 MAGNETIC PROPERTIES..........................................................................16 OPTICAL PROPERTIES ..............................................................................23 REFERENCE.................................................................................................28 CHAPTER 2. SYNTHESIS, STRUCTURE, AND MAGNETIC PROPERTIES OF THE NONCENTROSYMMETRIC TERNARY RARE-EARTH ANTIMONY POLYSULFIDE Eu6Sb6S17 ................................................................... 34 ABSTRACT..................................................................................................34 INTRODUCTION ........................................................................................35 EXPERIMENTAL........................................................................................36 SYNTHESES...................................................................................36 CRYSTALLOGRAPHIC STUDIES ...............................................36 MAGNETISM .................................................................................41 RESULTS AND DISCUSSION...................................................................41 x STRUCTURE .................................................................................41 MAGNETIC SUSCEPTIBILITY...................................................43 REFERENCES ............................................................................................50 CHAPTER 3. SYNTHESES, STRUCTURES, AND MAGNETIC PROPERTIES OF THE EUROPIUM(II) SELENIDO PNICTOGENATES(III), EuPnSe3 (Pn = Sb, Bi) .......................................................................................................52 ABSTRACT..................................................................................................52 INTRODUCTION ........................................................................................53 EXPERIMENTAL........................................................................................54 SYNTHESES..................................................................................54 EuPnSe3..................................................................................................................................... 54 CRYSTALLOGRAPHIC STUDIES ..............................................55 MAGNETISM ................................................................................56 151Eu AND 121Sb M?ssbauer SPECTROSCOPY ..........................56 RESULTS AND DISCUSSION...................................................................56 STRUCTURE of EuPnSe3 (Pn = Sb, Bi)........................................56 MAGNETIC SUSCEPTIBILITY...................................................68 151Eu M?ssbauer SPECTROSCOPY ..............................................68 121Sb M?ssbauer SPECTROSCOPY .............................................72 CONCLUSIONS...........................................................................................73 REFERENCES ............................................................................................78 CHAPTER 4. SYNTHESES, STRUCTURE, MAGNETISM, AND OPTICAL PROPERTIES OF THE ORDERED MIXED-LANTHANIDE xi SULFIDES, ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)..................................................81 ABSTRACT..................................................................................................81 INTRODUCTION ........................................................................................82 EXPERIMENTAL........................................................................................83 MATERIALS..................................................................................83 SYNTHESES of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)........84 CRYSTALLOGRAPHIC STUDIES ..............................................84 POWDER X-RAY DIFFRACTION ..............................................85 MAGNETIC SUSCEPTIBILITY MEASUREMENT ...................88 UV-vis-NIR DIFFUSE REFLECTANCE SPECTROSCOPY ..........................................................................88 RESULTS AND DISCUSSION...................................................................88 EFFECTS OF SYNTHETIC PARAMETERS ON PRODUCT COMPOSITION AND STRUCTURE........................88 STRUCTURAL FEATURES OF ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)....................................91 MAGNETIC PROPERTIES OF ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)....................................98 OPTICAL PROPERTIES OF ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)..................................103 CONCLUSIONS.........................................................................................105 REFERENCES ...........................................................................................109 CHAPTER 5. SYNTHESES, STRUCTURE, MAGNETISM, AND xii OPTICAL PROPERTIES OF THE INTERLANTHANIDE SULFIDES ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd)........................................................................................114 ABSTRACT ............................................................................................114 INTRODUCTION .....................................................................................115 EXPERIMENTAL.....................................................................................116 STARTING MATERIALS...........................................................116 SYNTHESES OF ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd).......................116 CRYSTALLOGRAPHIC STUDIES ............................................117 POWDER X-RAY DIFFRACTION ............................................123 MAGNETIC SUSCEPTIBILITY MEASUREMENT .................123 UV-vis-NIR DIFFUSE REFLECTANCE SPECTROSCOPY.....123 RESULTS AND DISCUSSION.................................................................124 STRUCTURES OF ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd)....................124 MAGNETIC SUSCEPTIBILITY.................................................128 OPTICAL PROPERTIES .............................................................133 CONCLUSIONS.........................................................................................133 REFERENCES ...........................................................................................136 CHAPTER 6. SYNTHESES, STRUCTURE, MAGNETISM, AND OPTICAL PROPERTIES OF LUTETIUM-BASED INTERLANTHANIDE SELENIDES .............................................................................138 ABSTRACT...............................................................................................138 INTRODUCTION .....................................................................................139 EXPERIMENTAL.....................................................................................140 xiii STARTING MATERIALS...........................................................140 SYNTHESES................................................................................140 CRYSTALLOGRAPHIC STUDIES ............................................141 POWDER X-RAY DIFFRACTION ............................................147 MAGNETIC SUSCEPTIBILITY MEASUREMENT .................147 UV-vis-NIR DIFFUSE REFLECTANCE SPECTROSCOPY.....147 RESULTS AND DISCUSSION.................................................................148 SYNTHESIS OF Ln/Ln'/Q using Sb2Q3 fluxes (Q = S, Se).........148 STRUCTURE OF LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd)......148 MAGNETIC SUSCEPTIBILITY.................................................159 OPTICAL PROPERTIES .............................................................170 CONCLUSIONS.........................................................................................170 REFERENCES ...........................................................................................173 CHAPTER 7. SYNTHESES, STRUCTURE, MAGNETISM, AND OPTICAL PROPERTIES OF THE PARTIALLY ORDERED QUATERNARY INTERLANTHANIDE SULFIDES PrLnYb2S6 (Ln = Tb, Dy)......................................176 ABSTRACT...............................................................................................176 INTRODUCTION .....................................................................................177 EXPERIMENTAL.....................................................................................178 STARTING MATERIALS...........................................................178 SYNTHESES................................................................................178 CRYSTALLOGRAPHIC STUDIES ............................................179 POWDER X-RAY DIFFRACTION ............................................185 xiv MAGNETIC SUSCEPTIBILITY MEASUREMENT .................185 UV-vis-NIR DIFFUSE REFLECTANCE SPECTROSCOPY.....185 RESULTS AND DISCUSSION.................................................................185 STRUCTURE OF PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) ...................185 MAGNETIC SUSCEPTIBILITY.................................................187 OPTICAL PROPERTIES .............................................................194 CONCLUSION...........................................................................................194 REFERENCES ..........................................................................................197 CHAPTER 8. SYNTHESES, STRUCTURE, MAGNETISM, AND OPTICAL PROPERTIES OF THE ORDERED INTERLANTHANIDE COPPER CHALCOGENIDES Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) .......................200 ABSTRACT................................................................................................200 INTRODUCTION ......................................................................................201 EXPERIMENTAL......................................................................................202 STARTING MATERIALS...........................................................202 SYNTHESES................................................................................202 CRYSTALLOGRAPHIC STUDIES ............................................203 POWDER X-RAY DIFFRACTION ............................................204 MAGNETIC SUSCEPTIBILITY MEASUREMENT. ................204 UV-vis-NIR DIFFUSE REFLECTANCE SPECTROSCOPY.....207 RESULTS AND DISCUSSION.................................................................207 STRUCTURE OF Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) ...........................................207 xv MAGNETIC SUSCEPTIBILITY..................................................217 OPTICAL PROPERTIES ..............................................................217 CONCLUSION...........................................................................................224 REFERENCES ...........................................................................................226 CHAPTER 9. PARTIALLY-FILLED MIXED-LANTHANIDE VARIANTS OF THE K2Tm23.33S36 STRUCTURE-TYPE: STRUCTURE AND PROPERTIES OF CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La-Nd, Sm-Yb; y = 0.26 ? 0.33).......................230 ABSTRACT................................................................................................230 INTRODUCTION ......................................................................................231 EXPERIMENTAL......................................................................................232 STARTING MATERIALS...........................................................232 Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Eu, Gd) ...............................................233 Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = Tb, Dy, Ho, Er, Tm, Yb).....................................................233 CRYSTALLOGRAPHIC STUDIES ............................................234 POWDER X-RAY DIFFRACTION ............................................235 MAGNETISM ..............................................................................235 X-RAY ABSORPTION NEAR EDGE SPECTROSCOPY (XANES) ........................................................239 UV-vis-NIR DIFFUSE REFLECTANCE SPECTROSCOPY......239 RESULTS AND DISCUSSION.................................................................239 CRYSTAL STRUCTURES..........................................................239 xvi MAGNETIC PROPERTIES..........................................................248 XANES ..........................................................................................254 OPTICAL PROPERTIES ..............................................................257 CONCLUSION...........................................................................................259 REFERENCES ...........................................................................................260 CHAPTER 10. SUMMARY...........................................................................................266 xvii LIST OF FIGURES Figure 1.1. Typical coordination geometries for lanthanides (green) bound to chalcogenides (yellow) .................................................................................10 Figure 1.2. Common connectivities between coordination polyhedra of lanthanide (green) chalcogenides (yellow)......................................................................11 Figure 1.3. General building units for structures of lanthanide chalcogenides: one- dimensional chains of lanthanide coordination polyhedra.............................12 Figure 1.4. Connectivities between one-dimensional chains of lanthanide coordination polyhedra.......................................................................................................13 Figure 1.5. A view of the two-dimensional structure of ?-LaYbS3 along the a axis. La-S bonds have been omitted for clarity......................................................14 Figure 1.6. An illustration of the three-dimensional channel structure of EuYb2S4 down the b axis. Eu-S bonds have been omitted for clarity ..................................15 Figure 1.7. Inverse molar magnetic susceptibility vs temperature for Gd1.87Lu2.13Se6 under an applied magnetic field of 0.1 T between 2 and 300 K......................................................................................20 Figure 1.8. The temperature dependence of the reciprocal molar magnetic susceptibility for ?-NdLuSe3 under an applied magnetic field of 0.1 T between 2 and 300 K........................................................................21 xviii Figure 1.9. Inverse molar magnetic susceptibility as a function of temperature for Sm1.82Lu2.18Se6 under an applied magnetic field of 0.1 T between 2 and 300 K....................................................................................22 Figure 1.10. Variation of the band gap Eg of Ln2Q3 (Q = O, S, Se) in the lanthanide series: 1) oxides; 2) sulphides; 3) selenides ...............................26 Figure 1.11. Variation of the dissociation (atomization) energy D of Ln2O3 in the lanthanide series..............................................................27 Figure 2.1. A view of the [Sb3S7]5? anions that consist of a trimer of corner-sharing SbS3 units in Eu6Sb6S17. 50% probability ellipsoids are depicted...................................................................................44 Figure 2.2. A depiction of the local environments of the six crystallographically unique Eu2+ cations in Eu6Sb6S17. 50% probability ellipsoids are depicted. ...................................................................................................47 Figure 2.3. An illustration of the three-dimensional channel structure of Eu6Sb6S17 viewed down the a axis. Eu?S bond have been omitted for clarity ..........................................................................................48 Figure 2.4. Plots of dc magnetic susceptibility and its inverse for Eu6Sb6S17. The straight line shows a fit of the inverse susceptibility to the Curie-Weiss law.............................................................................................49 Figure 3.1. Illustrations of the nine-coordinate tricapped trigonal prismatic environments around the Eu centers in EuSbSe3 .............................................................. 61 Figure 3.2. A view of the one-dimensional rectangular columns formed from xix Pn (Pn = Sb, Bi) and Se in EuPnSe3........................................................................................... 66 Figure 3.3. A depiction of the Pn/Se columns in EuPnSe3 (Pn = Sb, Bi) that are formed from two opposing nets of square pyramidal PnSe5 units that are linked by PnSe6 octahedra .............................................67 Figure 3.4. Plots of dc magnetic susceptibility (?) and its inverse (?) for EuSbSe3 .................................................................................................................................................. 69 Figure 3.5. Plots of dc magnetic susceptibility (?) and its inverse (?) for EuBiSe3. ........70 Figure 3.6. Experimental and simulated 151Eu M?ssbauer spectra of EuSbSe3 at various temperatures...............................................................75 Figure 3.7. Experimental and simulated 151Eu M?ssbauer spectra of EuBiSe3 at various temperatures....................................................................76 Figure 3.8. Experimental and simulated 121Sb M?ssbauer spectrum of EuSbSe3 at 77 K.............................................................................................77 Figure 4.1. Views of the structures of a) ?-LnLn'S3, b) ?-LnLn'S3, and c) ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)........................................92 Figure 4.2. A depiction of an individual 2? [Ln'3S9]9? (Ln' = Er, Tm, Yb) layer viewed down the a axis in ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) .....................93 Figure 4.3. Illustrations of the coordination environments for the larger lanthanides, La3+ and Ce3+, in ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) ......................94 Figure 4.4. Temperature dependence of the reciprocal molar magnetic susceptibility for ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) under an applied magnetic field of 0.1 T .....................................................100 xx Figure 4.5. a) A view of the layers in Sm-based triangles in ?-SmYbSe3. b) A depiction of the interconnection of the Sm layers by Yb3+ ions.........101 Figure 4.6. a) An illustration of the square and triangular networks in ?-CeYbS3. b) A drawing of the complex three-dimension network in ?-CeYbS3.......... 102 Figure 4.7. UV-vis diffuse reflectance spectra of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)..................................................................107 Figure 5.1. A view down the b axis shows the complex three-dimensional structure of ?-Ce1.30Lu0.70S3........................................................................................................... 125 Figure 5.2. A plot of inverse molar cerium magnetic susceptibility for ?-Ce1.30Lu0.70S3 between 2 and 300 K..........................................................129 Figure 5.3. Temperature dependence of the reciprocal molar praseodymium magnetic susceptibility for ?-Pr1.29Lu0.71S3 under an applied magnetic field of 0.1 T between 2 and 300 K.............................................130 Figure 5.4. Inverse molar neodymium magnetic susceptibility vs. T for ?-Nd1.33Lu0.67S3 under an applied magnetic field of 0.1 T between 2 and 300 K ..........................................................................131 Figure 5.5. UV-vis diffuse reflectance spectra of ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71)...............................................................135 Figure 6.1. An illustration of the three-dimensional structure of La3LuSe6 down the c axis.............................................................................................151 Figure 6.2. Unit cell of ?-PrLuSe3 viewed along the a axis ...........................................154 Figure 6.3. A depiction of an individual LuSe6 octahedra layer viewed down the b axis in ?-PrLuSe3 ......................................................................155 xxi Figure 6.4. A view of the three-dimensional channel structure of Sm1.82Lu2.18Se6 along the b axis...................................................................156 Figure 6.5. Illustrations of the coordination environments for Pr ions in ?-PrLuSe3 and Sm(1)/Lu(1) ions in Sm1.82Lu2.18Se................................157 Figure 6.6. The temperature dependence of the reciprocal molar magnetic susceptibility for ?-PrLuSe3 and ?-NdLuSe3, under an applied magnetic field of 0.1 T between 2 and 300 K..............................................161 Figure 6.7. Molar magnetic susceptibility vs temperature between 2 and 300 K for Sm1.82Lu2.18Se6................................................................................................................................ 163 Figure 6.8. Inverse molar magnetic susceptibility vs temperature for Gd1.87Lu2.13Se6 under an applied magnetic field of 0.1 T between 2 and 300 K...................................................................................165 Figure 6.9. The magnetization for Gd1.87Lu2.13Se6 as a function of applied field at 2 K......................................................................................166 Figure 6.10. Inverse molar magnetic susceptibility as a function of temperature for Ce3LuSe6 under an applied magnetic field of 0.1 T between 2 and 300 K..................................................................................167 Figure 6.11. Molar magnetic susceptibility as a function of temperature for Ce3LuSe6 under ZFC and FC conditions with an applied magnetic field of 0.01 T between 2 and 25 K...........................................168 Figure 6.12. The magnetization for Ce3LuSe6 as a function of applied field at 2 K....................................................................................169 Figure 6.13. UV-vis diffuse reflectance spectra of LnxLuySez xxii (Ln = La, Ce, Pr, Nd, Sm, Gd)...................................................................172 Figure 7.1. An illustration of the three-dimensional structure of PrTbYb2S6 along the b axis.........................................................................186 Figure 7.2. Bicapped trigonal prismatic coordination environment of the Pr ions in PrTbYb2S6......................................................................................................................... 188 Figure 7.3. Inverse molar magnetic susceptibility plotted against temperature between 2 and 300 K for Pr1.34Yb2.66S6................................................................................ 190 Figure 7.4. The plot of the inverse molar magnetic susceptibility vs T for PrTbYb2S6 under an applied magnetic field of 0.1 T between 2 and 300 K .................191 Figure 7.5. The temperature dependence of the reciprocal molar magnetic susceptibility for PrDyYb2S6 under an applied magnetic field of 0.1 T between 2 and 300 K.....................................................................192 Figure 7.6. UV-vis diffuse reflectance spectra of PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) ...................................................................................196 Figure 8.1. A view the three-dimensional structure of La2YbCuS5 along the b axis. La-S bonds have been omitted for clarity..................................208 Figure 8.2. Illustrations of the coordination environments for La ions in La2YbCuS5..............................................................................................209 Figure 8.3. Illustrations of the coordination environments for Cu ions in La2YbCuS5........................................................................................................................................... 210 Figure 8.4. Depictions of various connectivities of [CuQ5] (Q = S, Se, Te) trigonal bipyramids in different compounds: a) La2YbCuS5; b) Gd3Cu2Te7; c) Sm3CuSe6; d) LaCu0.28Te2 .................................................................... 215 xxiii Figure 8.5. Unit cell of La3CuO2S3 viewed along the b axis ..........................................216 Figure 8.6. Molar magnetic susceptibility vs temperature between 2 and 300 K for La2YbCuS5, Ce2YbCuS5, and Pr2YbCuS5 .................................................................. 218 Figure 8.7. The temperature dependence of the reciprocal molar magnetic susceptibility for Nd2YbCuS5 under an applied magnetic field of 0.1 T between 2 and 300 K .............................................................219 Figure 8.8. Molar magnetic susceptibility as a function of temperature for La2YbCuSe5 and Ce2YbCuS5 under an applied magnetic field of 0.1 T between 2 and 300 K..............................................220 Figure 8.9. UV-vis diffuse reflectance spectra of Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) ...........................................................223 Figure 9.1. A view down the c axis shows the three-dimensional channel structure of Cs0.14La0.30YbS2....................................................................................... 240 Figure 9.2. a) A polyhedral representation of EuYb2S4 structure projected along the b axis; b) A polyhedral view of Cs0.14La0.30YbS2 structure projected along the c axis. S(3) atoms have been removed for clarity ......................242 Figure 9.3. Depictions of the CsS9 and LaS9 tricapped trigonal prismatic geometries in Cs0.14La0.30YbS2 and distorted EuS9 tricapped trigonal prismatic enviroment in Cs0.16Eu0.33YbS2 ......................................................... 243 Figure 9.4. Unit cell volumes (?3) and Yb~Yb distances (?) vs the number of f electrons for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm)..............................247 Figure 9.5. a) An illustration of the Yb3+ cations network along the b axis xxiv with Yb-Yb distances labeled for EuYb2S4. b) An drawing of the Yb3+ cations network along the c axis with Yb-Yb distances labeled for Cs0.14La0.30YbS2 ......................................................249 Figure 9.6. a) Plots of dc inverse molar magnetic susceptibility for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Gd, Tb) under an applied field of 0.5 T.....................................................................251 Figure 9.6. b) Plots of dc inverse molar magnetic susceptibility for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = Dy, Ho, Er, Tm, Yb) under an applied field of 0.5 T................................................................................252 Figure 9.7. Ce, Nd, and Sm LIII edge x-ray absorption spectra of (a) Cs0.14Ce0.30YbS2, (b) Cs0.14Nd0.29YbS2, and (c) Cs0.15Sm0.29YbS2, respectively......................255 Figure 9.8. (a) Yb LIII edge spectra of the same samples in Fig. 7, and (b) the derivative of these spectra ................................................................256 Figure 9.9. UV-vis diffuse reflectance spectra of Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Gd, Tb, Dy, Ho, Er, Tm) ...................................258 xxv LIST OF TABLES Table 1.1. The ground-state electron configurations and radius of the lanthanides ........................................................................................................2 Table 1.2. Melting points (?C) of common fluxes used in solid state reactions.................6 Table 1.3. Crystal structures of Ln2S3. ?: orthorhombic Gd2S3-type; ?: tetragonal Pr10S14O-type; ?: cubic Th3P4-type; ?: monoclinic Er2S3-type; ?: rhombohedral Al2O3-type; ?: cubic Tl2O3-type ................................................9 Table 1.4. Calculated and measured effective magnetic moments (?) for the trivalent lanthanide ions............................................................................17 Table 2.1. Crystallographic data for Eu6Sb6S17.............................................................................................. 38 Table 2.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for Eu6Sb6S17..................................................................................................................... 39 Table 2.3. Eu?S, Sb?S, and S?S Bond Distances (?) for Eu6Sb6S17.............................................. 45 Table 3.1. Crystallographic Data for EuSbSe3 and EuBiSe3 ................................................................ 57 Table 3.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for EuSbSe3. .................................................................................58 Table 3.3. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for EuBiSe3...................................................................................59 Table 3.4. Selected Bond Distances (?) for EuSbSe3 ............................................................................... 62 xxvi Table 3.5. Selected Bond Distances (?) for EuBiSe3.......................................................64 Table 3.6. Fitting parameters of 151Eu M?ssbauer measurements of EuBiSe3 and EuSbSe3. ....................................................................................74 Table 4.1. Crystallographic Data for ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)...........86 Table 4.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?-LaYbS3...................................................................................................................... 87 Table 4.3. Selected Bond Distances (?) for ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb).....................................................................96 Table 4.4. Comparisons of Ternary Interlanthanide Sesquichalcogenides that Form in a Sb2Q3 (Q = S, Se) Flux at 1000 ?C...............................................108 Table 5.1. Crystallographic Data for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71)................................................................118 Table 5.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?-Ce1.30Lu0.70S3..................................................................................................... 119 Table 5.3. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?-Pr1.29Lu0.71S3......................................................................120 Table 5.4. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?-Nd1.33Lu0.67S3.....................................................................121 Table 5.5. Selected Bond Distances (?) for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71)................................................................126 Table 5.6. Magnetic Parameters for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71)................................................................132 Table 6.1. Crystallographic Data for LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd)...........143 xxvii Table 6.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for La3LuSe6...............................................................................144 Table 6.3. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?-PrLuSe3 ................................................................................................................. 145 Table 6.4. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for Sm1.82Lu2.18Se6......................................................................146 Table 6.5. Ternary Interlanthanide Sulfides Prepared Using Sb2S3 Flux at 1000 ?C. ....................................................................................................149 Table 6.6. Ternary Interlanthanide Selenides Prepared Using Sb2Se3 Flux at 1000 ?C. ....................................................................................................150 Table 6.7. Selected Bond Distances (?) for Ln3LuSe6 (Ln = La, Ce)............................153 Table 6.8. Selected Bond Distances (?) for ?-LnLuSe3 (Ln = Pr, Nd) ..........................158 Table 6.9. Selected Bond Distances (?) for LnxLu4-xSe6 (Ln = Sm, Gd; x = 1.82, 1.87).......................................................................160 Table 6.10. Magnetic Parameters for Ce3LuSe6, ?-PrLuSe3, ?-NdLuSe3, and Gd1.87Lu2.13Se6......................................................................................162 Table 7.1. Crystallographic Data for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy).........................181 Table 7.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for Pr1.34Yb2.66S6.........................................................................182 Table 7.3. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for PrTbYb2S6. ...........................................................................183 Table 7.4. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for PrDyYb2S6............................................................................184 xxviii Table 7.5. Selected Bond Distances (?) for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) .............189 Table 7.6. Magnetic Parameters for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy)..........................193 Table 8.1. Crystallographic Data for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) ...........................................................205 Table 8.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for La2YbCuS5. ..........................................................................206 Table 8.3. Selected Bond Distances (?) and Angles (deg) for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) ...........................................................211 Table 8.4. Magnetic Parameters for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) ...........................................................221 Table 9.1. a. Crystallographic Data for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Eu, Gd) ...............................................................236 b. Crystallographic Data for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = Tb, Dy, Ho, Er, Tm, Yb).....................................................................237 Table 9.2. Positional and Thermal parameters for Cs0.14La0.30YbS2 and Cs0.16Eu0.33YbS2. ..........................................................238 Table 9.3. a. Selected Bond Distances (?) for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Eu, Gd) ...............................................................244 b. Selected Bond Distances (?) for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = Tb, Dy, Ho, Er, Tm, Yb).....................................................................245 Table 9.4. Magnetic Parameters for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Gd, Tb, Dy, Ho, Er, Tm, Yb) .............................253 Table 10.1 A list of new compounds and some of their properties, contained in this xxix dissertation. .................................................................................................267 1 CHAPTER 1 INTRODUCTION Lanthanide chalcogenides have been the focus of intense interest not only because of their remarkably complex structures,1 but also because of the variety of physical properties that they exhibit including tunable band gaps,2-13 atypical magnetism,14-20 superconductivity,21-25 charge-density waves,26-29 and thermoelectricity.30,31 The most striking fact about lanthanide chemistry is the extraordinary homogeneity in properties of all the elements across the series. This arises became the 4f shell of electrons is being filled along the lanthanide series, while the number of outer valence electrons remains unchanged. The 4f atomic orbitals are deeply buried and thus play a small role in chemical bonding. The ionization beyond the Ln3+ ion is usually not energetically favorable leading to the characteristic trivalent state across the whole series. As shown in Table 1.1, the electronic configurations for trivalent lanthanides are 4fn, where n goes from 0 to 14, regardless of which configuration, 4f n5d6s2 or 4f n+16s2, is the ground state of the neutral atom. The trivalent lanthanides show a general decrease of ionic radius along the series of elements La-Lu. This is usually referred to as lanthanide contraction and is due to the imperfect shielding of one electron by another in the same sub-shell. Three configurations for trivalent lanthanides are especially stable. The first and most stable is that of La3+ with its unfilled 4f orbitals and a 1S ground state. The second 2 Table 1.1. The ground-state electron configurations and radiis of the lanthanides. Element Electronic configuration Radius/? Ln Ln2+ Ln3+ Ln Ln3+ La lanthanum [Xe]6s25d1 [Xe] 4f0 1.88 1.16 Ce cerium [Xe]4f16s25d1 [Xe]4f1 1.83 1.14 Pr praseodymium [Xe]4f36s2 [Xe]4f2 1.82 1.13 Nd neodymium [Xe]4f46s2 [Xe]4f3 1.81 1.11 Pm promethium [Xe]4f56s2 [Xe]4f4 1.81 1.09 Sm samarium [Xe]4f66s2 [Xe]4f6 [Xe]4f5 1.80 1.08 Eu europium [Xe]4f76s2 [Xe]4f7 [Xe]4f6 1.99 1.07 Gd gadolinium [Xe]4f76s25d1 [Xe]4f7 1.80 1.05 Tb terbium [Xe]4f96s2 [Xe]4f8 1.78 1.04 Dy dysprosium [Xe]4f106s2 [Xe]4f9 1.77 1.03 Ho holmium [Xe]4f116s2 [Xe]4f10 1.76 1.02 Er erbium [Xe]4f126s2 [Xe]4f11 1.75 1.00 Tm thulium [Xe]4f136s2 [Xe]4f13 [Xe]4f12 1.74 0.99 Yb ytterbium [Xe]4f146s2 [Xe]4f14 [Xe]4f13 1.94 0.99 Lu lutetium [Xe]4f146s25d1 [Xe]4f14 1.73 0.98 3 stable electronic arrangement is that of Gd3+ with the 4f shell half filled (4f7). According to Hund?s rule this arrangement has the maximum multiplicity possible (2S + 1 = 8) and must consequently be an S state (L = 0) through the Pauli Exclusion Principle. The third is that of Lu3+ with a completed 4f shell and as a consequence, also a 1S state. It has been known for some time that an oxidation state other than trivalent is possible for many rare earths. Because of the stability of full and half-filled shells, Yb and Eu might appear in the divalent form while Ce and Tb may be quadrivalent as well. Samarium would be expected to have a less stable divalent state than Eu. In the chalcogenides, no tetravalent lanthanide has been observed owing to the low eletronegativity of the anions. In contrast, the divalent state can be stabilized for the same reason. For example, in case of europium, for which the regular oxide is Eu2(III)O3, the corresponding Eu2(III) S3 does not exist. Instead, Eu(II)S can be easily obtained.32,33 However, mixed-valent europium ions were found in some chalcogenides, especially in ternary phases including Eu2BiSe4,34 Eu2CuS3,35,36 EuPd3S4,37 Eu5Sn3S12,38 and Eu4Sn2S9.39 With ytterbium, a similar behavior is observed, but with a greater stability of the trivalent state. The ubiquitous sulfide is Yb2(III)S3,40 but it is easily dissociated by heat into mixed-valent Yb3S4.41 Ytterbium selenide is also stable in the form of Yb3Se4,42 while only one binary telluride is known, Yb(II)Te.43 SOLID STATE SYNTHESIS Lanthanide chalcogenides have been synthesized through a variety of solid state synthetic methods. Because the starting materials are usually solids, classical solid-state 4 synthesis requires high-temperature (> 600 ?C) to cause sufficient diffusion for a reaction to occur. As a result of the high reaction temperatures, thermodynamically stable products are generally formed. These products are usually the simplest binary or ternary compounds with high lattice stability. They become synthetic road blocks that are hard to avoid. Furthermore, the high temperature reactions only involve building blocks on the atomic level. Molecules, as starting materials, will be reduced back to atoms at high temperature. Therefore, solid-state synthesis has much less predictability than other areas of chemical synthesis. There have been many attempts to develop methodology that allow for greater diffusion of reactants and lower reaction temperatures to circumvent thermodynamic traps. Solid state reactions performed at low temperatures are able to produce new kinetically stable compounds and can also employ molecular assemblies as building blocks to incorporate into solid state structures. Hydro(solvo)thermal synthesis44 and chemical vapor deposition (CVD)45 are two major synthetic methods that are performed at low temperatures. The hydro(solvo)thermal synthesis employs various solvents under high pressure that enhance the solubility of the reactants in the reaction matrix. The reaction temperatures are usually between the boiling temperature and the critical point of solvents. Polyatomic species can be used as building units for reaction products. The solubility and the diffusion of reactants can be further improved by adding mineralizers or using supercritical solvents.46 CVD is commonly used to synthesize known solid-state compounds as thin film. The chemical process involves the intimate gas-phase mixing of 5 volatile precursors leading, upon pyrolysis, to deposition of solid state intermetallic materials on various substrates. Therefore, CVD is limited to using elemental reactants. Molten salts can also be used as solvents to increase the diffusion of reactants and lower the reaction temperature.47 Compared to CVD and hydro(solvo)thermal techniques, reactions using molten salts are easier to conduct. Most salts are soluble in many common polar solvents. This allows for easy isolation of products since excess flux can simply be washed away. One group of low-melting salts that have been rapidly developed over the last twenty years is the alkali metal polychalcogenides, A2QX (A = alkali metal; Q = chalcogenide).48,49 They are usually referred to as reactive fluxes owing to their high reactivity with reactants. The low melting points of A2QX fluxes have led to the isolation of many new kinetically stable multinary polychalcogenides. Another class of well-known fluxes that have been employed for over 100 years for high-temperature single-crystal growth is alkali metal halides.47 Although many halides are high-melting species, as shown in Table 1.2, eutectic combinations of different salts often have melting points well below the temperatures of classical solid-state synthesis, making possible their use in the exploration of new chemistry at intermediate temperatures. In some cases, alkali metal halides act not only as solvents, but also as reactants, providing species that can be incorporated into the final product. A method of using Sb2Q3 (Q = S, Se) fluxes has been developed in our group to prepare interlanthanide chalcogenides.50 The desired products can be separated from the flux during the cooling process by slightly tilting the furnace, which allows the flux to flow to the bottom of the tubes leaving crystals behind, minimizing the need for manual separation of solids. Molten Sb2Q3 (Q = S, Se) fluxes turn out to be valuable media to 6 Table 1.2. Melting points (?C) of common fluxes used in solid state reactions. Flux Melting points Flux Melting points Flux Melting points Na2S2 490 LiCl 605 RbI 647 Na2S3 228.8 NaCl 801 CsCl 646 Na2Se2 495 NaBr 747 CsBr 636 K2S2 470 KCl 773 CsI 626 K2S4 145 KBr 743 SrCl2 873 K2Se2 460 KI 681 BaCl2 963 Rb2S 530 RbCl 718 Sb2S3 550 Rb2S4 160 RbBr 693 Sb2Se3 611 7 access ternary and quaternary interlanthanide chalcogenides. Eight different structure types and more than fifty compounds have been identified for ternary interlanthanide chalcogenides prepared by employing these fluxes. The limitations of using Sb2Q3 (Q = S, Se) fluxes to prepare LnLn'Q include: 1) Attempts to make interlanthanide tellurides have not succeeded; 2) It is difficult to get high yield and high-quality single crystals when the ionic radii of the two Ln3+ ions approach equality; 3) Occasionally, distinguishing and separating products from Sb2Q3 (Q = S, Se) fluxes proves tricky. STRUCTURES Lanthanide chalcogenides exhibit a large number of crystal structure types that are related to the variety of coordination environments of lanthanide ions and diverse connectivities between these coordination polyhedra.1 The structural complexity of lanthanide chalcogenides can be exemplified by simple binary sesquisulfides, which are found in at least six different modifications, as shown in Table 1.3. These include orthorhombic Gd2S3-type (?),51 tetragonal Pr10S14O-type (?),52-55 cubic Th3P4-type (?),56 monoclinic Er2S3-type (?),40,57-59 rhombohedral Al2O3-type (?),40,60,61 and cubic Tl2O3- type (?).62-64 Figure 1.1 presents the common coordination environments for lanthanides found in lanthanide chalcogenides with coordination number ranging from 6 to 9. Coordination polyhedra for smaller lanthanide ions, e.g. monocapped trigonal prismatic or octahedral Yb3+, have lower coordination number and are usually edge-sharing or corner-sharing with neighbors. In contrast, larger ions prefer higher coordination 8 numbers, and their coordination polyhedra, such as bicapped or tricapped trigonal prisms, can even share faces with each other (Figure 1.2). Most structures for lanthanide chalcogenides, except cubic phases, are constructed from one-dimensional chains of lanthanide coordination polyhedra along the crystal growth axis. As shown in Figure 1.3, small polyhedra are usually edge-sharing and large polyhedra are face-sharing down the chain-propagation direction. The polyhedra can share corners or edges with each other to form two-dimensional or three-dimensional structures (Figure 1.4). Figures 1.5 and 1.6 illustrate the two-dimensional layer structure of ?-LaYbS365,66 and three-dimensional channel structure of EuYb2S4.67,68 The structure of ?-LaYbS3 includes YbS6 octahedra layers that alternate with layers of LaS8 bicapped trigonal prisms along the b axis. The YbS6 octahedra are edge-sharing down the a axis to form one-dimensional chains that are corner-sharing along the c axis. The La-based layers consist of face-sharing one-dimensional chains of LaS8 bicapped trigonal prisms that share edges with identical neighbors. The structure of EuYb2S4 is constructed from two different edge-shared double rutile chains of YbS6 octahedra along the b axis. Each double chain is joined at the vertices to four other chains to form open channels of capped trigonal prismatic sites wherein Eu2+ ions reside. The introduction of other metal ions into the lanthanide chalcogenide system may result in even more complex structures. D-block and main-group metals normally have smaller radii than lanthanides ions, and therefore lower coordination numbers. Typical coordination environments include trigonal planar, tetrahedral, and octahedral moieties. In contrast, alkali metals and alkaline-earth metals are more loosely bound to chalcogenides and have higher coordination numbers. Because of their distinct oxidation 9 Table 1.3. Crystal structures of Ln2S3. ?: orthorhombic Gd2S3-type; ?: tetragonal Pr10S14O-type; ?: cubic Th3P4-type; ?: monoclinic Er2S3-type; ?: rhombohedral Al2O3- type; ?: cubic Tl2O3-type. La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu ? ? ? ? ? ? 10 6-coordinate Octahedron Trigonal prism 7-coordinate Monocapped trigonal prism 8-coordinate Bicapped trigonal prism Square antiprism 9-coordinate Tricapped trigonal prism Figure 1.1. Typical coordination geometries for lanthanides (green) bound to chalcogenides (yellow). 11 Corner-sharing Edge-sharing Face-sharing Figure 1.2. Common connectivities between coordination polyhedra of lanthanide (green) chalcogenides (yellow). 12 Edge-sharing octahedral chain Face-sharing bicapped trigonal prismatic chain. Figure 1.3. General building units for structures of lanthanides chalcogenide: one- dimensional chains of lanthanide coordination polyhedra. 13 Corner-sharing between one-dimensional octahedral chains Face-sharing between one-dimensional bicapped trigonal prismatic chains Figure 1.4. Connectivities between one-dimensional chains of lanthanide coordination polyhedra. 14 Figure 1.5. A view of the two-dimensional structure of ?-LaYbS3 along the a axis. La-S bonds have been omitted for clarity. b c 15 Figure 1.6. An illustration of the three-dimensional channel structure of EuYb2S4 down the b axis. Eu-S bonds have been omitted for clarity. a c 16 states (I or II), these metals usually occupy different crystallographic sites from lanthanides, even though their coordination geometries are quite similar. MAGNETIC PROPERTIES One of the outstanding properties of the lanthanides is the strong paramagnetism displayed by the metals and their compounds. This trait can be attributed to the presence of unpaired 4f electrons. Not only have magnetic methods been valuable for identification and the determination of purity, but also they have also cast considerable light on the configurations of the ions and possible magnetic interactions between ions in the solid state. Most lanthanides have unpaired 4f electrons, as shown in Table 1.4, except diamagnetic La3+ and Lu3+ ions. The magnetic moment of each unpaired electron includes spin angular momentum and orbital angular momentum. For transition metals, the partially filled d-shells are the outermost electronic shells, and are therefore strongly influenced by their environment. The orbital angular momentum in the case of many first row d-block metal ions is actually quenched by the crystal field. This is a particular example of a general phenomenon known as crystal-field splitting. The magnetic moment, ?, can be expressed as the so-called spin-only formula (equation 1.1).69 ? (spin-only) = 2 [S(S+1)]1/2 (1.1) In contrast, crystal field splitting is unimportant for lanthanides, because the partially filled 4f shells lie deep inside the ion (beneath filled 5s and 5p shells). The effective 17 Table 1.4. Calculated and measured effective magnetic moments (?) for the trivalent lanthanide ions Element Electronic configuration Ground-state term ?cal/?B ?mea/?B La lanthanum [Xe] 4f0 1S 0.00 diamagnetic Ce cerium [Xe]4f1 2F5/4 2.54 2.4 Pr praseodymium [Xe]4f2 3H4 3.58 3.5 Nd neodymium [Xe]4f3 4I9/2 3.62 3.5 Pm promethium [Xe]4f4 5I4 2.68 - Sm samarium [Xe]4f5 6H5/2 0.84 1.5 Eu europium [Xe]4f6 7F0 0.00 3.4 Gd gadolinium [Xe]4f7 8S7/2 7.94 8.0 Tb terbium [Xe]4f8 7F6 9.72 9.5 Dy dysprosium [Xe]4f9 6H15/2 10.63 10.6 Ho holmium [Xe]4f10 5I8 10.60 10.4 Er erbium [Xe]4f11 4I15/2 9.59 9.5 Tm thulium [Xe]4f12 3H6 7.57 7.3 Yb ytterbium [Xe]4f13 2F7/2 4.54 4.5 Lu lutetium [Xe]4f14 1S 0.00 diamagnetic 18 magnetic moment, ?eff, for lanthanides, given by equation 1.2, takes into account both spin angular momentum and orbital angular momentum. ?eff = gJ [J(J+1)]1/2 (1.2) where gJ is the Zeeman factor, gJ = 3/2 + [(S(S+1) ? L(L+1))/2J(J+1)]. ?eff can be obtained from the experimentally measured molar magnetic susceptibility, ?m, and is expressed in Bohr magnetons (?B) according to equation 1.3. ?eff = [ 3kB?mT/(L?0?B2)]1/2 (1.3) where kB is Boltzmann constant; L is Avogadro?s number; ?0 is vacuum permeability; T is temperature in Kelvin. Table 1.4 presents the calculated effective magnetic moments, based on equation 1.2, that generally agree well with experimental values. For most of the trivalent lanthanides, the energy difference between 2S+lLJ free-ion ground state and the first excited state is much larger than the thermal energy (kBT), therefore only the ground state is thermally populated at room temperature and below. In the free-ion approximation, the molar magnetic susceptibility is then given by equation 1.4.70 ?m = L(gJ?B)2J(J+1)/3kBT + 2L?B2(gJ -1)(gJ -2)/3? (1.4) where ? is the spin-orbit coupling constant. The second term in the above equation is a temperature independent component due to the coupling between ground and excited states through the Zeeman perturbation. Equation 1.4 can be rewritten as modified Curie- Weiss law (equation 1.5), where C is Curie constant, C = L(gJ?B)2J(J+1)/3kB; ? is Weiss constant; ?0 = 2L?B2(gJ -1)(gJ -2)/3?. ?m = ?0 + C/(T-?) (1.5) 19 The magnetic susceptibility behavior for lanthanides does not always follow the Curie-Weiss law, especially at low temperatures. Because of the crystal-field effect that splits the 2S+1LJ ground state into several Stark sublevels, at high temperatures all Stark sublevels are thermally populated, while as the temperature decreases, a depopulation of these sublevels occurs causing the deviation from Curie-Weiss law. For Eu3+ and Sm3+ ions, their first excited states are thermally populated, because the energy difference between the first excited state and ground state is comparable to the thermal energy. Considering the contribution from excited states, the magnetic susceptibility for Eu3+ and Sm3+ would not follow equation 1.5. Gd1.87Lu2.13Se6, ?-NdLuSe3, and Sm1.82Lu2.18Se6 are given as examples to illustrate the typical magnetic susceptibility behaviors of lanthanide chalcogenides.71 Since the Lu3+ ions are diamagnetic, Gd3+, Nd3+, and Sm3+ are the only magnetic ions in these compounds. Gd3+ ion is rather unique in the lanthanides series with a ground state of 8S7/2 (L = 0). It does not have orbital angular momentum, so the ground state is not perturbed by crystal field effect. Furthermore, its first excited state 6P7/2 is fully depopulated even at room temperature. The temperature dependence of magnetic susceptibility for Gd3+ follows the Curie-Weiss law, as shown in Figure 1.7. Gd1.87Lu2.13Se6 undergoes an antiferromagnetic transition around 4 K. It is worth noting that some Gd-based compounds have some small crystal field splitting at very low temperature because of the spin-orbital coupling between 8S7/2 and 6P7/2. The ground state of Nd3+ is 4I9/2 that would be influenced by the crystal field. Figure 1.8 shows the inverse molar magnetic susceptibility as a function of temperature 20 Figure 1.7. Inverse molar magnetic susceptibility vs temperature for Gd1.87Lu2.13Se6 under an applied magnetic field of 0.1 T between 2 and 300 K. The solid line represents the fit to Curie-Weiss law in the range of 100-300 K. Inset shows the molar magnetic susceptibility at low temperature. 0 50 100 150 200 250 300 0 4 8 12 16 2 4 6 8 10 1.5 2.0 2.5 ???? Temperature Gd1.87Lu2.13Se6 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) 21 0 50 100 150 200 250 300 0 30 60 90 120 150 180 ?-NdLuSe3 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) Figure 1.8. The temperature dependence of the reciprocal molar magnetic susceptibility for ?-NdLuSe3 under an applied magnetic field of 0.1 T between 2 and 300 K. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. 22 Figure 1.9. Inverse molar magnetic susceptibility as a function of temperature for Sm1.82Lu2.18Se6 under an applied magnetic field of 0.1 T between 2 and 300 K. 0 50 100 150 200 250 300 0 200 400 600 800 1000 1200 Sm1.82Lu2.18Se6 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) 23 for ?-NdLuSe3. It deviates from the Curie-Weiss law below 40 K owing to depopulation of Stark sublevels of higher energy. For Sm3+ ions, the difference between the ground state (6H5/2) and the first excited state (6H7/2) is not large compared to thermal energy (kBT). Therefore, the excited states make significant contributions to the magnetic susceptibility at high temperature. The magnetic susceptibility of Sm3+ is given by equation 1.6. ?m = [C2.14 + 3.67/? + (42.9 + 0.82/ ?) e-7? + (142 - 0.33/ ?) e-16? + ?????] /[T(3 + 4e-7? + 5e-16?)] (1.6) where C = L?B2/(3kB); ? is the ratio between the multipet width and the thermal energy, ? = ?/( kBT).The magnetic susceptibility of Sm1.82Lu2.18Se6 (Figure 1.9) shows a typical van Vleck paramagnetic behavior that does not follow the Curie-Weiss law.72 OPTICAL PROPERTIES One of the most promising applications of the lanthanide chalcogenides is the ability to systematically change the band gap of these semiconductors. The variation of their optical gaps across the series often exhibits certain periodicity. The well-studied compounds, Ln2Q3 (Q = O, S, Se), are typical examples of this character.73,74 The band gap (Eg) for binary lanthanide chalcogenides is the gap between the valence band, usually composed of the np states of chalcogenides, and conduction band that is formed by the 5d(6s) states of lanthanides ions. Figure 1.10 shows the variation of the band gap for Ln2Q3 (Q = O, S, Se).74 There are two apparent minima for cerium and terbium. Both of them have one excess 4f electron from the most stable 4f0 and 4f7 configurations. Ce-based compounds have the smallest gaps in the whole series because 24 of their high-energy 4f1 state that overlaps the forbidden band. The absorption edge is determined by optical transition from the 4f band to the conduction band. The energies of the 4f-shells start to decrease after Ce3+ causing the increase of band gaps in the sequence of Ce2O3-Pr2O3-Nd2O3-Sm2O3. The same thing happens for Tb2O3-Dy2O3- Ho2O3. La-, Gd-, and Lu-containing compounds have the maximum gaps. The empty 4f band for La3+ is located much higher than the bottom of the conduction band, while the 4f bands for Gd3+ and Lu3+ are buried deeply in the valence band owing to the high stability of half- and fully-filled 4f shells. The band gaps for Er and Tm oxides are close to the values for La, Gd, and Lu oxides because, as the 4f band enters the valence band, the Eg become constant. The Eg minima for Eu2O3 and Yb2O3 are because of their low dissociation energy (D), which is presented in Figure 1.11.74 The lower D values for these two compounds may indicate some bivalent character of Eu and Yb ions. Because the sequence of energy levels of p orbitals is 2p (O) < 3p (S) < 4p (Se), the periodicity of the variation of band gaps becomes less obvious in sulfides and selenides. The 4f bands in lanthanide oxides that lie in the forbidden band might drop into the valence band in the cases of sulfides and selenides. The band gaps of lanthanide chalcogenides with same composition highly depend on their structures as well. For instance, ?-LaYbS3 crystals are yellow,66 while ?-LaYbS3 crystals are dark red with smaller band gaps.50 Both ?-LaYbS3 and ?-LaYbS3 adopt two- dimensional layer structures, but the structure of ?-LaYbS3 is more condensed. Including other metals will have a substantial effect on the band gaps of lanthanide chalcogenides. Alkali metal and alkaline-earth cations are more electropositive than lanthanides and have almost no contribution to the electronic states 25 near the Fermi level, although they usually lift the energy of conduction band and give rise to larger band gaps.75-77 In contrast, transition metal ions and main group metal ions are more electronegative than lanthanides and form more covalent bonds with chalcogenides. The band gaps are usually smaller and are dominated by the interactions between transition metal ions (or main group metal ions) and chalcogenides.78-80 26 Figure 1.10. Variation of the band gap Eg of Ln2Q3 (Q = O, S, Se) in the lanthanide series: 1) oxides; 2) sulfides; 3) selenides. 27 Figure 1.11. Variation of the dissociation (atomization) energy D of Ln2O3 in the lanthanide series. 28 REFERENCES 1. Gschneidner Jr., K. A.; Eyring, L. R. Eds. Handbook on the Physics and Chemistry of Rare Earths; North-Holland Physics Publishing: New York, 1979; Vol. 4, pp 1-88. 2. Isaacs, T. J.; Hopkins, R. H.; Kramer, W. E. J. Electron. Mater. 1975, 4, 1181. 3. Hautala, J.; Taylor, P. C. J. Non-Cryst. Solids 1992, 141, 24. 4. Lowe-Ma, C. K.; Vanderah, T. A.; Smith, T. E. J. Solid State Chem. 1995, 117, 363. 5. Huang, F. Q.; Mitchell, K.; Ibers, J. A. Inorg. Chem. 2001, 40, 5123. 6. Inoue, S.; Ueda, K.; Hosono, H.; Hamada, N. Phys. Rev. B 2001, 64, 245211. 7. Ueda, K.; Inoue, S.; Hosono, H.; Sarukura, N.; Hirano, H. Appl. Phys. Lett. 2001, 78, 2333. 8. Mitchell, K.; Ibers, J. A. Chem. Rev. 2002, 102, 1929-1952. 9. Mitchell, K.; Haynes, C. L.; McFarland, A. D.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2002, 41, 1199. 10. Mitchell, K.; Huang, F. Q.; McFarland, A. D.; Haynes, C. L.; Somers, R. C.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2003, 42, 4109. 11. Ueda, K.; Takafuji, K.; Hiramatsu, H.; Ohta, H.; Kamiya, T.; Hirano, M.; Hosono, H. Chem. Mater. 2003, 15, 3692. 12. Yao, J.; Deng, B.; Sherry, L. J.; McFarland, A. D.; Ellis, D. E.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2004, 43, 7735. 29 13. Mitchell, K.; Huang, F. Q.; Caspi, E. N.; McFarland, A. D.; Haynes, C. L.; Somers, R. C.; Jorgensen, J. D.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2004, 43, 1082. 14. Furuuchi, F.; Wakeshima, M.; Hinatsu, Y. J. Solid State Chem. 2004, 177, 3853. 15. Sutorik, A. C.; Albritton-Thomas, J.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Mater. 1996, 8, 751. 16. Tougait, O.; Ibers, J. A. Inorg. Chem. 2000, 39, 1790. 17. Bronger, W.; Miessen, H.-J.; Schmitz, D. J. Less-Common Met. 1983, 95, 275. 18. Wakeshima, M.; Hinatsu, Y.; Oikawa, K.; Shimojo, Y.; Morii, Y. J. Mater. Chem. 2000, 10, 2183. 19. Meng, F.; Hughbanks, T. Inorg. Chem. 2001, 40, 2482. 20. Wakeshima, M.; Hinatsu, Y.; Ishii, Y.; Shimojo, Y.; Morii, Y. J. Mater. Chem. 2002, 12, 631. 21. Bozorth, R. M.; Holtzberg, F.; Methfessel S. Phys. Rev. Lett. 1965, 14, 952. 22. Westerholt, K.; Bach, H.; Wendemuth, R.; Methfessel, S. Solid State Commun. 1979, 31, 961. 23. Westerholt, K.; Timmer, F.; Bach, H. Phys. Rev. B 1985, 32, 2985. 24. Fischer, O. Springer Ser. Solid-State Sci. 1990, 90, 96-112. 25. Kang, J. -S.; Nahm, K.; Kim, C. K.; Olson, C. G.; Pelzl, J.; Shim, J. H.; Min, B. I. Phys. Rev. B 2002, 66, 075108. 26. Foran, B.; Lee, S; Aronson, M. C. Chem. Mater. 1993, 5, 974. 27. DiMasi, E.; Foran, B.; Aronson, M. C.; Lee, S. Chem. Mater. 1994, 6, 1867. 30 28. Prodan, A.; Marinkovic, V.; Boswell, F. W.; Bennett, J. C.; Remskar, M. J. Alloys Compd. 1995, 219, 69-72. 29. Malliakas, C. D.; Kanatzidis, M. G. J. Am. Chem. Soc. 2006, 128, 12612. 30. Aliev, O. M.; Rustamov, P. G.; Guseinov, G. G.; Guseinov, M. S.; Nauk SSSR, Izvest. Akad. Neorgan. Mater. 1978, 14, 1346. 31. Tritt, T. M. Science 1999, 283, 804-805. 32. Nowacki, W. Naturwissenschaften 1938, 26, 495. 33. Gschneidner Jr., K. A.; Eyring, L. R. Eds. Handbook on the Physics and Chemistry of Rare Earths; North-Holland Physics Publishing: New York, 1979; Vol. 2, pp 507-575. 34. Lemoine, P.; Guittard, M. Acta Cryst. 1982, B38, 727. 35. Lemoine, P.; Carr?, D.; Guittard, M. Acta Cryst. 1986, C42, 390. 36. Furuuchi, F.; Wakeshima, M.; Hinatsu, Y. J. Solid State Chem. 2004, 177, 3853. 37. Wakeshima, M.; Fujino, T.; Sato, N.; Yamada, K.; Masuda, H. J. Solid State Chem. 1997, 129, 1. 38. Jaulmes, S.; Julien-Pouzol, M. Acta Cryst. 1977, C33, 1191. 39. Flahaut, J.; Laruelle, P.; Guittard, M.; Jaulmes, S.; Julien-Pouzol, M.; Lavenant, C. J. Solid State Chem. 1979, 29, 125. 40. Flahaut, J.; Domange, L.; Pardo, M. P. Inorg. Chem. 1968, 7, 2282. 41. Chevalier, R.; Laruelle, P.; Flahaut, J. Bulletin de la Societe Francaise de Mineralogie et de Cristallographie 1967, 90, 564. 42. Pouzol, J.; Jaulmes, M.; Wintenberger, M.; Guittard, M. Mater. Res. Bull. 1987, 22, 95. 31 43. Noda, Y.; Matsumoto, K.; Ohba, S.; Saito, Y. Acta Crystallogr. 1987, 43, 1433. 44. Rabenau, A. Angew. Chem. Int. Ed. Engl. 1985, 24, 1026. 45. Steigerwald, M. L. Inorganomet. Chem. 1992, 333. 46. Wood, P. T.; Pennington,W. T.; Kolis, J. W. J. Am. Chem. Soc. 1992, 114, 9233. 47. Elwell, D.; Scheel, H. J. Crystal Growth from High-Temperature Solutions; Academic, New York, 1975. 48. Sunshine, S. A.; Kang, D.; Ibers, J. A. J. Am. Chem. Soc. 1987, 109, 6202. 49. Kanatzidis, M. G.; Sutorik, A. C. Prog. Inorg. Chem. 1995, 43, 151. 50. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. Chem. Mater. 2007, 19, 567. 51. Prewitt, C. T.; Sleight, A. W. Inorg. Chem. 1968, 7, 1090. 52. Besancon, P.; Laruelle, P. C. R. Acad. Sci. Paris, Ser. C 1969, 268, 48. 53. Besancon, P. J. Solid State Chem. 1973, 7, 232. 54. Carre, D.; Laruelle, P.; Besancon, P. C. R. Acad. Sci. Paris, Ser. C 1970, 270, 537. 55. Schleid, T.; Lissner, F. J. Less-Common Met. 1991, 175, 309. 56. Zachariasen, W. H. Acta Crystallogr. 1949, 2, 57. 57. White, J. G.; Yocom, P. N.; Lerner, S. Inorg. Chem. 1967, 6, 1872. 58. Fang, C. M.; Meetsma, A.;Wiegers, G. A.; Boom, G. J. Alloys Compd. 1993, 201, 255. 59. Landa-Canovas, A. R.; Amador, U.; Otero-Diaz, L. C. J. of Alloys Compd. 2001, 323, 91. 32 60. Flahaut, J.; Domange, L.; Pardo, M. P. Comptes Rendus Hebdomadaires des Seances de l'Academie des Sciences 1964, 258, 594. 61. El Fadli, Z.; Lemoine, P.; Guittard, M.; Tomas, A. Acta Crystallogr. 1994, C50, 166. 62. Kuzmicheva, G. M.; Efremov, V. A.; Khlyustova, S. Yu.; Eliseev, A. A. Zhurnal Neorganicheskoi Khimii 1986, 31, 2210. 63. Kuz'micheva, G. M.; Smarina, E. I.; Khlyustova, S. Yu.; Chernyshev, V. V. Zhurnal Neorganicheskoi Khimii 1990, 35, 869. 64. Schleid, T.; Lissner, F. J. Alloys Compd. 1992, 189, 69. 65. Rodier, N.; Julien, R.; Tien, V. Acta Crystallogr. C 1983, 39, 670. 66. Mitchell, K.; Somers, R. C.; Huang, F. Q.; Ibers, J. A. J. Solid State Chem. 2004, 177, 709. 67. Hulliger, F.; Vogt, O. Phys. Lett. 1966, 21, 138. 68. Lugscheider, W.; Pink, H.; Weber, K.; Zinn, W. Zeitschrift fuer Angewandte Physik 1970, 30, 36. 69. Kittel, C. Introduction to Solid State Physics; 6th Ed., Wiley, New York, 1986. 70. Andruh, M.; Bakalbassis, E.; Kahn, O.; Trombe, J. C.; Porcher, P. Inorg. Chem. 1993, 32, 1616. 71. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. Inorg. Chem. submitted. 72. Van Vleck, J. H. The Theory of Electric and Magnetic Susceptibilities; Oxford University Press: London, 1932. 33 73. Prokofiev, A. V.; Shelykh, A. I.; Golubkov, A. V.; Smirnov, I. A. J. Alloys Compd. 1995, 219, 172. 74. Prokofiev, A. V.; Shelykh, A. I.; Melekh, B. T. J. Alloys Compd. 1996, 242, 41. 75. Bronger, W.; Eyck, J.; Kruse, K.; Schmitz, D. Eur. J. Solid State Inorg. Chem. 1996, 33, 213. 76. Sutorik, A. C.; Kanatzidis, M. G. Chem. Mater. 1997, 9, 387. 77. Deng, B.; Ellis, D. E.; Ibers, J. A. Inorg. Chem. 2002, 41, 5716. 78. Ueda, K.; Hosono, H.; Hamada, N. J. Phys.: Condens. Matter 2004, 16, 5179. 79. Chan, G. H.; Deng, B.; Bertoni, M.; Ireland, J. R.; Hersam, M. C.; Mason, T. O.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2006, 45, 8264. 80. Liu, M. L.; Wu, L. B.; Huang, F. Q.; Chen, L. D.; Ibers, J. A. J. Solid State Chem. 2007, 180, 62. 34 CHAPTER 2 SYNTHESIS, STRUCTURE, AND MAGNETIC PROPERTIES OF THE NONCENTROSYMMETRIC TERNARY RARE-EARTH ANTIMONY POLYSULFIDE Eu6Sb6S17 ABSTRACT Eu6Sb6S17 is isostructural with Sr6Sb6S17 and can be treated as consisting of two isolated [Sb3S7]5? anions and a S32? polysulfide anion that are joined together by Eu2+ cations. There are six crystallographically unique Eu2+ cations bound by sulfide and polysulfide anions in seven-, eight-, and nine-coordinate environments. Each of the [Sb3S7]5? anions consists of a trimer of corner-sharing SbS3 units. The [Sb3S7]5? anions are connected by long (~3.1 ?) Sb?S interactions, forming one-dimensional ribbons running down the a axis and packed with opposing directions of polarity. The bent S32? anions stack in columns along the a axis, oriented in opposing directions with respect to one another. The overall structure is three-dimensional and has channels running down the a axis to house the stereochemically active lone-pair of electrons on the Sb3+ centers. The presence of Eu2+ ions is supported by both magnetic measurements and bond valence calculations. Crystallographic data (193 K, Mo K?, ? = 0.71073): orthorhombic, space 35 group P212121, a = 8.236(2) ?, b = 15.237(3) ?, c = 22.724(5) ?, Z = 4, R(F) = 3.11% for 264 parameters with 7062 reflections with I > 2?(I).) INTRODUCTION Ternary rare-earth thioantimonites have been known for more than twenty years since the initial report of the synthesis and crystal structure of Eu3Sb4S9.1 Despite this early success, the chemistry of thioantimonites is relatively undeveloped compared to that of thiophosphates,2 and there is still a dearth of understanding in the structure-property relationships in this system. There has been renewed interest in ternary and quaternary chalcoantimonites and chalcobismuthites owing to their potential applications as thermoelectric materials. Based on structural analogies with Pn2Q3 phases (Pn = Sb, Bi; Q = S, Se, Te), a series of ternary phases such as EuSb2Q4, EuSb2Q7, and EuBi2Te4 has been prepared and their transport properties measured.3 EuBi2Te4 displays a thermoelectric figure of merit comparable with that of Bi2Te3.3 Eu2BiS4 is also of particular interest because it contains Eu(II) and Eu(III) in different coordination environments.4 Other ternary rare-earth thioantimonites such as Pr8Sb2S15,5 Ln3Sb3S10 (Ln = La, Ce),6 and Ln6Sb8S21 (Ln = La, Ce)7 have been characterized from single-crystal and powder X-ray diffraction data. EuSb4S7 has also been reported, but only its lattice constants are known.8 More recently there has been attention paid to preparing quaternary rare-earth chalcoantimonites and chalcobismuthites, e.g. K2(Ln2-x)Sb4+xSe12 (Ln = La, Ce, Pr, Gd),9 K2Gd2Sb2Se9,10 K2La2Sb2S9,10 Na9Gd5Sb8S26,11 BaLaBi2S6,12 ALn1?xBi4?xS8 (A = K, Rb; La, Ce, Pr, Nd),13 and Pb2Eu2Bi6Se13.14 These compounds possess complex structures 36 owing to the combination of the high coordination numbers found for lanthanides with discrete PnQ33?, PnQ3+13?, and PnQ45? anions as well as extended networks of chalcoantimonites and chalcobismuthites. Herein we report the preparation, crystal structure, and magnetic properties of Eu6Sb6S17, a new ternary thioantimonite that is isostructural Sr6Sb6S17 ,15 and which contains Eu(II) and a polysulfide linkage. The information in this chapter has been published as a full paper in Acta Crystallographica.16 EXPERIMENTAL Syntheses. The single crystal used for X-ray diffraction experiments was isolated from the reaction of Eu foil (99.9%, Alfa-Aesar), Sb (99.5%, Alfa-Aesar), and S (99.5%, Alfa-Aesar) that were loaded in a fused-silica tube in a molar ratio of 3:5:12 with a total mass of 0.250 g. The following heating profile was used: 5 ?C/min to 850 ?C for 1 d, 5 ?C/min to 1000 ?C for 7 d, 0.5 ?C/min to 600 ?C for 5 d, and 0.25 ?C/min to 22 ?C. The product consisted of small black block-shaped crystals of Eu6Sb6S17 and acicular crystals of Sb2S3. The yields were generally low (~10%), and were not improved by quenching the reactions. Semi-quantitative SEM/EDX analysis was performed on crystals of Eu6Sb6S17 using a JEOL 840/Link Isis instrument. Eu, Sb, and S percentages were calibrated against standards, and a Eu:Sb:S ratio of close to 1:1:2.8 (20%:23%:57%) was found. Crystallographic Studies. A single crystal of Eu6Sb6S17 was mounted on a glass fiber with epoxy and aligned on a Bruker SMART APEX CCD X-ray diffractometer. Intensity measurements were performed using graphite-monochromated Mo K? radiation from a sealed tube with a monocapillary collimator. SMART was used for preliminary 37 determination of the cell constants and data-collection control. The intensities of reflections of a sphere were collected by a combination of three sets of exposures (frames). Each set had a different ? angle for the crystal and each exposure covered a range of 0.3? in ?. A total of 1800 frames were collected with an exposure time per frame of 30 s. Determination of integrated intensities and global cell refinement were performed with the Bruker SAINT (v 6.02) software package using a narrow-frame integration algorithm. A face-indexed absorption (Gaussian) correction was initially applied using XPREP.17 Individual shells of unmerged data were corrected and exported in the same format. These files were subsequently treated with a semi-empirical absorption correction by SADABS.18 The program suite SHELXTL (v 5.1) was used for space group determination (XPREP), direct methods structure solution (XS), and least-squares refinement (XL).17 The final refinements included anisotropic displacement parameters for all atoms and a secondary extinction parameter. Some crystallographic details are listed in Table 2.1. Atomic coordinates and equivalent isotropic displacement parameters are given in Table 2.2. Inspection of the atomic positions (Table 2.2) reveals that they come in symmetry-related pairs (e.g., Eu1?Eu2, Eu3?Eu4, Eu5?Eu6) and the presence of additional pseudosymmetry elements is confirmed by PLATON.19 The structure can be refined in Pmnb to a residual of R(F) = 0.11 if a large number of systematic absence exceptions are excluded. The structure in Pmnb is identical to the structure in P212121, except that the S32? groups are in a disordered arrangement. We conclude, in agreement 38 Table 2.1. Crystallographic data for Eu6Sb6S17. Compound Eu6Sb6S17 Formula mass (amu) 2187.28 Color and habit black, prism Crystal system orthorhombic Space group P212121 (No. 19) a (?) 8.236(2) b (?) 15.237(3) c (?) 22.724(5) V (?3) 2851.7(11) Z 4 T (?C) ?80 ? (?) 0.71073 2?max 56.64 ?calcd (g cm?3) 5.095 ?(Mo K?) (cm?1) 198.01 Flack parameter 0.699(16) R(F) for Fo2 > 2?(Fo2) a 0.0311 Rw(Fo2) b 0.0615 a ( )R F F F F= ?? ? o c o . b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 39 Table 2.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for Eu6Sb6S17. Atom x y z Ueq (?2) a Eu(1) -0.24890(7) -0.44990(4) 0.21293(3) 0.01187(13) Eu(2) 0.25359(6) -0.44487(4) 0.21124(2) 0.01138(13) Eu(3) 0.25743(6) -0.55805(4) 0.04396(2) 0.01101(13) Eu(4) 0.75051(6) -0.56076(4) 0.04524(2) 0.01075(12) Eu(5) 0.75622(6) -0.66597(4) -0.12281(2) 0.01029(13) Eu(6) 0.24942(6) -0.66659(4) -0.12393(2) 0.01054(13) Sb(1) -0.02334(8) -0.20669(4) 0.23203(2) 0.01337(14) Sb(2) 0.01222(9) -0.31758(4) 0.08537(2) 0.01276(12) Sb(3) 0.49364(9) -0.32492(4) 0.05958(2) 0.01261(12) Sb(4) 0.53432(7) -0.42215(4) -0.09216(3) 0.01265(14) Sb(5) 0.01938(8) -0.42766(4) -0.10816(2) 0.01144(13) Sb(6) 0.99997(9) -0.80859(3) -0.25011(2) 0.01199(12) S(1) 0.0001(4) -0.34890(13) 0.28152(9) 0.0120(4) S(2) -0.2157(3) -0.26825(19) 0.15715(13) 0.0126(6) S(3) 0.2297(3) -0.2702(2) 0.14998(14) 0.0139(6) S(4) 0.0029(3) -0.47102(14) 0.11941(8) 0.0114(4) S(5) 0.5040(3) -0.46228(14) 0.11727(8) 0.0110(4) S(6) 0.2759(3) -0.3737(2) -0.00482(13) 0.0140(6) S(7) 0.7240(3) -0.37670(19) -0.00959(13) 0.0122(6) S(8) 0.5070(3) -0.56906(13) -0.05339(8) 0.0099(4) S(9) 0.0089(3) -0.56270(14) -0.05055(8) 0.0117(4) 40 S(10) 0.7675(3) -0.4741(2) -0.16414(14) 0.0131(6) S(11) 0.2424(3) -0.4778(2) -0.17007(13) 0.0133(6) S(12) 0.7490(3) -0.84651(18) -0.19195(12) 0.0109(5) S(13) 0.2119(3) -0.85439(19) -0.18205(13) 0.0139(6) S(14) 0.9982(4) -0.65447(13) -0.22023(8) 0.0104(4) S(15) 0.5012(3) -0.79586(13) -0.07184(8) 0.0116(4) S(16) 0.3593(3) -0.75150(16) -0.00157(10) 0.0113(5) S(17) 0.4964(3) -0.70617(13) 0.06972(8) 0.0120(4) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 41 with Choi et al.,15 that the correct space group is indeed P212121. The sole atom that breaks the mirror symmetry is S(16). Magnetism. A 15-mg sample consisting of ground aggregates of Eu6Sb6S17 crystals was collected for magnetic studies. Magnetic susceptibility measurements were made between 2 and 300 K under zero-field-cooled conditions in an applied field of 0.1 T on a Quantum Design 9T?PPMS dc magnetometer/ac susceptometer. The susceptibility was corrected for contributions from the holder diamagnetism and the underlying sample diamagnetism. (Magnetic susceptibility measurements were performed by Shane J. Crerar and Arthur Mar at University of Alberta) RESULTS AND DISCUSSION Structure. Although three-dimensional in nature, the structure of Eu6Sb6S17 can be treated as consisting of two isolated [Sb3S7]5? anions and a S32? polysulfide anion that are joined together by Eu2+ cations. Analogous to Sr6Sb6S17,15 to which Eu6Sb6S17 is isostructural, charge balance is attained in a straightforward manner by assuming the presence of six Eu2+ cations, six Sb3+ cations, fourteen S2? anions, and one S32? anion. The [Sb3S7]5? anion consists of a trimer of corner-sharing SbS3 units as shown in Figure 2.1. Other anions formed from the linking of SbSn units include [Sb2S4]2? in BaSb2S420 and [Sb2S5]4? in Sr2Sb2S5?15H2O.21 The Sb?S bond distances are given in Table 2.3 and range from 2.416(2) to 2.644(3) ?. As in Sr6Sb6S17,15 the [Sb3S7]5? anions are connected by long (3.105(3) to 3.106(3) ?) Sb?S interactions to form one-dimensional ribbons running down the a axis. Each of the Sb3+ cations possesses a stereochemically active lone-pair of electrons. Although space group P212121, in which this compound 42 crystallizes, is enantiomorphic and not polar, there is approximate alignment of the lone pair of electrons on the individual Sb3+ cations along the b axis. However, the [Sb3S7]5? anions are packed with opposing directions of polarity (i.e. the lone-pairs on different anions are aligned in opposite directions). The S32? anion is bent with S?S bond distances of 2.091(3) and 2.092(3) ?, and an S?S?S bond angle of 113.34(15)?. Although the polysulfide anions do stack in oriented columns along the a-axis, the columns are oriented in opposing directions with respect to one another, canceling any polarity. The six crystallographically unique Eu2+ cations are found in a variety of coordination environments, as shown in Figure 2.2. Eu?S bond distances are given in Table 2.3. Eu(1) and Eu(2) are found in seven-coordinate monocapped trigonal prismatic environments with Eu?S bond distances ranging from 2.984(3) to 3.087(3), and 2.962(3) to 3.209(3) ?, respectively. Eu(3) is bound by eight sulfide anions with bond distances ranging from 2.968(2) to 3.235(3) ?. The geometry around Eu(3) is heavily distorted from a bicapped trigonal prism, a square antiprism, or a dodecahedron because two of the sulfide nearest neighbors are part of the polysulfide linkage creating a very short edge on the polyhedron. Eu(4), Eu(5), and Eu(6) are all found in nine-coordinate geometries that are quite distorted from idealized tricapped trigonal prismatic symmetry because of the presence of a polysulfide linkage within their inner spheres. Eu(5)?S, Eu(6)?S, Eu(7)?S bond distances range from 3.005(3) to 3.382(3) ?, 2.980(2) to 3.208(2) ?, and 2.988(3) to 3.197(3) ?, respectively. Bond-valence sum calculations provide values ranging from 1.87 to 2.10 for the Eu centers in Eu6Sb6S17, which are consistent with Eu2+ in this compound.22,23 A packing diagram for the complete structure of Eu6Sb6S17 is shown in Figure 2.3. When the structure is viewed down the a-axis, it becomes apparent that 43 Eu6Sb6S17 adopts a channel structure. These channels house the stereochemically active lone pair of electrons on the Sb3+ centers. A similar channel motif is also observed for Eu3Sb4S9.1 Magnetic Susceptibility. The dc magnetic susceptibility of Eu6Sb6S17 between 2 and 300 K at an applied field of 0.1 T is shown in Figure 2.4. In the high-temperature regime, the susceptibility can be fit to the Curie-Weiss law with ? = ?3.3 K and C = 35.5 cm3 mol?1 K. The effective magnetic moment is 7.0 ?B/Eu, which is somewhat lower than the expected free-ion moment of 7.9 ?B for Eu2+. We attribute this to the presence of minor amounts of diamagnetic Sb2S3, which is difficult to separate cleanly from Eu6Sb6S17, as revealed by a powder X-ray diffraction analysis of the sample used for the magnetic measurements. The negative Weiss constant is consistent with antiferromagnetic coupling of the Eu2+ ions (the closest Eu?Eu separations are 4.1 ?), and is close to the 3.2 K transition observed in the susceptibility curve. 44 Figure 2.1. A view of the [Sb3S7]5? anions that consist of a trimer of corner-sharing SbS3 units in Eu6Sb6S17. 50% probability ellipsoids are depicted. 45 Table 2.3. Eu?S, Sb?S, and S?S Bond Distances (?) for Eu6Sb6S17. Eu(1)?S(1) 3.000(3) Eu(4)?S(8) 3.010(2) Eu(1)?S(2) 3.056(3) Eu(4)?S(9) 3.044(2) Eu(1)?S(4) 2.987(3) Eu(4)?S(13) 3.382(3) Eu(1)?S(5) 2.984(3) Eu(4)?S(15) 3.066(3) Eu(1)?S(10) 3.028(3) Eu(4)?S(16) 3.157(3) Eu(1)?S(13) 3.081(3) Eu(4)?S(17) 3.098(3) Eu(1)?S(14) 3.009(3) Eu(5)?S(1) 3.039(3) Eu(2)?S(1) 3.008(3) Eu(5)?S(8) 2.980(2) Eu(2)?S(3) 3.010(3) Eu(5)?S(9) 3.082(2) Eu(2)?S(4) 2.962(3) Eu(5)?S(10) 3.073(3) Eu(2)?S(5) 2.981(2) Eu(5)?S(12) 3.168(3) Eu(2)?S(11) 2.944(3) Eu(5)?S(14) 2.984(3) Eu(2)?S(12) 3.209(3) Eu(5)?S(15) 3.110(3) Eu(2)?S(14) 2.982(3) Eu(5)?S(16) 3.208(2) Eu(3)?S(4) 3.015(3) Eu(5)?S(17) 3.027(2) Eu(3)?S(5) 3.005(3) Eu(6)?S(1) 2.988(3) Eu(3)?S(6) 3.023(3) Eu(6)?S(8) 3.046(2) Eu(3)?S(8) 3.024(2) Eu(6)?S(9) 3.035(2) Eu(3)?S(9) 2.968(2) Eu(6)?S(11) 3.063(3) Eu(3)?S(15) 3.132(3) Eu(6)?S(13) 3.167(3) Eu(3)?S(16) 3.235(3) Eu(6)?S(14) 3.017(3) Eu(3)?S(17) 3.051(3) Eu(6)?S(15) 3.095(2) Eu(4)?S(4) 3.005(3) Eu(6)?S(16) 3.197(3) 46 Eu(4)?S(5) 3.009(2) Eu(6)?S(17) 3.101(3) Eu(4)?S(7) 3.077(3) Sb(1)?S(1) 2.449(2) Sb(4)?S(8) 2.416(2) Sb(1)?S(2) 2.507(3) Sb(4)?S(10) 2.644(3) Sb(1)?S(12) 2.664(3) Sb(5)?S(9) 2.440(2) Sb(2)?S(2) 2.598(3) Sb(5)?S(10) 2.534(3) Sb(2)?S(3) 2.426(3) Sb(5)?S(11) 2.436(3) Sb(2)?S(4) 2.464(2) Sb(6)?S(12) 2.520(3) Sb(3)?S(5) 2.471(2) Sb(6)?S(13) 2.434(3) Sb(3)?S(6) 2.431(3) Sb(6)?S(14) 2.445(2) Sb(3)?S(7) 2.587(3) S(15)?S(16) 2.091(3) Sb(4)?S(7) 2.537(3) S(16)?S(17) 2.092(3) 47 Figure 2.2. A depiction of the local environments of the six crystallographically unique Eu2+ cations in Eu6Sb6S17. 50% probability ellipsoids are depicted. 48 Figure 2.3. An illustration of the three-dimensional channel structure of Eu6Sb6S17 viewed down the a axis. Eu?S bond have been omitted for clarity. 49 Figure 2.4. Plots of dc magnetic susceptibility and its inverse for Eu6Sb6S17. The straight line shows a fit of the inverse susceptibility to the Curie-Weiss law. 0 50 100 150 200 250 300 0 2 4 6 8 10 12 14 16 18 Su sc ep tib ilit y, M/ H (em u/f .u. ) Temperature, T (K) 0 1 2 3 4 5 6 7 8 9 Eu6Sb6S17 Inv ers e s us ce pti bil ity , H /M (f. u./ em u) 0 2 4 6 8 10 0 4 8 12 16 M/ H (em u/f .u. ) T (K) 50 REFERENCES 1. Lemoine, P.; Carr?, D.; Guittard, M. Acta Crystallogr. 1981, B37, 1281. 2. a) Evenson, C. R.; Dorhout, P. K. Inorg. Chem. 2001, 40, 2884. b) Kanatzidis, M. G. Curr. Opin. Solid State Mater. Sci. 1997, 2, 139. 3. a) Godzhaev, ?. M.; Rustamov, P. G.; Guseinov, M. S.; Aliev, O. M. Inorg. Mater. 1997, 13, 512. (b) Harman, T. C.; Honig, J. M. Thermoelectric and Thermomagnetic Effects and Applications; McGraw Hill, New York, 1967. 4. Carr?, D.; Guittard, M.; Jaulmes, S.; Julien-Pouzul, M.; Lemoine, P.; Laurelle, P.; Flahaut, J. J. Less-Common Met. 1985, 110, 349. 5. Guseinov, G. G.; Mamedov, F. K.; Amiraslanov, I. R.; Mamedov, Kh. S. Kristallografiya 1981, 26, 831. 6. Gao, J.-Z.; Nakai, I.; Nagashima, K. Bull. Chem. Soc. Jpn. 1983, 56, 2615. 7. Gao, J.-Z.; Nakai, I.; Nagashima, K. Bull. Chem. Soc. Jpn. 1984, 57, 875. 8. Aliev, O. M.; Rustamov, P. G.; Guseinov, G. G.; Guseinov, M. S. Izv. Akad. Nauk SSSR, Neorg. Mater. 1978, 14, 1346. 9. Chen, J. H.; Dorhout, P. K. J. Alloys and Compd. 1997, 249, 199. 10. Choi, K.-S.; Hanko, J. A.; Kanatzidis, M. G. J. Solid State Chem. 1999, 147, 309. 11. Park, S.; Kim, S.-J. J. Solid State Chem. 2001, 161, 129. 12. Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G. Inorg. Chem. 1997, 36, 3804. 13. Iordanidis, L.; Schindler, J. L.; Kannewurf, C. R.; Kanatzidis, M. G. J. Solid State Chem. 1999, 143, 151. 51 14. Chung, D.-Y.; Iordanidis, L.; Choi, K.-S.; Kanatzidis, M. G. Bull. Kor. Chem. Soc. 1998, 19, 1283. 15. Choi, K.-S.; Kanatzidis, M. G. Inorg. Chem. 2000, 39, 5655. 16. Jin, G. B.; Wells, D. M.; Crerar, Shane J.; Shehee, T. C.; Mar, A.; Albrecht- Schmitt, T. E. Acta Crystallogr. 2005, E61, i116. 17. Sheldrick, G. M. SHELXTL PC, Version 6.12, An Integrated System for Solving, Refining, and Displaying Crystal Structures from Diffraction Data; Siemens Analytical X-Ray Instruments, Inc.: Madison, WI 2001. 18. Sheldrick, G. M. SADABS 2001, Program for absorption correction using SMART CCD based on the method of Blessing: Blessing, R. H. Acta Crystallogr. 1995, A51, 33. 19. Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7. 20. Cordier, G.; Sch?fer, H. Z. Naturforsch. 1979, 34b, 1053. 21. Schwidetzky, C. W. D. Zur Struckturchemie der Alkali- und Erdalkali- Thio und Selenopnictate (III); Tech. Hochschule Darmstadt, 1985. 22? Brown, I. D.; Altermatt, D. Acta Crystallogr. 1985, B41, 244. 23? Brese, N. E.; O?Keeffe, M. Acta Crystallogr. 1991, B47, 192. 52 CHAPTER 3 SYNTHESES, STRUCTURES, AND MAGNETIC PROPERTIES OF THE EUROPIUM(II) SELENIDO PNICTOGENATES(III), EuPnSe3 (Pn = Sb, Bi) ABSTRACT EuPnSe3 (Pn = Sb, Bi) have been synthesized through the reaction of Eu with Pn2Se3 (Pn = Sb, Bi) and Se at 850 to 900 ?C. These compounds are isotypic with SrPnSe3 (Pn = Sb, Bi) and consist of square pyramidal PnSe5 units and distorted PnSe6 octahedra that form hollow columns that extend along the c axis. These columns are separated by Eu2+ cations that occur as nine-coordinate tricapped trigonal prisms. There are also additional V-shaped triselenide Se32? anions between the columns that bind the Eu2+ cations. The Se???Se contacts (in EuSbSe3) in these units are 2.4584(11) and 2.4359(11) ?, which are consistent with Se?Se single bonds. The overall structure is chiral. Bond-valence sum calculations indicate that this compound contains Eu2+. Magnetic susceptibility measurements provide values of 7.66 ?B/Eu for EuSbSe3 and 7.64 ?B/Eu for EuBiSe3, which are close to the expected free-ion moment for Eu2+. These compounds follow essentially Curie behavior from 300 K to 5 K, and undergo an apparently antiferromagnetic transition below 5 K. 151Eu and 121Sb M?ssbauer spectra of EuSbSe3 and EuBiSe3 were measured at different temperatures. The 53 presence of divalent europium and trivalent antimony were confirmed. The largely negative values of the isomer shift in 151Eu spectrum show highly ionic bonding within these two compounds. Both of them show magnetic hyperfine field splitting at 4.2 K, which indicates a change in the orientation of the EFG principal axis with respect to the magnetic hyperfine field direction. EuSbSe3 has slightly smaller electron density at the antimony nuclei, compared to Sb2Se3. Crystallographic data: EuSbSe3, orthorhombic, space group P212121, a = 32.936(2) ?, b = 15.406(1) ?, c = 4.2622(3) ?, V = 2162.7(2) ?3, Z = 16, R(F) = 2.63% for 183 parameters and 5095 reflections with I > 2?(I); EuBiSe3, orthorhombic, space group P212121, a = 33.307(2) ?, b = 15.5804(9) ?, c = 4.2274(2) ?, V = 2193.7(2) ?3, Z = 16, R(F) = 2.68% for 183 parameters and 4895 reflections with I > 2?(I). INTRODUCTION Ternary europium pnictogen chalcogenide compounds form a rich group that is currently represented by EuPn2Q4 (Pn = Sb, Bi; Q = S, Se, Te) ,1,2 EuSb4S7,2 Eu3Sb4S9,3 Eu1.1Bi2S4,4 Eu2BiS4,5 EuPn4Q7 (Pn = Sb, Bi; Q = S, Se),6 Eu3Pn4Q9 (Pn = Sb, Bi; Q = S, Se, Te),6 Eu6Sb6S17,7 and the metal-rich phase Eu4Bi2Te.8 The structures and properties of many of these compounds have been reviewed by Carr? and co-workers.9 In addition to adopting novel structure types, these compounds possess interesting electronic properties that are often associated with the divalent or mixed-valent character of the Eu ions. Many of the aforementioned compounds contain Eu(II), which can be rationalized by the stability of the half-filled 4f7 shell. Eu2BiS4 contains both Eu(II) and Eu(III) in different crystallographic sites.5 The transport and magnetic properties of many of these 54 compounds have been measured. In this regard, one of the more interesting phases is Eu2BiS4, which displays semi-metallic behavior ascribed to the mobility of electrons inside hexagonal channels within the structure.9 While europium pnictogen sulfides have good representation, less is known about selenides and tellurides. In an effort to better understand this system we have prepared the new phases EuPnSe3 (Pn = Sb, Bi), which proved to be isotypic with SrPnSe3 (Pn = Sb, Bi),10,11 determined their crystal structures, and measured their magnetic properties and 151Eu and 121Sb M?ssbauer spectra, which are reported herein. The information in this chapter has been reported as full papers in Journal of Solid-State Chemistry.12,13 EXPERIMENTAL Syntheses. Eu (99.9%, Alfa-Aesar), Sb (99.5%, Alfa-Aesar), Bi (99.5%, Alfa- Aesar), and Se (99.5%, Alfa-Aesar) were used as received. Sb2Se3 and Bi2Se3 were prepared from the direct reaction of the elements in sealed fused-silica ampoules at 850 ?C. EuPnSe3. EuSbSe3: Eu (0.0595 g, 0.392 mmol), Sb2Se3 (0.0941 g, 0.196 mmol), and Se (0.0464 g, 0.588 mmol) or for EuBiSe3: Eu (0.0508 g, 0.334 mmol), Bi2Se3 (0.1095 g, 0.167 mmol), and Se (0.0396 g, 0.501 mmol) were loaded into fused-silica ampoules that were then sealed under vacuum. The following heating profiles were used: EuSbSe3 - 2 ?C/min from room temperature to 500 ?C where it was held for 1 h, 0.5 ?C/min to 850 ?C where it was held for 6 d, 0.04 ?C/min to 400 ?C where it was held 2 d, and 0.5 ?C/min to 24 ?C; EuBiSe3: 2 ?C/min to 600 ?C where it was held for 1 h, 0.5 ?C/min to 900 ?C where it was held for 4 d, 0.03 ?C/min to 500 ?C where it was held for 55 1 d, and 0.5 ?C/min to 24 ?C. In both cases the products consisted of thin black needles up to 2 mm in length. PXRD measurements confirmed phase purity by comparison with powder patterns calculated from the single crystal X-ray structures. Semi-quantitative SEM/EDX analyses were performed using a JEOL 840/Link Isis instrument. Eu, Sb, Bi, and Se percentages were calibrated against standards. The Eu:Pn(Sb, Bi):Se ratio determined from EDX analyses was approximately 1:1:3. Crystallographic Studies. Single crystals of EuSbSe3 and EuBiSe3 were mounted on glass fibers and aligned on a Bruker SMART APEX CCD X-ray diffractometer and cooled to 193 K using an Oxford Cryostat. Intensity measurements were performed using graphite monochromated Mo K? radiation from a sealed tube and monocapillary collimator. SMART (v 5.624) was used for preliminary determination of the cell constants and data-collection control. The intensities of reflections of a sphere were collected by a combination of 3 sets of exposures (frames). Each set had a different ? angle for the crystal and each exposure covered a range of 0.3? in ?. A total of 1800 frames were collected with an exposure time per frame of 30 s for both compounds. For EuSbSe3 and EuBiSe3 determination of integrated intensities and global refinement were performed with the Bruker SAINT (v 6.02) software package using a narrow-frame integration algorithm. A face-indexed numerical absorption correction was initially applied using XPREP.14 These data were subsequently treated with a semiempirical absorption correction by SADABS.15 The program suite SHELXTL (v 6.12) was used for space group determination (XPREP), direct methods structure solution (XS), and least-squares refinement (XL).14 The final refinements included anisotropic displacement parameters for all atoms and secondary extinction. Some crystallographic 56 details are given in Table 3.1. Atomic coordinates and equivalent isotropic displacement parameters for EuSbSe3 and EuBiSe3 can be found in Tables 3.2 and 3.3, respectively. Magnetism. Magnetic susceptibility measurements were made between 2 and 300 K under zero-field-cooled conditions in an applied field of 0.5 T on a Quantum Design 9T?PPMS dc magnetometer / ac susceptometer. The susceptibility was corrected for contributions from the holder diamagnetism and the underlying sample diamagnetism. (Magnetic susceptibility measurements were performed by Shane J. Crerar and Arthur Mar at University of Alberta) 151Eu and 121Sb M?ssbauer spectroscopy. The 21.53 keV transition of 151Eu with an activity of 130 MBq (2% of the total activity of a 151Sm:EuF3 source) and a Ba121mSnO3 source were used for the M?ssbauer spectroscopic experiments, which were conducted in the usual transmission geometry. The measurements were performed with a commercial helium bath cryostat. The temperature of the absorber was varied between 4.2 K and room temperature, while the source was kept at room temperature. The temperature was controlled by a resistance thermometer (?0.5 K accuracy). The samples were placed within a thin-walled PVC container at a thickness corresponding to about 10 Eu/cm2, respectively. (The M?ssbauer spectroscopic experiments were conducted by Falko M. Schappacher and Rainer P?ttgen at Westf?lische Wilhelms-Universit?t M?nster) RESULTS AND DISCUSSION Structures of EuPnSe3 (Pn = Sb, Bi). EuSbSe3 and EuBiSe3 are isotypic with SrPnSe3 (Pn = Sb, Bi).10,11 The structure of EuSbSe3 will be discussed with values for the Bi analog given parenthetically where important. There are four crystallographically 57 Table 3.1. Crystallographic Data for EuSbSe3 and EuBiSe3. Formula EuSbSe3 EuBiSe3 Formula Mass 510.59 597.82 Color and habit Black needle black needle Crystal System orthorhombic orthorhombic Space group P212121 (No. 19) P212121 (No. 19) a (?) 32.936(2) 33.307(2) b (?) 15.406(1) 15.5804(9) c (?) 4.2622(3) 4.2274(2) V (?3) 2162.7(2) 2193.7(2) Z 16 16 T (K) 193 193 ? (?) 0.71073 0.71073 Maximum 2? (deg.) 56.74 56.58 R(int) 0.0430 0.0494 Reflections (total) 21782 22502 Reflections (ind.) 5410 5470 Parameter 182 183 Weighting scheme 0.0298 0.0190 Res. electron den. (min,max) -3.029, 2.044 -1.721, 2.856 ?calcd (g cm?3) 6.273 7.240 ? (Mo K?) (cm?1) 365.10 631.10 R(F) for Fo2 > 2?(Fo2) 0.0263 0.0268 Rw(Fo2) b 0.0612 0.0538 a ( )R F F F F= ?? ? o c o . b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 58 Table 3.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for EuSbSe3. Atom (site) x y z Ueq (?2) a Eu(1) 0.576979(10) 0.57063(2) 0.49241(10) 0.01218(8) Eu(2) 0.296366(10) 0.13685(2) 0.02213(11) 0.01185(8) Eu(3) 0.328241(11) 0.78806(2) 0.55498(9) 0.01190(8) Eu(4) 0.540637(11) 0.22202(3) 0.98005(11) 0.01640(9) Sb(1) 0.863350(19) 0.94069(4) 0.97073(16) 0.02636(13) Sb(2) 0.73758(2) 0.87685(4) 0.95831(17) 0.03376(17) Sb(3) 0.522404(14) 0.96358(3) 0.48936(14) 0.01751(11) Sb(4) 0.612807(16) 0.80824(3) 0.95840(15) 0.01861(12) Se(1) 0.58106(2) 0.02102(6) 0.9515(3) 0.0280(2) Se(2) 0.68091(2) 0.84877(5) 0.4793(3) 0.0260(2) Se(3) 0.80227(2) 0.91220(6) 0.4644(4) 0.0344(3) Se(4) 0.28659(2) 0.98044(4) 0.5007(2) 0.01259(14) Se(5) 0.63578(2) 0.65000(4) 0.9772(2) 0.01173(14) Se(6) 0.26013(2) 0.78192(5) 0.03648(19) 0.01229(15) Se(7) 0.43833(2) 0.27777(5) 0.0203(2) 0.01714(16) Se(8) 0.38870(2) 0.71799(5) 0.0379(2) 0.01347(15) Se(9) 0.39990(2) 0.92243(5) 0.5091(2) 0.01271(14) Se(10) 0.51079(2) 0.61894(5) 0.0084(2) 0.01391(14) Se(11) 0.47580(2) 0.89768(5) 0.9407(2) 0.0219(2) Se(12) 0.34292(2) 0.94859(5) 0.14871(18) 0.01230(16) 59 Table 3.3. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for EuBiSe3. Atom (site) x y z Ueq (?2) a Eu(1) 0.423100(13) 0.43651(3) 0.50871(15) 0.01152(10) Eu(2) 0.703237(13) 0.86713(3) 0.97886(15) 0.01120(10) Eu(3) 0.673097(14) 0.21559(3) 0.44798(12) 0.01090(11) Eu(4) 0.457513(14) 0.78660(3) 0.01679(16) 0.01635(11) Bi(1) 0.139593(10) 0.05145(2) 0.01155(12) 0.01252(8) Bi(2) 0.265127(10) 0.12156(2) 0.01405(12) 0.01250(8) Bi(3) 0.478387(10) 0.04741(2) 0.51028(11) 0.01252(8) Bi(4) 0.389701(10) 0.19634(2) 0.01923(12) 0.01215(8) Se(1) 0.41803(3) 0.98808(6) 0.0436(3) 0.0124(2) Se(2) 0.32141(3) 0.15262(6) 0.5316(3) 0.0127(2) Se(3) 0.20035(3) 0.08477(6) 0.5422(3) 0.0118(2) Se(4) 0.71366(2) 0.02172(6) 0.5002(3) 0.01103(18) Se(5) 0.36444(3) 0.35747(6) 0.0229(3) 0.01081(19) Se(6) 0.74097(3) 0.21586(6) 0.9655(3) 0.0109(2) Se(7) 0.55806(3) 0.73215(6) 0.9767(3) 0.01188(19) Se(8) 0.61263(3) 0.28598(6) 0.9668(3) 0.0116(2) Se(9) 0.60217(3) 0.08292(6) 0.4901(3) 0.01185(19) Se(10) 0.48925(3) 0.38643(6) -0.0049(3) 0.01159(18) Se(11) 0.52751(3) 0.11325(6) 0.0509(2) 0.0119(2) Se(12) 0.65795(3) 0.05528(7) -0.1459(2) 0.0106(2) 60 unique Eu atoms in the structures of EuPnSe3. All four Eu sites are nine-coordinate and occur as tricapped trigonal prisms. Eu?Se bond distances range from 3.0846(9) to 3.2950(9) ? (3.0907(12) to 3.3240(11) ?), 3.0516(8) to 3.3496(9) ? (3.0587(11) to 3.4417(11) ?), 3.0419(8) to 3.5712(9) ? (3.0449(11) to 3.5773(12) ?), and 3.1052(9) to 3.7031(10) ? (3.1042(12) to 3.7447(10) ?) for Eu(1), Eu(2), Eu(3), and Eu(4), respectively. These polyhedra are shown in Figure 3.1. Selected bond distances can be found in Tables 3.4 and 3.5. Bond-valence sum calculations provide values for the four Eu sites ranging from 1.60 to 2.27, implying Eu(II) character.16,17 The coordination environments around the Sb centers are challenging to describe owing to the extreme variability in the Sb?Se bond lengths. For example, for Sb(3) there are three relatively short Sb?Se bonds of 2.6312(9), 2.6621(10), 2.8974(12) ?. If these are the sole contacts used then the geometry could be described as a distorted trigonal pyramid with a stereochemically active lone pair of electrons. However, there are three longer contacts ranging from 2.9759(11) to 3.1414(9) ?. It is probably best to describe this unit as a distorted octahedron. For Sb(1), Sb(2), and Sb(4) the sixth potential Sb???Se contact exceeds 3.4 ?, and these units are probably best described as square pyramids. The Sb and Se atoms form rectangular columns that extend down the c axis as is shown in Figure 3.2. These columns are formed from two opposing nets of square pyramidal SbSe5 units that are linked by the SbSe6 units (Figure 3.3). The lone pair of electrons on the Sb(III) centers appear to be contained within these columns. These rock-salt-like fragments are known from other ternary antimony and bismuth chalcogenide phases such as CsBi3.67Se6,18 Sr4Bi6Se13,19 A1+xPb4-2xSb7+xSe15 (A = K, Rb; 0 2?(Fo2). b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 87 Table 4.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?- LaYbS3. Atom (site) x y z Ueq (?2)a La(1) 0.15941(4) 0.2500 0.24448(3) 0.00942(16) La(2) -0.16994(4) 0.2500 0.41369(3) 0.00593(15) La(3) -0.19293(4) -0.2500 0.08024(3) 0.00787(15) Yb(1) 0.06664(3) 0.2500 0.42039(2) 0.00756(13) Yb(2) -0.09466(3) -0.2500 0.25016(3) 0.01124(14) Yb(3) 0.05686(3) -0.2500 0.06883(2) 0.00886(13) S(1) -0.03553(18) 0.2500 0.31730(14) 0.0080(6) S(2) 0.20923(18) 0.2500 0.48254(14) 0.0083(6) S(3) 0.04299(18) 0.2500 0.55274(14) 0.0082(6) S(4) -0.1778(2) -0.7500 0.18492(14) 0.0124(6) S(5) 0.05953(18) -0.2500 -0.05774(14) 0.0091(6) S(6) 0.15831(18) -0.7500 0.06464(15) 0.0104(6) S(7) -0.22911(18) -0.2500 0.32202(14) 0.0076(6) S(8) 0.14585(18) -0.2500 0.35127(14) 0.0087(6) S(9) 0.05085(19) -0.2500 0.19232(14) 0.0101(6) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 88 Magnetic Susceptibility Measurements. Magnetism data were measured on powders in gelcap sample holders with a Quantum Design MPMS 7T magnetometer/susceptometer between 2 and 300 K and in applied fields up to 7 T. DC temperature dependent susceptibility measurements were made under zero-field-cooled conditions with an applied field of 0.1 T. Susceptibility values were corrected for the sample diamagnetic contribution according to Pascal?s constants47 as well as for the sample holder diamagnetism. ?p values were obtained from extrapolations from fits between 100 to 300 K. (Magnetic susceptibility measurements were performed by Eun Sang Choi and James S. Brooks at Florida State University) UV-vis-NIR Diffuse Reflectance Spectroscopy. The diffuse reflectance spectra for ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) were measured from 200 to 1500 nm using a Shimadzu UV3100 spectrophotometer equipped with an integrating sphere attachment. The Kubelka-Monk function was used to convert diffuse reflectance data to absorption spectra.48 RESULTS AND DISCUSSION Effects of Synthetic Parameters on Product Composition and Structure. ?- LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) were synthesized via the reaction of the respective lanthanides with elemental sulfur in a Sb2S3 flux at 1000 ?C. Conveniently, when the reactions are maintained at a slight angle, on cooling, the majority of the flux moves to the bottom of the ampoules leaving isolated crystals of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). In sharp contrast, in the absence of the flux, microcrystalline mixed- 89 lanthanide sulfides phases form, but the yield is very low, and the crystals are far too small to investigate using single crystal X-ray diffraction. Furthermore, because there are several recognized phases for mixed-lanthanide sulfides with compositions close to 1:1:3,37,41-43 the products of these direct reactions are also ambiguous. The choice of flux has proven to be critical in this system as demonstrated by the replacement of Sb2S3 with CsCl, which instead results in Cs+ incorporation, and the formation of the quaternary phases Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La ? Yb).49 A KI flux has been used in the preparation of ?-LnYbQ3 (Ln = La ? Sm; Q = S, Se).43 However, in contrast to the selenides, the yield of ?-LaYbS3 was very low, inhibiting detailed measurements of its electronic properties.43 It has been previously noted even in binary sesquisulfides that preparative conditions play a dramatic role in the structure type adopted.27-29 For example, four different forms of Er2S3 can be synthesized by varying temperature, pressure, and crystal growth methods.30 Product composition and structure were investigated as a function of the size of Ln, Ln', and by the substitution of Se for S. When the size of Ln is decreased slightly (on the order of 0.02 ?) on going from Ce to Pr, ternary phases with the CeYb3S6 (F-Tm2S3) structure-type29 were found to form. It is important to note that the trivalent ions in this phase are highly disordered.31,33 For the La-containing phases, when Ln' is increased in size by transitioning from Er to Ho, mixtures of LaHo3S6 with the CeYb3S6 structure and ?-LaHo'S3 crystallize. In contrast, for the Ce-containing phases, when the same substitution of Ho for Er is performed, the new phase ?-CeHoS3 (CeTmS3-type37) is found. The structure and properties of ?-CeHoS3 are quite distinct from that of ?-LnLn'S3 90 (Ln = La, Ce; Ln' = Er, Tm, Yb), and will be the subject of a subsequent report. In general, when Ln and Ln' become too close in size, we have found that for those reactions that occur in Sb2S3 fluxes, that the yield of the desired ternary phases becomes very low and crystals often do not form. The substitution of Se for S results in a highly complex LnxLn'ySez system. For the La/Yb/Se reaction in a Sb2Se3 flux, La5Yb5-xSe7 with the Y5S7 structure-type forms.50,51 The lanthanide ions in this phase are highly disordered. In the (Ce ? Nd)/Yb/Se series the ?-LnYbS3 phases are found instead.43 When the size of Ln' is increased on substituting Tm for Yb the Ln1+xTm3-xSe6 (Ln = La ? Sm) compounds form, which adopt the disordered CeYb3S6 structure.31 When the size of Ln' is increased again by replacing Tm with Er, in the La/Er/Se system, LaErSe3 with the ?-CeHoS3 structure crystallizes. As in the sulfide reactions, if the difference in the size of the two lanthanide ions becomes too small, as is the case in Nd/Er/Se, the disordered CeYb3S6 structure is adopted again (e.g. in Nd1+xEr3-xSe6). As can be gleaned from the above discussion, there can be sharp demarcations between neighboring lanthanide ions in ternary mixed-lanthanide sulfides and selenides. Notably we have yet to mention the Lu-containing phases in the above discussion. This is because our observations are that these reactions do not follow previously observed trends for mixed-lanthanide phases where there is substantial size mismatch. Instead, we have isolated (in low yield) LaLu3S6 with the CeYb3S6 structure, which would normally result from having lanthanides of more similar size, as well as ?-LnLuS3 (Ln = Ce, Pr, 91 Nd). To reiterate, both of structure-types have lanthanide site positional disorder that is unexpected in this system because these trivalent ions differ by approximately 0.17 ?. Structural Features of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). The isotypic series, ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb), crystallize in the centrosymmetric orthorhombic space group Pnma. While this structure is a dense layered network, we will describe it in terms of lower-dimensional substructures so that the subtleties of bonding in these complex materials can be better understood. It is important to note from the outset that the structure of ?-LnLn'S3 is dramatically different from that of both ?- and ?-LnLn'S3.41-43 ?-LnLn'S3 adopts the same space group as ?- LnLn'S3. However, there is only one distinct crystallographic site for each of the Ln and Ln' centers (two total sites). The same is also true for the layered structure of ?-LnLn'S3 (UFeS3-type52).43 In contrast, ?-LnLn'S3 can also be expressed as Ln3Ln'3S9 because there are three crystallographically unique sites for both the Ln and Ln' atoms. Ordering of two different Ln3+ ions in a lattice is often difficult to achieve owing to the similarities in the structural chemistry of the trivalent lanthanide ions (vide supra). In ?-, ?-, and ?- LnLn'S3, ordering of the Ln and Ln' sites is accomplished by choosing lanthanide ions at opposite ends of the series. Typically earlier, larger lanthanides favor higher coordination numbers than the later, smaller ions. In ?-LnLn'S3 the two different lanthanide ions are actually both six-coordinate; however the larger ions are found in trigonal prisms, whereas the small ones are found in octahedral environments. In contrast, the larger trivalent lanthanides in ?-LnLn'S3 have eight close neighbors, and the Yb3+ ions only have six. ?-LnLn'S3 follows the same general trends as ?-LnLn'S3 by 92 Figure 4.1. Views of the structures of a) ?-LnLn'S3, b) ?-LnLn'S3, and c) ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). ?-LnLn'S3 adopts a layered structure with two-dimensional 2 ? [Ln'3S9] 9? (Ln' = Er, Tm, Yb) slabs that extend in the [bc] plane that contain the smaller lanthanide ions that are separated by larger Ln3+ ions, La3+ and Ce3+, as is depicted in Figure 4.1c. a c c b a c a) b) c) 93 Figure 4.2. A depiction of an individual 2? [Ln'3S9]9? (Ln' = Er, Tm, Yb) layer viewed down the a axis in ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). This layer is constructed from edge-sharing double chains of [Ln'S7] monocapped trigonal prisms (A) and double chains of [Ln'S6] octahedra (B) that are linked by single chains of [Ln'S6] octahedra (C) in the manner of ACBCA. The central C chains share corners with the A and B chains. c b A A B C C 94 Figure 4.3. Illustrations of the coordination environments for the larger lanthanides, La3+ and Ce3+, in ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). We use ?-LaYbS3 to represent this family of compounds here. 95 placing the larger ions in eight- and nine-coordinate environments, and the smaller ions are seven- and six-coordinate geometries. We choose here to adopt the same convention as used for ?-LnLn'S3 to describe the structure of ?-LnLn'S3, which is as two-dimensional 2? [Ln'3S9]9? (Ln' = Er, Tm, Yb) layers that extend in the [bc] plane that contain the smaller lanthanide ions that are separated by larger Ln3+ ions, La3+ and Ce3+, as is depicted in Figure 4.1c. For comparison, ?- and ?-LnLn'S3 are shown in Figures 4.1a and 4.1b. There are no S?S bonds in ?-, ?-, or ?-LnLn'S3, and therefore the oxidation states in these compounds can be assigned as +3/+3/-2. This designation is confirmed by both bond-valence sum calculations53,54 and by magnetic susceptibility measurements (vide infra). An individual 2? [Ln'3S9]9? (Ln' = Er, Tm, Yb) layer viewed down the a axis is shown in Figure 4.2. As can be seen in this sketch, this layer is constructed from edge- sharing double chains of [Ln'S7] monocapped trigonal prisms (A) and double chains of [Ln'S6] octahedra (B) that are linked by single chains of [Ln'S6] octahedra (C) in the manner ACBCA. The central C chains share corners with the A and B chains. The 2 ? [Ln'3S9] 9? layers in ?-LnLn'S3 are more buckled than those in ?-LnLn'S3, because the chains of [Ln'S6] octahedra (C) share both axial and equatorial corners with the double chains (A and B). In ?-LnLn'S3 only axial (trans) atoms are corner-sharing. For reference, in ?-LaYbS3 the Yb?S bond distance for Yb(1), Yb(2), and Yb(3) range from 2.5979(19) to 2.783(2) ?, and are normal (see Table 4.3). As shown in Figure 4.3, the larger lanthanides La3+ and Ce3+ (Ln(1), Ln(2), and Ln(3)) ions are found in coordination environments with more neighbors than the Er3+, 96 Table 4.3. Selected Bond Distances (?) for ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). Formula ?-LaErS3 ?-LaTmS3 ?-LaYbS3 ?-CeErS3 ?-CeTmS3 ?-CeYbS3 Ln(1)-S(1) 3.613(3) 3.579(3) 3.552(3) 3.661(3) 3.647(3) 3.591(3) Ln(1)-S(4) 3.039(3) 3.048(3) 3.062(3) 2.988(3) 2.995(3) 3.003(3) Ln(1)-S(6) 3.825(3) 3.814(3) 3.813(3) 3.803(3) 3.781(3) 3.801(3) Ln(1)-S(7) ?2 3.058(2) 3.051(2) 3.050(2) 3.0248(19) 3.018(2) 3.0145(19) Ln(1)-S(8) ?2 3.035(2) 3.028(2) 3.024(2) 3.0096(19) 3.007(2) 2.9967(19) Ln(1)-S(9) ?2 2.884(2) 2.885(2) 2.892(2) 2.8616(19) 2.863(2) 2.866(2) Ln(2)-S(1) 3.003(3) 3.004(3) 3.007(3) 2.974(3) 2.977(3) 2.976(3) Ln(2)-S(2) ?2 3.036(2) 3.034(2) 3.037(2) 3.0145(18) 3.013(2) 3.0094(19) Ln(2)-S(3) ?2 2.980(2) 2.977(2) 2.969(2) 2.9580(18) 2.956(2) 2.9533(19) Ln(2)-S(6) 2.862(3) 2.858(3) 2.856(3) 2.844(3) 2.845(3) 2.838(3) Ln(2)-S(7) ?2 2.9450(19) 2.944(2) 2.948(2) 2.9256(18) 2.926(2) 2.9274(19) Ln(3)-S(2) ?2 2.8863(19) 2.8855(19) 2.884(2) 2.8593(18) 2.864(2) 2.8586(18) Ln(3)-S(4) ?2 2.989(2) 2.989(2) 2.993(2) 2.9610(19) 2.960(2) 2.964(2) Ln(3)-S(5) ?2 3.002(2) 2.998(2) 2.998(2) 2.9858(19) 2.984(2) 2.9764(19) 97 Ln(3)-S(6) 3.128(3) 3.128(3) 3.124(3) 3.098(3) 3.107(3) 3.082(3) Ln(3)-S(8) 3.024(3) 3.019(3) 3.018(3) 3.008(2) 3.012(3) 2.996(3) Ln'(1)-S(1) 2.785(3) 2.771(3) 2.755(3) 2.789(2) 2.786(3) 2.758(2) Ln'(1)-S(2) 2.716(3) 2.697(3) 2.685(3) 2.720(2) 2.713(3) 2.683(3) Ln'(1)-S(3) ?2 2.7594(18) 2.7497(18) 2.744(2) 2.7613(17) 2.7602(19) 2.7367(18) Ln'(1)-S(3) 2.837(3) 2.825(3) 2.833(3) 2.822(2) 2.822(3) 2.812(3) Ln'(1)-S(8) ?2 2.8045(19) 2.7976(19) 2.794(2) 2.7988(18) 2.798(2) 2.7854(18) Ln'(2)-S(1) ?2 2.6429(17) 2.6352(17) 2.6339(19) 2.6300(16) 2.6239(18) 2.6178(16) Ln'(2)-S(4) ?2 2.815(2) 2.797(2) 2.783(2) 2.8155(19) 2.806(2) 2.7852(19) Ln'(2)-S(7) 2.698(3) 2.691(3) 2.681(3) 2.691(2) 2.684(3) 2.673(3) Ln'(2)-S(9) 2.705(3) 2.689(3) 2.684(3) 2.720(3) 2.705(3) 2.685(3) Ln'(3)-S(5) ?2 2.7946(18) 2.7791(18) 2.770(2) 2.7869(18) 2.779(2) 2.7616(18) Ln'(3)-S(5) 2.720(3) 2.702(3) 2.684(3) 2.725(3) 2.708(3) 2.691(3) Ln'(3)-S(6) ?2 2.6149(17) 2.6036(17) 2.5979(19) 2.6142(16) 2.6056(18) 2.5928(17) Ln'(3)-S(9) 2.656(3) 2.640(3) 2.620(3) 2.650(3) 2.644(3) 2.627(3) 98 Tm3+, and Yb3+ ions. There is some difficulty in precisely describing the coordination number of the Ln(1) centers. These Ln3+ cations are located within highly distorted tricapped trigonal prismatic environments. Two of the Ln???S capping contacts, however, are substantially longer than those that define the trigonal prism. For example, in ?- LaYbS3 the short La?S bonds range from 2.892(2) to 3.062(3) ?, whereas the longer contacts are 3.552(3) and 3.813(3) ? (see Table 4.3). The longest contact should probably not be considered as important. The Ln(2) and Ln(3) atoms, however, are clearly eight-coordinate, and occur as bicapped trigonal prisms with La?S bonds ranging from 2.856(3) to 3.037(2) ? for La(2), and from 2.844(2) to 3.124(3) ? for La(3) in ?- LaYbS3 (see Table 4.3). The transitions between ?-LnLn'S3, ?-LnLn'S3, and ?-LnLn'S3 represent systematic increases in the average coordination numbers of the lanthanides, leading to more efficient packing and increased densities in this single compositional family. Magnetic Properties of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). At first glance the magnetic behavior of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) seem relatively uninteresting as there is no indication of long-range magnetic ordering down to 2 K, as shown in Figure 4.4. However, the ?p parameters are atypical for some of these compounds. The ?p values are ?0.4, ?6.7, ?32.1, ?18.7, ?14.6, ?37.5 K, for ?-LaErS3, ?- LaTmS3, ?-LaYbS3, ?-CeErS3, ?-CeTmS3, and ?-CeYbS3, respectively. The ?p parameters for ?-LaYbS3 and ?-CeYbS3 are quite negative, indicating relatively strong antiferromagnetic coupling between Ln3+ ions. In the case of ?-LaYbS3, this interaction must be between Yb3+ centers. We observed similar phenomena in Cs0.14-0.17Ln0.26- 99 0.33YbS2 (Ln = La ? Yb).49 Despite this strong coupling, short- and long-range order are apparently absent. This behavior is consistent with a geometrically spin-frustrated system. It should be noted that ?-CeYbSe3 and ?-SmYbSe3 have very large ?p values of ?44.6(9) K and ?107.6(4) K without any indication of long-range order down to 5 K.43 The origin of potential geometric spin-frustration may lie in the layered substructures found in ?-LnLn'S3 and ?-LnLn'S3. The Sm network for ?-SmYbSe3 is shown in Figure 4.5a. This drawing shows a layer constructed from triangles. The asymmetry of these triangles is capable of preventing ordering in two dimensions. Figure 4.5b shows these layers rotated by 90? with the Yb3+ ions further connecting the network to create a system that might be frustrated in all three dimensions. In contrast, the source of geometric spin-frustration in ?-LnLn'S3 is less obvious. In these phases, there are networks of both squares and triangles within the Yb layers as well as the potential for higher-order frustration via the interlayer Ce3+ ions as shown in Figures 4.6a and b, respectively. We propose that if spin-frustration is occurring in these compounds that it is within the Yb layers, because ?-LaYbS3 and ?-CeYbS3 have large ?p values that are quite similar in magnitude, and the La3+ ions obviously can not contribute to this behavior. The Yb???Yb distances vary considerably between ?-SmYbSe3 and ?-LaYbS3, and are different depending on the substructures that they are present in. In ?-LaYbS3 and ?-LaYbS3, the shortest Yb???Yb distances of 3.9238(8) and 3.9847(3) ?, respectively, are similar. In the absence of appropriate trigonal or hexagonal symmetry the triangular networks in ?-LnLn'S3 and ?-LnLn'S3 are necessarily asymmetric. As such, one of the 100 0 50 100 150 200 250 300 0 20 40 60 80 100 120 140 160 180 ?-CeErS3 ?-CeTmS3 ?-LaErS3 ?-LaTmS3 ?-CeYbS3 ?-LaYbS3 ?-1 (m ol/ em u) Temperature (K) Figure 4.4. Temperature dependence of the reciprocal molar magnetic susceptibility for ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) under an applied magnetic field of 0.1 T. 101 Figure 4.5. a) A view of the layers in Sm-based triangles in ?-SmYbSe3. b) A depiction of the interconnection of the Sm layers by Yb3+ ions. In both cases the chalcogenide ions have been omitted for clarity. a c c b a) b) 102 Figure 4.6. a) An illustration of the square and triangular networks in ?-CeYbS3. b) A drawing of the complex three-dimension network in ?-CeYbS3. In both cases the chalcogenide ions have been omitted for clarity. c b a c a) b) 103 requirements of spin-frustration may not be present, and there may be alternative mechanisms for explaining the large negative ?p values. The first of these is that this might be an effect of crystal-field splitting of the ground state of the Yb3+ ions. A second, and much more complex, mechanism has also been proposed for the frustrated pyrochlore antiferromagnet, Tb2Ti2O7.55 In this compound there is strong evidence supporting extra perturbative exchange coupling beyond nearest neighbors as well as dipolar interactions that may be the cause of the lack of long-range ordering of the moments.56 The calculated free-ion moments for ?-LaErS3, ?-LaTmS3, ?-LaYbS3, ?-CeErS3, ?- CeTmS3, and ?-CeYbS3 are 9.59, 7.57, 4.54, 9.92, 7.98, and 5.20 ?B, respectively.57 Not surprisingly the observed moments (8.35, 6.65, 3.98, 9.87, 7.28, and 4.50 ?B) are notably smaller than the ideal values, most likely because of splitting of the ground state terms for these ions. It would require substantially larger fields than 7 T to obtain full moment values for these compounds. The above calculated values are only true if Ce is trivalent. The single 4f electron in Ce is tenuous at best, and tetravalent Ce carries no moment, in which case the Ce-based compounds would have the same effective moment as the La- based compounds. The isothermal magnetization measurements also show that the saturation moment is considerably less than the free-ion value. Isothermal magnetization measurements carried out to 7 T at 2 K support the conclusions of the effective moment measurements, i.e., that the moments are low. Optical Properties of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). The optical band gaps of ternary and quaternary chalcogenides have been shown to be tunable based 104 on the choice of lanthanide ion and chalcogen (e.g. CsLnZnSe3 (Ln = Sm, Tb, Dy, Ho, Er, Tm, Yb, and Y)).9,11,13,14,58 Based on electronic structure calculations, CsLnZnSe3 are thought to be direct band gap semiconductors.9 The band gaps for binary lanthanide sesquisulfides are generally indirect.59 Therefore, both direct and indirect band gap semiconductors are recognized in lanthanide chalcogenides, and they are highly dependent on the structure adopted by the phase.13,14,56,59 The optical band gaps for ?- LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) were measured using UV-vis-NIR spectroscopy (see Figure 4.7). Unfortunately, the band gaps of ?- and ?-LnLn'S3 materials have not been reported, although ?-LaYbS3 is reported to be yellow.43 Qualitatively, the ?-LaLn'S3 (Ln' = Er, Tm, Yb) compounds are dark red, whereas the ?-CeLn'S3 (Ln' = Er, Tm, Yb) compounds are black. Congruent with these visual observations, the measured band gaps for ?-LaLn'S3 (Ln' = Er, Tm, Yb) are all approximately 1.6 eV. The replacement of La by Ce results in a smaller gap for ?-CeLn'S3 (Ln' = Er, Tm, Yb) of ~1.3 eV. The apparent fine-structure in these spectra are actually f-f transitions for the lanthanide ions. The band gaps observed for ?-LnLn'S3 are comparable to the indirect band gaps reported for Nd3+:doped ?-Gd2S3 (A-type), ?-La2S3, and ?-Y2S3 (D-type).59 The Ce compounds often show the smallest band gaps of the series owing the high energy of 4f1 electron.60,61 The structural modification that occurs for ?-LnLn'S3 clearly alters the band gap substantially from that observed for the binary phases. 105 CONCLUSIONS The role of reaction conditions (pressure, temperature, flux, etc.) in determining composition and structure in both binary and ternary interlanthanide sesquichalcogenides cannot be overemphasized. We focus here on the use of the new flux, Sb2S3, which has been used to prepare ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) in high yield. In contrast, when KI is employed as a flux, ?-LaYbS3 can only be prepared in low yield, and this medium is more amenable for the syntheses of selenide compounds. Some fluxes, such as CsCl, are in fact reactive, resulting in Cs+ incorporation into the lanthanide chalcogenide phases, and the formation of quaternary compounds that are illustrated by Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La ? Yb).49 The use of fluxes provides access to mixed-lanthanide compounds with novel structure types, and is a means of understanding structure-property relationships in these intricate materials. The structure of ?-LnLn'S3 diverges substantially from previously reported interlanthanide sesquisulfides in that there are three crystallographically unique Ln sites in the structure, whereas ?- and ?-LnLn'S3 have one site for each type of lanthanide ion.41-43 Ordering of the Ln3+ ions is achieved by placing the larger ions in eight- and nine-coordinate environments, and the smaller ions are seven- and six-coordinate geometries. Table 4.4 provides a listing of mixed-lanthanide sesquichalcogenides that form under the conditions described in this work, indicating whether ordering of the Ln3+ ions is achieved, and the coordination environments of the Ln3+ ions. These data apply to solids that form from a Sb2Q3 (Q = S, Se) flux at 1000 ?C. 106 Currently, the most promising application of the mixed-lanthanide chalcogenides is the ability to change, and perhaps systematically vary, the band gap of these semiconductors. This can be accomplished by both changing the lanthanide, as we demonstrated in this report, and by changing the chalcogenide, whereby the band gap should decrease when heavier chalcogenides are substituted into the structure.58 Doing this allows materials to be prepared that range from nearly colorless to black. This property is not unique to the present system, but rather applies to many lanthanide compounds. We suggest that the possibility exists for finding interesting electronic phenomena by preparing ordered quaternary interlanthanide sesquichalcogenides with three different lanthanides, which might be accomplished by exploiting structures with three different coordination environments for the Ln3+ ions. 107 Figure 4.7. UV-vis diffuse reflectance spectra of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb). (*Denotes detector change.) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.8 1.8 2.8 3.8 4.8 5.8 Energy (eV) ?/s (a rbi tar y u nit s) * ?-LaYbS3 ?-LaTmS3 ?-LaErS3 ?-CeErS3 ?-CeTmS3 ?-CeYbS3 108 Table 4.4. Comparisons of Ternary Interlanthanide Sesquichalcogenides that Form in a Sb2Q3 (Q = S, Se) Flux at 1000 ?C. Structure type Compounds Ordered/Disordered Coordination ?-LnLn?S3 (UFeS3-type) CeYbSe3, PrYbSe3, NdYbSe3 Ordered Ln ? 8 Ln? ? 6 ?-LnLn?S3 LnLn?S3 (Ln=La, Ce, Ln?=Er-Yb) Ordered Ln ? 8, 9 Ln? ? 6, 7 ?-LnLn?S3 (CeTmS3-type) La5-xEr3+xSe12, Ce5-xHo3+xSe12, Ln5-xLu3+xS12 (Ln = Ce, Pr, Nd) Disordered Ln ? 7, 8 Ln? ? 6, 7 F-Ln2S3 La1+xLu3-xS6, Pr1+xTm3-xS6 Ln1+xYb3-xS6 (Ln= Pr-Gd), Ln1+xYb3-xSe6 (Ln= Sm, Gd) Ln1+xTm3-xSe6 (Ln= La-Sm), Nd1+xEr3-xSe6 Disordered Ln ? 7, 8 Ln? ? 6, 7 Y5S7 La5Yb5-xSe7 Disordered Ln ? 7 Ln? ? 6, 7 109 REFERENCES 1. Isaacs, T. J.; Hopkins, R. H.; Kramer, W. E. J. Electron. Mater. 1975, 4, 1181. 2. Hautala, J.; Taylor, P. C. J. Non-Cryst. Solids 1992, 141, 24. 3. Nadler, M. P.; Lowe-Ma, C. K.; Vanderah, T. A. Mater. Res. Bull. 1993, 28, 1345. 4. Kumta P. N.; Risbud, S. H. J. Mater. Sci. 1994, 29, 1135. 5. Lowe-Ma, C. K.; Vanderah, T. A.; Smith, T. E. J. Solid State Chem. 1995, 117, 363. 6. Ueda, K.; Inoue, S.; Hosono, H.; Sarukura, N.; Hirano, H. Appl. Phys. Lett. 2001, 78, 2333. 7. Inoue, S.; Ueda, K.; Hosono, H.; Hamada, N. Phys. Rev. B 2001, 64, 245211. 8. Huang, F. Q.; Mitchell, K.; Ibers, J. A. Inorg. Chem. 2001, 40, 5123. 9. Mitchell, K.; Haynes, C. L.; McFarland, A. D.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2002, 41, 1199. 10. Mitchell, K.; Ibers, J. A. Chem. Rev. 2002, 102, 1929. 11. Mitchell, K.; Huang, F. Q.; McFarland, A. D.; Haynes, C. L.; Somers, R. C.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2003, 42, 4109. 12. Ueda, K.; Takafuji, K.; Hiramatsu, H.; Ohta, H.; Kamiya, T.; Hirano, M.; Hosono, H. Chem. Mater. 2003, 15, 3692. 13. Mitchell, K.; Huang, F. Q.; Caspi, E. N.; McFarland, A. D.; Haynes, C. L.; Somers, R. C.; Jorgensen, J. D.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2004, 43, 1082. 110 14. Yao, J.; Deng, B.; Sherry, L. J.; McFarland, A. D.; Ellis, D. E.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2004, 43, 7735. 15. Bronger, W.; Miessen, H.-J.; Schmitz, D. J. Less-Common Met. 1983, 95, 275. 16. Sutorik, A. C.; Albritton-Thomas, J.; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Mater. 1996, 8, 751. 17. Givord, D.; Courtois, D. J. Magn. Magn. Mater. 1999, 196-197, 684. 18. Wakeshima, M.; Hinatsu, Y. J. Solid State Chem. 2000, 153, 330. 19. Tougait, O.; Ibers, J. A. Inorg. Chem. 2000, 39, 1790. 20. Wakeshima, M.; Hinatsu, Y.; Oikawa, K.; Shimojo, Y.; Morii, Y. J. Mater. Chem. 2000, 10, 2183. 21. Meng, F.; Hughbanks, T. Inorg. Chem. 2001, 40, 2482. 22. Wakeshima, M.; Hinatsu, Y.; Ishii, Y.; Shimojo, Y.; Morii, Y. J. Mater. Chem. 2002, 12, 631. 23. Lau, G. C.; Ueland, B. G.; Freitas, R. S.; Dahlberg, M. L.; Schiffer, P.; Cava, R. J. Phys. Rev. B 2006, 73, 012413. 24. Sleight, A. W.; Prewitt, C. T. Inorg. Chem. 1968, 7, 2283. 25. Patrie, M. Bull. Soc. Chim. Fr. 1969, 5, 1600. 26. Bergman, H., Ed. Gmelin Handbook of Inorganic Chemistry, Sc, Y, La-Lu (Rare Earth Elements), System Number 39, C7: Sulfides; Springer: Berlin, 1983. 27. Flahaut, J.; Guittard, M.; Patrie, M.; Pardo, M. P.; Golabi, S. M.; Domange, L. Acta Cryst. 1965, 19, 14. 28. Range, K. J.; Leeb, R. Z. Naturf. B. 1975, 30, 889. 29. Schleid, T.; Lissner, F. J. Alloys Compd. 1992, 189, 69. 111 30. Fang, C. M.; Meetsma, A.; Wiegers, G. A. J. Alloys Compd. 1993, 201, 255. 31. Rodier, N.; Tien, V. C. R. Seances Acad. Sci. Ser. C 1974, 279, 817. 32. Rodier, N.; Firor, R. L.; Tien, V.; Guittard, M. Mat. Res. Bull. 1976, 11, 1209. 33. Rodier, N.; Laruelle, P. Bull. Soc. fr. de Mineral. Cristallogr. 1973, 96, 30. 34. Hamilton, D. S.; Gayen, S. K.; Pogatshnik, G. J.; Ghen, R. D. Phys. Rev. B 1989, 39, 8807. 35. Cheung, Y. M.; Gayen, S. K. Phys. Rev. B 1994, 49, 14827. 36. Schweizer, T.; Mobert, P. E.-A.; Hector, J. R.; Hewak, D. W.; Brocklesby, W. S.; Payne, D. N.; Huber, G. Phys. Rev. Lett. 1998, 80, 1537. 37. Rodier, N. Bull. Soc. fr. de Mineral. Cristallogr. 1973, 96, 350. 38. Marezio, M.; Remeika, J. P.; Dernier, P. D. Acta Crystallogr. B 1970, 26, 2008. 39. Rodier, N.; Laruelle, P. C. R. Seances Acad. Sci. Ser. C 1970, 270, 2127. 40. Ijdo, D. J. W. Acta Crystallogr. B 1980, 36, 2403. 41. Rodier, N.; Julien, R.; Tien, V. Acta Crystallogr. C 1983, 39, 670. 42. Carre, D.; Laruelle, P. Acta Crystallogr. B 1974, 30, 952. 43 Mitchell, K.; Somers, R. C.; Huang, F. Q.; Ibers, J. A. J. Solid State Chem. 2004, 177, 709. 44. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. Chem. Mater. 2007, 19, 567. 45. Sheldrick, G. M. SHELXTL PC, Version 6.12, An Integrated System for Solving, Refining, and Displaying Crystal Structures from Diffraction Data; Siemens Analytical X-Ray Instruments, Inc.: Madison, WI 2001. 112 46. Sheldrick, G. M. SADABS 2001, Program for absorption correction using SMART CCD based on the method of Blessing: Blessing, R. H. Acta Crystallogr. 1995, A51, 33. 47. Mulay, L. N.; Boudreaux, E. A. Theory and Applications of Molecular Diamagnetism; Wiley?Interscience: New York, 1976. 48. Wendlandt, W. W.; Hecht, H. G.: Reflectance Spectroscopy; Interscience Publishers: New York, 1966. 49. Jin, G. B.; Booth, C. H.; Albrecht-Schmitt, T. E. Phys. Rev. B manuscript in preparation. 50. Adolphe, C. C. R. Seances Acad. Sci. 1961, 252, 3266. 51. Rodier, N.; Laruelle, P. Bull. Soc. Fr. Mineral. Cristallogr. 1972, 95, 548. 52. Noel, H.; Padiou, J. Acta Crystallogr. B 1976, 32, 1593. 53. Brown, I. D.; Altermatt, D. Acta Crystallogr. 1985, B41, 244. 54. Brese, N. E.; O?Keeffe, M. Acta Crystallogr. 1991, B47, 192. 55. Gardner, J. S.; Dunsinger, S. R.; Gaulin, B. D.; Gingras, M. J. P.; Greedan, J. E.; Kiefl, R. F.; Lumsden, M. D.; MacFarlane, W. A.; Raju, N. P.; Sonier, J. E.; Swainson, I.; Tun, Z. Phys. Rev. Lett. 1999, 82, 1012. 56. Gingras, M. J. P.; den Hertog, B. C.; Faucher, M.; Gardner, J. S.; Dunsinger, S. R.; Chang, L. J.; Gaulin, B. D.; Raju, N. P.; Greedan, J. E. Phys. Rev. B 2000, 62, 6496. 57. Kittel, C. Introduction to Solid State Physics; 6th Ed.; Wiley: New York, 1986. 58. Deng, B.; Ellis, D. E.; Ibers, J. A. Inorg. Chem. 2002, 41, 5716. 59. Leiss, M. J. Phys. C: Solid St. Phys. 1980, 13, 151. 113 60. Prokofiev, A. V.; Shelykh, A. I.; Golubkov, A. V.; Smirnov, I. A. J. Alloys Compd. 1995, 219, 172. 61. Prokofiev, A. V.; Shelykh, A. I.; Melekh, B. T. J. Alloys Compd. 1996, 242, 41. 62. Ito, K.; Tezuka, K.; Hinatsu, Y. J. Solid State Chem. 2001, 157, 173. 114 CHAPTER 5 SYNTHESES, STRUCTURE, MAGNETISM, AND OPTICAL PROPERTIES OF THE INTERLANTHANIDE SULFIDES ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd) ABSTRACT ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) compounds have been synthesized through the reaction of elemental rare earth metals and S using Sb2S3 flux at 1000 ?C. These compounds are isotypic with CeTmS3, which has a complex three- dimensional structure. It includes four larger Ln3+ sites in eight- and nine-coordinate environments, two disordered seven-coordinate Ln3+/Lu3+ positions, and two six- coordinate Lu3+ ions. The structure is constructed from one-dimensional chains of LnSn (n = 6 ? 9) polyhedra that extend along the b axis. These polyhedra share faces or edges with two neighbors within the chains, while in the [ac] plane they share edges and corners with other chains. Least squares refinements gave rise to the formulas of ?- Ce1.30Lu0.70S3, ?-Pr1.29Lu0.71S3 and ?-Nd1.33Lu0.67S3, which are consistent with the EDX analysis and magnetic susceptibility data. ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) show no evidence of magnetic ordering down to 5 K. Optical-property measurements show that the band gaps for ?-Ce1.30Lu0.70S3, ?-Pr1.29Lu0.71S3, and ?- Nd1.33Lu0.67S3 are 1.25 eV, 1.38 eV, and 1.50 eV, respectively. Crystallographic data: ?- 115 Ce1.30Lu0.70S3, monoclinic, space group P21/m, a = 11.0186(7), b = 3.9796(3), c = 21.6562(15) ?, ? = 101.6860(10), V = 929.93(11), Z = 8; ?-Pr1.29Lu0.71S3, monoclinic, space group P21/m, a = 10.9623(10), b = 3.9497(4), c = 21.5165(19) ?, ? = 101.579(2), V = 912.66(15), Z = 8; ?-Nd1.33Lu0.67S3, monoclinic, space group P21/m, a = 10.9553(7), b = 3.9419(3), c = 21.4920(15) ?, ? = 101.5080(10), V = 909.47(11), Z = 8. INTRODUCTION Ternary interlanthanide chalcogenides display a wide variety of structures that can possess both ordered and disordered Ln3+ sites.1-14 The ordering of two different Ln3+ cations over two or more crystallographic sites can be achieved by maximizing the difference in the size of the Ln3+ ions. Typical examples of ordered phases include ?- LnLn'S3 (GdFeO3-type)15,1-4 ?-LnLn'Q3 (Q = S, Se) (UFeS3-type)16,3,5,6 and ?-LnLn'S3.7 In these compounds, there are sharp demarcations in coordination numbers and bond distances between the larger ions and smaller ions, which inhibit the disordering of two lanthanides. In contrast, mixed site occupancies are found in disordered structures, e.g. F-Ln2S3 (CeYb3S6)9,17,18 CeTmS3 ,10 and Y5S7 (Sc2Er3S7)11.19 When the difference in size between the two Ln3+ ions is too small, disorder is unavoidable owing to the strong similarities in the structural chemistry of lanthanides. This is best represented by F-Ln2S3 type compounds, which contain an eight-coordinate environment for larger Ln3+ ion (A), a seven-coordinate intermediate site (B), and two six-coordinate octahedral sites for the smaller Ln3+ ion (C). In case of F-GdLu3S6, the position B is occupied by both metal ions.9 For F-Er3ScS6, both C sites are disordered.8 116 In addition to the remarkable structural flexibility of mixed-lanthanide sulfides that gives rise to a myriad of structure types, these compounds also display important physical properties including tunable band gaps. In an effort to understand the structure- property relationships in interlanthanide chalcogenides, we present the preparation, structure determination, magnetism, and optical properties of the partially disordered ?- Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) (CeTmS3-type) compounds. The information in this chapter has been published as a full paper in Journal of Solid-State Chemistry.20 EXPERIMENTAL Starting Materials. Ce (99.9%, Alfa-Aesar), Pr (99.9%, Alfa-Aesar), Nd (99.9%, Alfa-Aesar), Lu (99.9%, Alfa-Aesar), S (99.5%, Alfa-Aesar), and Sb (99.5%, Alfa-Aesar) were used as received. The Sb2S3 flux was prepared from the direct reaction of the elements in sealed fused-silica ampoules at 850 ?C. Syntheses of ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71). Reaction mixtures included 0.2000 g of Ln, Lu, S, and Sb2S3 in a ratio of 1:1:3:0.5 mmol. They were loaded into fused-silica ampoules in an argon-filled glovebox. The ampoules were sealed under vacuum and heated in the following profile using a programmable tube furnace.: 2 ?C/min to 500 ?C (held for 1 h), 0.5 ?C/min to 1000 ?C (held for 5 d), 0.04 ?C/min to 550 ?C (held for 2 d), and 0.5 ?C/min to 24 ?C. High yields of black crystals of ?-Ce1.30Lu0.70S3 and dark red crystals of ?-Pr1.29Lu0.71S3 and ?-Nd1.33Lu0.67S3 were isolated manually. Powder X-ray diffraction measurements were used to confirm phase purity by comparing the powder patterns calculated from the single crystal X-ray 117 structures with the experimental data. Semi-quantitative SEM/EDX analyses were performed using JEOL 840/Link Isis or JEOL JSM-7000F instruments. Ln, Ln', and S percentages were calibrated against standards. Sb was not detected in the crystals. The Ln:Ln':S ratios were determined to be approximately 2:1:4.5 from EDX analyses. Crystallographic Studies. Single crystals of ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) were mounted on glass fibers with epoxy and optically aligned on a Bruker APEX single crystal X-ray diffractometer using a digital camera. Initial intensity measurements were performed using graphite monochromated Mo K? (? = 0.71073 ?) radiation from a sealed tube and monocapillary collimator. SMART (v 5.624) was used for preliminary determination of the cell constants and data-collection control. The intensities of reflections of a sphere were collected by a combination of 3 sets of exposures (frames). Each set had a different ? angle for the crystal and each exposure covered a range of 0.3? in ?. A total of 1800 frames were collected with exposure times per frame of 10 or 20 seconds depending on the crystal. For ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71), determination of integrated intensities and global refinement were performed with the Bruker SAINT (v 6.02) software package using a narrow-frame integration algorithm. These data were treated first with a face-indexed numerical absorption correction using XPREP,21 followed by a semi-empirical absorption correction using SADABS.22 The program suite SHELXTL (v 6.12) was used for space group determination (XPREP), direct methods structure solution (XS), and least-squares refinement (XL).21 The final refinements included anisotropic displacement parameters for all atoms and secondary extinction. Some crystallographic details are given in Table 5.1. Atomic coordinates and equivalent isotropic displacement 118 Table 5.1. Crystallographic Data for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71). Formula ?-Ce1.30Lu0.70S3 ?-Pr1.29Lu0.71S3 ?-Nd1.33Lu0.67S3 Fw 400.86 402.31 405.13 Color Black Dark red Dark red Crystal system monoclinic monoclinic monoclinic Space group P21/m (No. 11) P21/m (No. 11) P21/m (No. 11) a (?) 11.0186(7) 10.9623(10) 10.9553(7) b (?) 3.9796(3) 3.9497(4) 3.9419(3) c (?) 21.6562(15) 21.5165(19) 21.4920(15) ? 101.6860(10) 101.579(2) 101.5080(10) V (?3) 929.93(11) 912.66(15) 909.47(11) Z 8 8 8 T (K) 193 193 193 ? (?) 0.71073 0.71073 0.71073 ?calcd (g cm?3) 5.726 5.856 5.918 ? (cm?1) 284.95 300.82 306.04 R(F)a 0.0288 0.0407 0.0300 Rw(Fo2) b 0.0656 0.0987 0.0762 a ( )R F F F F= ?? ? o c o for Fo 2 > 2?(Fo2). b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 119 Table 5.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?- Ce1.30Lu0.70S3. Atom (site) x y z Ueq (?2)a Ce1 0.19847(5) 0.25 0.76193(3) 0.00853(13) Ce2 0.58871(5) 0.25 0.86078(3) 0.00753(13) Ce3 0.80925(5) 0.25 0.72906(3) 0.00609(12) Ce4 0.69987(5) 0.25 0.53486(3) 0.00697(13) Ce5 0.02154(10) 0.25 0.09339(5) 0.0113(2) Lu1 0.94671(4) 0.25 0.41832(2) 0.00977(11) Lu2 0.52820(5) 0.25 0.35266(3) 0.00965(13) Ce/Lu 0.67261(5) 0.25 0.04880(2) 0.00988(18) Lu4 0.0588(5) 0.25 0.0804(2) 0.0176(10) S1 0.4135(2) 0.25 0.03938(12) 0.0088(5) S2 0.2948(2) 0.25 0.36659(11) 0.0069(5) S3 0.7532(2) 0.25 0.32707(12) 0.0094(5) S4 0.8554(2) 0.25 0.96392(12) 0.0107(5) S5 0.9936(2) 0.25 0.22765(12) 0.0089(5) S6 0.2529(2) 0.25 0.17836(12) 0.0080(5) S7 0.6208(2) 0.25 0.16832(12) 0.0080(5) S8 0.1437(2) 0.25 0.51372(12) 0.0089(5) S9 0.4433(2) 0.25 0.55520(12) 0.0109(5) S10 0.5364(2) 0.25 0.72589(12) 0.0083(5) S11 0.9553(2) 0.25 0.63164(12) 0.0104(5) S12 0.1680(2) 0.25 0.89765(12) 0.0117(5) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 120 Table 5.3. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?- Pr1.29Lu0.71S3. Atom (site) x y z Ueq (?2)a Pr1 0.19691(8) 0.25 0.76308(4) 0.0084(2) Pr2 0.58798(8) 0.25 0.86070(4) 0.0075(2) Pr3 0.80815(8) 0.25 0.72974(4) 0.00628(19) Pr4 0.69981(8) 0.25 0.53557(4) 0.00693(19) Pr5 0.0220(4) 0.25 0.09241(17) 0.0087(4) Lu1 0.94660(6) 0.25 0.41786(3) 0.00951(17) Lu2 0.52939(6) 0.25 0.35258(3) 0.01012(18) Pr/Lu 0.67394(7) 0.25 0.04788(3) 0.0094(3) Lu4 0.051(2) 0.25 0.0793(9) 0.015(3) S1 0.4140(4) 0.25 0.03935(19) 0.0092(8) S2 0.2947(3) 0.25 0.36625(18) 0.0071(7) S3 0.7549(3) 0.25 0.32577(19) 0.0090(8) S4 0.8567(4) 0.25 0.96457(19) 0.0113(8) S5 0.9947(3) 0.25 0.22627(19) 0.0095(8) S6 0.2531(3) 0.25 0.17760(18) 0.0076(7) S7 0.6222(4) 0.25 0.16747(19) 0.0090(7) S8 0.1442(3) 0.25 0.51375(19) 0.0086(7) S9 0.4424(4) 0.25 0.55476(19) 0.0107(8) S10 0.5369(4) 0.25 0.72688(18) 0.0092(7) S11 0.9543(4) 0.25 0.6323(2) 0.0111(8) S12 0.1657(4) 0.25 0.89822(19) 0.0112(8) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 121 Table 5.4. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?- Nd1.33Lu0.67S3. Atom (site) x y z Ueq (?2)a Nd1 0.19518(5) 0.25 0.76324(3) 0.00894(14) Nd2 0.58694(5) 0.25 0.85947(3) 0.00822(14) Nd3 0.80693(5) 0.25 0.72945(3) 0.00681(13) Nd4 0.70009(5) 0.25 0.53580(3) 0.00763(13) Nd5 0.0234(3) 0.25 0.09212(12) 0.0082(3) Lu1 0.94665(4) 0.25 0.41779(2) 0.01036(12) Lu2 0.53089(4) 0.25 0.35315(2) 0.01107(13) Nd/Lu 0.67466(5) 0.25 0.04720(2) 0.00911(19) Lu4 0.049(2) 0.25 0.0816(10) 0.013(3) S1 0.4146(2) 0.25 0.04108(12) 0.0094(5) S2 0.2955(2) 0.25 0.36661(12) 0.0069(5) S3 0.7555(2) 0.25 0.32486(13) 0.0100(5) S4 0.8566(3) 0.25 0.96404(13) 0.0110(5) S5 0.9960(2) 0.25 0.22621(13) 0.0094(5) S6 0.2545(2) 0.25 0.17833(12) 0.0087(5) S7 0.6232(2) 0.25 0.16816(12) 0.0091(5) S8 0.1452(2) 0.25 0.51344(12) 0.0091(5) S9 0.4426(2) 0.25 0.55385(13) 0.0115(5) S10 0.5367(2) 0.25 0.72630(12) 0.0090(5) S11 0.9535(3) 0.25 0.63255(13) 0.0106(5) S12 0.1628(2) 0.25 0.89773(12) 0.0103(5) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 122 parameters for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) are given in Tables 5.2- 5.4. The structure of CeTmS3 was previously determined to be ordered, with four eight- coordinate Ce3+ ions, two seven-coordinate Tm3+ and two six-coordinate Tm3+ ions.10 However, elemental analysis for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) indicated the ratio of Ln3+: Lu3+ is approximately 2:1. Considerable disordering of Ce/Lu in the Lu3+ positions should be present. Reexamining the Tm-S bond distances in CeTmS3, the average values of Tm(3)S7 and Tm(4)S7 are 2.77 and 2.86 ? respectively. Compared to Shannon?s data,23 in which TmS7 and CeS7 are 2.77 and 2.91 ?, Tm(4) site is more likely disordered. For ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71), the four eight-coordinate positions were assigned as Ln3+ and the Tm(1), Tm(2) and Tm(3) sites were assigned as Lu3+ at the beginning of the refinement cycles. Two large residual electron density peaks approximately 0.5 ? away from each other were found in the original place of Tm(4). Each of these sites was assigned as a Ln3+ position. The one with longer Ln-S bond distances was named as Ln(5) (Ln = Ce, Pr, or Nd), and the other one was assigned as Lu(4). It has to be mentioned, switching the Ln(5) and Lu(4) positions as well as disordering of Ln(5)/Lu(4) at one site were both tried, and gave poorer residuals. The refinement of occupancies of Ln(5) and Lu(4) showed 3:2 ratio of Ln3+:Lu3+, which requires more disordering in other Lu3+ sites. In the next step, Ln/Lu were both assigned at Lu(3) positions. This lowered the R1 value and the weighting scheme to an even greater extent. The final refinements gave rise to formulas of ?- Ce1.30Lu0.70S3, ?-Pr1.29Lu0.71S3, and ?-Nd1.33Lu0.67S3, which are consistent with the Ln:Lu ratios from calibrated EDX results. The standard deviation on the composition from the refinements is 0.01. 123 Powder X-ray Diffraction. Powder X-ray diffraction patterns were collected with a Rigaku Miniflex powder X-ray diffractometer using Cu K? (? = 1.54056 ?) radiation. Magnetic Susceptibility Measurements. Magnetism data were measured on powders in gelcap sample holders with a Quantum Design MPMS 7T magnetometer/susceptometer between 2 and 300 K and in applied fields up to 7 T. DC susceptibility measurements were made under zero-field-cooled conditions with an applied field of 0.1 T. Susceptibility values were corrected for the sample diamagnetic contribution according to Pascal?s constants24 as well as for the sample holder diamagnetism. ?p values were obtained from extrapolations from fits between 100 and 300 K. In addition, ZFC and FC data were collected as follows: The samples were first zero field cooled from room temperature and the susceptibility was measured at 100 Oe to 300 K. Then the sample was field cooled at 100 Oe to 2 K and the same measurement was done by increasing with increasing temperatures. There are no differences between ZFC and FC data at the measured temperatures. (Magnetic susceptibility measurements were performed by Eun Sang Choi and James S. Brooks at Florida State University) UV-vis-NIR Diffuse Reflectance Spectroscopy. The diffuse reflectance spectra for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) were measured from 200 to 1500 nm using a Shimadzu UV3100 spectrophotometer equipped with an integrating sphere attachment. The Kubelka-Munk function was used to convert diffuse reflectance data to absorption spectra.25 124 RESULTS AND DISCUSSION Structures of ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71). The ?-Ln2- xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) series are isotypic with CeTmS3,10 which crystallizes in P21/m space group with a very complex three-dimensional structure. A view of the unit cell is illustrated in Figure 5.1. There are nine crystallographically unique lanthanides sites and twelve sulfide positions. Ln(1) are coordinated to nine S atoms in a tricapped trigonal prismatic environment with two long capping Ln???S contacts. For instance, in ?-Ce1.30Lu0.70S3 the short Ce-S bond distances range from 2.8940(19) ? to 3.025(3) ?, while the longer contacts are 3.476(3) ? and 3.957(3) ?. The longest of these can probably be disregarded. Ln(2), Ln(3), and Ln(4) have similar coordination geometries. All of them bond to eight S atoms and occur as bicapped trigonal prisms. Ln(5) and Ln/Lu sites are seven-coordinate in a monocapped trigonal prismatic arrangement. It is worth noting that Ln/LuS7 has intermediate bond distances. For example, the average value of Ce/LuS7 is 2.81 ?, which is between 2.91 ? for CeS7 and 2.75 ? for LuS7 according to the radii reported by Shannon.23 Compared to larger lanthanides, Lu3+ ions have fewer S neighbors. Both Lu(1) and Lu(2) atoms are bound to six S atoms in octahedral environments. Seven-coordinate Lu(4) atoms are found to have a highly distorted monocapped trigonal prismatic geometry, with two short bonds and three long contacts. In case of ?-Ce1.30Lu0.70S3, Lu(4)-S bond distances are 2.478(3) ? ?2, 2.690(6) ?, 3.016(6) ?, 3.304(4) ? ?2, and 3.408(5) ?. The selected bond distances are listed in Table 5.5. 125 Figure 5.1. A view down the b axis shows the complex three-dimensional structure of ?- Ce1.30Lu0.70S3. a c 126 Table 5.5. Selected Bond Distances (?) for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71). Formula ?- Ce1.30Lu0.70S3 ?-Pr1.29Lu0.71S3 ?-Nd1.33Lu0.67S3 Ln(1)-S(3) ?2 2.8940(19) 2.870(3) 2.858(2) Ln(1)-S(5) ?2 2.9466(19) 2.925(3) 2.918(2) Ln(1)-S(7) ?2 2.9978(19) 2.980(3) 2.974(2) Ln(1)-S(10) 3.957(3) 3.957(4) 3.976(3) Ln(1)-S(11) 3.476(3) 3.462(4) 3.454(3) Ln(1)-S(12) 3.025(3) 2.995(4) 2.981(3) Ln(2)-S(1) ?2 2.9421(19) 2.923(3) 2.910(2) Ln(2)-S(4) 3.312(3) 3.321(4) 3.336(3) Ln(2)-S(6) ?2 2.8850(18) 2.862(3) 2.8507(19) Ln(2)-S(7) ?2 3.0162(19) 3.002(3) 2.998(2) Ln(2)-S(10) 2.861(3) 2.821(4) 2.805(3) Ln(3)-S(2) ?2 2.9331(18) 2.916(3) 2.9122(19) Ln(3)-S(5) ?2 2.9551(19) 2.941(3) 2.938(2) Ln(3)-S(6) ?2 3.0007(19) 2.977(3) 2.966(2) Ln(3)-S(10) 2.993(3) 2.962(4) 2.948(3) Ln(3)-S(11) 2.901(3) 2.882(4) 2.872(3) Ln(4)-S(2) ?2 2.9098(18) 2.884(3) 2.8715(19) Ln(4)-S(8) ?2 2.9615(19) 2.946(3) 2.932(2) Ln(4)-S(9) ?2 3.001(2) 2.982(3) 2.973(2) Ln(4)-S(9) 2.948(3) 2.933(4) 2.923(3) 127 Ln(4)-S(11) 3.150(3) 3.128(4) 3.119(3) Ln(5)-S(4) ?2 2.832(2) 2.798(5) 2.776(3) Ln(5)-S(4) 3.022(3) 2.975(5) 2.985(4) Ln(5)-S(5) 2.985(3) 2.955(6) 2.957(4) Ln(5)-S(6) 2.825(3) 2.813(4) 2.823(4) Ln(5)-S(12) ?2 2.920(2) 2.888(4) 2.876(3) Lu(1)-S(3) 2.597(3) 2.583(4) 2.590(3) Lu(1)-S(8) ?2 2.7771(18) 2.765(3) 2.7702(19) Lu(1)-S(8) 2.677(3) 2.675(4) 2.678(3) Lu(1)-S(11) ?2 2.6020(17) 2.592(3) 2.5933(17) Lu(2)-S(2) 2.649(2) 2.647(4) 2.652(3) Lu(2)-S(3) 2.647(3) 2.649(4) 2.649(3) Lu(2)-S(9) ?2 2.7904(19) 2.779(3) 2.7808(19) Lu(2)-S(10) ?2 2.6210(16) 2.618(3) 2.6160(17) Ln/Lu-S(1) ?2 2.7880(18) 2.762(3) 2.7749(19) Ln/Lu-S(1) 2.820(3) 2.819(4) 2.826(3) Ln/Lu-S(4) 2.989(3) 2.944(4) 2.930(3) Ln/Lu-S(7) 2.761(3) 2.744(4) 2.768(3) Ln/Lu-S(12) ?2 2.7500(18) 2.742(3) 2.7595(19) Lu(4)-S(4) ?2 2.478(3) 2.487(16) 2.513(16) Lu(4)-S(4) 3.016(6) 2.921(15) 2.951(18) Lu(4)-S(5) 3.408(5) 3.34(2) 3.27(2) Lu(4)-S(6) 2.690(6) 2.736(15) 2.741(18) Lu(4)-S(12) ?2 3.304(4) 3.20(2) 3.145(19) 128 The complex three-dimensional structure of ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) is constructed from one-dimensional chains of LnSn (n = 6 ? 9) polyhedra that extend along the b axis. Chains constructed from LnS9 or LnS8 polyhedra share opposite trigonal faces with two neighbors along the direction of chain propagation. The small six- and seven coordinate Ln3+-containing units only share edges within the chains. Each LnSx or LuSy polyhedron shares edges and corners with others in the [ac] plane. They are normal compared to Shannon?s data.23 Because there are no S?S bonds in ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71), the oxidation states in these compounds can be assigned as +3/+3/-2. This designation is confirmed by both bond-valence sum calculations27, 28 and by magnetic susceptibility measurements (vide infra). Magnetic Susceptibility. The inverse molar Ln magnetic susceptibilities for ?- Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) in the range of 2-300 K are shown in Figures 5.2-5.4. All three compounds show a deviation from the Curie-Weiss law near 100 K. Magnetic parameters, which are presented in Table 5.6, were determined from the fit from the Curie-Weiss regions. These compounds do not show evidence of long- range magnetic ordering down to 5 K. ?-Ce1.30Lu0.70S3 has very similar magnetic behavior with ?-LnLn'S36 and ?-LnLn'S37 with a large negative value of ?p of -34(1) K. The 1/? data for ?-Pr1.29Lu0.71S3 show a positive, rather than negative, deviation from Curie-Weiss behavior at 120 K. The gradual change in the slope and the negative value of ?p (-6.0(5) K) may indicate short-range antiferromagnetic ordering. ?- Nd1.33Lu0.67S3 acts like an intermediate state of ?-Ce1.30Lu0.70S3 and ?-Pr1.29Lu0.71S3. The curvature of the plot starts as upward at 120 K. At lower temperature, it becomes 129 Figure 5.2. A plot of inverse molar cerium magnetic susceptibility for ?-Ce1.30Lu0.70S3 between 2 and 300 K. Data were taken under an applied magnetic field of 0.1 T. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. 0 50 100 150 200 250 300 0 100 200 300 400 ?-Ce1.30Lu0.70S3 ?-1 (O e m ol Ce /em u) Temperature, T (K) 130 Figure 5.3. Temperature dependence of the reciprocal molar praseodymium magnetic susceptibility for ?-Pr1.29Lu0.71S3 under an applied magnetic field of 0.1 T between 2 and 300 K. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. 0 50 100 150 200 250 300 0 50 100 150 200 ?-Pr1.29Lu0.71S3 ?-1 (O e m ol Pr/ em u) Temperature, T (K) 131 Figure 5.4. Inverse molar neodymium magnetic susceptibility vs. T for ?-Nd1.33Lu0.67S3 under an applied magnetic field of 0.1 T between 2 and 300 K. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. 0 50 100 150 200 250 300 0 50 100 150 200 ?-Nd1.33Lu0.67S3 ?-1 (O e m ol Nd /em u) Temperature, T (K) 132 Table 5.6. Magnetic Parameters for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71). Formula Pcal/?B Peff/?B ?p/K R2 ?-Ce1.30Lu0.70S3 2.54 2.438(5) -34(1) 0.99952 ?-Pr1.29Lu0.71S3 3.58 3.583(4) -6.0(5) 0.99989 ?-Nd1.33Lu0.67S3 3.62 3.413(8) -8(1) 0.99959 a Pcal and Peff : calculated26 and experimental effective magnetic moments per Ln ion. b Weiss constant (?p) and goodness of fit (R2) obtained from high temperature (100-300 K) data. 133 negative. Crystal-field effects and short-range ordering may both contribute to this behavior. The ?p of ?-Nd1.33Lu0.67S3 is -8(1) K. The experimental effective magnetic moments per Ln ion based on the formulas proposed are very close to the theoretical value of the free Ln3+ ions as shown in Table 5.6. As a reference, the experimental moments using the formulas as LnLuS3 (Ln = Ce, Pr, Nd) are 2.82 ?B for CeLuS3, 4.05 ?B for PrLuS3, and 3.98 ?B for NdLuS3, which are larger than the accepted values.26 This provides further supporting evidence for the disorder refinements. Optical Properties. The UV-vis-NIR diffuse reflectance spectra are presented in Figure 5.5 for ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71). The measured band gaps for ?-Ce1.30Lu0.70S3, ?-Pr1.29Lu0.71S3, and ?-Nd1.33Lu0.67S3 are 1.25 eV, 1.38 eV, and 1.50 eV respectively. They are consistent with the observed colors. ?-Ce1.30Lu0.70S3 is black, while both ?-Pr1.29Lu0.71S3 and ?-Nd1.33Lu0.67S3 are dark red. In addition to the observed band gap for ?-Nd1.33Lu0.67S3, f-f transitions are also apparent in the spectrum of this compound. The band gap results are comparable to the value reported for ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb).7 ?-Ce1.30Lu0.70S3 has smaller band gap due to the enhanced energy of the 4f1 electron. Much like other mixed-lanthanide chalcogenides, the electronic structures of ?-LnLuS3 are tunable based on the choice of lanthanide. Conclusions ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71) were prepared using a Sb2S3 flux and their structures determined by single crystal X-ray diffraction. These compounds crystallize in the disordered CeTmS3 structure-type. EDX analyses and magnetic 134 measurements support the proposed formulas as ?-Ce1.30Lu0.70S3, ?-Pr1.29Lu0.71S3, and ?- Nd1.33Lu0.67S3. The UV-vis-NIR diffuse reflectance measurements show these compounds to be wide band-gap semiconductors. 135 Figure 5.5. UV-vis diffuse reflectance spectra of ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67 ? 0.71). 1.0 1.5 2.0 2.5 3.0 3.5 0.0 0.5 1.0 1.5 2.0 2.5 ?-Pr1.29Lu0.71S3 ?-Nd1.33Lu0.67S3 ?-Ce1.30Lu0.70S3 ?/ s ( arb ita ry un its ) Energy (eV) 136 REFERENCES 1. Rodier, N.; Laruelle, P. C. R. Seances Acad. Sci. Ser. 1970, C270, 2127. 2. Ijdo, D.J.W. Acta Crystallogr. 1980, B36, 2403. 3. Rodier, N.; Julien, R.; Tien, V. Acta Crystallogr. 1983, C39, 670. 4. Range, K.-J.; Gietl, A.; Klement, U. Z. Kristallogr. 1993, 207, 147. 5. Carr?, D.; Laruelle, P. Acta Crystallogr. 1974, B30, 952. 6. Mitchell, K.; Somers, R. C.; Huang, F. Q.; Ibers, J. A. J. Solid State Chem. 2004, 177, 709. 7. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. Chem. Mater. 2007, 19, 567. 8. Rodier, N.; Laruelle, P. Bull. Soc. fr. de Mineral. Cristallogr. 1973, 96, 30. 9. Rodier, N.; Firor, R. L.; Tien, V.; Guittard, M. Mat. Res. Bull. 1976, 11, 1209. 10. Rodier, N. Bull. Soc. fr. de Mineral. Cristallogr. 1973, 96, 350. 11. Rodier, N.; Laruelle, P. Bull. Soc. fr. de Mineral. Cristallogr. 1972, 95, 548. 12. Carr?, D.; Laruelle, P. Acta Crystallogr. 1973, B29, 70. 13. Rodier, N.; Tien, V. Bull. Soc. fr. de Mineral. Cristallogr. 1975, 98, 30. 14. Vovan, T.; Guittard, M.; Rodier, N. Mat. Res. Bull. 1979, 14, 597. 15. Marezio, M.; Remeika, J.P.; Dernier, P.D. Acta Crystallogr. 1970, B26, 2008. 16. No?l, H.; Padiou, J. Acta Crystallogr. 1976, B32, 1593. 17. Schleid, T.; Lissner, F. J. Alloys Compd. 1992, 189, 69. 18. Fang, C. M.; Meetsma, A.; Wiegers, G. A. J. Alloys Compd. 1993, 201, 255. 19. Adolphe, C. Annales de Chimie (Paris) 1965, 271. 137 20. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. J. Solid State Chem. in press. 21. Sheldrick, G. M. SHELXTL PC, Version 6.12, An Integrated System for Solving, Refining, and Displaying Crystal Structures from Diffraction Data; Siemens Analytical X-Ray Instruments, Inc.: Madison, WI 2001. 22. Sheldrick, G. M. SADABS 2001, Program for absorption correction using SMART CCD based on the method of Blessing: Blessing, R. H. Acta Crystallogr. 1995, A51, 33. 23. Shannon, R. D. Acta Crystallogr. 1976, A32, 751. 24. Mulay, L. N.; Boudreaux, E. A. Theory and Applications of Molecular Diamagnetism; Wiley?Interscience: New York, 1976. 25. Wendlandt, W. W.; Hecht, H. G. Reflectance Spectroscopy; Interscience Publishers: New York, 1966. 26. Kittel, C. Introduction to Solid State Physics, 6th Edition; Wiley: New York, 1986. 27. Brown, I. D.; Altermatt, D. Acta Crystallogr. 1985, B41, 244. 28. Brese, N. E.; O?Keeffe, M. Acta Crystallogr. 1991, B47, 192. 138 CHAPTER 6 SYNTHESES, STRUCTURE, MAGNETISM, AND OPTICAL PROPERTIES OF LUTETIUM-BASED INTERLANTHANIDE SELENIDES ABSTRACT Ln3LuSe6 (Ln = La, Ce), ?-LnLuSe3 (Ln = Pr, Nd), and LnxLu4-xSe6 (Ln = Sm, Gd; x = 1.82, 1.87) have been synthesized using a Sb2Se3 flux at 1000 ?C. Ln3LuSe6 (Ln = La, Ce) adopt the U3ScS6-type three-dimensional structure, which is constructed from two-dimensional 2? [Ln3Se6]3- slabs with the gaps between these slabs filled by octahedrally coordinated Lu3+ ions. The series of ?-LnLuSe3 (Ln = Pr, Nd) are isotypic with UFeS3. Their structures include layers formed from LuSe6 octahedra that are separated by eight-coordinate larger Ln3+ ions in bicapped trigonal prismatic environments. Sm1.82Lu2.18Se6 and Gd1.87Lu2.13Se6 crystallize in the disordered F-Ln2S3 type structure with the eight-coordinate bicapped trigonal prismatic Ln(1) ions residing in the one-dimensional channels formed by three different double chains via edge and corner sharing. These double chains are constructed from Ln(2)Se7 monocapped trigonal prisms, Ln(3)Se6 octahedra, and Ln(4)S6 octahedra, respectively. The magnetic susceptibilities of ?-PrLuSe3 and ?-NdLuSe3 follow the Curie-Weiss law. Sm1.82Lu2.18Se6 shows van Vleck paramagnetism. Magnetic measurements show that 139 Gd1.87Lu2.13Se6 undergoes an antiferromagnetic transition around 4 K. Ce3LuSe6 exhibits ferromagnetic ordering below 5 K. The optical band gaps for La3LuSe6, Ce3LuSe6, ?- PrLuSe3, ?-NdLuSe3, Sm1.82Lu2.18Se6, and Gd1.87Lu2.13Se6 are 1.26, 1.10, 1.56, 1.61, 1.51, and 1.56 eV, respectively. INTRODUCTION Interlanthanide chalcogenides have been the focus of intense interest owing to their remarkably complex structures, tunable band gaps, and in some cases atypical magnetism.1-21 New developments in this area have been aided by the use of fluxes for the synthesis and crystal growth of new compounds. These fluxes have included a variety of alkali metal halides as well as Sb2Q3 (Q = S, Se).14,15,20-22 Ternary interlanthanide chalcogenides usually include one large ion (Ln) and one small ion (Ln') from opposite ends of the lanthanide series with different coordination environments to avoid possible disordering. Ln/Yb/Q (Q = S, Se) phases have been extensively studied owing to the potential mixed-valency of Yb (II, III) ions, which might lead to interesting structures and electronic properties. The group of Ln/Yb/Q (Q = S, Se) is represented by ?-LaYbS312 (GdFeO3 type [19]),23 ?-LnYbQ3 (Q = S, Se)12-14 (UFeS3 type),24 ?-LnYbS3 (Ln = La, Ce),15 and LnYb3S616,17 (F-Ln2S3 type).25,26 There are some Er or Tm containing ternary compounds prepared and characterized, including CeTmS3,18 La10Er9S27,19 ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm),15 SmEr3Q6 (Q = S, Se)20 (F-Ln2S3 type),25,26 and Sm0.88Er1.12Se320 (U2S3 type).27 Both Er3+ and Tm3+ ions are paramagnetic. Their substitutions in smaller Ln' sites may result in different magnetic performance from 140 the parent compounds. In contrast, Ln/Lu/Q phases are much simpler systems in term of magnetism and less investigated, because the Lu3+ ion is 4f14. The full 4f shell of the Lu3+ ion can be advantageous when the magnetic behavior of other Ln ions needs to be probed without the interference from the Ln' ions. Recently, we reported the synthesis and characterization of interlanthanide sulfides ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd).21 These compounds crystallized in the disordered CeTmS3 structure-type with tunable band gaps. The magnetic susceptibility of ?-Ce1.30Lu0.70S3 deviates from the Curie-Weiss law at low temperature, due to the crystal- field effects. ?-Pr1.29Lu0.71S3 shows some short-range antiferromagnetic ordering. While ?-Nd1.33Lu0.67S3 acts like an intermediate state of ?-Ce1.30Lu0.70S3 and ?-Pr1.29Lu0.71S3. In the present study, we disclose the synthesis, structure, optical, and magnetic properties of Ln3LuSe6 (Ln = La, Ce), ?-LnLuSe3 (Ln = Pr, Nd), and LnxLu4-xSe6 (Ln = Sm, Gd; x = 1.82, 1.87). The information in this chapter has been submitted as a full paper in Inorganic Chemistry.28 EXPERIMENTAL Starting Materials. La (99.9%, Alfa-Aesar), Ce (99.9%, Alfa-Aesar), Pr (99.9%, Alfa-Aesar), Nd (99.9%, Alfa-Aesar), Sm (99.9%, Alfa-Aesar), Gd (99.9%, Alfa-Aesar), Lu (99.9%, Alfa-Aesar), Se (99.5%, Alfa-Aesar), and Sb (99.5%, Alfa-Aesar) were used as received. The Sb2Se3 flux was prepared from the direct reaction of the elements in sealed fused-silica ampoules at 850 ?C. Syntheses. Ln3LuSe6 (Ln = La, Ce) were synthesized from a reaction of 150 mg of stoichiometric Ln, Lu, and Se, and 100 mg of Sb2Se3. For ?-LnLuSe3 (Ln = Pr, Nd) 141 and LnxLu4-xSe6 (Ln = Sm, Gd; x = 1.82, 1.87), the reaction mixtures include 150 mg of Ln, Lu, and Se in a molar ratio of 1:1:3, and 100 mg of Sb2Se3. All these starting materials were loaded into fused-silica ampoules under argon atmosphere in a glovebox. The ampoules were flame sealed under vacuum and heated in programmable tube furnaces. The following heating profile was used for all reactions: 2 ?C/min to 500 ?C (held for 1 h), 0.5 ?C/min to 1000 ?C (held for 14 d), 0.04 ?C/min to 550 ?C (held for 2 d), and 0.5 ?C/min to 24 ?C. Major title products were found as large black chunks that were well separated from the Sb2Se3 flux. They were isolated and ground for powder X-ray diffraction measurements, which were used to confirm phase purity by comparing the powder patterns calculated from the single crystal X-ray structures with the experimental data. Semi-quantitative SEM/EDX analyses were performed on several single crystals for each compound using JEOL 840/Link Isis or JEOL JSM-7000F instruments. Ln, Lu, and Se percentages were calibrated against standards. Sb was not detected in the crystals. The Ln:Lu:Se ratios for Ln3LuSe6 (Ln = La, Ce) were determined to be approximately 3:1:6 from EDX analyses. While the ratios of Ln:Lu:Se were close to 1:1:3 for ?- LnLuSe3 (Ln = Pr, Nd) and LnxLu4-xSe6 (Ln = Sm, Gd; x = 1.82, 1.87). Crystallographic Studies. Single crystals of LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd) were mounted on glass fibers with epoxy and optically aligned on a Bruker APEX single crystal X-ray diffractometer using a digital camera. Initial intensity measurements were performed using graphite monochromated Mo K? (? = 0.71073 ?) radiation from a sealed tube and monocapillary collimator. SMART (v 5.624) was used for preliminary determination of the cell constants and data collection control. The intensities of reflections of a sphere were collected by a combination of 3 sets of 142 exposures (frames). Each set had a different ? angle for the crystal and each exposure covered a range of 0.3? in ?. A total of 1800 frames were collected with exposure times per frame of 10 or 20 seconds depending on the crystal. For LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd), determination of integrated intensities and global refinement were performed with the Bruker SAINT (v 6.02) software package using a narrow-frame integration algorithm. These data were treated first with a face-index numerical absorption correction using XPREP,29 followed by a semi-empirical absorption correction using SADABS.30P The program suite SHELXTL (v 6.12) was used for space group determination (XPREP), direct methods structure solution (XS), and least-squares refinement (XL).29 The final refinements included anisotropic displacement parameters for all atoms and secondary extinction. Some crystallographic details are given in Table 6.1. Atomic coordinates and equivalent isotropic displacement parameters for La3LuSe6, ?-PrLuSe3, and Sm1.82Lu2.18Se6 are given in Tables 6.2-6.4. . The structures of Ln3LuSe6 (Ln = La, Ce) and ?-LnLuSe3 (Ln = Pr, Nd) are ordered. For these compounds, the assignments of cation positions were straightforward. LnxLu4-xSe6 (Ln = Sm, Gd; x = 1.82, 1.87) compounds crystallize in the F-Ln2S325,26 type structure, which is highly disordered. All four cation sites, including one eight- coordinate position (Ln (1)), one seven-coordinate position (Ln (2)), and two octahedral positions (Ln (3) and Ln (4)), were assumed to be occupied by both metals at the beginning of the refinements. The final refinements shown that the occupancies of Lu atoms in Ln (1), Ln (2), Ln (3), and Ln (4) positions are 0.02, 0.32, 0.88, and 0.95 respectively for the Sm-based compound and 0.04, 0.40, 0.84 and 0.86 for the Gd case. 14 3 Table 6.1. Crystallographic Data for LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd). Formula La3LuSe6 Ce3LuSe6 ?-PrLuSe3 ?-NdLuSe3 Sm1.82Lu2.18Se6 Gd1.87Lu2.13Se6 fw 1065.46 1069.09 552.76 556.09 1128.83 1140.50 Color black black black black black black Crystal System orthorhombic orthorhombic orthorhombic orthorhombic monoclinic monoclinic Space group Pnnm (No. 58) Pnnm (No. 58) Cmcm (No. 63) Cmcm (No. 63) P21/m (No. 11) P21/m (No. 11) a (?) 14.6195(10) 14.5020(9) 4.0052(10) 3.9946(5) 11.3925(13) 11.4274(12) b (?) 17.5736(12) 17.4954(11) 12.996(3) 13.0015(17) 4.0483(5) 4.0542(4) c (?) 4.1542(3) 4.1129(3) 9.865(3) 9.8583(13) 11.6844(14) 11.7160(12) ? / / / / 108.915(2) 109.005(2) V (?3) 1067.29(13) 1043.52(12) 513.5(2) 512.00(11) 509.79(11) 513.20(9) Z 4 4 4 4 2 2 T (K) 193 193 193 193 193 193 ? (?) 0.71073 0.71073 0.71073 0.71073 0.71073 0.71073 ?calcd (g cm?3) 6.631 6.805 7.150 7.214 7.354 7.381 ? (cm?1) 413.24 430.70 495.66 503.36 525.94 534.34 R(F)a 0.0273 0.0212 0.0429 0.0226 0.0300 0.0308 Rw(Fo2) b 0.0645 0.0485 0.1049 0.0601 0.0822 0.0817 a ( )R F F F F= ?? ? o c o for Fo 2 > 2?(Fo2). b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 144 Table 6.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for La3LuSe6. Atom (site) x y z Ueq (?2)a La1 0.04906(4) 0.21859(3) 0 0.00650(14) La2 0.74828(4) 0.59977(3) 0 0.00737(14) La3 0.18086(3) 0.65120(3) 0 0.00551(14) Lu1 0.5 0.5 0 0.00765(14) Lu2 0.5 0 0 0.00796(15) Se1 0.09090(6) 0.92147(5) 0 0.0070(2) Se2 0.37451(6) 0.60983(5) 0 0.0088(2) Se3 0.39505(7) 0.25435(5) 0 0.0079(2) Se4 0.31246(6) 0.02363(5) 0 0.0075(2) Se5 0.02329(6) 0.39487(5) 0 0.0077(2) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 145 Table 6.3. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for ?- PrLuSe3. Atom (site) x y Z Ueq (?2)a Pr1 0 0.74823(8) 0.25 0.0104(4) Lu1 0 0 0 0.0094(4) Se1 0 0.35622(11) 0.06234(15) 0.0098(4) Se2 0 0.08646(16) 0.25 0.0103(5) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 146 Table 6.4. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for Sm1.82Lu2.18Se6. Atom (site) x y Z Ueq (?2)a Sm1/Lu1 0.54854(5) 0.25 0.19591(5) 0.00802(16) Sm2/Lu2 0.18189(5) 0.25 0.00181(5) 0.00928(16) Sm3/Lu3 0.93774(4) 0.25 0.33530(4) 0.00899(15) Sm4/Lu4 0.65835(4) 0.25 0.58350(4) 0.00947(15) Se1 0.41771(9) 0.25 0.59567(9) 0.0079(2) Se2 0.89302(10) 0.25 0.56010(10) 0.0095(2) Se3 0.23926(9) 0.25 0.77152(9) 0.0074(2) Se4 0.30802(10) 0.25 0.25392(9) 0.0093(2) Se5 0.98403(10) 0.25 0.11953(10) 0.0104(2) Se6 0.61334(9) 0.25 0.96386(9) 0.0070(2) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 147 These results gave rise to the formula of Sm1.82Lu2.18Se6 and Gd1.87Lu2.13Se6, which are consistent with the 1:1:3 ratios from EDX analysis. It is worth noting that Gd1.87Lu2.13Se6 has a higher degree of disorder than Sm1.82Lu2.18Se6 as the size of the cations get closer, even though they have similar Ln:Lu ratios. Powder X-ray Diffraction. Powder X-ray diffraction patterns were collected with a Rigaku Miniflex powder X-ray diffractometer using Cu K? (? = 1.54056 ?) radiation. Magnetic Susceptibility Measurements. Magnetism data were measured on powders in gelcap sample holders with a Quantum Design MPMS 7T magnetometer/susceptometer between 2 and 300 K and in applied fields up to 7 T. DC susceptibility measurements were made under zero-field-cooled conditions with an applied field of 0.1 T. Magnetic susceptibility for Ce3LuSe6 under zero-field-cooled (ZFC) and field-cooled (FC) conditions were measured with 0.01 T applied field between 2 and 25 K. Susceptibility values were corrected for the sample diamagnetic contribution according to Pascal?s constants31 as well as for the sample holder diamagnetism. Experimental effective magnetic moments and Weiss constants for Ce3LuSe6, ?-PrLuSe3, ?-NdLuSe3, and Gd1.87Lu2.13Se6 were obtained from extrapolations from fits between 100 and 300 K. (Magnetic susceptibility measurements were performed by Eun Sang Choi and James S. Brooks at Florida State University) UV-vis-NIR Diffuse Reflectance Spectroscopy. The diffuse reflectance spectra for LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd) were measured from 200 to 2500 nm using a Shimadzu UV3100 spectrophotometer equipped with an integrating sphere attachment. 148 The Kubelka-Munk function was used to convert diffuse reflectance data to absorption spectra.32 RESULTS AND DISCUSSION Synthesis of Ln/Ln'/Q using Sb2Q3 fluxes (Q = S, Se). The proven method of using Sb2Q3 (Q = S, Se) fluxes to prepare ternary interlanthanide chalcogenides has been very effective in this present study. Eight different structure types have been identified for ternary interlanthanide chalcogenides prepared by employing these fluxes at 1000 ?C. These compounds are given in Tables 6.5 and 6.6. The structures of Ln/Ln'/Q phases depend highly on the choices of Ln and Ln'. This is especially true for La/Ln'/Se, which can adopt five different structures with the variation of Ln'. It is important to note that the choices of flux and temperature could be critical as well; these will not be discussed here. Ordered phases can usually be found in bottom left corner of Tables 6.5 and 6.6, where Ln and Ln' ions have larger difference in size. When it reaches to the opposite corner, disordered compounds are often present owing to the similar structural chemistry of lanthanides. There are some exceptions, e.g. LaLu3S6 and LaxYb5-xSe7. The limitations of using Sb2Q3 (Q = S, Se) fluxes to prepare LnLn'Q include: 1) Attempts to prepare interlanthanide tellurides have not succeeded; 2) It is difficult to achieve high yield and high-quality single crystals when the ionic radii of the two Ln3+ ions approach equality; 3) Occasionally, distinguishing and separating products from Sb2Q3 (Q = S, Se) fluxes proves tricky. Structures of LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd). The Ln3LuSe6 (Ln = La, Ce) compounds adopt U3ScS633 type structure. The unit cell of La3LuSe6, projected 149 Table 6.5. Ternary Interlanthanide Sulfides Prepared Using Sb2S3 flux at 1000 ?C. Eu2+ La3+ Ce3+ Pr3+ Nd3+ Sm3+ Gd3+ Tb3+ ? Dy3+ ? ?? Ho3+ ? ?? ? Er3+ ? ? ? Tm3+ ? ? ? ? Yb3+ ? ? ? ? ? ? ? Lu3+ ? ? ? ? ? ? Ordered structure types: ? ?-LnLn?S3,12-14 ? ?- LnLn?S3,15 ? U3ScS6,33 ? CaFe2O4.34 Disordered structure types: ? F-Ln2S3,25,26 ? ?-LnLn?S3,18,21 ? U2S3,27 ? Y5S7.35 150 Table 6.6. Ternary Interlanthanide Selenides Prepared Using Sb2Se3 flux at 1000 ?C. Eu2+ La3+ Ce3+ Pr3+ Nd3+ Sm3+ Gd3+ Tb3+ ? ? Dy3+ ? ? Ho3+ ? Er3+ ? ? ? Tm3+ ? ? ? ? ? ? ? Yb3+ ? ? ? ? ? ? ? Lu3+ ? ? ? ? ? ? ? Ordered structure types: ? ?-LnLn?S3,12-14 ? ?- LnLn?S3,15 ? U3ScS6,33 ? CaFe2O4.34 Disordered structure types: ? F-Ln2S3,25,26 ? ?-LnLn?S3,18,21 ? U2S3,27 ? Y5S7.35 151 Figure 6.1. An illustration of the three-dimensional structure of La3LuSe6 down the c axis. a b 152 along the c axis, is shown is Figure 6.1. There are three crystallographically unique Ln sites (4g) and two octahedral Lu positions (2d, 2b) in the structure. Both Ln(1) and Ln(2) atoms are surrounded by eight Se atoms and occur as bicapped trigonal prisms. Ln(3) sites are seven-coordinate in a monocapped trigonal prismatic environment. LnSe8 and LnSe7 polyhedra share faces or edges with each other to form two-dimensional slabs extending in the [ac] plane. Furthermore these slabs connect at Se(4) positions to produce a three-dimensional structure. The gaps between these slabs are filled by isolated one-dimensional edge-sharing LuSe6 octahedral chains running down the c axis. The bond distances for these two compounds, which can be found in Supporting Information, are normal compared to average values reported by Shannon.36 In the case of La3LuSe6, the bond distances for the LaSe8, LaSe7, and LuSe6 polyhedra range from 2.9982(8) to 3.3408(11) ?, 2.9229(11) to 3.1116(11) ?, and 2.6629(9) to 2.8257(6) ?, respectively (see Table 6.7). The series of ?-LnLuSe3 (Ln = Pr, Nd) are isotypic with UFeS3.24 The structure, as shown in Figure 6.2, is constructed from two-dimensional LuSe6 octahedra layers, which are separated by Ln3+ ions. The local environment of Ln3+ ions can be found in Figure 6.5. They coordinate to eight Se atoms with a bicapped trigonal prismatic geometry. The connectivities within LuSe6 layers are illustrated in Figure 6.3. The LuSe6 octahedral units are edge sharing along a axis and corner sharing along c axis. The bond lengths within these two compounds are regular. For example, Pr-Se and Lu-Se distances in compound ?-PrLuSe3 are in the range of 2.9035(18) and 3.3670(17) ?, and 2.7102(10) and 2.8072(11) ?, respectively (see Table 6.8). 153 Table 6.7. Selected Bond Distances (?) for Ln3LuSe6 (Ln = La, Ce). Formula La3LuSe6 Ce3LuSe6 Ln(1)-Se(1) 3.2007(11) 3.009(2) Ln(1)-Se(2) ?2 3.0357(8) 3.0044(7) Ln(1)-Se(3) ?2 3.1001(8) 3.0757(7) Ln(1)-Se(5) 3.1207(11) 3.0818(9) Ln(1)-Se(6) ?2 3.1500(8) 3.1241(7) Ln(2)-Se(1) ?2 3.1221(8) 3.0924(6) Ln(2)-Se(3) 3.3110(11) 3.2896(9) Ln(2)-Se(4) ?2 3.1460(8) 3.1239(7) Ln(2)-Se(5) 3.3408(11) 3.3114(9) Ln(2)-Se(6) ?2 2.9982(8) 2.9678(6) Ln(3)-Se(2) 2.9229(11) 2.8915(9) Ln(3)-Se(3) ?2 2.9718(7) 2.9415(6) Ln(3)-Se(4) ?2 3.0578(8) 3.0352(7) Ln(3)-Se(5) 3.0925(11) 3.0650(9) Ln(3)-Se(6) 3.1116(11) 3.0872(9) Lu(1)-Se(1) ?4 2.8257(6) 2.8169(5) Lu(1)-Se(2) ?2 2.6629(9) 2.6620(8) Lu(2)-Se(4) ?2 2.7731(9) 2.7684(8) Lu(2)-Se(5) ?4 2.8007(6) 2.7883(5) 154 Figure 6.2. Unit cell of ?-PrLuSe3 viewed along the a axis. Pr-Se bonds have been omitted for clarity. c b 155 Figure 6.3. A depiction of an individual LuSe6 octahedra layer viewed down the b axis in ?-PrLuSe3. a c 156 Figure 6.4. A view of the three-dimensional channel structure of Sm1.82Lu2.18Se6 along the b axis. 157 Figure 6.5. Illustrations of the coordination environments for Pr ions in ?-PrLuSe3 and Sm(1)/Lu(1) ions in Sm1.82Lu2.18Se. 158 Table 6.8. Selected Bond Distances (?) for ?-LnLuSe3 (Ln = Pr, Nd). Formula ?-PrLuSe3 ?-NdLuSe3 Ln(1)-S(1) ?4 3.0671(13) 3.0577(6) Ln(1)-S(1) ?2 3.3670(17) 3.3677(9) Ln(1)-S(2) ?2 2.9035(18) 2.8935(9) Lu(1)-Se(1) ?4 2.8072(11) 2.8078(6) Lu(1)-Se(2) ?2 2.7102(10) 2.7138(5) 159 Sm1.82Lu2.18Se6 and Gd1.87Lu2.13Se6 crystallize in F-Ln2S325,26 type structure, which will be detailed in Chapter 7.22 As shown in Figure 6.4, the eight-coordinate bicapped trigonal prismatic Ln(1) ions (Figure 6.5) sit in the one-dimensional channels formed by three different double chains via edge- and corner- sharing. These double chains, all running down the b axis, are constructed from Ln(2)Se7 monocapped trigonal prisms, Ln(3)Se6 octahedra, and Ln(4)S6 octahedra, respectively. Within each double chain, the building polyhedra share edges with each other both in the direction of chain propagation and with adjacent chains. For Sm1.82Lu2.18Se6, as shown in Table 6.9, the average distances for Ln(1)S8, Ln(2)Se7, Ln(3)Se6, and Ln(4)Se6 polyhedra are 3.0406(10) ?, 2.9230(10) ?, 2.8064(10) ?, and 2.7791(9) ?, respectively, which are comparable to Shannon?s data,35 3.05 ? for SmSe8, 3.00 ? for SmSe7, 2.90 ? for LuSe7, and 2.84 ? for LuSe6. Magnetic Susceptibility. Figure 6.6 shows the temperature dependence of the inverse molar magnetic susceptibilities for ?-LnLuSe3 (Ln = Pr, Nd). Both compounds are paramagnetic and deviate from the Curie-Weiss law below 40 K. The effective magnetic moment and Weiss constant were obtained by fitting the high-temperature susceptibility data to the Curie-Weiss law. As shown in Table 6.10, the effective magnetic moments are close to calculated values for free Ln3+ ions. The negative ?p values indicate antiferromagnetic coupling between magnetic ions. The magnetic susceptibility of Sm1.82Lu2.18Se6 shows a typical van Vleck paramagnetic behavior similar to Sm metal, which is displayed in Figure 6.7. There is no magnetic transition down to 2 K and susceptibility data do not follow the Curie-Weiss law. The difference between the ground state (6H5/2) and the first excited state (6H7/2) of 160 Table 6.9. Selected Bond Distances (?) for LnxLu4-xSe6 (Ln = Sm, Gd; x = 1.82, 1.87). Formula Sm1.82Lu2.18Se6 Gd1.87Lu2.13Se6 Ln(1)-S(1) ?2 3.0932(9) 3.0955(9) Ln(1)-S(3) ?2 3.0806(9) 3.0803(9) Ln(1)-S(4) 3.0284(12) 3.0191(12) Ln(1)-S(6) ?2 2.9582(8) 2.9472(8) Ln(1)-S(6) 3.0323(11) 3.0326(11) Ln(2)-S(3) 2.9678(11) 2.9666(11) Ln(2)-S(4) 2.8281(12) 2.8227(11) Ln(2)-S(5) ?2 2.8182(9) 2.8152(9) Ln(2)-S(5) 2.9972(12) 3.0112(12) Ln(2)-S(6) ?2 3.0156(8) 3.0269(9) Ln(3)- S(2) ?2 2.7895(8) 2.8027(8) Ln(3)-S(2) 2.8331(12) 2.8460(12) Ln(3)-S(3) ?2 2.8441(8) 2.8582(8) Ln(3)-S(5) 2.7379(12) 2.7544(11) Ln(4)-S(1) ?2 2.8380(8) 2.8505(8) Ln(4)-S(1) 2.7909(11) 2.8071(11) Ln(4)-S(2) 2.7741(11) 2.7827(11) Ln(4)-S(4) ?2 2.7169(8) 2.7308(8) 161 Figure 6.6. The temperature dependence of the reciprocal molar magnetic susceptibility for ?-PrLuSe3 and ?-NdLuSe3, under an applied magnetic field of 0.1 T between 2 and 300 K. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. 0 50 100 150 200 250 300 0 30 60 90 120 150 180 210 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) ?-PrLuSe3 ?-NdLuSe3 162 Table 6.10. Magnetic Parameters for Ce3LuSe6, ?-PrLuSe3, ?-NdLuSe3, and Gd1.87Lu2.13Se6. Formula Pcal/?B Peff/?B ?p/K R2 Ce3LuSe6 4.40 4.56(1) -20(1) 0.99959 ?-PrLuSe3 3.58 3.509(3) -15.6(4) 0.99993 ?-NdLuSe3 3.62 3.913(9) -25(1) 0.99961 Gd1.87Lu2.13Se6 10.86 11.77(1) -4.4(4) 0.99991 a Pcal and Peff : calculated37 and experimental effective magnetic moments per formula unit. b Weiss constant (?p) and goodness of fit (R2) obtained from high temperature (100- 300 K) data. 163 Figure 6.7. Molar magnetic susceptibility vs temperature between 2 and 300 K for Sm1.82Lu2.18Se6. Data were taken under an applied magnetic field of 0.1 T. 0 50 100 150 200 250 300 0.00 0.01 0.02 0.03 0.04 0.05 0.06 Sm1.82Lu2.18Se6 ? ( em u/O e m ol for mu la un it) Temperature, T (K) 164 Sm3+ is not large compared to thermal energy (kBT). Therefore, the excited states make significant contributions to the magnetic susceptibility at high temperature.38 The experimental effective magnetic moment of Sm3+ can be determined approximately using ?eff = [ 3kB?mT/(L?0?B2)]1/2, where kB is Boltzmann constant, L is Avogadro?s number, ?0 is vacuum permeability, T is temperature in Kelvin, ?m is molar susceptibility. At T = 300 K, ?eff = 1.02 ?B, which is smaller than the calculated value (1.55 ?B) for free Sm3+ ions using the van Vleck formula.37 It is probably caused by crystal-field effects. The magnetic susceptibility of Gd1.87Lu2.13Se6 obeys the Curie-Weiss law above the temperature around 4 K where it undergoes a sharp antiferromagnetic transition, as shown in Figure 6.8. The effective magnetic moment and Weiss constant were obtained to be 11.77(1) ?B per formula unit and -4.4(4) K. The magnetization measurement was performed at 2 K and the results are presented in Figure 6.9. The saturation moment per Gd3+ ion is 7.5 ?B, which is close to the value for free Gd3+ ion (7.94 ?B) assuming g = 2. There is a weak spin reorientation transition at approximately H = 0.5 T. Ce3LuSe6 shows a deviation from the Curie-Weiss law near 70 K owing to crystal-field effects. A magnetic transition was observed below 5 K, which is illustrated in Figure 6.10. A small divergence on the ZFC-FC measurements (Figure 6.11) below this temperature may indicate a ferromagnetic component of the transition or a small temperature difference. In order to investigate the magnetic transition in detail, the magnetization measurements at 2 K were conducted as well. Figure 6.12 shows the field- dependent magnetizations for Ce3LuSe6. M(H) increases abruptly at low field, which is consistent with ferromagnetic behavior. The saturation moment per Ce3+ ion (1.13 ?B) is 165 Figure 6.8. Inverse molar magnetic susceptibility vs temperature for Gd1.87Lu2.13Se6 under an applied magnetic field of 0.1 T between 2 and 300 K. The solid line represents the fit to Curie-Weiss law in the range of 100-300 K. Inset shows the molar magnetic susceptibility at low temperature. 0 50 100 150 200 250 300 0 4 8 12 16 2 4 6 8 10 1.5 2.0 2.5 ???? Temperature Gd1.87Lu2.13Se6 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) 166 Figure 6.9. The magnetization for Gd1.87Lu2.13Se6 as a function of applied field at 2 K. Inset shows the M(H) curve between 0 and 1 T. Red and green lines are linear fits extended from zero field and from 1T, respectively. Slight increasing of the slope and the weak spin reorientation transition field at the junction (up arrow) can be observed. 0 1 2 3 4 5 6 7 0 20000 40000 60000 80000 M (em u/m ol for mu la un it) H (T) 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.5 1.0 1.5 2.0 M /10 4 H Gd1.87Lu2.13Se6 T = 2 K 167 Figure 6.10. Inverse molar magnetic susceptibility as a function of temperature for Ce3LuSe6 under an applied magnetic field of 0.1 T between 2 and 300 K. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. Inset shows the inverse molar magnetic susceptibility at low temperature. 0 50 100 150 200 250 300 0 20 40 60 80 100 120 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) 2 4 6 8 10 0 1 2 3 4 ????-1 Temperature Ce3LuSe3 H = 0.1 T 168 0 5 10 15 20 25 0 5 10 15 20 25 30 Ce3LuSe3 H = 0.01 T ? ( em u/O e m ol for mu la un it) Temperature, T (K) ZFC FC Figure 6.11. Molar magnetic susceptibility as a function of temperature for Ce3LuSe6 under ZFC and FC conditions with an applied magnetic field of 0.01 T between 2 and 25 K. 169 0 1 2 3 4 5 6 0 4000 8000 12000 16000 20000 M (em u/m ol for mu la un it) H (T) Ce3LuSe6 T = 2 K Figure 6.12. The magnetization for Ce3LuSe6 as a function of applied field at 2 K. 170 substantially smaller than the moment for free Ce3+ ion (2.54 ?B) assuming g = 6/7. It may be because of cystal-field splitting the ground state of Ce3+ ion (2F5/2). No magnetic hysteresis was found at 2 K, which is consistent with the behavior for a soft ferromagnet. In contrast, the Weiss constant for Ce3LuSe6 was determined to be -20(1) K, indicating antiferromagnetic interactions between Ce3+ ions. It is worth noting that |?p| value may be enlarged due to the crystal-field splitting of the full J=5/2 multiplet for Ce3+. To conclude, Ce3LuSe6 orders ferromagnetically with a weak antiferromagnetic component that might be due to canted spins of Ce3+. Optical Properties. There are few interlanthanide selenides reported in the literature. SmEr3Se6, a red compound, was determined to be a wide direct semiconductor with a band gap of 2.0 eV.20 The series of compounds, ?-LnYbSe3 (Ln = La, Ce, Pr, Nd, Sm), are black in color, as are the title compounds.14 The UV-vis-NIR diffuse reflectance spectra (Fig. 13) of LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd) reveal band gaps for La3LuSe6, Ce3LuSe6, ?-PrLuSe3, ?-NdLuSe3, Sm1.82Lu2.18Se6, and Gd1.87Lu2.13Se6 of 1.26, 1.10, 1.56, 1.61, 1.51, and 1.56 eV, respectively. The more condensed structure that Ln3LuSe6 (Ln = La, Ce) adopts may be the reason for their considerably smaller gaps compared to the Pr, Nd, Sm, and Gd containing phases. The smaller value for Ce3LuSe6 is due to the relatively low energy of the 4f1 ? 4f05d1 transition for cerium. The fine- structure observed in the spectra for ?-PrLuSe3, ?-NdLuSe3, and Sm1.82Lu2.18Se6 is actually due to f-f transitions within the lanthanide ions. CONCLUSIONS 171 Molten Sb2Q3 (Q = S, Se) fluxes have been valuable media to access ternary interlanthanide chalcogenides. The composition and structure of the products depends highly on the choices of lanthanides and chalcogenides. In this present study, we detailed the synthesis of lutetium-based interlanthanide selenides LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd) using a Sb2Se3 flux. All of these compounds show diverse structures and physical properties as a function of the Ln ions. They adopt three different structures types including U3ScS6 for Ln = La and Ce, UFeS3 for Ln = Pr and Nd, and F-Ln2S3 for Ln = Sm and Gd. Ln3LuSe6 (Ln = La, Ce) have a very condensed three-dimensional structure that is constructed from two-dimensional 2? [Ln3Se6]3- slabs with the gaps between these slabs filled by octahedral Lu3+ ions. The structure of ?-LnLuSe3 (Ln = Pr, Nd) includes two-dimensional LuSe6 octahedra layers that are separated by eight- coordinate larger Ln3+ ions, whereas Sm1.82Lu2.18Se6 and Gd1.87Lu2.13Se6 have a three- dimensional channel structure. Magnetic measurements have shown that both ?-PrLuSe3 and ?-NdLuSe3 are Curie-Weiss type paramagnets. Sm1.82Lu2.18Se6 exhibits van Vleck paramagnetism. Gd1.87Lu2.13Se6 was found to have an antiferromagnetic transition around 4 K, whereas Ce3LuSe6 has ferromagnetic ordering with a weak antiferromagnetism below 5 K. LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd) are semiconductors with tunable band gaps. 172 Figure 6.13. UV-vis diffuse reflectance spectra of LnxLuySez (Ln = La, Ce, Pr, Nd, Sm, Gd). 1 2 3 4 0 1 2 3 Gd1.87Lu2.13Se6 Sm1.82Lu2.18Se6 ?-NdLuSe3 ?-PrLuSe3 Ce3LuSe6 La3LuSe6 ?/ s ( arb ita ry un its) Energy (eV) 173 REFERENCES 1. Moreau, J. M. Mater. Res. Bull. 1968, 3, 427. 2. Moreau, J. M.; Mareschal, J.; Bertaut, E. F. Solid State Commun. 1968, 6, 751. 3. Mareschal, J.; Moreau, J. M.; Ollivier, G.; Pataud, P.; Sivardiere, J. Solid State Commun. 1969, 7, 669. 4. Rodier, N.; Laruelle, P. C. R. Seances Acad. Sci. Ser. C 1970, 270, 2127. 5. Coutures, J.; Coutures, J. P. J. Solid State Chem. 1976, 19, 29. 6. M?ller-Buschbaum, H.; Graebner, P. -H. Z. Anorg. Allg. Chem. 1971, 386, 158. 7. Ijdo, D. J. W. Acta Crystallogr. B 1980, 36, 2403. 8. Ito, K.; Tezuka, K.; Hinatsu, Y. J. Solid State Chem. 2001, 157, 173. 9. Deepa, M.; Varadaraju, U. V. Mater. Res. Soc. Symp. Proc. 1998, 527, 507. 10. Berndt, U.; Maier, D.; Keller, C. J. Solid State Chem. 1976, 16, 189. 11. Ito, K.; Tezuka, K.; Hinatsu, Y. J. Solid State Chem. 2001, 157, 173. 12. Rodier, N.; Julien, R.; Tien, V. Acta Crystallogr. C 1983, 39, 670. 13. Carre, D.; Laruelle, P. Acta Crystallogr. B 1974, 30, 952. 14. Mitchell, K.; Somers, R. C.; Huang, F. Q.; Ibers, J. A. J. Solid State Chem. 2004, 177, 709. 15. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. Chem. Mater. 2007, 19, 567. 16. Rodier, N.; Tien, V. C. R. Seances Acad. Sci. Ser. C 1974, 279, 817. 17. Rodier, N.; Firor, R. L.; Tien, V.; Guittard, M. Mater. Res. Bull. 1976, 11, 1209. 18. Rodier, N. Bull. Soc. Fr. Mineral. Cristallogr. 1973, 96, 350. 174 19. Carr?, D.; Laruelle, P. Acta Cryst. 1973, B 29, 70. 20. Gray, D. L.; Rodriguez, B. A.; Chan, G. H.; Van Duyne, R. P.; Ibers, J. A. J. Solid State Chem. in press. 21. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. J. Solid State Chem. in press. 22. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Booth, C. H.; Albrecht- Schmitt, T. E. PrLnYb2S6 (Ln = Tb, Dy); J. Solid State Chem. in press. 23. Marezio, M.; Remeika, J. P.; Dernier, P. D. Acta Crystallogr. B 1970, 26, 2008. 24. Noel, H.; Padiou, J. Acta Crystallogr. B 1976, 32, 1593. 25. Schleid, T.; Lissner, F. J. Alloys Compd. 1992, 189, 69. 26. Fang, C. M.; Meetsma, A.; Wiegers, G. A. J. Alloys Compd. 1993, 201, 255. 27. Zachariasen, W.H. Acta Crystallogr. 1949, 2, 291. 28. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Booth, C. H.; Albrecht- Schmitt, T. E. Inorg. Chem. submitted. 29. Sheldrick, G. M. SHELXTL PC, Version 6.12, An Integrated System for Solving, Refining, and Displaying Crystal Structures from Diffraction Data; Siemens Analytical X-Ray Instruments, Inc.: Madison, WI, 2001. 30. Sheldrick, G. M. SADABS 2001, Program for absorption correction using SMART CCD based on the method of Blessing: Blessing, R. H. Acta Crystallogr. 1995, A51, 33. 31. Mulay, L. N.; Boudreaux, E. A. Theory and Applications of Molecular Diamagnetism; Wiley-Interscience: New York, 1976. 175 32. Wendlandt, W. W.; Hecht, H. G. Reflectance Spectroscopy; Interscience Publishers: New York, 1966. 33. Rodier, N.; Tien, V. Acta Crystallogr. 1976, 32, 2705. 34. Becker, D. F.; Kasper, J. S. Acta Crystallogr. 1957, 10, 332. 35. Adolphe, C. Annales de Chimie (Paris) 1965, 10, 271. 36. Shannon, R. D. Acta Cryst.1976, A32, 751. 37. Kittel, C. Introduction to Solid State Physics, 6th Ed., Wiley, New York, 1986. 38. Van Vleck, J. H. The Theory of Electric and Magnetic Susceptibilities; Oxford University Press: London, 1932. 176 CHAPTER 7 SYNTHESES, STRUCTURE, MAGNETISM, AND OPTICAL PROPERTIES OF THE PARTIALLY ORDERED QUATERNARY INTERLANTHANIDE SULFIDES PrLnYb2S6 (Ln = Tb, Dy) ABASTRACT Dark red single crystals of PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) have been synthesized through the reaction of elemental rare earth metals and S using a Sb2S3 flux at 1000 ?C. These isotypic compounds adopt the F-Ln2S3 three-dimensional open channel structure type. Eight-coordinate Pr3+ ions sit in the channels that are constructed from three different edge-shared double chains running down the b axis that contain Yb(1)S6 octahedra, Yb(2)S6, octahedral, and LnS7 monocapped trigonal prisms, respectively. Each double chain connects to four other neighbors by sharing vertices and edges. Considerable disordering in Ln positions was observed in single X-ray diffraction experiments only in the case of Pr/Yb. Least square refinements gave rise to the formulas of Pr1.34Yb2.66S6, PrTbYb2S6, and PrDyYb2S6 that were confirmed by elemental analysis and magnetic susceptibility measurements. Pr1.34Yb2.66S6, PrTbYb2S6, and PrDyYb2S6 are paramagnetic down to 2 K without any indications of long-range magnetic ordering. 177 The optical transitions for Pr1.34Yb2.66S6, PrTbYb2S6, and PrDyYb2S6 are at approximately 1.6 eV. Crystallographic data: Pr1.34Yb2.66S6, monoclinic, space group P21/m, a = 10.960(2), b = 3.9501(8), c = 11.220(2) ?, ? = 108.545(3), V = 460.54(16), Z = 2; PrTbYb2S6, monoclinic, space group P21/m, a = 10.9496(10), b = 3.9429(4), c = 11.2206(10) ?, ? = 108.525(2), V = 459.33(7), Z = 2; PrDyYb2S6, monoclinic, space group P21/m, a = 10.9384(10), b = 3.9398(4), c = 11.2037(10) ?, ? = 108.612(2), V = 457.57(7), Z = 2. INTRODUCTION There have been numerous studies on ternary interlanthanide chalcogenides in terms of their diverse structural chemistry and interesting physical properties.1-18 ?- LnLn'S3 (GdFeO3 type)19,1-4 CeYb3S6 (F-Ln2S3 type)20,21,5-6 Sc2Er3S77 (Y5S7 type)22 and EuLn2Q48-10 (CaFe2O4 type)23 possess three-dimensional open-channel structures, wherein the larger Ln3+ ions reside in channels formed by smaller lanthanide chalcogenides polyhedra. While ?-LnLn'Q3 (Q = S, Se) (UFeS3 type)24,3,11,12 and ?- LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)13 have layers of Ln'Qx polyhedra, separated by larger Ln3+ ions. ?-LnLuS3 (Ln = Ce, Pr, Nd)14 (CeTmS3 type)15 have a very condensed three-dimensional structure. Recent work has shown that the electronic and magnetic properties of these materials highly depend on the structures they adopt and the choices of lanthanides. For example, the optical band gaps of ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb) are approximately 1.3 ? 1.6 eV,13 while SmEr3S6 is 2.4 ? 2.6 eV.16 ?-LnLuS3 (Ln = Pr, Nd) exhibit possible short-range antiferromagnetic ordering at low temperatures.14 178 In contrast, there are no existing ordered quaternary interlanthanide chalcogenides that possess three different lanthanide elements. Instead of making new ordered quaternary phases, they can be prepared using intermediate lanthanides to substitute in the disordered sites in already known ternary structures. F-Ln2S3 type ternary compounds are probably the best candidates to achieve this goal.17,18 This structure type has three different coordination environments for Ln3+ as octahedral, and mono- and bicapped trigonal prisms. The seven-coordinate sites are often disordered. By carefully choosing three different metals, ordered quaternary phases can be accessed. In this paper, we present the syntheses, structure, optical, and magnetic properties of the first two partially ordered quaternary interlanthanide chalcogenides, PrTbYb2S6 and PrDyYb2S6. As a reference, Pr1.34Yb2.66S6 is also included in the discussion. The information in this chapter has been published as a full paper in Journal of Solid-State Chemistry.25 EXPERIMENTAL Starting Materials. Pr (99.9%, Alfa-Aesar), Tb (99.9%, Alfa-Aesar), Dy (99.9%, Alfa-Aesar), Yb (99.9%, Alfa-Aesar), S (99.5%, Alfa-Aesar), and Sb (99.5%, Alfa-Aesar) were used as received. The Sb2S3 flux was prepared from the direct reaction of the elements in sealed fused-silica ampoules at 850 ?C. Syntheses. PrLnYb2S6 (Ln = Tb, Dy) were prepared through the reaction of Pr (0.17 mmol), Ln (0.17 mmol), Yb (0.34 mmol), S (1.02 mmol), and Sb2S3 (0.17 mmol). For Pr1.34Yb2.66S6, the reaction mixture consists of Pr (0.23 mmol), Yb (0.45 mmol), S (1.02 mmol), and Sb2S3 (0.17 mmol). All of the reactants were loaded into fused-silica ampoules under an argon atmosphere in a glovebox. The ampoules were sealed under 179 vacuum and heated in a programmable tube furnace. The following heating profile was used: 2 ?C/min to 500 ?C (held for 1 h), 0.5 ?C/min to 1000 ?C (held for 5 d), 0.04 ?C/min to 550 ?C (held for 2 d), and 0.5 ?C/min to 24 ?C. Powder X-ray diffraction measurements were used to confirm phase purity by comparing the powder patterns calculated from the single crystal X-ray structures with the experimental data. Semi- quantitative SEM/EDX analyses were performed using JEOL 840/Link Isis or JEOL JSM-7000F instruments. Pr, Ln, Yb, and S percentages were calibrated against standards. Sb was not detected in the crystals. Pr:Ln:Yb:S ratios of close to 1:1:2:6 were found for PrLnYb2S6 (Ln = Tb, Dy), while the Pr:Yb:S ratios in Pr1.34Yb2.66S6 samples are approximately 1:2:4.5 from EDX analyses. Crystallographic Studies. Single crystals of PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) were mounted on glass fibers with epoxy and optically aligned on a Bruker APEX single crystal X-ray diffractometer using a digital camera. Initial intensity measurements were performed using graphite monochromated Mo K? (? = 0.71073 ?) radiation from a sealed tube and monocapillary collimator. SMART (v 5.624) was used for preliminary determination of the cell constants and data collection control. The intensities of reflections of a sphere were collected by a combination of 3 sets of exposures (frames). Each set had a different ? angle for the crystal and each exposure covered a range of 0.3? in ?. A total of 1800 frames were collected with exposure times per frame of 10 or 20 seconds depending on the crystal. For PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy), determination of integrated intensities and global refinement were performed with the Bruker SAINT (v 6.02) software package using a narrow-frame integration algorithm. These data were treated first with a face- 180 index numerical absorption correction using XPREP,26 followed by a semi-empirical absorption correction using SADABS.27 The program suite SHELXTL (v 6.12) was used for space group determination (XPREP), direct methods structure solution (XS), and least-squares refinement (XL).26 The final refinements included anisotropic displacement parameters for all atoms and secondary extinction. Some crystallographic details are given in Table 7.1. Atomic coordinates and equivalent isotropic displacement parameters for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) are given in Table 7.2-7.4. The formula of F-Ln2S3 type compounds can be expressed as (AVIII)(BVII)(CVI)2S6. In the case of Pr1.34Yb2.66S6, eight-coordinate A sites were assigned as Pr atoms and both seven-coordinate B and octahedral C positions were named as Yb at the beginning of the refinement. However the average bond distance of YbS7 is longer than the accepted value, according to Shannon?s radii data,28 and its thermal parameter is larger than the other Yb atoms. The elemental analysis showed that the ratio of Pr:Yb is 1:2. This evidence suggests that there should be some disordering on the B sites. The refinement of occupancy lowered the residual and weighting scheme and gave rise to the final formula of Pr1.34(1)Yb2.66(1)S6. For PrLnYb2S6 (Ln = Tb, Dy), Pr, Ln, Yb atoms were put in A, B, and C positions respectively. This gave excellent residuals in the refinements, and the suggested formulas as PrLnYb2S6 (Ln = Tb, Dy) are consistent with the EDX results. Considering the similarity among Pr, Ln, and Yb, small amount of disordering on B and C, and even A sites can not be excluded. 181 Table 7.1. Crystallographic Data for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy). Formula Pr1.34Yb2.66S6 PrTbYb2S6 PrDyYb2S6 fw 841.31 838.27 841.85 Color dark red dark red dark red Crystal System Monoclinic monoclinic monoclinic Space group P21/m (No. 11) P21/m (No. 11) P21/m (No. 11) a (?) 10.960(2) 10.9496(10) 10.9384(10) b (?) 3.9501(8) 3.9429(4) 3.9398(4) c (?) 11.220(2) 11.2206(10) 11.2037(10) ? 108.545(3) 108.525(2) 108.612(2) V (?3) 460.54(16) 459.33(7) 457.57(7) Z 2 2 2 T (K) 193 193 193 ? (?) 0.71073 0.71073 0.71073 ?calcd (g cm?3) 6.067 6.061 6.110 ? (cm?1) 349.71 342.64 348.32 R(F)a 0.0233 0.0330 0.0243 Rw(Fo2) b 0.0657 0.1104 0.0597 a ( )R F F F F= ?? ? o c o for Fo 2 > 2?(Fo2). b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 182 Table 7.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for Pr1.34Yb2.66S6. Atom (site) x y z Ueq (?2)a Pr1 0.55123(4) 0.25 0.19606(4) 0.00843(13) Pr/Yb 0.18137(4) 0.25 0.00138(4) 0.01093(16) Yb1 0.94227(3) 0.25 0.33496(3) 0.00997(13) Yb2 0.65971(3) 0.25 0.58575(3) 0.01011(13) S1 0.41859(19) 0.25 0.59374(19) 0.0086(4) S2 0.8937(2) 0.25 0.5587(2) 0.0100(4) S3 0.23343(19) 0.25 0.76947(19) 0.0090(4) S4 0.3064(2) 0.25 0.25444(19) 0.0098(4) S5 0.9805(2) 0.25 0.1160(2) 0.0121(4) S6 0.61659(19) 0.25 0.96234(19) 0.0084(4) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 183 Table 7.3. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for PrTbYb2S6. Atom (site) x y z Ueq (?2)a Pr1 0.55027(7) 0.25 0.19511(7) 0.0064(2) Tb1 0.18125(6) 0.25 0.00196(6) 0.0075(2) Yb1 0.94104(5) 0.25 0.33390(5) 0.0098(2) Yb2 0.65988(5) 0.25 0.58556(5) 0.0096(2) S1 0.4185(3) 0.25 0.5947(3) 0.0073(6) S2 0.8943(3) 0.25 0.5591(3) 0.0085(6) S3 0.2357(3) 0.25 0.7711(3) 0.0071(6) S4 0.3064(3) 0.25 0.2532(3) 0.0095(6) S5 0.9800(3) 0.25 0.1144(3) 0.0101(7) S6 0.6159(3) 0.25 0.9625(3) 0.0081(6) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 184 Table 7.4. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for PrDyYb2S6. Atom (site) x y z Ueq (?2)a Pr1 0.54996(5) 0.25 0.19506(5) 0.00797(13) Dy1 0.18100(4) 0.25 0.00156(4) 0.00847(13) Yb1 0.94108(4) 0.25 0.33369(4) 0.01048(13) Yb2 0.66021(4) 0.25 0.58597(4) 0.01040(13) S1 0.4181(2) 0.25 0.5945(2) 0.0087(5) S2 0.8943(2) 0.25 0.5590(2) 0.0098(5) S3 0.2348(2) 0.25 0.7715(2) 0.0088(4) S4 0.3051(2) 0.25 0.2525(2) 0.0100(5) S5 0.9799(2) 0.25 0.1140(2) 0.0101(5) S6 0.6165(2) 0.25 0.9628(2) 0.0095(5) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 185 Powder X-ray Diffraction. Powder X-ray diffraction patterns were collected with a Rigaku Miniflex powder X-ray diffractometer using Cu K? (? = 1.54056 ?) radiation. Magnetic Susceptibility Measurements. Magnetism data were measured on powders in gelcap sample holders with a Quantum Design MPMS 7T magnetometer/susceptometer between 2 and 300 K and in applied fields up to 7 T. DC susceptibility measurements were made under zero-field-cooled conditions with an applied field of 0.1 T. Susceptibility values were corrected for the sample diamagnetic contribution according to Pascal?s constants29 as well as for the sample holder diamagnetism. ?p values were obtained from extrapolations from fits between 100 and 300 K. (Magnetic susceptibility measurements were performed by Eun Sang Choi and James S. Brooks at Florida State University) UV-vis-NIR Diffuse Reflectance Spectroscopy. The diffuse reflectance spectra PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) were measured from 200 to 1500 nm using a Shimadzu UV3100 spectrophotometer equipped with an integrating sphere attachment. The Kubelka-Munk function was used to convert diffuse reflectance data to absorption spectra.30 RESULTS AND DISCUSSION Structures of PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy). The isotypic series of PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) have the F-Ln2S3 type structure with Pr3+ ions sitting on eight-coordinate positions, Ln3+ ions in seven-coordinate positions, and Yb3+ ions occupying two octahedral sites. As shown in Figure 7.1, the structure of these 186 Figure 7.1. An illustration of the three-dimensional structure of PrTbYb2S6 along the b axis. 187 compounds is constructed from three different edge-shared double chains running down the b axis, which contain Yb(1)S6 octahedra, Yb(2)S6 octahedra, and LnS7 monocapped trigonal prisms, respectively. Each double chain connects to four other neighbors by sharing vertices and edges to form the channels where Pr3+ ions reside. For example, Yb(1)S6 double chains are bound to two Yb(2)S6 double chains via corner-sharing and two LnS7 double chains via edge-sharing. The PrS8 polyhedra can be viewed as a bicapped trigonal prism, which is shown in Figure 7.2. Selected bond distances for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) are listed in Table 7.5. Pr-S bond distances rang from 2.8755(19) ? to 3.0127 ?, which are comparable to Shannon?s radii data of 2.966 ?.28 The average bond distances for Pr/YbS7, TbS7, and DyS7 are 2.807(2) ?, 2.800(2) ?, and 2.792(2) ? respectively. Compared to the accepted values for PrS7 (2.89 ?), YbS7 (2.765 ?), TbS7 (2.82 ?), and DyS7 (2.81 ?), they are all reasonable.28 The bond distances for YbS6 octahedra are in the range of 2.6134(14) ? and 2.759(2) ?. Magnetic Susceptibility. The magnetic susceptibilities for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy), in the range of 2-300 K, are presented in Figures 7.3-7.5. There are no indications of long-range magnetic orderings down to 2 K. Pr1.34Yb2.66S6 deviates from the ideal Curie-Weiss law below 70 K due to crystal-field splitting of lanthanide ions. PrTbYb2S6 shows a pure Curie-Weiss paramagnetic behavior in the whole temperature range, while the 1/? plot for PrDyYb2S6 exhibits a deviation from the Curie-Weiss law and the onset of upward curvature at low temperature. This may indicate a short-range antiferromagnetic ordering, which has also been observed in ?-Pr1.29Lu0.71S3.14 Table 7.6 shows the magnetic parameters for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy), which were 188 Figure 7.2. Bicapped trigonal prismatic coordination environment of the Pr ions in PrTbYb2S6. 189 Table 7.5. Selected Bond Distances (?) for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy). Formula Pr1.34Yb2.66S6 PrTbYb2S6 PrDyYb2S6 Pr(1)-S(1) ?2 3.0127(16) 3.009(2) 3.0058(19) Pr(1)-S(3) ?2 3.0068(16) 2.993(2) 3.0004(19) Pr(1)-S(4) 2.958(2) 2.943(3) 2.947(3) Pr(1)-S(6) ?2 2.8919(16) 2.877(2) 2.8755(19) Pr(1)-S(6) 2.930(2) 2.919(3) 2.918(2) Ln-S(3) 2.836(2) 2.837(3) 2.821(2) Ln-S(4) 2.732(2) 2.715(3) 2.704(3) Ln-S(5) ?2 2.7012(16) 2.692(2) 2.6832(18) Ln-S(5) 2.881(2) 2.865(3) 2.863(2) Ln-S(6) ?2 2.8977(15) 2.901(2) 2.8947(18) Yb(1)- S(2) ?2 2.6819(15) 2.684(2) 2.6810(18) Yb(1)-S(2) 2.726(2) 2.734(3) 2.731(2) Yb(1)-S(3) ?2 2.7471(15) 2.752(2) 2.7426(18) Yb(1)-S(5) 2.620(2) 2.630(3) 2.628(2) Yb(2)-S(1) ?2 2.7557(15) 2.759(2) 2.7565(17) Yb(2)-S(1) 2.673(2) 2.678(3) 2.680(2) Yb(2)-S(2) 2.677(2) 2.675(3) 2.672(2) Yb(2)-S(4) ?2 2.6134(14) 2.621(2) 2.6194(16) 190 Figure 7.3. Inverse molar magnetic susceptibility plotted against temperature between 2 and 300 K for Pr1.34Yb2.66S6. Data were taken under an applied magnetic field of 0.1 T. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. 0 50 100 150 200 250 300 0 10 20 30 40 Pr1.34Yb2.66S6 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) 191 Figure 7.4. The plot of the inverse molar magnetic susceptibility vs T for PrTbYb2S6 under an applied magnetic field of 0.1 T between 2 and 300 K. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. 0 50 100 150 200 250 300 0 5 10 15 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) PrTbYb2S6 192 Figure 7.5. The temperature dependence of the reciprocal molar magnetic susceptibility for PrDyYb2S6 under an applied magnetic field of 0.1 T between 2 and 300 K. The straight line represents the fit to Curie-Weiss law in the range of 100-300 K. 0 50 100 150 200 250 300 0 5 10 15 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) PrDyYb2S6 193 Table 7.6. Magnetic Parameters for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy). Formula Pcal/?B Peff/?B ?p/K R2 Pr1.34Yb2.66S6 8.48 7.91(1) -36.9(8) 0.99979 PrTbYb2S6 12.19 11.82(1) -3.1(4) 0.99992 PrDyYb2S6 12.92 11.38(2) -0.2(7) 0.99979 a Pcal and Peff : calculated31 and experimental effective magnetic moments per formula unit. b Weiss constant (?p) and goodness of fit (R2) obtained from high temperature (100- 300 K) data. 194 obtained from fitting the data in the range of 100 K and 300 K into the Curie-Weiss law. All of compounds have negative value of ?p, which indicates antiferromagnetic interactions between cations. The experimental effective magnetic moments for these compounds are close to the accepted values.31 This provides further supporting evidence for the proposed formula from the single crystal X-ray experiments. Optical Properties. The UV-vis-NIR diffuse reflectance spectra of PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) are shown in Figure 7.6. They are very similar to each other. This suggests that the substitutions using different lanthanides ions in seven-coordinate Ln positions causes only small changes in the band structures near the Fermi level for PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy). The optical transition mainly results from the interactions among eight-coordinate Pr3+ cations, six-coordinate Yb3+ cations, and S2- anions. It is also possible that the 4f-band of Pr, Yb, Tb, and Dy lie deep in the valence band. So the optical transitions are determined by the same gap between the [S]3p valence band and 5d(6s) conduction band.31,32 The band gaps of PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy) are approximately 1.6 eV, which are consistent with the dark red color they possess. They are also close to the values we reported for ?-LnLn'S3 (Ln = La, Ce; Ln' = Er, Tm, Yb)13 and ?-Ln2-xLuxS3 (Ln = Ce, Pr, Nd; x = 0.67-0.71).14 CONCLUSIONS ?he first two partially ordered quaternary interlanthanide sulfides PrLnYb2S6 (Ln = Tb, Dy) were prepared and characterized. They adopt the same F-Ln2S3 type structure as the parent disordered Pr1.34Yb2.66S6 phase. All three compounds are paramagnetic on the range from 2 to 300 K. The UV-vis-NIR diffuse reflectance spectra show that these 195 compounds have very similar electronic structures near the Fermi level with wide band gaps. The elemental analysis and magnetic susceptibility measurements are consistent with the proposed formula. 196 Figure 7.6. UV-vis diffuse reflectance spectra of PrLnYb2S6 (Ln = Pr/Yb, Tb, Dy). 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0.0 0.5 1.0 1.5 ?/ s ( arb ita ry un its ) Energy (eV) PrDyYb2S6 PrTbYb2S6 Pr1.34Yb2.66S6 197 REFERENCES 1. Rodier, N.; Laruelle, P.; Seances, C. R. Acad. Sci. Ser. 1970, C 270, 2127. 2. Ijdo, D. J. W. Acta Crystallogr. 1980, B36, 2403. 3. Rodier, N.; Julien, R.; Tien, V. Acta Crystallogr. 1983, C39, 670. 4. Range, K.-J.; Gietl, A.; Klement, U. Z. Kristallogr. 1993, 207, 147. 5. Rodier, N.; Firor, R. L.; Tien, V.; Guittard, M. Mat. Res. Bull. 1976, 11, 1209. 6. Rodier, N.; Tien, V. C. R. Acad. Sc. Paris Serie 1974, C279, 817. 7. Rodier, N.; Laruelle, P. Bull. Soc. fr. de Mineral. Cristallogr. 1972, 95, 548. 8. Hulliger, F.; Vogt, O. Phys. Lett. 1966, 21, 138. 9. Lugscheider, W.; Pink, H.; Weber, K.; Zinn, W. Zeitschrift fuer Angewandte Physik 1970, 30, 36. 10. Lemoine, P.; Carre, D.; Guittard, M. Acta Crystallogr. 1985, C41, 667. 11. Carr?, D.; Laruelle, P. Acta Crystallogr. 1974, B30, 952. 12 Mitchell, K.; Somers, R. C.; Huang, F. Q.; Ibers, J. A. J. Solid State Chem. 2004, 177, 709. 13. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. Chem. Mater. 2007, 19, 567. 14. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. J. Solid State Chem. 2007, in press. 15. Rodier, N. Bull. Soc. fr. de Mineral. Cristallogr. 1973, 96, 350. 16. Gray, D. L.; Rodriguez, B. A.; Chan, G. H.; Van Duyne, R. P.; Ibers, J. A. J. Solid State Chem. 2007, in press. 17. Carr?, D.; Laruelle, P. Acta Crystallogr. 1973, B29, 70. 198 18. Rodier, N.; Tien, V. Bull. Soc. fr. de Mineral. Cristallogr. 1975, 98, 30. 19. Marezio, M.; Remeika, J. P.; Dernier, P. D. Acta Crystallogr. 1970, B26, 2008. 20. Schleid, T.; Lissner, F. J. Alloys Compd. 1992, 189, 69. 21. Fang, C. M.; Meetsma, A.; Wiegers, G. A. J. Alloys Compd. 1993, 201, 255. 22. Adolphe, C. Annales de Chimie (Paris), 1965, 271. 23. Becker, D. F.; Kasper, J. S. Acta Crystallogr. 1957, 10, 332. 24. No?l, H.; Padiou, J. Acta Crystallogr. 1976, B32, 1593. 25. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Booth, C. H.; Albrecht- Schmitt, T. E. J. Solid State Chem. in press. 26. Sheldrick, G. M. SHELXTL PC, Version 6.12, An Integrated System for Solving, Refining, and Displaying Crystal Structures from Diffraction Data; Siemens Analytical X-Ray Instruments, Inc.: Madison, WI 2001. 27. Sheldrick, G. M. SADABS 2001, Program for absorption correction using SMART CCD based on the method of Blessing: Blessing, R. H. Acta Crystallogr. 1995, A51, 33. 28. Shannon, R. D. Acta Crystallogr. 1976, A32, 751. 29. Mulay, L. N.; Boudreaux, E. A. Theory and Applications of Molecular Diamagnetism; Wiley?Interscience: New York, 1976. 30. Wendlandt, W. W.; Hecht, H. G. Reflectance Spectroscopy; Interscience Publishers: New York, 1966. 31. Kittel, C. Introduction to Solid State Physics, 6th Edition; Wiley: New York, 1986. 199 32. Prokofiev, A. V.; Shelykh, A. I.; Golubkov, A. V.; Smirnov, I. A. J. Alloys Compd. 1995, 219, 172. 33. Prokofiev, A. V.; Shelykh, A. I.; Melekh, B. T. J. Alloys Compd. 1996, 242, 41. 200 CHAPTER 8 SYNTHESES, STRUCTURE, MAGNETISM, AND OPTICAL PROPERTIES OF THE ORDERED INTERLANTHANIDE COPPER CHALCOGENIDES Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) ABSTRACT Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) were prepared using Sb2Q3 (Q = S, Se) fluxes at 900 ?C. The structure of Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) consists of one-dimensional [YbCuQ5]6- ribbons extending along the b axis that are separated by larger Ln3+ ions. Each one-dimensional [YbCuQ5]6- ribbon is constructed from two single [YbQ6] octahedral chains with one double [CuQ5] trigonal bipyramidal chain in the middle. All three chains connect with each other via edge sharing. Two crystallographically unique eight-coordinate Ln atoms, one octahedral Yb site, and two disordered Cu positions are found in the structure. These two Cu sites reside in the trigonal bipyramidal cavities formed by Q2- anions. Ce2YbCuSe5, La2YbCuS5, Ce2YbCuS5, and Pr2YbCuS5 are Curie-Weiss paramagnets. La2YbCuSe5 and Nd2YbCuS5 have short-range antiferromagnetic ordering at low temperature. The band gaps of La2YbCuSe5, Ce2YbCuSe5, La2YbCuS5, Ce2YbCuS5, Pr2YbCuS5, Nd2YbCuS5, and 201 Sm2YbCuS5 are 1.15 eV, 1.05 eV, 1.45 eV, 1.37 eV, 1.25 eV, 1.35 eV, and 1.28 eV, respectively. INTRODUCTION Rare-earth copper chalcogenides have been the source of considerable interest not only because of their tunable structures and electronic properties, but also because of some unusual coordination geometries and oxidation states of copper that they exhibit.1-11, 19-39 High-temperature solid-state syntheses often produce thermodynamic products, in which there are no Q-Q (Q = S, Se, Te) bonds. They include LnCuQ2 (Ln = rare-earth element, Sc, Y; Q = S, Se, Te), 1-11 Ln0.66Cu2S2 (Ln = Gd, Er), 12-15 and Eu2CuQ3 (Q = S, Se).16-18 LnCuQ2 compounds adopt at least three different structures (monoclinic, orthorhombic, and trigonal) depend on the relative size of Ln ions. Mixed-valence Eu2CuQ3 phases have Eu2+ and Eu3+ ions occupying two crystallographically independent sites and Eu2CuS3 17shows a ferromagnetic transition at 3.4K owing to the coupling of Eu2+ ions. Kinetic products, in which polychalcogenides can exist, can be prepared using alkali-metal/lanthanide halide fluxes or reactive fluxes, thereby allowing for lower reaction temperatures. Examples include La2CuS4,20 Sm3CuSe6,21,22 EuCu0.66Te2,23 Gd3Cu2Te7,24 LaCu0.28Te2,25 and LnCuxTe2 (Ln = La, Nd, Sm, Gd and Dy).26 La2CuS4 features unusual discrete [S3Cu???S-S???CuS3]12- units, which contain two nearly planar [CuS3]5- triangles bridged by a disulfide anions. EuCu0.66Te2 contains a flat square net of Te atoms, while all the latter three have linear Te chains. Partial substitutions of chalcogenides have been tried to access new phases with different physical properties from the parent ternary phase LnCuQ2, e.g. LaCuSTe27 and 202 SmCuSTe.27 Lanthanide copper oxychalcogenides including LnCuOQ (Ln = La, Ce, Pr, Nd; Q = S, Se, Te),28-38 La5Cu6O4S7,39 and La3CuO2S340 have been extensively studied. LnCuOQ adopt a structure with alternately stacking PbO-like [Cu2Q2]2- layers and anti- PbO-like [Ln2O2]2+ layers, which is similar to that of copper-based high-Tc superconducting oxides. LaCuOQ are wide band gap p-type semiconductors, and are considered as potential transparent conductive materials. In contrast, La5Cu6O4S7 is metallic and the average oxidation state of the Cu atoms is +7/6. To our knowledge, EuLnCuS3 (Ln = Y, Gd-Lu) 19 are the only known examples of interlanthanide copper chalcogenides, which were made by replacing the Eu3+ ions in Eu2CuS3 compound using other trivalent Ln atoms. The purpose of this study is to prepare new ordered quaternary interlanthanide copper chalcogenides phases by including two Ln atoms from opposite ends of lanthanide series, which tend to have different coordination geometries. Here we present the syntheses, structure, optical, and magnetic properties of new quaternary interlanthanide copper chalcogenides, Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se). EXPERIMENTAL Starting Materials. La (99.9%, Alfa-Aesar), Ce (99.9%, Alfa-Aesar), Pr (99.9%, Alfa-Aesar), Nd (99.9%, Alfa-Aesar), Sm (99.9%, Alfa-Aesar), Yb (99.9%, Alfa-Aesar), Se (99.5%, Alfa-Aesar), S (99.5%, Alfa-Aesar), and Sb (99.5%, Alfa-Aesar) were used as received. The Sb2Q3 (Q = S, Se) fluxes were prepared from the direct reaction of the elements in sealed fused-silica ampoules at 850 ?C. Syntheses. Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) were prepared 203 using Sb2Q3 (Q = S, Se) fluxes. 150 mg of Ln, Yb, Cu, and S were stoichiometrically mixed with 0.075mg Sb2Q3 into fused-silica ampoules in an Ar-filled glovebox. The ampoules were sealed under vacuum and placed in a programmable tube furnace. The following heating profile was used: 2 ?C/min to 500 ?C (held for 1 h), 0.5 ?C/min to 900 ?C (held for 7 d), 0.04 ?C/min to 500 ?C (held for 2 d), and 0.5 ?C/min to 24 ?C. In each reaction the major phases included high yields of black crystals of desired products and unreacted Sb2Q3 lying in the bottom of the tubes. Powder X-ray diffraction measurements were used to confirm phase purity by comparing the powder patterns calculated from the single-crystal X-ray structures with the experimental data. Semi- quantitative SEM/EDX analyses were performed using JEOL 840/Link Isis or JEOL JSM-7000F instruments. Ln, Yb, Cu, and Q percentages were calibrated against standards. Sb was not detected in the crystals. The Ln:Yb:Cu:S ratios were determined to be approximately 2:1:1:5 from EDX analyses. Crystallographic Studies. Single crystals of Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) were mounted on glass fibers with epoxy and optically aligned on a Bruker APEX single crystal X-ray diffractometer using a digital camera. Initial intensity measurements were performed using graphite-monochromated Mo K? (? = 0.71073 ?) radiation from a sealed tube and monocapillary collimator. SMART (v 5.624) was used for preliminary determination of the cell constants and data collection control. The intensities of reflections of a sphere were collected by a combination of 3 sets of exposures (frames). Each set had a different ? angle for the crystal and each exposure covered a range of 0.3? in ?. A total of 1800 frames were collected with exposure times per frame of 10 or 20 seconds depending on the crystal. 204 For Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se), determination of integrated intensities and global refinement were performed with the Bruker SAINT (v 6.02) software package using a narrow-frame integration algorithm. These data were treated first with a face-indexed numerical absorption correction using XPREP,41 followed by a semi-empirical absorption correction using SADABS.42 The program suite SHELXTL (v 6.12) was used for space group determination (XPREP), direct methods structure solution (XS), and least-squares refinement (XL).41 The final refinements included anisotropic displacement parameters for all atoms and secondary extinction. Some crystallographic details are given in Table 8.1. As an example, atomic coordinates and equivalent isotropic displacement parameters for La2YbCuS5 are given in Table 8.2. Powder X-ray Diffraction. Powder X-ray diffraction patterns were collected with a Rigaku Miniflex powder X-ray diffractometer using Cu K? (? = 1.54056 ?) radiation. Magnetic Susceptibility Measurements. Magnetic data were measured on powders in gelcap sample holders with a Quantum Design MPMS 7T magnetometer/susceptometer between 2 and 300 K and in applied fields up to 7 T. DC susceptibility measurements were made under zero-field-cooled conditions with an applied field of 0.1 T. Susceptibility values were corrected for the sample diamagnetic contribution according to Pascal?s constants43 as well as for the sample holder diamagnetism. ?p values were obtained from extrapolations from fits between 100 to 300 K. (Magnetic susceptibility measurements were performed by Eun Sang Choi and James S. Brooks at Florida State University) 20 5 Table 8.1. Crystallographic Data for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se). Formula La2YbCuSe5 Ce2YbCuSe5 La2YbCuS5 Ce2YbCuS5 Pr2YbCuS5 Nd2YbCuS5 Sm2YbCuS5 fw 909.20 911.62 674.70 677.12 678.70 685.36 697.58 Color black black black black black black black Crystal System orthorhombic orthorhombic orthorhombic orthorhombic orthorhombic orthorhombic orthorhombic Space group Pnma (No. 62) Pnma (No. 62) Pnma (No. 62) Pnma (No. 62) Pnma (No. 62) Pnma (No. 62) Pnma (No. 62) a (?) 12.1326(11) 12.1113(7) 11.615(4) 11.5616(13) 11.547(2) 11.5466(8) 11.5323(8) b (?) 4.1119(4) 4.0780(3) 3.9662(13) 3.9304(4) 3.9071(8) 3.8927(3) 3.8531(3) c (?) 17.6653(16) 17.5714(11) 16.923(6) 16.8423(18) 16.795(3) 16.7597(11) 16.6470(12) V (?3) 881.29(14) 867.85(10) 779.6(5) 765.34(14) 757.7(3) 753.31(9) 739.71(9) Z 4 4 4 4 4 4 4 T (K) 193 193 193 193 193 193 193 ? (?) 0.71073 0.71073 0.71073 0.71073 0.71073 0.71073 0.71073 ?calcd (g cm?3) 6.853 6.977 5.748 5.877 5.950 6.043 6.264 ? (cm?1) 429.51 442.61 265.47 277.73 28.8.98 299.16 323.04 R(F)a 0.0369 0.0290 0.0261 0.0242 0.0317 0.0250 0.0263 Rw(Fo2) b 0.0966 0.0757 0.0628 0.0588 0.0736 0.0566 0.0665 a ( )R F F F F= ?? ? o c o for Fo 2 > 2?(Fo2). b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 206 Table 8.2. Atomic Coordinates and Equivalent Isotropic Displacement Parameters for La2YbCuS5. Atom (site) x y z Ueq (?2)a La(1) 0.97977(4) -0.25 0.82545(3) 0.00691(14) La(2) 0.86754(4) 0.25 0.40038(3) 0.00686(14) Yb(1) 0.80240(3) 0.25 0.63683(2) 0.01060(13) Cu(1) 0.6657(2) -0.25 0.51632(13) 0.0144(4) Cu(2) 0.5960(3) -0.25 0.50288(17) 0.0177(6) S(1) 1.01210(17) 0.25 0.69895(13) 0.0066(4) S(2) 0.81356(17) -0.75 0.88370(13) 0.0072(4) S(3) 0.88403(18) -0.25 0.53444(13) 0.0104(4) S(4) 0.59555(18) 0.25 0.57265(13) 0.0084(4) S(5) 0.73861(18) -0.25 0.72916(13) 0.0076(4) a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 207 UV-vis-NIR Diffuse Reflectance Spectroscopy. The diffuse reflectance spectra for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) were measured from 200 to 2500 nm using a Shimadzu UV3100 spectrophotometer equipped with an integrating sphere attachment. The Kubelka-Munk function was used to convert diffuse reflectance data to absorption spectra.44 RESULTS AND DISCUSSION Structures of Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se). The compounds Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) are isotypic and crystallize in the centrosymmetric orthorhombic space group Pnma. As illustrated in Figure 8.1, the structure can be described as composed of one-dimensional [YbCuQ5]6- ribbons running down the [010] direction and are separated by Ln3+ ions. It includes two crystallographically unique Ln atoms, one octahedral Yb site, and two Cu positions. The polyhedra of Ln and Cu atoms are shown in Figure 8.2 and 8.3. Both Ln3+ are eight- coordinate and occur as bicapped trigonal prisms. Two Cu atoms are very close to each other with the Cu(1)-Cu(2) distance being 0.841(3) ? in case of La2YbCuS5 (see Table 8.3). Obviously, these two sites cannot be occupied simultaneously. The occupancy of Cu(1) ranges from 0.19 to 0.54. Each Cu has a highly distorted tetrahedral environment. For example, the four Cu(1)-S distances for La2YbCuS5 are 2.257(3) ?, 2.3464(18) ?, 2.3464(18) ?, and 2.544(4) ?, and S-Cu(1)-S angles are in the range 90.78(11)- 116.16(8)? (see Table 8.3). These bond distances are closer to the average values for Cu with triangular coordination (e.g. 2.33 ? in Cu2S45,46) than the ones for Cu with tetrahedral coordination (2.44 ?, according to Shannon47). This similar 208 Figure 8.1. A view the three-dimensional structure of La2YbCuS5 along the b axis. La-S bonds have been omitted for clarity. a c 209 Figure 8.2. Illustrations of the coordination environments for La ions in La2YbCuS5. 210 Figure 8.3. Illustrations of the coordination environments for Cu ions in La2YbCuS5. 21 1 Table 8.3. Selected Bond Distances (?) and angles (deg) for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se). Formula La2YbCuSe5 Ce2YbCuSe5 La2YbCuS5 Ce2YbCuS5 Pr2YbCuS5 Nd2YbCuS5 Sm2YbCuS5 Ln(1)-Q(1) ?2 3.0560(12) 3.0293(8) 2.9421(18) 2.9108(15) 2.894(2) 2.8834(16) 2.8574(17) Ln(1)-Q(2) ?2 3.0613(12) 3.0428(9) 2.9380(17) 2.9185(16) 2.899(2) 2.8959(16) 2.8691(18) Ln(1)-Q(4) ?2 3.0701(13) 3.0462(9) 2.9522(18) 2.9280(16) 2.908(2) 2.8969(17) 2.8689(18) Ln(1)-Q(5) 3.2994(18) 3.2871(12) 3.145(2) 3.150(2) 3.150(3) 3.110(2) 3.017(2) Ln(1)-Q(5) 3.3571(18) 3.3352(12) 3.241(2) 3.188(2) 3.171(3) 3.201(2) 3.268(2) Ln(2)-Q(1) ?2 3.0678(12) 3.0368(8) 2.9518(17) 2.9234(15) 2.909(2) 2.8954(16) 2.8620(17) Ln(2)-Q(2) ?2 3.0231(12) 3.0037(8) 2.9046(16) 2.8820(16) 2.872(2) 2.8597(16) 2.8373(17) Ln(2)-Q(3) ?2 3.1265(13) 3.1107(9) 3.0193(19) 2.9998(17) 2.975(2) 2.9587(18) 2.9345(19) Ln(2)-Q(3) 3.2154(17) 3.1962(12) 3.089(2) 3.064(2) 3.050(3) 3.046(2) 3.023(2) Ln(2)-Q(5) 3.2789(17) 3.2486(12) 3.149(2) 3.116(2) 3.104(3) 3.096(2) 3.067(2) Yb(1)-Q(1) 2.7830(16) 2.7696(11) 2.653(2) 2.644(2) 2.646(3) 2.648(2) 2.645(2) Yb(1)-Q(3) ?2 2.9185(12) 2.8951(8) 2.7989(17) 2.7824(16) 2.780(2) 2.7769(17) 2.7543(18) Yb(1)-Q(4) 2.7597(16) 2.7547(11) 2.637(2) 2.630(2) 2.631(3) 2.633(2) 2.632(2) 21 2 Yb(1)-Q(5) ?2 2.7572(11) 2.7538(7) 2.6311(16) 2.6274(14) 2.625(2) 2.6240(15) 2.6193(16) Cu(1)-Q(2) 2.380(4) 2.360(4) 2.257(3) 2.244(3) 2.238(5) 2.226(4) 2.205(7) Cu(1)-Q(3) 2.583(4) 2.591(4) 2.554(4) 2.556(4) 2.567(6) 2.610(6) 2.656(11) Cu(1)-Q(4) ?2 2.466(2) 2.468(2) 2.3464(18) 2.3459(18) 2.351(3) 2.347(3) 2.345(5) Cu(2)-Q(2) 2.404(4) 2.404(2) 2.274(3) 2.277(3) 2.278(4) 2.287(3) 2.281(3) Cu(2)-Q(4) ?2 2.405(2) 2.4142(13) 2.3079(19) 2.3018(16) 2.302(2) 2.3043(16) 2.3081(16) Cu(2)-Q(4) 2.590(5) 2.564(3) 2.566(4) 2.526(4) 2.516(5) 2.502(3) 2.497(3) Q(2)-Cu(1)-Q(3) 92.51(13) 91.59(13) 90.78(11) 90.12(12) 89.30(19) 87.47(17) 85.4(3) Q(2)-Cu(1)-Q(4) 114.14(10) 115.09(10) 116.16(8) 116.85(9) 117.52(14) 118.66(12) 120.2(2) Q(3)-Cu(1)-Q(4) 110.67(11) 111.07(11) 107.23(9) 107.82(10) 108.37(16) 107.91(15) 107.9(3) Q(4)-Cu(1)-Q(4) 112.93(15) 111.39(15) 115.38(13) 113.80(14) 112.4(2) 112.06(18) 110.5(3) Q(2)-Cu(2)-Q(4) 88.36(14) 87.97(8) 87.63(13) 87.19(11) 86.47(14) 86.25(10) 85.45(10) Q(2)-Cu(2)-Q(4) 115.55(11) 115.49(6) 117.06(9) 117.29(8) 117.88(10) 117.92(8) 118.57(7) Q(4)-Cu(2)-Q(4) 107.53(13) 109.48(7) 104.63(11) 105.60(10) 106.04(12) 106.82(9) 108.07(8) Q(4)-Cu(1)-Q(4) 117.45(19) 115.26(10) 118.47(16) 117.24(13) 116.09(18) 115.27(13) 113.16(12) 213 feature of Cu atoms have been found in other lanthanide copper chalcogenides including Sm3CuSe6,21,22 Gd3Cu2Te7,24 LaCu0.28Te2,25 and LnCuxTe2 (Ln = La, Nd, Sm, Gd and Dy),26 in which two disordered Cu sites are equivalent and the tetrahedra are less distorted. In order to better describe the connectivity among the polyhedra, Cu(1) and Cu(2) atoms are considered as one average Cu site sitting in the cavities of a distorted trigonal bipyramid, as shown in Figure 8.3. Then the one-dimensional [YbCuQ5]6- ribbons, which are shown Figure 8.4a, consist of two single [YbQ6] octahedral chains with one double [CuQ5] trigonal bipyramidal chain in the middle. All three chains connect with each other via edge sharing. Within the [YbQ6] octahedral chain, each unit shares edges with two neighbors along the chain direction, while each [CuQ5] trigonal bipyramid shares corner along the b axis and shares edge with adjacent identical chains. The connectivity of [CuQ5] trigonal bipyramids in these compounds is quite different from other known examples. For Gd3Cu2Te7,24 the [CuTe5] trigonal bipyramids share corners with four [CuTe4] tetrahedra within two-dimensional [Cu2Te5] layers (Figure 8.4b). In Figure 8.4c, [CuSe5] trigonal bipyramids in compound Sm3CuSe621 shares edges to form one-dimensional single chains along the [010] direction, while Figure 8.4d shows a two-dimensional [CuxTe2] layer of LaCu0.28Te225 or LnCuxTe2 (Ln = La, Nd, Sm, Gd and Dy)26, constructed from [CuTe5] trigonal bipyramids sharing edges with four close neighbors and sharing corners with other four bipyramidal units. It is noted that La3CuO2S340 adopts similar formula, condensed three-dimensional structure, and space group as Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se). But the structure of La3CuO2S3, which is illustrated in Figure 8.5, has three eight-coordinate La3+ ions with bicapped trigonal prismatic geometry and one ordered tetrahedral 214 Figure 8.4. Depictions of various connectivities of [CuQ5] (Q = S, Se, Te) trigonal bipyramids in different compounds: a) La2YbCuS5; b) Gd3Cu2Te724; c) Sm3CuSe621; d) LaCu0.28Te225. a) b) 215 Figure 8.4. Depictions of various connectivities of [CuQ5] (Q = S, Se, Te) trigonal bipyramids in different compounds: a) La2YbCuS5; b) Gd3Cu2Te724; c) Sm3CuSe621; d) LaCu0.28Te225. c) d) 216 Figure 8.5. Unit cell of La3CuO2S340 viewed along the b axis. a c 217 Cu+ site. It has higher symmetry in terms of [CuS4] tetrahedra. The bond distances of LnQ8 and YbQ6 for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) are normal, shown in Table 8.3. For example, in case of La2YbCuS5, La-S distances range from 2.9046(16) ? to 3.241(2) ? and Yb-S distances are in the range of 2.6311(16) ? to 2.7989(17) ?. They are comparable to Shannon?s data,47 3.00 ? for LaS8 and 2.70 ? for YbS6. Since there are no Q-Q bonds and Cu-Cu interactions, the oxidation states of Ln2YbCuQ5 can be assigned as +3/+3/+1/-2. Magnetic Susceptibility. The inverse molar Ln magnetic susceptibilities for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd; Q = S, Se) in the range of 2-300 K are shown in Figure 8.6-8.8. All of three compounds show deviation from the Curie-Weiss law around 70 K. This could be due to the crystal-field splitting of the ground state of magnetic ions. There is no evidence of magnetic ordering for Ce2YbCuSe5, La2YbCuS5, Ce2YbCuS5, and Pr2YbCuS5. In contrast, the magnetic susceptibilities for La2YbCuSe5 and Nd2YbCuS5 show gradual changes in the slope at low temperature, which may indicate the short- range antiferromagnetic ordering. Magnetic measurements have shown that Sm2YbCuS5 exhibits spin glass behavior below 7 K. In order to draw a conclusion, further investigations, e.g. heat capacity measurements, are underway. Therefore the magnetic properties of Sm2YbCuS5 will not be discussed here. Table 8.4 presents the magnetic parameters that were determined from the fit from the Curie-Weiss regions. The Weiss constants for these compounds are all quite negative; that might be due to the crystal field effect as well. The experimental effective magnetic moments per formula unit are very close to the theoretical values based on the free Ln3+ ions.52 Optical Properties. The electronic structures of interlanthanide copper 218 Figure 8.6. Molar magnetic susceptibility vs temperature between 2 and 300 K for La2YbCuS5, Ce2YbCuS5, and Pr2YbCuS5. Data were taken under an applied magnetic field of 0.1 T. The straight line represents the fit to Curie-Weiss law in the range of 100- 300 K. 0 50 100 150 200 250 300 0 40 80 120 160 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) La2YbCuS5 Ce2YbCuS5 Pr2YbCuS5 219 0 50 100 150 200 250 300 0 15 30 45 60 2 4 6 8 102 4 6 Temperature ????-1 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) Nd2YbCuS5 Figure 8.7. The temperature dependence of the reciprocal molar magnetic susceptibility for Nd2YbCuS5 under an applied magnetic field of 0.1 T between 2 and 300 K. Inset shows the molar magnetic susceptibility at low temperature. 220 Figure 8.8. Molar magnetic susceptibility as a function of temperature for La2YbCuSe5 and Ce2YbCuS5 under an applied magnetic field of 0.1 T between 2 and 300 K. Inset shows the molar magnetic susceptibility for La2YbCuSe5 at low temperature. 0 50 100 150 200 250 300 0 40 80 120 160 2 4 6 8 10 4 6 8 10 Temperature ????-1 ?-1 (O e m ol for mu la un it/e mu ) Temperature, T (K) La2YbCuSe5 Ce2YbCuSe5 221 Table 8.4. Magnetic Parameters for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se). Formula Pcal/?B Peff/?B ?p R2 La2YbCuSe5 4.54 4.229(9) -38(1) 0.99959 Ce2YbCuSe5 5.79 5.76(1) -54(1) 0.99962 La2YbCuS5 4.54 4.372(6) -68.4(8) 0.99981 Ce2YbCuS5 5.79 5.31(3) -47(2) 0.99794 Pr2YbCuS5 6.80 7.00(2) -44(2) 0.99926 Nd2YbCuS5 6.84 6.87(1) -46.0(8) 0.99979 a Pcal and Peff : calculated52 and experimental effective magnetic moments per formula unit. b Weiss constant (?p) and goodness of fit (R2) obtained from high temperature (100- 300 K) data 222 chalcogenides are expected to be different from the parent interlanthanide chalcogenides after introducing more soft and electronegative Cu+ ions into the system. Cu+ ions prefer to bind larger chalcogenides to form more covalent bonds. This is best exhibited by the LnCuOQ (Ln = La, Ce, Pr, Nd; Q = S, Se, Te) 28-38 series that consists of alternately stacked [Cu2Q2]2- layers and [Ln2O2]2+ layers. The optical properties of LnCuOQ (Ln = La, Pr, Nd; Q = S, Se, Te) are mainly determined by [Cu2Q2]2- layers, e.g. the valence band of LaCuOTe38 is filled with Cu 3d and Te 5p states and the conduction band is composed of Cu 4s, Te 5p and La 5d states; La 4f states are well above the Fermi energy. LnCuOQ (Ln = La, Pr, Nd; Q = S, Se) are determined to be direct allowed p-type semiconductors with wide band gaps, while corresponding tellurides have indirect-type gaps.35,38 In contrast, recent studies have shown that Ce 4f states in CeCuOS and CeCu0.75OS compounds are fully spin-polarized and delocalized result in black colors and much smaller band gaps.36 Another series compounds with layered structures, LnCuS2 (Ln = La, Nd, Sm, Gd, Dy, Ho, Yb, Lu, Y), are wide band gap p-type semiconductors too.2 Substitution of larger chalcogenides narrows the band gaps by increasing the covalency in the Cu-Q bonds to lift the Fermi levels.27,35 The optical properties for Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) were measured by UV-vis-NIR diffuse reflectance spectroscopy. The spectra are presented in Figure 8.9. The band gaps of La2YbCuSe5, Ce2YbCuSe5, La2YbCuS5, Ce2YbCuS5, Pr2YbCuS5, Nd2YbCuS5, and Sm2YbCuS5 are 1.15 eV, 1.05 eV, 1.45 eV, 1.37 eV, 1.25 eV, 1.35 eV, and 1.28 eV respectively. Apparently selenides have smaller band gaps than sulfides due to the higher energy of Se 4p orbitals. The two lanthanum compounds have somewhat larger values than the rest. This means 4f states of other lanthanides 223 Figure 8.9. UV-vis diffuse reflectance spectra of Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se). 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 ?/ s ( arb ita ry un its ) Energy (eV) La2YbCuSe5 Ce2YbCuSe5 La2YbCuS5 Ce2YbCuS5 Pr2YbCuS5 Nd2YbCuS5 Sm2YbCuS5 224 besides La have some contribution to the electronic structures around the Fermi levels. Unlike most lanthanide and interlanthanide chalcogenides, among which cerium compounds have the smallest band gaps due to the high energy of 4f1 electron,48-51 Ce2YbCuS5 has slight larger values than the Pr, Nd, and Sm cases. This indicates that the band structures of Ln2YbCuS5 are more controlled by Cu energy levels as expected, because of the more covalent Cu-S bonds. The band gaps of Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) are consistent with the observed black colors and are reasonable compared to SmCuS2 (2.1 eV)6 and La3CuO2S3 (2.01 eV)40, which have less condensed structure and lower energy O 2p orbitals respectively. Overall, the electronic structures of Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) are tunable based on the choices of Ln and Q. CONCLUSIONS A new series of ordered quaternary interlanthanide copper chalcogenides, Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se), have been synthesized using Sb2Q3 (Q = S, Se) fluxes at 900 ?C. Compared to other known lanthanide copper chalcogenides, these compounds crystallize in a new structure type that is realized by including two different lanthanides with large size difference, which tend to have distinct coordination environments. The three-dimensional complex structure of Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) include two crystallographically unique eight-coordinate Ln atoms, one octahedral Yb site, and two Cu positions. These two Cu sites closely reside in the trigonal bipyramidal cavities formed by Q2- anions that cannot be occupied simultaneously. The structure includes one-dimensional [YbCuQ5]6- ribbons along the b 225 axis that are separated by larger Ln3+ ions. Ce2YbCuSe5, La2YbCuS5, Ce2YbCuS5, and Pr2YbCuS5 are Curie-Weiss paramagnets. La2YbCuSe5 and Nd2YbCuS5 have short range antiferromagnetic ordering at low temperature. The UV-vis-NIR diffuse reflectance measurements show these compounds to be wide band-gap semiconductors. 226 REFERENCES 1. Julien-Pouzol, M.; Jaulmes, S.; Mazurier, A.; Guittard, M. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1981, 37, 1901. 2. Murugesan, T.; Gopalakrishnan, J. Indian J. Chem. 1983, 22A, 469. 3. Guseinov, G. G.; Amirov, A. S.; Mamedov, K. S. Dokl. Akad. Nauk Az. SSR 1984, 40, 62. 4. Gschneidner, K. A., Jr., Eyring, L. R., Eds. Handbook on the Physics and Chemistry of Rare Earths; North-Holland Physics Publishing: New York, 1984; Vol. 6. 5. Wang, Yarong; Sato, Nobuaki; Fujino, Takeo Mater. Res. Bull. 2001, 36, 1029. 6. Llanos, Jaime; Mujica, Carlos; Sanchez, Victor; Schnelle, Walter; Cardoso-Gil, Raul. J. Solid State Chem. 2004, 177, 1388. 7. Ijjaali, I.; Mitchell, K.; Ibers, J. A. J. Solid State Chem. 2004, 177, 760. 8. Julien-Pouzol, M.; Guittard, M.; Mazurier, A. C. R. Acad. Sci. 1970, 271, 1317. 9. Julien-Pouzol, M.; Guittard, M. Ann. Chim. (Paris) 1972, 7, 253. 10. Lauxmann, Petra; Schleid, Thomas. Z. Anorg. Allg. Chem. 2000, 626, 1608. 11. Dismukes, J. P.; Smith, R. T.; White, J. G. J. Phys. Chem. Solids 1971, 32, 913. 12. Ballestracci, R.; Bertaut, E. F. C. R. Acad. Sci. Paris 1965, 261, 5064. 13. Ballestracci, R.; Bertaut, E. F. Bull. Soc. Miner. Crystallogr. 1965, 88, 575. 14. Guymont, M.; Thomas, A.; Julien-Pouzol, M.; Jaulmes, S.; Guittard, M. Phys. Status Solidi A 1990, 121, 21. 15. Onoda, M.; Chen, X. A.; Sato, A.; Wada, H. J. Solid State Chem. 2000, 152, 332. 16. Lemoine, P.; Carr?, D.; Guittard, M. Acta Cryst. 1986, C42, 390. 227 17. Furuuchi, F.; Wakeshima, M.; Hinatsu, Y. J. Solid State Chem. 2004, 177, 3853. 18. Eu2CuSe3, isostructure to Eu2CuS3, unpublished results. 19. Wakeshima, M.; Furuuchi, F. Hinatsu, Y. J. Phys.: Condens. Matter 2004, 16 5503. 20. Strobel, S.; Schleid, T. Angew. Chem. Int. Ed. 2003, 42, 4911. 21. Strobel, S.; Schleid, T. J. Solid State Chem. 2003, 171, 424. 22. Ijjaali, I.; Ibers, J. A. Acta Cryst.2004, E39, i89. 23. Patschke, R.; Brazis, P.; Kannewurf, C. R.; Kanatzidis, M. G. J. Mater. Chem. 1999, 9, 2293. 24. Huang, F. Q.; Ibers, J. A. J. Solid State Chem. 2001, 159, 186. 25. Dung, N.-H.; Pardo, M.-P.; Boy, P. Acta Cryst.1983, C39, 668. 26. Huang, F. Q.; Brazis, P.; Kannewurf, C. R.; Ibers, J. A. J. Am. Chem. Soc. 2000, 122, 80. 27. Huang, F. Q.; Ibers, J. A. Inorg. Chem. 1999, 38, 5978. 28. Popovkin, B. A.; Kusainova, A. M.; Dolgikh, V. A.; Aksel?rud, L. G. Russ. J. Inorg. Chem. (Transl. of Zh. Neorg. Khim.) 1998, 43, 1471. 29. Charkin, D. O.; Akopyan, A. V.; Dolgikh, V. A. Russ. J. Inorg. Chem. (Transl. of Zh. Neorg. Khim.) 1999, 44, 833. 30. Ueda, K.; Inoue, S.; Hirose, S.; Kawazoe, H.; Hosono, H. Appl. Phys. Lett. 2000, 77, 2701. 31. Ueda, K.; Inoue, S.; Hosono, H.; Sarukura, N.; Hirano, H. Appl. Phys. Lett. 2001, 78, 2333. 32. Inoue, S.; Ueda, K.; Hosono, H.; Hamada, N. Phys. Rev. B 2001, 64, 245211. 228 33. Ueda, K.; Takafuji, K.; Hiramatsu, H.; Ohta, H.; Kamiya, T.; Hirano, M.; Hosono, H. Chem. Mater. 2003, 15, 3692. 34. Ueda, K.; Takafuji, K.; Hosono, H. J. Solid State Chem. 2003, 170, 182. 35. Ueda, K.; Hosono, H.; Hamada, N. J. Phys.: Condens. Matter 2004, 16, 5179. 36. Chan, G. H.; Deng, B.; Bertoni, M.; Ireland, J. R.; Hersam, M. C.; Mason, T. O.; Van Duyne, R. P.; Ibers, J. A. Inorg. Chem. 2006, 45, 8264. 37. Ueda, K.; Hiramatsu, H.; Hirano, M.; Kamiya, T.; Hosono, H. Thin Solid Films 2006, 496, 8. 38. Liu, M. L.; Wu, L. B.; Huang, F. Q.; Chen, L. D.; Ibers, J. A. J. Solid State Chem. 2007, 180, 62. 39. Huang, F. Q.; Brazis, P.; Kannewurf, C. R.; Ibers, J. A. J. Solid State Chem. 2000, 155, 366. 40. Ijjaali, I.; Haynes, C. L.; Mcfarland, A. D.; Van Duyne, R. P.; Ibers, J. A. J. Solid State Chem. 2003, 172, 257. 41. Sheldrick, G. M. SHELXTL PC, Version 6.12, An Integrated System for Solving, Refining, and Displaying Crystal Structures from Diffraction Data; Siemens Analytical X-Ray Instruments, Inc.: Madison, WI 2001. 42. Sheldrick, G. M. SADABS 2001, Program for absorption correction using SMART CCD based on the method of Blessing: Blessing, R. H. Acta Crystallogr. 1995, A51, 33. 43. Mulay, L. N.; Boudreaux, E. A. Theory and Applications of Molecular Diamagnetism; Wiley?Interscience: New York, 1976. 229 44. Wendlandt, W. W.; Hecht, H. G.: Reflectance Spectroscopy; Interscience Publishers: New York, 1966. 45. Evans, H. T., Jr. Nature 1971, 232, 69. 46. Evans, H. T., Jr. Z. Kristallogr. 1979, 150, 299. 47. Shannon, R. D. Acta Cryst.1976, A32, 751. 48. Prokofiev, A. V.; Shelykh, A. I.; Golubkov, A. V.; Smirnov, I. A. J. Alloys Compd. 1995, 219, 172. 49. Prokofiev, A. V.; Shelykh, A. I.; Melekh, B. T. J. Alloys Compd. 1996, 242, 41. 50. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. Chem. Mater. 2007, 19, 567. 51. Jin, G. B.; Choi, E. S.; Guertin, R. P.; Brooks, J. S.; Bray, T. H.; Booth, C. H.; Albrecht-Schmitt, T. E. J. Solid State Chem. in press. 52. Kittel, C. Introduction to Solid State Physics, 6th Edition; Wiley: New York, 1986. 230 CHAPTER 9 PARTIALLY-FILLED MIXED-LANTHANIDE VARIANTS OF THE K2Tm23.33S36 STRUCTURE-TYPE: STRUCTURE AND PROPERTIES OF CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La-Nd, Sm-Yb; y = 0.26 ? 0.33) ABSTRACT The mixed-lanthanide sulfides, CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb; y = 0.26 ? 0.33), have been grown as single crystals from a CsCl reactive flux at 900?C. Single-crystal X-ray diffraction experiments on all of these compounds reveal that they crystallize in the hexagonal space group P63/m with a structure similar to that found for K2Tm23.33S36. CsxLnyYbS2 adopts a three- dimensional channel structure with two different hexagonal channels. One of these channels is nearly filled with Cs+ cations while the other is only partially filled. The Ln3+ cations are bound by bridging S2? anions to create two distinct local environments: one where the smaller Yb3+ cations preferentially reside in six-coordinate distorted octahedral sites, and a second where the larger Ln3+ are located in a nine-coordinate tricapped trigonal prismatic site. Owing to the hexagonal lattice that is adopted by these compounds, spin-frustration was investigated by means of magnetic susceptibility measurements on polycrystalline samples. The ?W values, some of which are large, are 231 all negative except for Tb, and range from -14.2(3) to -45(2) K. For Cs0.15Tb0.29YbS2, ferromagnetic interactions are found with ? = 3.4(4) K. XANES measurements for Cs0.14Ce0.30YbS2, Cs0.14Nd0.29YbS2, and Cs0.15Sm0.29YbS2 have shown that all lanthanides in these samples are trivalent. The optical band gaps for CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb; y = 0.26 ? 0.33) are in the range of 2.1 and 2.4 eV. INTRODUCTION The arrangement of paramagnetic ions in triangular arrays gives rise to the possibility of geometric spin-frustration, where long-range magnetic ordering is not predicted to occur.1,2 Spin-frustration is particularly well-known for systems with kagom? lattices, but is also well recognized in other high-symmetry structure types containing triangular moieties such as spinels3-14 and pyrochlores.15-36 In the partially zinc- substituted paratacamite, Zn0.33Cu3.67(OH)6Cl2, which possesses kagom?-type layers, spin-frustration is exceedingly large with a spin-frustration parameter f > 157 (f = |?CW|/Tc).37-39,40 Despite well-established predictions to the contrary, long-range magnetic ordering is known to occur in certain phases with triangular sublattices, such as Fe- jarosites, AFe3(OH)6(SO4)2 (A = Na, K, Rb, NH4+, H3O+, Ag+, Tl+, and Pb2+), where the Dzyaloshinksy-Moriya (DM)41-43 interaction allows for weak in-plane ferromagnetism and long-range ordering.44-48 These compounds are of particular interest because they can be prepared as large single crystals that have allowed for the discernment of spin chirality via neutron scattering measurements.45 More recently, materials with one- 232 dimensional arrays of triangular units have been investigated such as the rare-earth olivines, ZnLn2S4 (Ln = Er, Tm, Yb), and these too show evidence for spin-frustration.49 The K2Tm23.33S36 structure type represents an unusual case for investigating spin- frustration in that there are two different interpenetrating honeycomb layers formed from Tm3+ in two different environments.50 In this structure, one of the rare-earth ions resides in a site with seven close sulfide neighbors, and the second in a site with six surrounding sulfide anions. This allows for the possibility of using two different rare-earth ions of different sizes to preferentially occupy the two different sites, with larger rare-earth ions potentially favoring the site with higher coordination. In addition, this allows for the investigation of the electronic properties in systems where there are different numbers of f electrons on the metal centers. To this end we have synthesized the partially-filled mixed-lanthanide variants of the K2Tm23.33S36 structure-type, CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb; y = 0.26 ? 0.33). We provide structural, magnetic susceptibility, XANES and optical data for these compounds. EXPERIMENT Starting Materials. La (99.9%, Alfa-Aesar), Ce (99.9%, Alfa-Aesar), Pr (99.9%, Alfa-Aesar), Nd (99.9%, Alfa-Aesar), Sm (99.9%, Alfa-Aesar), Eu (99.9%, Cerac), Gd (99.9%, Alfa-Aesar), Tb (99.9%, Alfa-Aesar), Dy (99.9%, Alfa-Aesar), Ho (99.9%, Alfa- Aesar), Er (99.9%, Alfa-Aesar), Tm (99.9%, Alfa-Aesar), Yb (99.9%, Alfa-Aesar), CsCl (99.9%, Alfa-Aesar), and S (99.5%, Alfa-Aesar) were used as received. 233 Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Eu, Gd). Ln, Yb, and S in a ratio of 1:3:6 with a total mass of 0.0750 g was mixed together with 0.1500 g of CsCl and then loaded into fused-silica ampoules in an Ar-filled glovebox. The tubes were sealed under vacuum and heated using the following profiles: 2 ?C/min to 500 ?C (1 h), 0.5 ?C/min to 900 ?C (5 d), 0.04 ?C/min to 400 ?C (2 d), and 0.5 ?C/min to 24 ?C. The acicular crystals were isolated from the excess CsCl flux by washing with water and methanol. The yields are close to 100% based on Yb except for Eu which is Cs0.16Eu0.33YbS2. The products are yellow (Ln= La, Pr, Nd, Sm), red (Ln= Ce, Eu), or brown (Ln= Gd) respectively. Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = Tb, Dy, Ho, Er, Tm, Yb). The same procedure as above was used to prepare Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = Tb, Dy, Ho, Er, Tm, Yb) except that the Ln, Yb, S/Yb, and S ratio of 1:2:5/1:2 was used, and the total mass of these reactants was 0.0500 g. These starting materials were mixed with 0.1500 g of CsCl. The following heating profile was followed: 2 ?C/min to 500 ?C (1 h), 0.5 ?C/min to 900 ?C (7 d), 0.02 ?C/min to 400 ?C (2 d), and 0.5 ?C/min to 24 ?C. The desired products are obtained as brown needles. All the compounds are stable in moist air for more than one year. Powder X-ray diffraction measurements confirmed the phase purity of CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La, Ce, Pr, Nd, Sm, Gd, Tb, Dy, Ho, Er, Tm, Yb; y = 0.26 ? 0.33) by comparison with powder patterns calculated from the single crystal X-ray structures. The major product of the reaction with Eu is EuYb2S4,53 and Cs0.16Eu0.33YbS2 is only isolated as a minor product. Semi-quantitative SEM/EDX analyses were performed using a JEOL 840/Link Isis instrument. Cs, Ln, Yb and S percentages were calibrated against standards. 234 Crystallographic Studies. Single crystals of CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb; y = 0.26 ? 0.33) were mounted on glass fibers using epoxy, and optically aligned on a Bruker SMART APEX CCD X-ray diffractometer. The diffraction data were collected at 193 K using an Oxford Cryostat. Intensity measurements were performed using graphite-monochromated Mo K? radiation from a sealed tube and monocapillary collimator. SMART (v 5.624) was used for preliminary determination of the cell constants and data collection control. The intensities of reflections of a sphere were collected by a combination of 3 sets of exposures (frames). Each set had a different ? angle for the crystal and each exposure covered a range of 0.3? in ?. A total of 1800 frames were collected with an exposure time per frame of 30 s for all compounds. Determination of integrated intensities and global refinement were performed with the Bruker SAINT (v 6.02) software package using a narrow-frame integration algorithm. A face-indexed numerical absorption correction was initially applied using XPREP.51 These files were subsequently treated with a multiscan absorption correction by SADABS.52 The program suite SHELXTL (v 6.12) was used for space group determination (XPREP), direct methods structure solution (XS), and least-squares refinement (XL).51 The final refinements included anisotropic displacement parameters for all atoms and secondary extinction. The formulae for these compounds were arrived at by refinement of the occupancy of each site, yielding the following: Cs0.14La0.30YbS2, Cs0.14Ce0.30YbS2, Cs0.14Pr0.30YbS2, Cs0.14Nd0.29YbS2, Cs0.15Sm0.29YbS2, Cs0.16Eu0.33YbS2, Cs0.15Gd0.29YbS2, Cs0.15Tb0.29YbS2, Cs0.16Dy0.28YbS2, Cs0.16Ho0.30YbS2, Cs0.17Er0.32YbS2, Cs0.15Tm0.26YbS2, 235 and Cs0.15Yb0.26YbS2. The cell reported herein represents a sub-cell of the K2Tm23.33S36 structure-type.50 This sub-cell can be inscribed within the larger cell by a 30? rotation and division of the a and b axes by ?2. Some, but not all of the above compounds can be indexed to the K2Tm23.33S36 cell. For the sake of consistency we have used the same cell to describe all compounds. These compounds suffer from the classic problems of disorder in hexagonal systems along the c axis that is most evident from the elongation of the Cs thermal parameter along this axis. It is important to note that while we have modeled these structures with the two different lanthanide sites each containing only one lanthanide element, the chance that there is partial occupancy of both lanthanide elements at both sites is significant, particularly as the size of the two different lanthanides becomes more similar. We have refined models with this type of disorder, and they do not result in improvements in the structures, all of which are already of high resolution. Some crystallographic details are given in Table 9.1a and Table 9.1b. Atomic coordinates and equivalent isotropic displacement parameters for these compounds can be found in Table 9.2. Powder X-ray Diffraction. Powder X-ray diffraction patterns were collected with a Rigaku Miniflex powder X-ray diffractometer using Cu K? (? = 1.54056 ?) radiation. Magnetism. Magnetic susceptibility measurements were made between 2 and 300 K under field-cooled conditions in an applied field of 5 kG on a Quantum Design 7T?MPMS SQUID magnetometer. ?p values were obtained from extrapolations from 23 6 Table 9.1a. Crystallographic Data for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Eu, Gd). Formula Cs0.14La0.30YbS2 Cs0.14Ce0.30YbS2 Cs0.14Pr0.30YbS2 Cs0.14Nd0.29YbS2 Cs0.15Sm0.29YbS2 Cs0.16Eu0.33YbS2 Cs0.15Gd0.29YbS2 fw 297.43 298.01 299.17 298.04 300.76 308.64 303.40 Color Yellow Red Yellow Yellow Yellow Red Brown Crystal System Hexagonal Hexagonal Hexagonal Hexagonal Hexagonal Hexagonal Hexagonal Space group P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) a (?) 12.2455(7) 12.2147(14) 12.2177(8) 12.2119(8) 12.1909(9) 12.2483(13) 12.1906(13) c (?) 3.9393(3) 3.9216(6) 3.9223(4) 3.9140(4) 3.9085(4) 3.9218(6) 3.8976(6) V (?3) 511.57(6) 506.71(11) 507.05(7) 505.50(7) 503.05(7) 509.53(11) 501.62(11) Z 6 6 6 6 6 6 6 T (K) 193 193 193 193 193 193 193 ?(?) 0.71073 0.71073 0.71073 0.71073 0.71073 0.71073 0.71073 ?calcd (g cm?3) 5.793 5.860 5.879 5.874 5.957 6.035 6.026 ? (cm?1) 334.72 340.58 344.06 345.81 354.30 361.56 362.36 R(F)a 0.0204 0.0216 0.0235 0.0229 0.0315 0.0355 0.0384 Rw(Fo2)b 0.0471 0.0545 0.0550 0.0581 0.0759 0.0907 0.0933 a ( )R F F F F= ?? ? o c o . b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 23 7 Table 9.1b. Crystallographic Data for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = Tb, Dy, Ho, Er, Tm, Yb). Formula Cs0.15Tb0.29YbS2 Cs0.16Dy0.28YbS2 Cs0.16Ho0.30YbS2 Cs0.17Er0.32YbS2 Cs0.15Tm0.26YbS2 Cs0.15Yb0.26YbS2 fw 303.36 304.07 307.52 312.89 300.29 302.09 Color Brown Brown Brown Brown Brown Brown Crystal System Hexagonal Hexagonal Hexagonal Hexagonal Hexagonal Hexagonal Space group P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) P63/m (No. 176) a (?) 12.1558(13) 12.1882(15) 12.2006(16) 12.2329(8) 12.1944(11) 12.1818(11) c (?) 3.8894(6) 3.8926(7) 3.8976(7) 3.9038(4) 3.9026(5) 3.8932(5) V (?3) 497.71(11) 500.78(13) 502.45(13) 505.91(7) 502.58(9) 500.33(9) Z 6 6 6 6 6 6 T (K) 193 193 193 193 193 193 ?(?) 0.71073 0.71073 0.71073 0.71073 0.71073 0.71073 ?calcd(g cm?3) 6.073 6.050 6.098 6.162 5.953 6.016 ? (cm?1) 368.38 369.02 374.76 383.12 371.74 377.95 R(F)a 0.0379 0.0325 0.0279 0.0284 0.0280 0.0253 Rw(Fo2)b 0.0910 0.0781 0.0631 0.0577 0.0706 0.0616 a ( )R F F F F= ?? ? o c o . b ( ) ( )R F w F F wF w o 2 o 2 c 2 2 o 4 1 2 = ???? ???? ???? ??? . 238 Table 9.2. Positional and Thermo parameters for Cs0.14La0.30YbS2 and Cs0.16Eu0.33YbS2 Atom (site) x Y z Ueq(?2)a Wychoff positon Cs0.14La0.30YbS2 Cs1 0 0 1/4 0.0798(14) 2a La1 2/3 1/3 1/4 0.0285(3) 2d Yb1 0.34938(3) -0.00448(3) 3/4 0.01486(14) 6h S1 0.60244(16) 0.13376(17) -1/4 0.0154(4) 6h S2 0.1314(7) -0.2140(7) -1/4 0.0119(11) 6h S3 0.1084(7) -0.1907(8) -1/4 0.0221(14) 6h Cs0.16Eu0.33YbS2 Cs1 0 0 1/4 0.0278(8) 2a Eu1 0.687(2) 0.333(6) 1/4 0.024(2) 6h Yb1 0.34926(6) -0.00529(6) 3/4 0.0193(3) 6h S1 0.6029(4) 0.1351(4) -1/4 0.0246(9) 6h S2 0.1270(13) -0.2166(11) -1/4 0.017(2) 6h S3 0.1081(13) -0.1848(12) -1/4 0.019(2) 6h a Ueq is defined as one-third of the trace of the orthogonalized Uij tensor. 239 fits between 100 and 300 K. (Magnetic susceptibility measurements were performed by Corwin H. Booth at Lawrence Berkeley National Laboratory) X-ray Absorption Near Edge Spectroscopy (XANES). Samples were ground and passed through a 20 ?m sieve, then brushed onto adhesive tape. The tape was cut into strips and stacked to achieve an edge step at the Yb LIII edge corresponding to a change of one absorption length. X-ray absorption spectra were collected at room temperature at the Stanford Synchrotron Radiation Laboratory (SSRL) on beamline 2-3 using a double Si(111) crystal monochromator. The monochromator was detuned by 50% to reduce harmonic contamination of the beam. Contributions to the absorption from other processes than the LIII-edge excitation are removed using a standard procedure71 by extrapolating the pre-edge absorption such that the subtracted data follow a Victoreen formula. (XANES measurements were performed by Corwin H. Booth from Lawrence Berkeley National Laboratory) UV-vis-NIR Diffuse Reflectance Spectroscopy. The diffuse reflectance spectra for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Gd, Tb, Dy, Ho, Er, Tm) were measured from 200 to 1500 nm using a Shimadzu UV3100 spectrophotometer equipped with an integrating sphere attachment. The Kubelka-Munk function was used to convert diffuse reflectance data to absorption spectra.56 RESULTS AND DISCUSSION Crystal Structures. The CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb; y = 0.26 ? 0.33) series are essentially isotypic, with the largest deviation being found when Ln = La. A view of part of the structure of 240 Figure 9.1. A view down the c axis shows the three-dimensional channel structure of Cs0.14La0.30YbS2. Cs-S and La-S bonds have been omitted for clarity. The black solid rhombohedral is the unit cell of Cs0.14La0.30YbS2 and the dashed one is the unit cell of K2Tm23.33S36. a b 241 CsxLnyYbS2 is shown in Figure 9.1. These compounds are partially-filled mixed- lanthanide variants of the K2Tm23.33S36 structure type.50 The structure of these compounds consists of edge-shared double rutile chains of [YbS6] octahedra. The double-octahedra unit is quite common in other ytterbium sulfide compounds, and is known from EuYb2S4 (CaFe2O4 structure-type).53 Polyhedral representations of CsxLnyYbS2 and EuYb2S4 are shown in Figure 9.2. In both cases, the basic framework is constructed from these double-octahedra. Each [YbS6] unit is joined at the vertices to four other units to form open channels of capped trigonal prismatic sites. Differing connectivity between neighboring units results in tunnels of different shapes. CsxLnyYbS2 has hexagonal and pseudo-triangular channels containing Cs+ ions and lanthanide ions respectively, while the Eu2+ ions reside in the slightly larger pseudo-triangular channels in EuYb2S4. They are consistent with the size of the cations (Cs+ > Eu2+ > Ln3+). The Cs+ cations are nine-coordinate and occur as tricapped trigonal prisms. Cs-S bond distances range from 3.648(7) ? to 3.745(6) ?, which are comparable to the crystal radii sum, 3.62 ?, of nine-coordinated Cs+ (1.92 ?) and six-coordinated S2? (1.70 ?), according to Shannon.54 This polyhedron is shown in Figure 9.3. Selected bond distances can be found in Table 9.3a and 9.3b. The structure is compressed along [100]. Three capping S(2) atoms adopt comparable or even shorter Cs?S distances in cases of Ln = La, Ce, and Eu, which is quite unusual. Cs+ cations sit on (0,0,1/4) and have abnormally large thermal parameters associated with disorder for the early examples in the series; the largest residual electron density is always found at (0,0,1/2). It is suspected that the Cs+ cations rattle along the channels. There could be some ionic conductivity along the c axis, but further measurements are required for confirmation. 242 Figure 9.2. a) A polyhedral representation of EuYb2S4 structure projected along the b axis; b) A polyhedral view of Cs0.14La0.30YbS2 structure projected along the c axis. S(3) atoms have been removed for clarity. c a b a) b) 243 Figure 9.3. Depictions of the CsS9 and LaS9 tricapped trigonal prismatic geometries in Cs0.14La0.30YbS2 and distorted EuS9 tricapped trigonal prismatic enviroment in Cs0.16Eu0.33YbS2. 24 4 Table 9.3a. Selected Bond Distances (?) for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Eu, Gd). Formula Cs0.14La0.30YbS2 Cs0.14Ce0.30YbS2 Cs0.14Pr0.30YbS2 Cs0.14Nd0.29YbS2 Cs0.15Sm0.29YbS2 Cs0.16Eu0.33YbS2 Cs0.15Gd0.29YbS2 Cs-S2?3 3.698(5) 3.665(5) 3.661(5) 3.694(5) 3.723(6) 3.686(9) 3.705(13) Cs-S3?6 3.768(6) 3.689(5) 3.654(5) 3.669(5) 3.656(6) 3.704(9) 3.630(10) Ln-S1?2 2.9238(13) 2.749(19) 2.724(13) 2.707(10) 2.662(7) 2.78(3) 2.636(10) Ln?S1?2 2.9238(13) 2.92(3) 2.92(3) 2.910(19) 2.902(12) 2.88(5) 2.90(2) Ln?S1?2 2.9238(13) 3.061(15) 3.062(13) 3.065(9) 3.079(6) 3.09(2) 3.064(12) Ln -S2 3.444(5) 3.25(2) 3.246(19) 3.193(14) 3.117(11) 3.24(4) 3.14(2) Ln -S2 3.444(5) 3.48(5) 3.47(4) 3.44(3) 3.404(18) 3.52(7) 3.40(3) Ln -S2 3.444(5) 3.67(3) 3.707(18) 3.681(15) 3.687(11) 3.67(3) 3.721(17) Yb-S1?2 2.7690(12) 2.7632(13) 2.7678(15) 2.7677(15) 2.768(2) 2.765(3) 2.770(4) Yb-S1 2.6875(17) 2.6886(17) 2.6983(19) 2.7007(19) 2.708(2) 2.696(4) 2.716(5) Yb-S2(3)?2 2.622(6) 2.626(4) 2.634(4) 2.627(4) 2.626(4) 2.626(9) 2.627(8) Yb-S2(3) 2.649(8) 2.647(6) 2.648(6) 2.646(5) 2.640(6) 2.658(14) 2.642(11) Yb ~ Yb 3.9393(3) 3.9216(6) 3.9223(4) 3.9140(4) 3.9085(4) 3.9218(6) 3.8976(6) Ln ~ Yb 3.9323(4) 3.673(14) 3.647(8) 3.627(7) 3.565(5) 3.80(5) 3.545(4) Ln ~ Ln 3.9393(3) 3.9216(6) 3.9223(4) 3.9140(4) 3.9085(4) 3.9218(6) 3.8976(6) 24 5 Table 9.3b. Selected Bond Distances (?) for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = Tb, Dy, Ho, Er, Tm, Yb). Formula Cs0.15Tb0.29YbS2 Cs0.16Dy0.28YbS2 Cs0.16Ho0.30YbS2 Cs0.17Er0.32YbS2 Cs0.15Tm0.26YbS2 Cs0.15Yb0.26YbS2 Cs-S2?3 3.711(8) 3.74(2) 3.73(2) 3.70(4) 3.703(6) 3.685(6) Cs-S3?6 3.616(7) 3.646(19) 3.64(2) 3.67(3) 3.670(6) 3.659(5) Ln-S1?2 2.639(8) 2.633(6) 2.637(5) 2.686(7) 2.682(9) 2.667(5) Ln?S1?2 2.859(12) 2.877(10) 2.886(9) 2.892(13) 2.908(19) 2.918(16) Ln?S1?2 3.073(5) 3.076(5) 3.079(5) 3.095(6) 3.063(11) 3.054(11) Ln -S2 3.086(10) 3.08(2) 3.10(2) 3.16(4) 3.166(16) 3.186(17) Ln -S2 3.427(17) 3.40(3) 3.41(3) 3.48(4) 3.42(3) 3.41(2) Ln -S2 3.684(12) 3.68(2) 3.71(2) 3.72(4) 3.686(13) 3.710(9) Yb-S1?2 2.764(2) 2.770(2) 2.772(2) 2.768(2) 2.7634(18) 2.7587(18) Yb-S1 2.710(3) 2.714(3) 2.717(3) 2.706(3) 2.699(2) 2.694(2) Yb-S2(3)?2 2.621(4) 2.622(7) 2.625(7) 2.625(10) 2.620(4) 2.618(4) Yb-S2(3) 2.636(6) 2.641(6) 2.648(6) 2.656(8) 2.643(4) 2.640(7) Yb ~ Yb 3.8894(6) 3.8926(7) 3.8976(7) 3.9038(4) 3.9026(5) 3.8932(5) Ln ~ Yb 3.535(6) 3.534(4) 3.540(3) 3.596(4) 3.595(5) 3.5819(15) Ln ~ Ln 3.8894(6) 3.8926(7) 3.8976(7) 3.9038(4) 3.9026(5) 3.8932(5) 246 When Ln = La, the cations reside in the center of the pseudo-triangular channel (or 2d site), and are coordinated to nine S atoms in a tricapped trigonal prism arrangement. However, the atoms are disordered over three 6h sites in a triangular shape for Ln = Ce ? Yb in order to fit into the channel, which is illustrated in Figure 9.3. This results in large distortions of the polyhedra. The Ln?S distances range from 3.097(3) ? (Ln = La) to 3.04(1) ? (Ln = Dy). These distances are longer than Shannon?s values especially for the later cases, 3.06 ? (X = La) to 2.88 ? (X = Yb), using Ln3+ in nine-coordinate S environments.54 The Yb(1) sites have distorted octahedral geometries with normal Yb?S distances ranging from 2.693(5) ? to 2.681(4) ?. The unit cell volume and Yb(1)?Yb(1) distances are plotted in Figure 9.4 to identify standard systematic changes associated with the lanthanide contraction. It is worth noting that the minor percentage of Ln in the compound and the confinement of the channel framework remarkably limit the effect of the lanthanide contraction on these parameters, so the experimental errors are considerable. Both plots are quite consistent with each other. For the early lanthanides, the data follow the trends of lanthanide contraction except Eu. In the later cases, large deviations are found. The large deviation at Eu can be ascribed to divalent or mixed-valent +2/+3 character, an occurrence that is common in chalcogenides, and consistent with the bond-valence sum of 2.50.55 Yb(2) could be divalent also, with a bond-valence sum of 2.19. However, it is much less likely for Ln = Dy?Tm to be anything other than +3, even though their valence sums were calculated to be 2.65, 2.68, 2.31, and 2.28, respectively. For these cases, disorder of Ln into the Yb(1) sites might be one of the reasons for the deviation, considering their similar sizes. It is far better to interpret the oxidation state of Ln from magnetic 247 Figure 9.4. Unit cell volumes (?3) and Yb~Yb distances (?) vs the number of f electrons for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm). 0 2 4 6 8 10 12 496 498 500 502 504 506 508 510 512 Number of f electrons Un it c ell vo lum es (? 3 ) 3.89 3.90 3.91 3.92 3.93 3.94 Yb -Y b d ist an ces (? ) 248 susceptibility and XANES measurements described in this work. S?S bonds are absent in the structure of CsxLnyYbS2, and therefore the formal oxidation states of Cs/X/Yb/S are ideally +1/+3/+3/?2, except when Ln = Eu. Magnetic Properties. Recently, AIILnIII2O4 (A = Sr, Ba), which crystallize in CaFe2O4 structure-type, were recognized as potentially geometrically frustrated systems because of their triangular Ln3+ sublattice and lack of long-range magnetic ordering at low temperatures.57,58 As previously mentioned, CsxLnyYbS2 and EuYb2S4 (CaFe2O4 structure-type) are structurally related. Both structures have similar three-dimensional Yb3+ triangles, as shown in Figure 9.5. In contrast to EuYb2S4, which contains two different double-octahedra chains, CsxLnyYbS2 has one double chain even though the Yb- Yb distances are comparable in these two systems. Another factor that may soften the geometric frustration is that the triangular coordination is not perfect: there is a distribution of Yb-Yb bond lengths and other asymmetries that will at least partially remove the requirement of equal magnetic interactions for frustration to occur. In any case, Yb-Yb distances vary with the size of cations in the channels in the series CsxLnyYbS2. The shortest Yb-Yb distances within the triangles range from 3.9393(3) ? (Cs0.14La0.30YbS2) to 3.8994(6) ? (Cs0.15Tb0.29YbS2). These distances can be reduced by using smaller alkali metal cations. For Cs0.14La0.30YbS2, the Yb-Yb distances along the spines of the chains are 0.2 ? shorter than that between Yb3+ in parallel spines. This may allow for one-dimensional magnetic interactions. The interchain Yb-Yb distances are approximately 0.6 ? longer than the intrachain ones. Magnetic coupling across the double-octahedra chains should be weaker. In addition to the Yb(1)3+ lattice, the larger Ln cations (except La) are also paramagnetic. There are potentially some Ln?Ln and Ln? 249 Figure 9.5. a) An illustration of the Yb3+ cations net work along the b axis with Yb-Yb distances labeled for EuYb2S4. b) An drawing of the Yb3+ cations network along the c axis with Yb-Yb distances labeled for Cs0.14La0.30YbS2. 250 Yb(1) interactions within the small channels too, which makes the CsxLnyYbS2 system more complicated than AIILnIII2O4 (A = Sr, Ba) 57,58. The Ln?Ln distances range from 3.9223(4) ? (in Cs0.14Pr0.30YbS2) to 3.8894(6) ? (in Cs0.15Tb0.29YbS2), while Ln?Yb distances are in the range of 3.80(5) ? (in Cs0.16Eu0.33YbS2) to 3.534(4) ? (in Cs0.16Dy0.28YbS2). The magnetic coupling in CsxLnyYbS2 compounds within the three- dimensional triangular Yb3+ lattice is potentially weakly frustrated. Magnetic susceptibility data were all collected at 5 kG applied field after verifying linearity of M vs. H. There is a decrease in slope at higher fields due to a Brillouin-type saturation of the paramagnetic moments. All the data show a deviation from Curie-Weiss behavior around 50 K, as shown in Figure 9.6. Except for the Cs0.15Tb0.29YbS2 data, all can be described as starting at RT from a full-moment state with a negative Weiss temperature, cooling through a state near 50 K where the moment and the Weiss temperature decrease. Table 9.4 shows the magnetic parameters for CsxLnyYbS2, which were obtained from fitting the data in the range of 100 K and 300 K to the Curie-Weiss law. The ?p values range from -14.2(3) K (for Cs0.15Gd0.29YbS2) to - 45(2) K (for Cs0.15Yb0.26YbS2) without any indication of long-rang magnetic ordering down to 2 K. This suppression is consistent with a geometrically spin-frustrated system. As a cautionary note the observed |?p| value may be large due to the deviations from the Curie-Weiss law caused by a crystal-field splitting of the full J=7/2 multiplet for the Yb3+, and possibly the other magnetic atoms. The experimental effective magnetic moments are all close to that expected for full moment Yb and the Ln-atom, except for some deviations were observed for Cs0.15Gd0.29YbS2, Cs0.16Dy0.28YbS2 and 251 Figure 9.6. a) Plots of dc inverse molar magnetic susceptibility for Cs0.14-0.17Ln0.26- 0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Gd, Tb) under an applied field of 0.5 T. The straight lines represent fits of the inverse susceptibility to the Curie-Weiss law in the high temperature rang of 100-300 K. 0 50 100 150 200 250 300 0 20 40 60 80 100 120 140 160 ?-1 (m ol for mu la un it/e mu ) Temperature (K) La Ce Pr Nd Sm Gd Tb 252 Figure 9.6. b) Plots of dc inversed molar magnetic susceptibility for Cs0.14-0.17Ln0.26- 0.33YbS2 (Ln = Dy, Ho, Er, Tm, Yb) under an applied field of 0.5 T. The straight lines represent fits of the inverse susceptibility to the Curie-Weiss law in the high temperature rang of 100-300 K. 0 10 20 30 40 50 60 70 80 90 100 110 120 ?-1 (m ol for mu la un it/e mu ) Temperature (K) Dy Ho Er Tm Yb 0 50 100 150 200 250 300 253 Table 9.4. Magnetic Parameters for Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Gd, Tb, Dy, Ho, Er, Tm, Yb). Formula Pcal/?B Peff/?B ?p R2 Cs0.14 La0.30YbS2 4.54 4.59(2) -36(1) 0.99938 Cs0.14Ce0.30YbS2 4.75 4.832(8) -32(1) 0.99955 Cs0.14Pr0.30YbS2 4.94 5.161(8) -44.2(9) 0.99974 Cs0.14Nd0.29YbS2 4.94 4.950(6) -25.2(6) 0.99986 Cs0.15Sm0.29YbS2 4.56 4.344(9) -44(1) 0.99966 Cs0.15Gd0.29YbS2 6.24 6.657(4) -14.2(3) 0.99997 Cs0.15Tb0.29YbS2 6.93 7.708(5) 3.4(4) 0.99993 Cs0.16Dy0.28YbS2 7.23 6.726(5) -24.8(3) 0.99996 Cs0.16Ho0.30YbS2 7.37 8.123(5) -14.9(3) 0.99997 Cs0.17Er0.32YbS2 7.07 7.042(6) -24.6(4) 0.99996 Cs0.15Tm0.26YbS2 5.96 5.898(7) -15.7(7) 0.9998 Cs0.15Yb0.26YbS2 5.10 4.99(2) -45(2) 0.99842 a Pcal and Peff : calculated and experimental effective magnetic moments per formula unit. b Weiss constant (?p) and goodness of fit (R2) obtained from high temperature (100-300 K) data. 254 Cs0.16Ho0.30YbS2. Splitting of the ground state terms for magnetic ions and disorder of Ln into the Yb(1) sites might be the reasons for such differences. Cs0.15Tb0.29YbS2 is different from the other compounds, with a slightly positive Weiss temperature 3.4(4) K in the high-temperature (HT) state, and an increase of the susceptibility as it is cooled through the ~50 K point. The observed moment (7.708(5) ?B) is considerably larger than the calculated value (6.93 ?B). All these behaviors could be due to strong ferromagnetic interactions among the Tb3+ and Yb3+ ions. One piece of supporting evidence for this argument is that Cs0.15Tb0.29YbS2 has the shortest distances between magnetic atoms, as is shown in Figure 9.4 and Table 9.3. XANES. Lanthanide LIII-edge X-ray absorption spectroscopy is commonly used to determine their valence state, and can be used to verify the results from magnetic susceptibility measurements. The focus is on the lanthanides that can be in more than one valence state, and we have chosen to focus on cerium, samarium, and ytterbium, along with neodymium for comparison. The normalized, energy-calibrated spectra of Cs0.14Ce0.30YbS2, Cs0.14Nd0.29YbS2, and Cs0.15Sm0.29YbS2 (Figure 9.7, 9.8) support the conclusion that all lanthanides in these samples are in their trivalent state. This result is particularly clear in the Ce LIII edge data because the character of spectra from formally tetravalent cerium samples (such as the CeO2 data in Figure 9.7(a)) includes a predominantly double-peaked structure with the first clear maximum at about 5727 eV, whereas trivalent cerium spectra contain a dominant single peak at about 5722 eV (Figure 9.7). There is no evidence of a peak at 5727 eV in the Ce LIII spectrum of Cs0.14Ce0.30YbS2 (Fig. 7(a)). Nd and Sm LIII edge spectra also are dominated by a single peak, and the measured peak positions of 6210 eV and 6718 eV are consistent only with a 255 Figure 9.7. Ce, Nd, and Sm LIII edge x-ray absorption spectra of (a) Cs0.14Ce0.30YbS2, (b) Cs0.14Nd0.29YbS2, and (c) Cs0.15Sm0.29YbS2, respectively. Also shown in (a) is the spectrum of CeO2 as a reference. All data are calibrated to the first inflection point of the Nd LIII edge of Nd2O3 at 6208 eV. The sulfide spectra are consistent with a fully trivalent state for the relevant lanthanide. 6700 6710 6720 6730 6740 6750 6760 0 1 2 3 (b) Cs0.15Sm0.29YbS2 E (eV) 6190 6200 6210 6220 6230 6240 6250 0 1 2 3 (b) Cs0.15Nd0.29YbS2 No rm ali ze d a bs orp tio n 5700 5710 5720 5730 5740 5750 5760 5770 0 1 2 3 (a) Cs0.14Ce0.30YbS2 CeO2 256 Figure 9.8. (a) Yb LIII edge spectra of the same samples in Fig. 7, and (b) the derivative of these spectra. These data are calibrated to the first inflection point of the Yb LIII edge of Yb2O3 at 8943 eV. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 N orm ali ze d a bs orp tio n (a) Cs0.14Ce0.30YbS2 Cs0.14Nd0.29YbS2 Cs0.15Sm0.29YbS2 Yb2O3 8930 8940 8950 8960 8970 -0.4 -0.2 0.0 0.2 0.4 0.6 (b) Cs 0.14Ce0.30YbS2 Cs0.14Nd0.29YbS2 Cs0.15Sm0.29YbS2 Yb2O3 De riv ati ve E (eV) 257 trivalent state.68 The LIII threshold energy is about 8 eV lower in Sm(II) than Sm(III),69 and we see no evidence of such a peak in the Cs0.15Sm0.29YbS2 spectrum. A similar shift is expected in Yb LIII spectra between Yb(II) and Yb(III), and would appear as a shoulder on the main absorption edge Figure 9.8(a).70 To emphasize this point, the derivative spectra in Figure 9.8(b) show no inflection point at the expected Yb(II) position of 8936 eV, below the main trivalent feature at 8943 eV. Optical Properties. Many ternary alkali metal rare-earth chalcogenides have been synthesized, but few of them have had their optical properties examined in detail.59- 67 Early work by B. Deng et al. have shown that RbLnSe2 ( Ln = La, Ce, Pr, Nd, Sm, Gd, Tb, Ho, Er, Lu ) are direct band gap semiconductors.67 The absorption in these compounds was attributed as the transition from a Se2- valence band to a Ln3+ conduction band. Therefore the band gaps in the series ALnQ2 is tunable based on the choice of lanthanide ion and chalcogen, which is also true for other ternary and quaternary chalcogenides. In this present study, only the larger lanthanide site in CsxLnyYbS2 has been systematically varied. The optical band gaps were measured using UV-vis-NIR spectroscopy. The gap of Cs0.14Ce0.30YbS2 was determined to be approximately 2.1 eV and Cs0.14La0.30YbS2, Cs0.14Pr0.30YbS2, Cs0.14Nd0.29YbS2, Cs0.15Sm0.29YbS2 are ~2.3 eV, which are consistent with their red and yellow colors. For the brown compounds, the measured values of Cs0.15Gd0.29YbS2, Cs0.15Tb0.29YbS2 and Cs0.16Dy0.28YbS2 are between 2.3 and 2.4 eV, while the band gap of Cs0.16Ho0.30YbS2, Cs0.17Er0.32YbS2, Cs0.15Tm0.26YbS2 are about 2.2 eV. The brown color might be due to the absorptions before 2.0 eV, which are not shown in Figure 9.9. These band gaps are comparable to the 258 Figure 9.9 UV-vis diffuse reflectance spectra of Cs0.14-0.17Ln0.26-0.33YbS2 (Ln = La, Ce, Pr, Nd, Sm, Gd, Tb, Dy, Ho, Er, Tm). 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 ?/ s ( arb ita ry un its ) Energy eV La Ce Pr Nd Sm Gd Tb Dy Ho Er Tm 259 values reported for red NaCeS3 (2.15 eV) and yellow NaLaS3 (2.61 eV).66 The Ce compounds often show the smallest band gaps of the series owing the high energy of the 4f1 electron. The fine structure in these spectra are f-f transitions for the lanthanide ions. CONCLUSIONS In this present work we have prepared a new family of interlanthanide sulfide, CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La-Nd, Sm-Yb; y = 0.26 ? 0.33), using CsCl flux. These compounds are partially-filled mixed-lanthanide variants of the K2Tm23.33S36 structure-type. The three-dimensional channel structure of CsxLnyYbS2 is constructed from one-dimensional edge-shared double rutile chains of [YbS6] octahedral. Each chain is connected to four other identical neighbors to form two different hexagonal channels. The larger channels are nearly filled with Cs+ cations while the other is only partially filled with lanthanide (Ln) ions. In all but one of these compounds the magnetic interactions are antiferromagnetic in nature. However, in Cs0.15Tb0.29YbS2 the coupling is ferromagnetic. The trivalancy of all lanthanides for Cs0.14Ce0.30YbS2, Cs0.14Nd0.29YbS2, and Cs0.15Sm0.29YbS2 have been confirmed by XANES measurements. The band gaps for these compounds are relatively constant except for the Ce-analog, which possess a notably smaller gap than the other compounds. 260 REFERENCES 1. Ramirez, A.P. in Handbook of Magnetic Materials, edited by K. J. H. Buschow (Elsevier, Amsterdam, 2001), Vol. 13. 2. Schiffer, P.; Ramirez, A. P. Comments Condens. Matter Phys. 1996, 18, 21. 3. Lau, G. C.; Freitas, R. S.; Ueland, B. G.; Schiffer, P.; Cava, R. J. Phys. Rev. B 2005, 72, 54411. 4. Buttgen, N.; Hemberger, J.; Fritsch, V.; Krimmel, A.; Mucksch, M.; Krug von Nidda, H. A.; Lunkenheimer, P.; Fichtl, R.; Tsurkan, V.; Loidl, A. New J. Phys. 2004, 6, 191. 5. Fritsch, V.; Hemberger, J.; Buttgen, N.; Scheidt, E. W.; Krug von Nidda, H. A.; Loidl, A.; Tsurkan, V. Phys. Rev. Lett. 2004, 92, 116401. 6. Tsurkan, V.; Mucksch, M.; Fritsch, V.; Hemberger, J.; Klemm, M.; Klimm, S.; Korner, S.; Krug von Nidda, H. A.; Samusi, D.; Scheidt, E. W.; Loidl, A.; Horn, S.; Tidecks, R. Phys. Rev. B 2003, 68, 134434. 7. Martinho, H.; Moreno, N. O.; Sanjurjo, J. A.; Rettori, C.; Garcia-Adeva, A. J.; Huber, D. L.; Oseroff, S. B.; Ratcliff II, W.; Cheong, S. W.; Pagliuso, P. G.; Sarrao, J. L.; Martins, G. B. Phys. Rev. B 2001, 64, 024408. 8. Wills, A. S.; Raju, N. P.; Morin, C. J.; Greedan, E. Chem. Mater. 1999, 11, 1936. 9. Hamedoun, M.; Silimani, M.; Sayouri, S.; Benyoussef, A. Physica Status Solidi A: Applied Research 1994, 144, 441. 10. Ostorero, J.; Mauger, A.; Guillot, M.; Derory, A.; Escorne, M.; Marchand, A. Phys. Rev. B 1989, 40, 391. 11. Pawlak, L.; Duczmal, M.; Zygmunt, A. J. Magn. Magn. Mater. 1988, 76-7, 199. 261 12. Bendor, L.; Shilo, I.; Felner, I. J. Solid State Chem. 1981, 39, 257. 13. Bendor, L.; Shilo, I. J. Solid State Chem. 1980, 35, 278. 14. Pokrzywnicki, S.; Pawlak, L.; Czopnik, A. Physica B & C 1977, 86, 1141. 15. Snyder, J.; Ueland, B. G.; Mizel, A.; Slusky, J. S.; Karunadasa, H.; Cava, R. J.; Schiffer, P. Phys. Rev. B 2004, 70, 184431. 16. Snyder, J.; Ueland, B. G.; Slusky, J. S.; Karunadasa, H.; Cava, R. J.; Schiffer, P. Phys. Rev. B 2004, 69, 064414. 17. Champion, J. D. M.; Harris, M. J.; Holdsworth, P. C. W.; Wills, A. S.; Balakrishnan, G.; Bramwell, S. T.; Cizmar, E.; Fennell, T.; Gardner, J. S.; Lago, J.; McMorrow, D. F.; Orendac, M.; Orendacova, A.; Paul, D. McK.; Smith, R. I.; Telling, M. T. F.; Wildes, A. Phys. Rev. B 2003, 68, 020401(R). 18. Sakakibara, T.; Tayama, T.; Hiroi, Z.; Matsuhira, K.; Takagi, S. Phys. Rev. Lett. 2003, 90, 207205. 19. Snyder, J.; Ueland, B. G.; Slusky, J. S.; Karunadasa, H.; Cava, R. J.; Mizel, A.; Schiffer, P. Phys. Rev. Lett. 2003, 91, 107201. 20. Taira, N.; Wakeshima, M.; Hinatsu, Y. J. Mater. Chem. 2002, 12, 1475. 21. Ramirez, A. P.; Shastry, B. S.; Hayashi, A.; Krajewski, J. J.; Huse, D. A.; Cava, R. J. Phys. Rev. Lett. 2002, 89, 067202. 22. Hodges, J. A.; Bonville, P.; Forget, A.; Yaouanc, A.; Dalmas de Reotier, P.; Andre, G.; Rams, M.; Krolas, K.; Ritter, C.; Gubbens, P. C. M.; Kaiser, C. T.; King, P. J. C.; Baines, C. Phys. Rev. Lett. 2002, 88, 077204. 23. Fukazawa, H.; Melko, R. G.; Higashinaka, R.; Maeno, Y.; Gingras, M. J. P. Phys. Rev. B 2002, 65, 054410. 262 24. Bramwell, S. T.; Gingras, M. J. P. Science 2001, 294, 1495. 25. Snyder, J.; Slusky, J. S.; Cava, R. J.; Schiffer, P. Nature 2001, 413, 48. 26. Matsuhira, K.; Hinatsu, Y.; Tenya, K.; Sakakibara, T. J. Phys.: Condens. Matter 2000, 12, L649. 27. Gingras, M. J. P.; den Hertog, B. C.; Faucher, M.; Gardner, J. S.; Dunsiger, S. R.; Chang, L. J.; Gaulin, B. D.; Raju, N. P.; Greedan, J. E. Phys. Rev. B 2000, 62, 6496. 28. Ramirez, A. P.; Hayashi, A.; Cava, R. J.; Siddharthan, R.; Shastry, B. S. Nature 1999, 399, 333. 29. Taira, N.; Wakeshima, M.; Hinatsu, Y. J. Phys.: Condens. Matter 1999, 11, 6983. 30. Harris, M. J.; Bramwell, S. T.; Holdsworth, P. C. W.; Champion, J. D. M. Phys. Rev. Lett. 1998, 81, 4496. 31. Bramwell, S. T.; Harris, M. J. J. Phys. Condens. Matter 1998, 10, L215. 32. Harris, M. J.; Bramwell, S. T.; McMorrow, D. F.; Zeiske, T.; Godfrey, K. W. Phys. Rev. Lett. 1997, 79, 2554. 33. Greedan, J. E.; Raju, N. P.; Maignan, A.; Simon, Ch.; Pedersen, J.S.; Niraimathi, A. N.; Melin, E. G.; Subramanian, M.A. Phys. Rev. B 1996, 54, 7189. 34. Harris, M. J.; Zinkin, M. P.; Tun, Z.; Wanklyn, B. M.; Swainson, I. P. Phys. Rev. Lett. 1994, 73, 189. 35. Gaulin, B. D.; Reimers, J. N.; Mason, T. E.; Greedan, J. E.; Tun, Z. Phys. Rev. Lett. 1992, 69, 3244. 263 36. Greedan, J. E.; Greedan, J. E.; Reimers, J. N.; Stager, C. V.; Penny, S. L. Phys. Rev. B 1991, 43, 5682. 37. J. E. Greedan, J. Mater. Chem. 2001, 11, 37. 38. Schiffer, P.; Ramirez, A. P. Comments Condens. Matter Phys. 1996, 18, 21. 39. Ramirez, A. P. Annu. Rev. Mater. Sci. 1994, 24, 453. 40. Shores, M. P.; Nytko, E. A.; Bartlett, B. M.; Nocera, D. G. J. Am. Chem. Soc. 2005, 127, 13462. 41. Elhajal, M.; Canals, B.; Sunyer, R.; Lacroix, C. Phys. Rev. B 2005, 71, 094420. 42. Elhajal, M.; Canals, B.; Lacroix, C. J. Phys.: Condens. Matter 2004, 16, S917. 43. Elhajal, M.; Canals, B.; Lacroix, C. Phys. Rev. B 2002, 66, 014422. 44. Bartlett, B. M.; Nocera, D. G. J. Am. Chem. Soc. 2005, 127, 8985. 45. Grohol, D.; Matan, K.; Cho, J.-H.; Lee, S.-H.; Lynn, J. W.; Nocera, D. G.; Lee, Y. S.; Nat. Mater. 2005, 4, 323. 46. Grohol, D.; Nocera, D. G.; Papoutsakis, D. Phys. Rev. B 2003, 67, 064401. 47. Inami, T.; Nishiyama, M.; Maegawa, S.; Oka, Y. Phys. Rev. B 2000, 61, 12181. 48. Wills, A. S.; Harrison, A.; Ritter, C.; Smith, R. I. Phys. Rev. B 2000, 61, 6156. 49. Lau, G. C.; Ueland, B. G.; Freitas, R. S.; Dahlberg, M. L.; Schiffer, P.; Cava, R. J. Phys. Rev. B 2006, 73, 012413. 50. Lemoine, Per P.; Tomas, A.; Carre, D.; Vovan Et M. Guittard, T. Acta Cryst. C 1989, 45, 350. 51. Sheldrick, G. M. SHELLXTL PC, Version 6.12, An Integrated System for Solving, Refining, and Displaying Crystal Structures from Diffraction Data; Siemens Analytical X-ray Instruments, Inc.: Madison, WI, 2001. 264 52. Sheldrick, G. M. SADABS 2001, Program for absorption correction using SMART CCD based on the method of Blessing: R. H. Blessing, Acta Crystallogr. 1995, A51, 33. 53. Lugscheider, W.; Pink, H.; Weber, K.; Zinn, W. Zeitschrift fuer Angewandte Physik 1970, 30, 36. 54. Shannon, R. D. Acta Crystallogr. 1976, A 32,751. 55. Brese, N. E.; Okeeffe, M. Acta Cryst. 1991, B47, 192. 56. Wendlandt, W. W.; Hecht, H. G. Reflectance Spectroscopy. Interscience Publishers: New York, 1966. 57. Karunadasa, H.; Huang, Q.; Ueland, B. G.; Lynn, J. W.; Schiffer, P.; Regan, K. A.; Cava, R. J. Phys. Rev. B 2005, 71, 144414. 58. Doi, Y.; Nakamori, W.; Hinatsu, Y. J. Phys. Condens. Matter 2006, 18, 333. 59. Tromme, M. C. R. Acad. Sci. Ser. C 1971, 273, 849. 60. Sato, M.; Adachi, G.; Shiokawa, J. Mater. Res. Bull. 1984, 19, 1215. 61. Ohtani, T.; Honjo, H.; Wada, H. Mater. Res. Bull. 1987, 22, 829. 62. Schleid, T.; Lissner, F. Eur. J. Solid State Inorg. Chem. 1993, 30, 829. 63. Foran, B.; Lee, S.; Aronson, M. Chem. Mater. 1993, 5, 974. 64. Bronger, W.; Bruggemann, W.; von der Ahe, M.; Schmitz, D. J. Alloys Compd. 1993, 200, 205. 65. Bronger, W.; Eyck, J.; Kruse, K.; Schmitz, D. Eur. J. Solid State Inorg. Chem. 1996, 33, 213. 66. Sutorik, A. C.; Kanatzidis, M. G. Chem. Mater. 1997, 9, 387. 67. Deng, B.; Ellis, D. E.; Ibers, J. A. Inorg. Chem. 2002, 41, 5716. 265 68. Nitani, H.; Nakagawa, T.; Yananouchi, M.; Osuki, T.; Yuya, M.; Yamamoto, A. Mater. Lett. 2004, 58, 2076. 69. Shimizugawa, Y.; Sawaguchi, N.; Kawamura, K.; Hirao, K. J. Appl. Phys. 1997, 81, 6657. 70. Sarrao, J. L.; Immer, C. D.; Fisk, Z.; Booth, C. H.; Figueroa, E.; Lawrence, J. M.; Modler, R.; Cornelius, A. L.; Hundley, M. F.; Kwei, G. H.; Thompson, J. D.; Bridges, F. Phys. Rev. B 1999, 59, 6855. 71. Li, G. G.; Bridges, F.; Booth, C. H. Phys. Rev. B 1995, 52,6332. 266 CHAPTER 10 SUMMARY Syntheses and characterizations of new lanthanide chalcogenides have been discussed in Chapter 2 through Chapter 9. All these compounds were prepared by solid- state reactions of corresponding elements using alkali metal halides or Sb2Q3 (Q = S, Se) fluxes. Single crystal X-ray diffraction, UV-vis-NIR diffuse reflectance spectroscopy, magnetic susceptibility measurements, M?ssbauer spectroscopy, and X-ray Absorption Near Edge Spectroscopy were used to determine their structures and physical properties. A list of new compounds and their crystallographic, magnetic, and optical properties are presented in Table 10.1a and Table 10.1b. Chapter 2 and Chapter 3 report three ternary europium pnictogen chalcogenide compounds, Eu6Sb6S17 and EuPnSe3 (Pn = Sb, Bi). All of them have three-dimensional complex structures in a chiral space group. These compounds follow essentially Curie behavior from 300 K to 5 K, and undergo an apparently antiferromagnetic transition below 5 K. Magnetic susceptibility measurements and 151Eu and 121Sb M?ssbauer spectroscopy studies have revealed the presence of divalent europium and trivalent antimony. Chapters 4 to Chapter 6 discuss the synthesis of ternary interlanthanide chalcogenides using Sb2Q3 (Q = S, Se) fluxes. Structures and physical properties of these compounds depend highly on the choices of lanthanides and chalcogenides. They have 26 7 Table 10.1a. A list of new compounds and some of their properties, contained in this dissertation. Formula Chapter Space Group Magnetism Color/Band Gap (eV) Eu6Sb6S17 2 P212121 Antiferromagnetism Black/N/A EuPnSe3 (Pn = Sb, Bi) 3 P212121 Antiferromagnetism Black/N/A ?-LaLn'S3 ( Ln' = Er, Tm, Yb) 4 Pnma Paramagnetism Dark Red/1.6 ?-CeLn'S3 ( Ln' = Er, Tm, Yb) 4 Pnma Paramagnetism Black/1.3 ?-Ce1.30Lu0.70S3 5 P21/m Paramagnetism Black/1.25 ?-Pr1.29Lu0.71S3 5 P21/m Paramagnetism Dark Red/1.38 ?-Nd1.33Lu0.67S3 5 P21/m Paramagnetism Dark Red/1.50 La3LuSe6 6 Pnnm Diamagnetism Black/1.26 Ce3LuSe6 6 Pnnm Ferromagnetism Black/1.10 ?-PrLuSe3 6 Cmcm Paramagnetism Black/1.56 26 8 Table 10.1b. A list of new compounds and some of their properties, contained in this dissertation. Formula Chapter Space Group Magnetism Color/Band Gap (eV) ?-NdLuSe3 6 Cmcm Paramagnetism Black/1.61 Sm1.82Lu2.18Se6 6 P21/m van Vleck Paramagnetism Black/1.51 Gd1.87Lu2.13Se6 6 P21/m Antiferromagnetism Black/1.56 Ln2YbCuS5 (Ln = La, Ce, Pr, Nd) 8 Pnma Paramagnetism Black/1.45, 1.37, 1.25, 1.35 Sm2YbCuQ5 8 Pnma Spin Glass? Black/1.28 Ln2YbCuSe5 (Ln = La, Ce) 8 Pnma Paramagnetism Black/1.05, 1.15 CsxLnyYbS2 ( Ln = La, Pr, Nd, Sm) 9 P63/m Paramagnetism Yellow/2.3 Cs0.14Ce0.30YbS2 9 P63/m Paramagnetism Red/2.1 CsxLnyYbS2 ( Ln = Gd, Tb, Dy) 9 P63/m Paramagnetism Brown/2.3-2.4 CsxLnyYbS2 ( Ln = Ho, Er, Tm) 9 P63/m Paramagnetism Brown/2.2 Ln2YbCuS5 (Ln = La, Ce, Pr, Nd) 8 Pnma Paramagnetism Black/1.45, 1.37, 1.25, 1.35 269 shown a variety of structures including ordered and disordered types under different reaction conditions. Most of compounds are semiconductors with wide tunable band gaps. Different magnetic behaviors have been found in these systems, namely Curie- Weiss type paramagnetism, van Vleck paramagnetism, antiferromagnetism, ferromagnetism, and spin glass performance. Possible spin-frustrations in some of these interlanthanide compounds were also discussed. Chapter 7 presents the first two partially ordered quaternary interlanthanide sulfides PrLnYb2S6 (Ln = Tb, Dy). They were prepared using intermediate lanthanides to substitute in the disordered sites in already known ternary F-Ln2S3 type structures. The elemental analysis and magnetic susceptibility measurements are consistent with the proposed formula. In Chapter 8, a new series of ordered quaternary interlanthanide copper chalcogenides, Ln2YbCuQ5 (Ln = La, Ce, Pr, Nd, Sm; Q = S, Se) are reported. These compounds crystallize in a new structure type that is realized by including two different lanthanides with large size difference, which tend to have distinct coordination environments. Magnetic measurements have shown that Ce2YbCuSe5, La2YbCuS5, Ce2YbCuS5, and Pr2YbCuS5 are Curie-Weiss paramagnets. La2YbCuSe5 and Nd2YbCuS5 have short-range antiferromagnetic ordering at low temperature. Sm2YbCuS5 exhibits an interesting magnetic transition below 7 K. The UV-vis-NIR diffuse reflectance measurements show these compounds to be wide band-gap semiconductors. Finally, Chapter 9 encloses a new family of interlanthanide sulfide, CsxLnyYbS2 (x = 0.14 ? 0.16; Ln = La-Nd, Sm-Yb; y = 0.26 ? 0.33. These compounds are partially- filled mixed-lanthanide variants of the K2Tm23.33S36 structure-type. In all but one of these 270 compounds the magnetic interactions are antiferromagnetic in nature. However, in Cs0.15Tb0.29YbS2 the coupling is ferromagnetic. The trivalancy of all lanthanides for Cs0.14Ce0.30YbS2, Cs0.14Nd0.29YbS2, and Cs0.15Sm0.29YbS2 have been confirmed by XANES measurements. The band gaps for these compounds are relatively constant except for the Ce analog, which possesses a notably smaller gap than the other compounds. Possible spin-frustrations in these interlanthanide compounds were also examined.