..............m. w... , ..............f. s... a...” 25%;...” rigs 1.1. a...“r r: n: . . . .2}... 9L... . vi: 3”” . , . .. , . . M ,L . . ... new. m2§..m:v.m¢.. . aawfimm...mq§m. a . . high... a”! . ~. 4.2:. 433$? 2.3.332..."- Wléau/ - . ~d -1.“- ‘HT LIBRA I: Michigan Staff; 2; University {.5 law ’mtrM' This is to certify that the thesis entitled A STRUCTURAL STUDY OF SOME COMPLEXES CONTAINING THE SCANDIUM(III) ION: SIX AND EIGHT COORDINATION presented by Thomas John Anders on has been accepted towards fulfillment of the requirements for Ph.D. Che 'st degree in LI), Wflfl» Major piéfessor September 25, 1974 / Date / 0—7 639 ABSTRACT A STRUCTURAL STUDY OF SOME COMPLEXES CONTAINING THE SCANDIUM(III) ION: SIX AND EIGHT COORDINATION BY Thomas John Anderson The crystal and molecular structures of some complexes containing the scandium(III) ion and uninegative bidentate oxygen donor ligands have been determined. Precession camera techniques have been used to determine space group information and lattice parameters and three—dimensional single—crystal intensity data have been collected by use of a four— circle, computer controlled diffractometer. The crystal and molecular structure of tris(acetylacetonato)- scandium(III), Sc(C5H702)3, has been determined. The crystal structure belongs to the orthorhombic space group Ebga, with a_= 15.38(3), h = 13.73(3) and g_= l6.72(4)A, with E = 8. The structure has been solved by Patterson and Fourier techniques with 1088 reflections refined by full-matrix least—squares to a final reliability factor, based on F, of 0.059. The structure consists of discrete Sc(C5H702)3 molecules with the six oxygen atoms forming a distorted octahedron about the scandium(III) ion. The compound tris(tropolonato)scandium(III), Sc(C7H502)3, crystallizes in the trigonal space group R3c, with a = lO.455(2) and g = 32.595(1) in the hexagonal setting, with Z_= 6. The structure was solved by Patterson and Fourier techniques and 783 reflections were refined to a final reliability factor, based on F, of 0.033. Thomas John Anderson The structure consists of discrete Sc(C7H502)3 molecules with crystallographically imposed D3 symmetry. The coordination environment about the scandium(III) ion is intermediate between octahedral stereochemistry and trigonal prismatic stereochemistry. The crystal and molecular structure of hydrogentetrakis(tropolonato)- scandium(III), HSc(C7H502)4, has been determined. The crystal structure belongs to the triclinic space group 3:, with a = 10.022(4), g = 11.515(3), g = 12.004(4)X, a = 72.74(1), s = 8458(1) and Y = 65.04°, with g = 2. The molecular structure has been solved by a combination of Patterson, direct method and Fourier techniques with.2539 reflections refined to a final reliability factor, based on F, of 0.027. The structure consists of a hydrogen bonded dimer, with halves of the dimer joined across the center of inversion. The hydrogen bond is nearly linear with the O-H...0 angles being l75.9°. The O—H bond length is 1.00A, while the O...H separation is 1.49A. Each scandium(III) ion is eight—coordinate. The stereochemistry of the scandium is intermediate between a D2d dodecahedron and a C2v bicapped trigonal prism. Factors controlling the stereochemistry for six— and eight- coordinate complexes of scandium(III) are discussed. A STRUCTURAL STUDY OF SOME COMPLEXES CONTAINING THE SCANDIUM(III) ION: SIX AND EIGHT COORDINATION By Thomas John Anderson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1974 ACKNOWLEDGEMENTS There have been many people, both teachers and students, whose desire to learn has encouraged me to work hard to expand my own knowledge. I regret that I cannot personally thank each one that I have met during my education, to a special few I would like to offer my deepest gratitude: I wish to express my most profound thanks to Dr. Gordon Melson, without his-prodding and direction I would still be trying to get some work done. A special thanks to my parents for their encouragement and believing in me. I would also like to thank Dr. Sam Paton, for if he had not been teaching a laboratory class across the hall from the one I was in charge of, I would never have met Charmaine. I would also like to thank my colleagues, past and present, for the good times at Aker's Golf Course, at the ”card table”, and trying to keep alive in room 317. A long overdue thanks is extended to Dr. Robert Echt, whom I have known since my high school chemistry class. I would like to thank those who have taught me the crystallography that I have learned and expanded my curiosity also: Dr. Mel Neuman and Dr. Bobby Barnett. Finally, I would like to thank the Department of Chemistry at Michigan State University for financial support in the form of a Teaching Assistantship for my years at MSU. ii Preface This thesis has been divided into four main divisions: l)the introduction, describing the chemistry of scandium, chemical similarities to the lanthanides, and reasons for the investigation itself; 2)chapters discussing the crystal and molecular structures of tristacetylacetonato)- scandium(III) and tris(tropolonato)scandium(III), and a chapter discussing factors controlling six—coordination and the relationships of these two structures to those factors; 3)a chapter presenting the crystal and molecular structure of hydrogentetrakis(tropolonato)— scandium(III) and eight-coordination; finally, 4)other work partially completed or work that has been attempted but was not successful. TABLE OF CONTENTS CHAPTER 1. Introduction History The Chemistry of Scandium and Its Relationship to Yttrium and the Lanthanides Complexes of Scandium Structural Studies of Scandium Complexes Outline of Investigation The Crystal and Molecular Structure of Tris(acetyl— acetonato)scandium(III) Introduction Experimental Section Determination of Space Group and Measurement of Intensity Data Solution and Refinement of the Structure Description of the Structure The Crystal and Molecular Structure of Tris(tropolo— nato)scandium(III) Introduction Experimental Section Measurement of Crystal and Intensity Data Solution and Refinement of the Structure Description of the Structure and Discussion Discussion of Six Coordination Introduction Discussion of [M(bidentate) ] Complexes The Ionic Radius of Scandium(III) The Reaction Path Model for Six-Coordination Conclusions for Six-Coordination The Crystal and Molecular Structure of Hydrogen- tetrakis(tropolonato)scandium(III) Introduction Experimental Section Measurement of Cell Dimensions and Intensity Data iv PAGE H O\O\-L\i-‘ 26 26 26 27 29 34 42 42 48 50 50 53 53 57 58 TABLE OF CONTENTS (CONTINUED) Solution and Refinement of Structure Description of the Structure and Discussion Coordination Polyhedron Hydrogen Bonding and Ligands 6. Discussion of Eight—Coordination and Further Work APPENDICES Appendix I Appendix II Appendix III Appendix V Appendix V H REFERENCES In Text General Computer Programs 100 101 102 103 104 105 109 109 LIST OF TABLES TABLE 1. Selected Parameters of Scandium, Yttrium and the Lanthanides 2 Crystal Data 33. Final Atomic Coordinates 3b. Final Anisotropic Thermal Parameters 4. Hydrogen Atom Positions (B = 8.0A2) 5. Final Calculated and Observed Structure Factor Amplitudes 6 Intramolecular Interatomic Distances (A) and Angles (deg) 7. Deviations from Ligand Planes (A) 8. Crystal Data 9. Final Calculated and Observed Structure Factor Amplitudes (X 10) 10a. Final Atomic Coordinates 10b. Final Thermal Parameters ll. Interatomic Distances (A) and Angles (deg) 12. Deviations from Ligand Plane (A) 13. Structural Details of Some M(acac)3 and M(trop)3 Complexes 14. Predicted O...O Contacts for Some M(acac)3 and M(trop)3 with Trigonal Prismatic Stereochemistry 15. Dihedral Angles for Sc(acac)3 and Sc(trop)3 (deg) 16. Infrared Spectra (nujol mull) vi PAGE ll 13 14 16 19 22 24 28 31 32 33 37 40 45 49 51 55 LIST 17. 18. 19. 20. 21. 223. 22b. 23. 24. 25. 26. 27. 28. 29. 30. 31. Al. OF TABLES (CONTINUED) X—Ray Photoelectron Spectra in the Oxygen (ls) Region for Several Compounds Crystal Data Starting Phases from MULTAN Statistical Analysis of lEl Values from FAME ( Rescaled to 1.00) Final Calculated and Observed Structure Factor Amplitudes (X 10) Final Atomic Coordinates Final Anisotropic Thermal Parameters Final Hydrogen Atom Coordinates and Isotropic Thermal Parameters Shortest Non—bonding Separations O Interatomic Distances (A) and Angles (deg) within the Coordination Polyhedron Least Squares Planes Shape Parameters for HSc(trop)4 Dihedral Angles for HSc(trop)4 and Ideal Polyhedra (deg) O Interatomic Distances (A) and Angles (deg) for the Tropolonato Ligands Normalized Ligand Bite Distances for Several Eight— Coordinate Complexes Crystal Data Analytical Data for the Compuonds Studied 56 59 61 63 67 69 71 74 77 80 82 86 88 91 97 99 104 10. 11. 12. 13. LIST OF FIGURES Structure of scandium formate, Sc(CHO ) , projected 2 3 on (010) View of Sc(acac) showing numbering scheme, adopted bond lengths, and the 20% probability envelopes of the anisotropic thermal ellipsoids Stereoscopic View of Sc(acac) down the C axis of the molecule, 20% probability envelopes, hydrogens omitted for clarity Stereoscopic View of the Unit Cell of Sc(acac) down the b axis, hydrogens omitted for clarity Stereoscopic view of tris(tropolonato)scandium(III) down the C3 axis showing the numbering scheme adopted Coordination environment of Sc(trop)3 Interatomic distances and angles Stereoscopic view of the molecular contents of the Unit Cell, 20% probability envelopes, hydrogens omitted for clarity Variation of ALFSE with the d_orbita1 occupation number, where ALFSE = LFSE(TAP) — LFSE(TP) (from Wentworth (47)) Difference Fourier map showing hydrogen (H) position, X, located between oxygen atoms 01(4) and 01(1). Section of map at b_ = 0.45. Atomic coordinates are related to those in Tables 22 and 23 by 1—x,l—y,1—z Stereoscopic View of half of the HSc(trop)4 dimer, 25% probability envelopes, ring hydrogens omitted for clarity Interpenetrating trapezoids and hydrogen bonding in HSc(trop)4 Coordination polyhedron for the scandium(III) ion in HSc(trop)4 viii PAGE 20 21 21 35 36 38 39 43 65 76 76 85 LIST OF FIGURES (CONTINUED) 14. Al. A2. A3. A4. Polyhedron edges defining the dihedral angles indicated by double lines ESCA spectrum in the O ls ESCA spectrum in the O ls ESCA spectrum in the 0 ls ESCA spectrum in the 0 Is region region region region for Htrop for Sc(trop)3 for HSc(trop)4 for Cs[Sc(trop)4] 89 100 101 102 103 CHAPTER 1 INTRODUCTION History The discovery of scandium (Sc) in 1879 by Nilson (l) and the subsequent determination of its elemental properties established it as the first member of the 3d transition series. The discovery also helped substantiate Mendelev 3 Periodic Law (2) by verifying the existence of an element he had predicted, 315., "ska-boron". The chemical properties of eka—boron were predicted to be intermediate between those of aluminum and yttrium in the boron group. The early chemical investigations of scandium dealt principally with the purification of the metal from its mineral sources, development of analytical techniques and the preparation of binary compounds (3). During the period 1879 until after World War II, only scattered reports of investigations of coordination compounds of scandium have appeared. The Chemistry of Scandium and Its Relationship to Yttrium and the Lanthanides The free scandium atom has an electronic configuration [Ar]3d14sz. Because of the similarity of their electronic configurations, the chemistry of scandium is often discussed along with that of yttrium and the lanthanides. Moeller, g£_§1, (5), have discussed these 1 similarities. The most common oxidation state of all these elements is +3, each with a noble gas configuration resulting in unfilled outer d_orbitals. For the lanthanides, the +3 oxidation state arises because electrons occupying the 4f orbitals are screened from the environment by overlying 55 and 5p shells, thereby rendering the 4f electrons relatively inert to the chemical surroundings. The electronegativities and other properties of these elements are listed in Table 1. The tervalent ions, in particular, exhibit the greatest affinity for ligands containing electronegative donor atoms, especially oxygen. Complexes containing chelating oxygen donor ligands may be isolated as anhydrous products of reactions that have taken place in aqueous solutions. Examples of these are B—diketonates, carboxylates and tropolonates. Complexes with neutral, unidentate oxygen donor ligands are thermodynamically unstable relative to the hydrated species. In these cases, insoluble adducts may be obtained from solutions of the hydrated meta1(III) salts in organic solvents, since a reduction of water concentration would reduce the concentration of [M(H20)n]+3 species. Hydration is an ever present problem when working with any of the tervalent salts of scandium (especially the halides). The preparation of all but the anhydrous fluoride must be completed under anhydrous conditions. A method for these preparations was reported by Stotz and Nelson (7). The degree of electrostatic contribution to the metal-ligand bond increases with decreasing electronegativity of the cation in the order La+3 2.5. The data were corrected for the Lorentz and polarization phenomena, but no correction for absorption was applied (u = 4.40cm_l). Solution and Refinement of the Structure Scattering factors for the neutral atoms, Sc, 0, and C were taken Table 2. Crystal Data Molecular Formula Molecular Weight Crystal Habit Crystal Size, mm Crystal Color p, cm—1 Space Group Systematic Absences U’ IU‘ W {>0 >0 >0 [N 0 .A3 c: l a Dexp’ g/ml Dcalc’ g/ml 11 Sc(C H7°2)3 5 342.29 Plate 0.3 x 0.6 x 0.08 g Clear 4.40 Pbca, Orthorhombic EEO, h_= 22_+ l hO£, E = 22_+ 1 @Lk=23+l 15.38(3) 13.73(3) l6.72(4) 8 3531 1.26 8‘Measured by flotation in carbon disulfide. 12 from Cromer and Waber (33) and those for Hydrogen from Stewart, §£_§l; (34). The positional coordinates of the scandium and six oxygen atoms were obtained from a three—dimensional point—atom sharpened Patterson map. Structure factors calculated on the basis of these coordinates and isotrOpic thermal parameters of 3.53.2 were used to generate a three— dimensional heavy atom Fourier map which yielded the positions of the carbon atoms. The positional and isotropic thermal parameters associated with all nonhydrogen atoms were refined by using full—matrix least-squares on all parameters for four cycles. A difference Fourier map was calculated at this time and it yielded positions of all methyl hydrogens. The y— carbon hydrogens were then positioned by assuming a unit vector beyond the vector from the scandium to the y-carbon. Isotropic thermal parameters of 8.0A2 were assigned to the hydrogen atoms, based upon the thermal parameters of the methyl carbons. The coordinates of all atoms and anisotrOpic thermal parameters of non— hydrogen atoms were refined to convergence. The final agreement factor, R1, was 0.05(Rl = EllFo—FCII/ZIFOI) while the weighted agreement factor, 1/2 Rw, was 0.067(RW = [Zw(F0—FC)2/E(FO)2] where w = 1/02(F). 0(F) = 0.05F for F>4Fmin and (F) = 0.20Fmin for F<4Fmin, Fmin = 9.46. (This is a modified Hughes' weighting scheme (35).) The final atomic coordinates of all non-hydrogen atoms and the anisotropic thermal parameters with their associated estimated standard deviations, e.s.d.'s, are listed in Table 3. The final hydrogen atom positions are listed in Table 4 along with their e.s.d.'s. The largest parameter shift in the final cycle of refinement was not larger than 0.2 of its associated e.s.d. A difference Fourier map of the final structure 13 Table 3a. Final Atomic Coordinates Atomsa ' x/a y/b z/c Sc 0.1434(1) 0.2687(1) 0.2418(1) 01(1) 0.0704(4) 0.3526(4) 0.3189(3) 02(1) 0.0386(3) 0.1749(4) 0.2539(4) 01(2) 0.0966(4) 0.3367(4) 0.1394(3) 02(2) 0.1979(4) 0.1760(4) 0.1569(5) 01(3) 0.2441(4) 0.3671(4) 0.2524(4) 02(3) 0.2125(4) 0.1968(4) 0.3293(3) 01(1) -0.0313(8) 0.4222(9) 0.4052(7) 02(1) -0.0015(7) 0.3396(7) 0.3519(5) 03(1) —0.0531(6) 0.2589(8) 0.3417(6) 04(1) -0.0304(6) 0.1796(6) 0.2944(6) 05(1) —0.0906(6) 0.0937(8) 0.2875(6) 01(2) 0.0535(8) 0.3832(9) 0.0107(6) 02(2) 0.0998(6) 0.3137(7) 0.0666(5) 03(2) 0.1424(7) 0.2339(8) 0.0363(5) 04(2) 0.1887(6) 0.1690(6) 0.0826(6) 05(2) 0.2368(8) 0.0864(9) 0.0425(6) 01(3) 0.3630(7) 0.4610(7) 0.2932(7) 02(3) 0.3088(6) 0.3716(6) 0.2964(6) 03(3) 0.3330(6) 0.2977(8) 0.3486(6) 04(3) 0.2851(7) 0.2139(7) 0.3607(5) 05(3) 0.3195(7) 0.1378(10) 0.4176(7) a . The numbers in parentheses refer to the ring number. % l4 Aevmaoo.ou AmeHoo.o- nevsooo.o novsmoo.o Amvmooo.o Asvaaoo.o Amvmaoo.o Asvoooo.o onsooo.o- Amvaaoo.o Amvoooo.o Amvmooo.on Amvoooo.o Amvmaoo.os Amvmaoo.01 ANVHoooo.o Mum Asvmooo.o- Asvoooo.o Asvwfioo.ou onmaoo.ou Amvmooo.o Aevmooo.o- Asvemoo.o Aevmaoo.o onemoo.o Amvoooo.ou Aevsooo.o- Anemooo.o Amvmooo.o- Anemooo.o Amvssoo.o Aquoooo.o mam Amvwooo.o- AsVsHoo.o Asvmaoo.ou Amvmooo.o Awesooo.o Amvmooo.o novwooo.o oneaoo.o Asymmoo.o Amveooo.o- Amvaooo.o- Amvmaoo.o AmeHoo.o Anvwooo.o Anymooo.o vaHsooo.o Nam Ascesoc.o Amemoo.o Assamoo.o Amvamoo.o Asvewoo.o Amvesoo.o Amvmsoo.o Aevomoo.o “mammoo.o Amvmsoo.o Amvmmoo.o AmVNmoo.o Amvsmoo.o AmVsmoo.o vammoo.o Amvmsmoo.o mmm p Asvaooo.o Amvsoao.o Asvssoo.o Aoavmsao.o Asvmwoo.o onmqoo.o Asvanoo.o Amvweoo.o Aoavmmao.o Asvomoo.o Amvmmoo.o Asymmoo.o Aevmsoo.o Aevsmoo.o Asvaooo.o Asvsseoo.o Amvmmoo.o “ovmwoo.o “evesoo.o Amvmsao.o Amvamoo.o Amvmmoo.o Amvmmoo.o onseoo.o Amvmaao.o Asymmoo.o Anvamoo.o Asvseoo.o Amvmsoo.o Amvosoo.o Amvsmoo.o onsasoo.o Ham mumuwamunm Hmfiumnw osnouuomfia< Hmcfim vaso vamo vauo ANVHU Aavmu Aavqo Aavmu Aavmo AHVHU Amvmo Amvao Amvmo Amvao Aflvmo Aavao 6m 334 .pm menus 15 + damamu + Mswfimm + Nammm + Nxmmm + NnHHmleaNm ”show mnu mo mum mumumamuma Hmaumsu camouuomfiam 03H vammoo.o Amvmaoo.o Anymooo.o Asvmaoo.ou Amvsaoo.o- novsmoo.ou mum onsmoo.ou Revsooo.ou Asvmaoo.o- Asveooo.o Asvwooo.o lovamoo.o mam Amvmooo.o onouoo.o Amvmooo.o- Amveooo.o- Asvomoo.ou Anymooo.01 Nam flavowoo.o Asvseoo.o Amvomoo.o Aevomoo.o onqwoo.o nmvonoo.o. mmm Amavcoao.o vasmoo.o vammoo.o asvmmoo.o Asvwmoo.o Amvaoao.o «mm Asvmwoo.o Asvssoo.o onomoo.o Amvaeoo.o Ascomoo.o Amvsoao.o Ham Auwnnwuaoov .HAsmemN n Ameo Amvso Amvmo Amvmo Amvso Auvmo 334 .nm wanna 16 °2 Table 4. Hydrogen Atom Positions (B = 8.0 A )a Atom b H1(1) H2(1) H3(1) H4(1) H5(1) H6(l) H7(l) H1(2) H2(2) H3(2) H4(2) H5(2) H6(2) H7(2) Hl(3) H2 (3) H3(3) x/a -0.092 —0.007 -0.033 —0.107 -0.073 -0.079 -0.152 0.094 0.006 0.070 0.147 0.213 0.285 0.215 0.426 0.373 0.359 7/61 0.422 0.400 0.487 0.251 0.053 0.064 0.109 0.427 0.398 0.379 0.216 0.061 0.095 0.034 0.448 0.483 0.486 0.406 0.449 0.375 0.368 0.330 0.234 0.308 0.004 0.030 —0.036 —0.019 -0.009 0.393 0.064 0.285 0.347 0.253 17 Table 4. (Continued) Atomb x/a y/a z/a H4(3) 0.400 0.303 0.373 H5(3) 0.281 0.134 0u465 H6(3) 0.374 0.152 0.434 H7(3) 0.292 0.091 0.405 aNo attempt was made to idealize methyl hydrogens. Estimated deviation t0.006 for all values. bThe atoms are labelled as: l-3 attached to Cl; 4 to C3; 5-7 to C5. The numbers_in parentheses correspond to the ring number. l8 0 showed no peaks larger than 0.35e /A3. The final calculated and yobserved structure factor amplitudes are listed in Table 5. Description of the Structure In Sc(acac)3, the b axis is nearly parallel with the pseudo-C3 axis of the molecule. A view down the pseudo-C3 axis is shown in Figure 2. This illustrates the numbering scheme, bond lengths and 20% probability envelopes for all non—hydrogen atoms of one molecule. Stereoscopic views of the molecule and the molecular contents of the unit cell (also down the b_axis) are shown in Figures 3 and 4, respectively. Intramolecular distances and angles together with their estimated standard deviations are listed in Table 6. The structure consists of discrete Sc(acac)3 molecules within the unit cell. The molecules have distorted D3 symmetry with the chelate rings showing only small deviations from planarity (Table 7). The non—planarity may be due to intermolecular forces which cause rotations about the 01...02 vector (folding) and rotations about the Sc...y-carbon (C3) of the ring. Thus, the former effect can account for the Sc atom being out of the plane for ring 2. A deviation of 0.15; of the metal ion from the plane corresponds to a rotation of approximately 5° about the 0...0 vector. The latter effect may explain why Cl and C5 atoms are markedly out of the plane of ring 3. These effects may also explain similar deviations found in other acac complexes (36—39). In comparison with close packed organic structures, the intermolecular contacts observed are generally long. The shortest H...H contact is 2.49;, while for C...H approaches the shortest is 2.95:. For 0 0...H approaches, one value at 2.56A is found while all others are close 19 Final Calculated and Observed Structure Factor Amplitudes Table 5. n at unflnninl n nnuuunnn unlaxnn n unnu- I )6I0. - .I:I . 10:41.: a 61-11)) uuonuununuuuUunn p) ur»nnln..uunyuon). puma-luau num.u~uu. . II” '0‘ quvIu-1).¢nunlxvvuuonr .Ivv: :v:vua-un-.~o~o‘.~ ..... nu. -- . ........- u an "unnuunnunflnn: 2 p. . .. . u-.......uuuuu-.u..unuun-.. . I .. .... .... ... . ..........................y.-. u .---.-.- .I........I..,.>-..................-..... m I . .nnuuunuu - I unfluuu-Iw . m I IlllplI: m . . l III I ID | I I II III III- Illll I I n a I .I--. -I.-..--.a u a . . ....uuuu-...uuu.. .. . . I run III you Ill l I I Inn ”A l I l l I I I n n b p . a .. - .- . . r IuII II n a 20 Figure 2. View of Sc(acac)3 showing numbering scheme, adopted bond lengths, and the 20% probability envelopes of the anisotropic thermal ellipsoids. 21 Figure 3. Stereoscopic View of Sc(acac)3 down the C3 axis of the molecule, 20% probability envelopes, hydrogens omitted for clarity. Figure 4. Stereoscopic view of the Unit Cell of Sc(acac)3 down the b_axis, hydrogens omitted for clarity. Table 6. Sc-Ol(1) 80-02(1) SC-Ol(2) Sc-02(2) 80—01(3) 30-02(3) Average 01(1)-02(1) 02(1)—04(1) 01(2)—02(2) 02(2)-04(2) 01(3)-02(3) 02(3)—04(3) Average 01(1)—02(1) c4(1)-05(1) 01(2)-02(2) 04(2)—05(2) Cl(3)—02(3) 04(3)—05(3) Average 22 Interatomic Distances (A) and Angles (deg)a O-O "bite" Polyhedron Edges C-C (Ring) Intramolecular Distances ngl 2.062(6) 01(1)-02(1) 2.073(5) 01(2)—02(2) 2.079(6) 01(3)—02(3) 2.082(6) Average 2.062(7) 2.061(6) 01(1)—01(2) 2.070(9)b 01(1)—01(3) g:§_ 01(2)-01(3) 1.248(10) 02(1)—02(2) 1.261(10) 02(1)—02(3) 1 259(10) 02(2)-02(3) 1.254(10) 01(1)-02(3) 1.240(10) 01(2)-02(1) 1.257(11) 01(3)-02(2) 1.253(11)b C—C (Me thyl) 1.514(14) 02(1)-03(1) 1.504(12) 03(1)—04(1) 1.513(14) 02(2)—03(2) 1.511(13) 03(2)—04(2) 1.485(12) 02(3)-03(3) 1.509(13) 03(3)—04(3) 1.506(14)b Average 2 2 2 2 3 2 N N N N U) U) U) H H H H H H H .717(7) .716(8) .713(8) .715(5)b .037(10) .900(9) .981(9) .938(9) .972(9) .907(10) .063(9) .065(9) .152(9) .373(l3) .390(l3) .373(l3) .379(12) .389(l3) .382(l3) .381(8)b Table 6. 01(1)-Sc—02(l) 01(2)-Sc—02(2) 01(3)—Sc-02(3) Sc-01(1)-C2(1) s6-02(1)-04(1) Sc—01(2)—02(2) Sc-02(2)-04(2) Sc-Ol(3)—C2(3) Sc—02(3)-04(3) 01(1)-02(1)—01(1) 02(1)-04(1)—05(1) 01(2)-02(2)—01(2) 02(2)—C4(2)—C5(2) 01(3)-02(3)—01(3) 02(3)—C4(3)—C5(3) (Continued) 82 81. 82. 132. 132. 123. 132. 133 131. 114. 115. 114. 116. 118. 116 .l(2) 5(2) 3(2) 7(6) 4(6) 1(6) 6(6) .3(6) 7(6) 9(9) 9(8) 9(9) 2(9) 1(9) .4(9) 23 Angles 01(1)—C2(l)-C3(1) 02(1)-C4(l)—C3(l) Ol(2)—C2(2)-C3(2) 02(2)—C4(2)-C3(2) Ol(3)—C2(3)-C3(3) 02(3)—C4(3)—C3(3) C2(l)—C3(1)-C4(1) CZ(2)-C3(2)-C4(2) C2(3)-C3(3)-C4(3) Cl(l)—C2(1)—C3(1) C3(l)—C4(l)—C5(l) Cl(2)—C2(2)—C3(2) C3(2)-C4(2)—C5(2) Cl(3)-C2(3)—C3(3) C3(3)-C4(3)—C5(3) 125. 123. 125. 124 123. 124 O 120. 119. 118. 120. 120. 120. 119. 118. a Errors referred to the last significant digit are in parentheses. b Errors for averages are computed using the method of small sample statistics. Analysis? Row, Peterson, and Co., 1957, p. 557J 0(9) 9(9) 1(8) .4(8) 5(8) 7(8) 1(9) 3(9) 9(9) 1(9) 2(9) 0(9) 3(9) .4(9) 9(9) (See Blaedel, W. and Meloche, V.,"Elementary Quantitative 24 Table 7. Deviations from Ligand Planes (A)8 Ring Sc C1 C5 1 -0.011 -0.011 +0.007 2 —0.l49 +0.07l +0.051 3 +0.031 -0.164 +0.105 aLigand planes defined by atoms 01, C2, C3, C4, and 02 of their respective ring. Estimated standard deviation is 0.02 A. Equations of planes: Ring 1 -0.4553x + 0.4351y - 0.77682 = —2.5289 Ring 2 0.8212x + 0.5569y — 0.1245z = 3.5116 Ring 3 0.4986x — 0.4758y - 0.72462 = —3.6160 X, y and z are coordinates (g) in an orthogonal system relative to the crystal axes a, b_and g, respectively. 25 o to 3A. The bond lengths and angles within the acac ligand are similar to thOSe found for other acac complexes (27). In Table 6, the oxygen— oxygen separations are listed for the molecule. The average intrachelate separation (2.715A) usually referred to as the "bite” of the ligand is similar to that found in other tervalent acac complexes (27). Two types of distances are found for the interchelate distances. For oxygen atoms in planes formed by the C3 axis, the separations average to 2.956;, while those between the planes average to 3.093;. It is thus apparent that considerable distortion from a regular octahedral environment is observed for the coordination environment of the scandium(III) ion. In contrast, for the Cr(acac)3 complex (40) a nearly regular octahedron is found with the intrachelate separations, 2.7152. The interchelate separations in Sc(acac)3, all being longer than the minimum van der Waals contact distances found in other acac complexes (27) (2a; 2.60A), indicate that the rings probably behave independently within the molecule. This nonrigid stereochemistry then allows the rotations described earlier to take place. CHAPTER 3 THE CRYSTAL AND MOLECULAR STRUCTURE OF TRIS(TROPOLONATO)SCANDIUM(III) Introduction The tropolonate ion has a shorter "bite" and the ligand itself is not easily distorted as the acetylacetonate ion. These two factors may influence the overall stereochemistry of a complex involving this ligand and scandium(III). For these reasons the second compound of scandium used for structural investigation was the tris—tropolonate of scandium, Sc(C7H502)3. Exper imental Sect ion Scandium(III) oxide, Sc203 (99.9%), was purchased from Research Organic/Inorganic Corp. Tropolone, C7H6O2 (98%), was purchased from Aldrich Chemical Co., Inc., and was used without further purification. Tris(tropolonate)scandium(III), Sc(trop)3. Hydrated scandium nitrate (0.34 g), prepared by evaporation of a solution of $0203 and dilute nitric acid, was dissolved in water (30 ml) and added to a solution of tropolone (0.46 g) in methanol (5 ml). The solution was stirred for 2-3 hours at room temperature and then filtered. The residue was dried under vacuum over P4010 at room temperature and consisted of pale-yellow HSc(trop)4 and red—brown Sc(trop)3 crystals. Crystals 0f Sc(trop)3 suitable for crystal structure studies were readily separated from the mixture. The procedure reported by Muetterties and Wright (14) for the preparation of Sc(trop)3 from ESc(trop)4 did not yield acceptable crystals. 26 27 Measurement of Crystal and Intensity Data Preliminary symmetry and space group determination was conducted by using the precession film technique and MoKo radiation. On the basis of systematic absences, the crystal was initially assigned to the space group 1215 or Ia with the spindle axis coincident with s the a axis for the precession geometry. This was transformed to the equivalent £215 or_gg space groups (correct orientation for monoclinic space) (a = 12.36, b_= 10.39, g_= 18.112; 8 = 119°). It was suspected then that Sc(trop)3 could be isomorphic with Al(trop)3 (41). However, further investigation of reciprocal space revealed six-fold symmetry with 3% coincident to the spindle axis of the camera. Dr. L. J. Guggenberger was kind enough to send the precession pictures of A1(trop)3 (43), which also encouraged further inspection of the symmetry involved as the intensity patterns were not similar.) As a result the crystal is assigned to the trigonal space group, in the hexagonal setting, 322 or REE (hkig, 7h +.§ +‘g = 3E5 hng, &.= 22, observed). The monoclinic to hexagonal transformation was as follows: The (200) reflection became the (006), the (020) relection became the (110), and the (002) reflection became the (112). The choice of the centrosymmetric space group, REE) was later justified by the refinement. Based upon this assignment, complete three-dimensional X—ray diffraction data were again measured by use of a Picker Four-circle computer controlled diffractometer. Conditions were the same as discussed in Chapter 2. The crystal was mounted with the (114) reflection approximately coincident with the ¢—axis of the diffractometer. Lattice parameters (Table 8) were determined from twelve hand—centered reflections at ambient temperature (24 1 2°). Table 8. Crystal Data Molecular Formula Molecular Weight Crystal Habit Crystal Color Crystal Dimensions, mm -1 “9 cm Space Group Systematic Absences >0 9? >0 U.) a Dexp’ g/ml Dcalc’ g/ml IN bIn hexagonal setting, i_= I , 28 Sc(C7H502)3 408.31 Trigonal Reddish-Brown 0.31 X 0.31 X 0.42 4.00 R30, Trigonal, Hexagonal Setting hkizb, —h+k+£=3g+l I?! IN 0 9 £I= 22 + l 10.455(2) 32.595(1) 3085 8Measured by flotation in methylene chloride. -e+s. 29 Intensities were determined from w—scans over a width of O.8° at a rate of 0.5°/min. Data were collected over the range 2.5°<20<60°. Backgrounds were measured by two 10 second counts, one at each end of the scan. Attenuators were automatically inserted when the count rate exceeded 10,000 counts per second. One monitor reflection was measured every 50 reflections. No decomposition nor motion of the crystal was observed, the average deviation being :12. An initial set of data was collected (to 26 = 30°)'(h,':k, :£) to verify the systematic absences for the space group and lattice type employed (Table 8). The only exception was the (0, 0, 13) reflection, this may be explained as a result of the Renninger effect (44), since the crystal was oriented in a manner to allow multiple reflections. The second data set, which was used in the refinement of the structure, was the unique (h, k, g) set. A total of 1075 independent relections were scanned of which 783 were found to be above background using the criterion I/o(I):3.0. Solution and Refinement of the Structure The positional coordinates for the scandium were obtained from a three—dimensional Patterson map. The scandium is located at the 22 special position (0, O, 1/4). Because of the complexity of a Patterson map in high symmetry space groups and the need to determine only one sixth of the molecule, a model was generated with Sc—O distances within the ligand and bond angles were used as found by Muetterties and Guggenberger (42). A rotation of 45° out of the 52 plane was assumed for the ligand (about the Sc—C(4) line). The coordinates of all non—hydrogen atoms were placed relative to the_§2 30 position. Isotropic thermal parameters of 2.5;.2 were initially assigned to all atoms. A structure factor calculation at this stage resulted in R1 = 0.46. The positional and istropic thermal parameters of those atoms allowed to vary were refined for three cycles of fullnmatrix least-squares. At this point, R1 = 0.096. Hydrogen atoms were then placed 0.96A from their bonded carbon atoms by assuming the corresponding angle bisector. Further refinement of the positional parameters, anisotropic thermal parameters for the hydrogen atoms resulted in a converged R1 = 0.033 and R.w = 0.057 (45). Early in the refinement, it became obvious several strong reflections were suffering from either extinction [(006) and (012)] or Renninger effects [(104) and (122)]. As a complete hemisphere of data had been taken, the inconsistencies of intensities could be verified easily. These reflections were omitted from refinement and are indicated by an asterisk in the final listing of calculated and observed strucutre factor amplitudes (X10) (Table 9). No correction for absorption (u = 4.00 cm—l) was made. The error of a measurement of unit weight was 1.14 (S2 = w(|FO] - [FCI)2/(m - n), m = the number of observations, n = the number of variables and w = weight applied to each reflection (see footnote 45). The error suggests that the weighting scheme employed was not incorrect, but could have been improved. The atomic positions and anisotropic thermal parameters of all non—hydrogen atoms and isotropic thermal parameters of the hydrogen atoms obtained from the final cycle of least—squares refinement are listed in Table 10 along with their estimated standard deviations, e. s. d.'s. A difference Fourier map of the final structure showed 31 Table 9. Final Calculated and Observed Structure Factor Amplitudes (X 10) IIIIIu-In -III|IIuI IIII-IIIII IIIIIIIIIII ll Irvmm ‘I-KI . I 'M'Ll I I “III:- I I 'I.‘ re. I I. mum. IIII I l‘ I III I I II II III .. .. .. II I I I P‘ III. III I I III I III III I IIII II I II III III I III I I n I I I4 I. In I II I n..." I I I. II. I... I I. I I II In I I I I I I I I I I a II I p I- I .. I... I II I II I I I I . I l I I I I II I. I I I I I I I II II I I I I r I I I I. I I I I I l‘ I r I I I I I I I I I I I I I I I I I I I .I.I I I I II I I I I I I I I I I I I I I I I. I I I I I I I I II I I I I I I .. III I r III III I L- I I I I I II I run I I I I I I Ian- I ’ ‘ 3 I I IIIIu I I I I | I I I l I“... I I I I I l I I I I I I I I I I I II II- I I I ~ I I I I- II I I I I I I I . I I I II I I I I I I I II I I I I II I Ir I I I I I I I I I II I I I I I I I I I I ) I I I I I I I I I I n I I I P I I I I I I I I I- I I II I I I I I I I I I I I a I I I I I I I I I I I I. l I I I I I II I I I I I I I I . r I I I I l - I I I I I I I I I I l ) I I I I I I I I I z 2 I I I I I I I I a I r I I I I I I . I I I I I r a r II I I I I I I II I I 7‘ v I I I I I . I . I I I I I II I II I I4 I I II I I I I I l I I I I ...... .I I I I .I ! I I .. I II ~ I I l I .I I )1 . II IIIIIIII I I .I- .I III I I I .I I. I» I IN. I I ) I III . I. II P4 III I I I I I I II I I II I I H A I I I . ~' ~ I I I- I -- - I II I II I I. . I I I I In I II P I I I I I II: I III I i I I I II I I I I II II III I II I I I I I III III III III n \r' I I I I I I III II \I I I I I NI III II III . \N I I II I I III I II - ., , I . I I III M. VII A) I I III . I > III II Table 108. Atom Sc 01 C1 C2 C3 C4 H2 H3 H4 x/a 0.00 0.18920(12) 0.29064(15) 0.41776(18) 0.53915(22) 0.56953(25) 0.4252(25) 0.6173(28) 0.6619(45) 32 Final Atomic Coordinates y/b 0.00 0.05680(13) 0.03308(15) 0.07189(20) 0.05670(28) 0.00 0.1221(25) 0.1077(27) 0.00 z/c 0.25 0.21466(3) 0.22962(4) 0.20662(5) 0.21557(7) 0.25 0.1808(6) 0.1943(7) 0.25 33 .Hfiaxmumm + unmamw + x: onmaaoo.o Aqvuhooo.o Amvmmooo.o Amvmnooo.o o.o mu NH Aqumooo.o Amvmaaocao Aqvmoooo.o AMVHNooo.o Amvomooo.o o.o mH .mHm*o.N u mmm mm + Nammm + Nxmmm + msfifivalmxw Aamvnmnoo.o Acavawqoo.o AHHmemoo.o AHHVmHmoo.o Nam .NNQIm.o I Nam .qu “am new 0 I Aqvmmfioo.o AquOHoo.o AHVooooo.o Afivmqooo.o Amvmmooo.o onwmqooo.o mmm mmm I mam .HHQIn.o I "om pow umnu ”Aoomavnqa .mm ..uw0HHmum%uu muo¢ .Eamm .mH. vmm wmpmumm .Z.<.h.3~ mwufidwmu huumaahmn “Epow mSu mo wum mumuwamumm Hmauonu camouuomfic camoomoquum LI “1,1 523m— npuznonaom. .mHE 2:.Eazuumfipuzndmnufitfi b. «v .m mHDMHm 36 A5335 "so Quake: "no “No/«xx; "so ANLA .x.>.x¢ “no ANL\_.>L>-5 "No AME: u 5 .mamouuvom mo uamsdoufi>cw :OHumckuoou .o whswwh 37 0 Table 11. Interatomic Distances (A) and Angles (deg)a Sc-Ol Ol-Cl Cl—C2 C2—C3 C3—C4 Cl—Cl' CZ—HZ C3-H3 C4-H4 Ol-Sc—OZ 01—Sc-03 Ol—Sc-O4 Ol-Sc—OS Ol—Sc-06 Ol—Cl—Cl' Ol-Cl-CZ 2.102(1) 1.299(2) 1.398(2) 1.387(3) 1.379(3) 1.457(3)c 0.97(2) 1.00(3) 0.97(3) 73.77(6) 158.32(6) 92.83(4) 104.56(7) 92.83(4) c 114.23(8) 119.45(12) Distances Angles 0102b 2 01-03 4 01-04 3 01—05 3 02—03 3 CZ—Cl—Cl' 126. Cl-CZ-C3 130. C2-C3-C4 129 C3‘C4—C3' 127 C1—C2-H2 115 H2-C2—C3 114 C2-C3-H3 110. aErrors referred to last significant digit are in parentheses. bFor relations between oxygens, see Figure 6. cPrimed atom related to unprimed atom (x,y,z) by (x-y. —y, 1/2-Z)- .lO(24) .523(2) .129(2) .045(2) .325(2) .045(2) 32(8)c 24(15) .89(18) C .3(l4) .4(14) 2(14) 38 / .mMde—m USN WOUHHWUWHHU UHSOUNHGUCH CVNOHN .n wHSMfim 39 Fi ur ' v e g e 8. Stereoscopic iew of the molecular contents of the Unit C 11 5 204 probability envelOpes, hydrogens omitted for clarity 40 Table 12. Deviations from Ligand Plane (A)a Atom Deviation Sc 0.0000 01 0.0060(9) Cl 0.0008(11) C2 —0.0102(l4) C3 -0.0048(20) C4 0.0000 H2 —0.0672(172) H3 —0.1250(l92) H4 0.0000 aLigand plane defined by all non-hydrogen atoms. Equation of plane: -0.4553 x — 0.7891 y ~ 0.4119 z = —3.3S67. x, y and z are coordinates 0 (A) in an orthogonal system defined by figs; b4 of, respectively. 41 contacts are in the ab plane with no short contacts in the £_direction. The bond lengths and angles within the tropolonato ligand conform well with those found by Muetterties and Guggenberger (42) for other structures containing the tropolonato ligand. The ligand "bite" (2.523;) falls within the narrow range observed for compounds containing chelating tropolonato ligands. Thus, the planarity and apparent rigidity of the chelating agent indicate little change in overall structure of the ligand upon coordination to the scandium(III) ion. The coordination environment of the scandium in Sc(trop)3 is best discussed by consideration of the intermolecular non-bonding O...0 distances. There are three distinct groups which define two planes normal to the C axis, viz., 01-06, 01—04, 04-06, and 02—03, 3 02—05, 03—05, separations being 3.045A; third, a group of distances between the two planes defined above, 01-05, 02—04, and 03—06, each having a value of 3.325A. These distances correspond to a trigonal distortion (elongation along the C axis) of the six donor oxygen 3 atoms about the scandium(III) ion. This distortion may be due to two factors: first, the restrictions placed on the molecule due to a short bite distance and the inflexible nature of the ligand and second, packing within the unit cell such that contacts have been increased in the ab plane but not along directions in g_thus, solid state energies may play an important role. CHAPTER 4 DISCUSSION OF SIX COORDINATION Introduction Studies of complexes exhibiting six-coordination have attracted considerable attention because of the characterization of octahedral (or trigonal-antiprismatic (TAP)), trigonal-prismatic (TP), and intermediate stereochemistries. For complexes where the ligands do not control stereochemistry ("innocent ligands"), Wentworth (47) has shown that for most metal ions the ligand field stabilization energy (LFSE) of a trigonal—prismatic complex is less than that for an octahedral complex (see Figure 9). Exceptions to this are predicted for ions with d0, d1, low spin d2, high spin d5 and d10 configurations where there is no preference between stereochemistries. A discussion of the structures of several [M(bidentate)3]+n complexes by Kepert (48) concludes that the stereochemistry of these complexes is principally dependent upon the ligand "bite", with trigonal- prismatic coordination being favored by a short ligand bite distance. The most widely studied complexes of this type contain either the acetylacetonate ion or the tropolonate ion. The tropolonate ion has a shorter bite and is a much more rigid ligand (42 and Chapter 3) than the acetylacetonate ion (27 and Chapter 2) and thus complexes containing the former ligand may have stereochemistries closer to trigonal—prismatic than the corresponding acetylacetonate complexes. This should be more the case when these ligands are coordinated to metal ions where no stereochemical preference is predicted from 43 AmHVMmmA I Am w o ZHNm 30A .AAnqv :uuosusoz aouwv .m muamfim u w u on I|*I1 o HSd'IV ZHMm mUHm 44 LFSE considerations. There have been several reports discussing a convenient way to differentiate between trigonal—prismatic and trigonal—antiprismatic stereochemistries, and ways to describe intermediate geometries (47—50). The simplest to use is the projected twist angle, ¢. This angle is defined as that made by projecting the two chelated donor atoms and the metal atom onto the same plane which is normal to the C3 axis of the molecule, viz.: 1L if‘ with the metal atom as the vertex and the two donor atom projections at the ends. Thus, the projected twist angle for an octahedron would be 60°, while that for a trigonal—prism would be 0°, since both threefold faces would be eclipsed and the donor atoms would be projected onto the same point. 6 may then be used to indicate the degree of distortion from octahedral geometry. Discussion of [M(bidentate)3] Complexes Structural details for some M(acac)3 and M(trop)3 complexes are listed in Table 13. First, note that for complexes where data 45 ¢H qH es mm m .umwso mom we mm: oer ms Amonv Nw mm mm an Amoov 020 moMH @«.N Nm.N mm.N mm.N Asv muss ow.H Ho.N oo.N 0H.N AMn(trop)3>Fe(trop)3> Sc(trop)3. Co(III) and Mn(III), with electronic configurations, 6 d and d4, respectively, have a preference toward trigonal—antiprismatic stereochemistry by consideration of ALFSE with the larger ALFSE being predicted for the Co(III) ion (47) (see Figure 9). Co(trop)3 (with a value of ¢ = 55°) has a stereochemistry close to octahedral, showing that in this case, the electronic preference almost completely overcomes the limitation placed on it by the ligand. For Fe(trop)3 and Sc(trop)3, with d5 and d0 configurations, respectively, there is no preference toward either stereochemistry from LFSE considerations. In these cases, restrictions placed on the molecule by the ligand bite distance are the important factors and considerable distortion toward trigonalcprismatic stereochemistry results. For Mn(trop)3 where there is a small ALFSE in favor of TAP stereochemistry, both the ligand restrictions and electronic considerations are important in determining the resultant stereochemistry. 47 The value of Q for Sc(trop)3 is the smallest observed for any M(trop)3 complex. The Al(trop)3 complex, also a do system, has ¢ = 48°, with an Al—O bond length of 1.89;. (In Sc(trop)3, this distance is 2.10;.) The observed crystal and molecular structures for Al(trop)3 and Sc(trop)3 result from a combination of inter— and intra—molecular contacts. In Al(trop)3, with a twist angle of 48°, the intra—molecular 0...0 nearest neighbor contacts are almost equal at 2.7;. If a twist angle of 33° is introduced (as found in Sc(trop)3) two sets of O...0 contacts are predicted, those within the plane normal to the C3 axis at 2.56A and those between the planes at 3.15;. Correspondingly, the contacts in Sc(trop)3 are 3.05 and 3.25;, respectively. Thus as Q decreases, i;g:, the molecule is distorted toward trigonal—prismatic stereochemistry, the "in—plane" O...O separations decrease while the "between plane" separations (b1 and b2) increase, viz.: Octahedron Trigonal Prism 1 b 48 In Table 14, predicted O...0 contacts for trigonal—prismatic stereo— chemistry are listed for some M(acac)3 and M(trop)3 complexes. (For acac-, the bite is assumed to be 2.8A, while for trop—, the value of 2.5; was used.) If one assumes a van der Waals separation 2 = 1.35A for O...O contacts of 2.702 (effective ionic radius of O— (54, 55)), it can be seen that for acac_ complexes, the O...O van der Waals contact is not attained until M—O bond lengths of 33. 2.10; are reached for TP stereochemistry while for trop_ complexes the value is reached for M—O bond lengths of 22' 2.0;. Further, the structure of Al(trop)3 may be considered as a deviation from TP stereochemistry due to the need for the molecule to increase intramolecular 0...O contacts from the calculated 2.42; to the observed values, which are close to the van der Waals contacts, 315., ~2.73;. However, the distortion of Sc(trop)3 from TP stero- chemistry appears to be the result of a different reason. The in— plane O...O separations already exceed the van der Waals values. Thus, although a TP stereochemistry could be stable for Sc(trop)3 on a geometric basis, in the crystalline lattice a distortion toward TAP geometry is observed. The conclusion is that the molecular structure in the crystal is the result of intermolecular compressive forces which tend to reduce the volume of the molecule. The Ionic Radius of Sc(III) The average Sc-O distances found in Sc(acac)3 and Sc(trop)3 are 2.070 and 2.102A, respectively. By using the assumed effective ionic radius of oxygen as 1.4OA (51), one calculates a radius of 0.67 and 0.70A for the scandium(III) in these structures. This is 49 Table 14. Predicted O...O Contacts for Some M(acac)3 and M(trop)3 with Trigonal Prismatic Stereochemistry M (+3 ionic M4) In Plane (A) Out of Plane (A) radius X) (A) ACAC TROP TROP A1 (0.5) 1.90 2.22 2.48 3.52 Fe (0.6) 2.00 2.47 2.70 3.68 Sc (0.7) 2.10 2.71 2.92 3.85 M (0.8) 2.20 2.94 3.14 4.01 50 in contrast to the Pauling value of 0.81A (4). This value is similar to that found in structures that have appeared since completion of this work where considerable covalent character in the Sc-O bond is expected (56-61) and is somewhat smaller than that determined for "ionic" sincoordinate species (54, 55). The Reaction Path Model for Six—Coordination Muetterties and Guggenberger (42) have developed another set of criteria to describe the coordination polyhedron. The criteria are based upon the dihedral angles formed by the intersection of adjacent planes formed by the donor atoms. There are eight faces and twelve edges determined by either an octahedron or trigonal prism; thus, there are twelve dihedral angles present in either of these polyhedra. The calculated values for Sc(acac)3 and Sc(trop)3 are listed in Table 15 along with the theoretical values for an ideal octahedron and a trigonal prism. For Sc(acac)3 with a ¢ of 47°, indicative of distorted octahedron, the dihedral angles do not differ greatly from the ideal octahedron dihedral angles. However, for Sc(trop)3 the dihedral angles indicate considerable distortion toward trigonal—prismatic stereochemistry, as does a ¢ value of 33°. Thus, both structures are indicated to be intermediate between TAP stereochemistry and TP stereochemistry. Conclusions for Six-Coordination In predicting the stereochemistry of M(bidentate)3+n complexes both the size and electronic configuration of the metal ion, the nature of the bidentate ligands, and the interactions between these Table 15. Dihedral Angles for Sc(acac)3 and Sc(trOp)3 (deg.)a b1 edges b2 edges Other edges Sc(acac)3 60.2, 62.3, 58.1 81.3, 79.7, 81.5 68.2, 68.5, 73.1 70.1, 72.2, 73.2 Sc(trop)3 3 at 57.5 3 at 156.5 6 at 66.5 Octahedron 3 at 70.5 3 at 70.5 6 at 70.5 Trigonal Prism 3 at 0.0 3 at 120.0 6 at 90.0 8The dihedral angles are defined as the angle between the normals of two adjacent planes. “'III"" b; Octahedron b5 Trigonal Prism b1 \ b: \\ bl bt \ 52 ligands must be considered since it is apparent that all these factors play important roles in determining the ultimate stereochemistry. If one can correlate twist angles as determined in the solid state with reaction pathways in solution, as suggested by Muetterties and coworkers (42, 59), Sc(trop)3 and its derivatives should undergo rapid intramolecular rearrangements due to an anticipated low barrier to rotations about the C axis of the molecule, or more precisely, 3 concerted rotations of the ligands about the C2 axes. CHAPTER 5 THE CRYSTAL AND MOLECULAR STRUCTURE OF HYDROGEN TETRAKIS(TROPOLONATO)SCANDIUM(III) Introduction Muetterties and Wright (14) reported the synthesis of metal complexes with the tropolonate anion as the ligand. Among these complexes was the "acid", HSc(trop)4. With only a few physical measurements being made, it was proposed that the scandium ion was in an eight—coordinate environment, surrounded by eight oxygen atoms from four tropolone ligands. Isomorphism was indicated with NaSc(trop)4 and the corresponding lanthanide complexes, which are known to exhibit high coordination. There is however, the possibility that HSc(trop)4 may instead be of the form Sc(trop)3.Htrop, that is, a tropolone adduct to the tris—complex. This possibility arises because the tris—chelated complex may be obtained from the "acid" complex in solution (14). The species ScQ3.HQ has been isolated (60) from the reaction of scandium(III) and 8—hydroxyquinoline, HQ. This species dissociates readily in solution to form ScQ3 and HQ (61). The scandium(III) ion is much smaller than any of the lanthanides (see Table l) which usually have high coordination numbers, thus making it difficult for a large number of ligands to coordinate to the scandium(III) ion. However, an example of a complex of scandium(III) in an eight—coordinate environment has recently been reported by Hansson (62). 53 54 The absorptions associated with v(C=O) and v(O-H) in the infrared spectra of HSc(trop)3, Sc(trop)3 and Htrop are shown in Table 16. The absence of any absorption at 1606 or 3500cm—1 in the spectrum of HSc(trop)4 suggests that no uncoordinated Htrop is present in this compound. X—ray photoelectron spectroscopy can be a useful tool for determining chemical differences of atoms in a compound. The X—ray photoelectron spectra of several scandium compounds have been determined (63) in order to ascertain differences in chemical environment between oxygen atoms of these complexes. Table 17 lists the results of X-ray photoelectron spectra obtained in the oxygen (ls) region for Htrop, Sc(trop)3 and HSc(trop)3. For Htrop, a broad spectrum with a width at half height of 2.7 eV is observed, suggesting two types of oxygen atoms present in this compound. The crystal structure of tropolone (64) verifies the conclusion that there are two types of oxygen atoms present. The molecule is a hydrogen bonded dimer, viz.; ifferences X-ray photoelectron spectroscopy is thus sensitive to small d in the chemical environment of several oxygen atoms in the same molecule. For Sc(trop)3, the symmetrical peak with a width at half height of 2.0 eV suggests all six oxygen atoms are equivalent. The 55 Table 16. Infrared Spectra (nujol mull)a Compound v(C=0), cm_l v(O—H), cm-l HSc(trop)A 1593 not seen Sc(trop)3 1593 ---— Htrop 1606 ‘3500 aObtained on a Perkin-Elmer 457 grating infrared spectrophotometer. 56 Table 17. X—ray Photoelectron Spectra in the Oxygen (ls) Region for Several Compounds Compound Htrop Sc(trop)3 1:1 Composite HSc(trop)4 Width of Peak at Half-Height (eV) 2.7 2.0 2.3 2.1 Comment Unsymmetrical Symmetrical Unsymmetrical Symmetrical 57 crystal structure of Sc(trop)3 verifies the equivalence. The spectrum of HSc(trop)4 is also symmetrical with a half height of 2.1 eV, suggesting that the eight oxygen atoms are equivalent. In order to simulate a mixture of Htrop and Sc(trop)3, a composite spectrum was constructed with the components of the spectra from each compound. This resulted in an unsymmetrical, broad spectrum (a width at half height of 2.3 eV), very unlike the observed spectrum for HSc(trop)4. Thus, the X-ray photoelectron data suggest HSc(trop)4 does not exist as an adduct, 315., Sc(trop)3.Htrop. Hamer, Tisley and Walton (65) have recently reported the X—ray photoelectron spectra in the Sc (2pl/2) and Sc (2p3/2) regions and in the 0 (ls), N (ls) and C (ls) regions for ten compounds of scandium. For HSc(trop)4 and Sc(trop)3 similar 0 (ls) spectra were obtained. Also, the Sc (2p1/2) and Sc (2p3/2) spectra were uninformative about the environment of the scandium ion. Presumably an increase in the number of metal—ligand bonds in HSc(trop)4 compensates for the increased binding energy expected for Sc 2p electrons by increasing the Sc—O bond distance, also expected for higher coordination. In order to ascertain the nature of the solid state environment of HSc(trop)4 a complete single crystal structural determination has been completed. Experimental Section Hydrogentetrakis(tropolonato)scandium(III). The synthesis of HSc(trop)4 is described in the experimental section of Chapter 3. In this case, the method described by Muetterties and Wright (14) also produced suitable crystals for X-ray studies. 58 Infrared Spectra. The infrared spectra were obtained from nujol mulls of the compounds on a Perkin—Elmer 457 Grating Spectrophotometer. X—ray Photoelectron Spectra. Dr. D. M. Hercules (63) provided the X—ray photoelectron spectra of the compounds. The spectra were obtained in the oxygen (ls) region (KE = 716.4 :_0.2 eV; BE = 532'0.i 0.2 eV) by irradiating the samples with MgKa radiation. Measurement of Cell Dimensions and Intensity Data Preliminary investigations were conducted by using the precession technique with MoKa radiation. The highest Laue Class symmetry found was Pi} thus the space group is either 21 or PE. Three- dimensional intensity data were collected by use of Picker Four-circle diffractometer controlled by a PDP-8/I computer. Monochromatic radiation (MoKal = 0.70926A) was obtained by initial diffraction of MoKa radiation by a graphite crystal with the (002) plane in diffracting orientation. Other parameters have been previously discussed (Chapters 2 and 3). The crystal (0.4 X 0.25 X 0.18 mm) was mounted with the gf axis approximately coincident with the ¢ axis of the diffractometer. Lattice parameters (Table 18) were determined from 12 hand centered reflections at room temperature (22 : 1°). The intensities were measured by using the m-scan technique with w being scanned over 0.8° at a rate of 0.5°/min. Data for the hemisphere (h, +k, +3) were collected between 2.8 and 45° for 26. Backgrounds were measured at each end of the scan for a total of 10 seconds each. The monitor reflections [(111), (014) and (213)] were measured every 100 reflections and showed a mean variation 0f :12, thus suggesting that no decomposition nor movement had taken Table 18. Crystal Data Molecular Formula Molecular Weight Crystal Habit Crystal Color Crystal Size, mm -1 U9 cm Space Group Systematic Absences .9) >0 0‘ >0 c, A a, deg. 8, deg. Y. deg- 9, A3 Dexp’ g/mla Dcalc’ g/ml E HSc(C7H502)4 530.43 Triclinic Pale Yellow 0.4 X 0.25 X 0.18 4.40 Pi, Triclinic None 10.020.) [ll.647]b ll.515(3) [12.004] 12.004(4) [10.022] 72.74(l) [95.42] 84.58(1) [116.32] 65.04(1) [102.25] 1198.6(1) 1.47 1.47 2 aMeasured by flotation in chloroform. b Lattice parameters in brackets correspond to the conventional orientation. 60 place during data collection. A total of 3144 independent reflections were measured of which 2539 were found to be above background using the criterion I/o(I) :_3.0. Solution and Refinement of the Structure Scattering factor amplitudes for the neutral Sc, 0 and C were obtained from Cromer and Weber (33), and those for H from Stewart, g£_al,, (34). The normal Lorentz and polarization corrections were made to the raw data, but no corrections were made for absorption nor anomalous dispersion. After data correction, the unscaled [F] values were used to calculate a three—dimensional Patterson map. Due to possible vector overlap, it was not obvious where the scandium atom was situated. For this reason, direct methods (symbolic phase addition) were used for at least a partial solution of the structure. A set of 275 [E] values was calculated by using the program FAME (66). After assignment of seven symbols, the programs MAGIC, LINK and SYMPL (66) failed to yield a non—trivial solution. At this time, MULTAN (67) became available. By using the IEI's calculated by FAME, a 22 listing was determined from which an initial set of three origin reflections and three symbols were determined by hand for a total of 29 reflections. The program then calculated eight sets of phases. The third set of phases (Table 19) yielded an E-map which gave atomic positions for the scandium and six oxygen atoms. A reinspection of the Patterson map verified peaks corresponding to the scandium and oxygens and also revealed coordinates for the other two oxygen atoms. Calculations of structure factor amplitudes and subsequent 61 Starting Phases from MULTAN Table 19. Reflection h k 2 1 —1 2 2 —1 3 6 -2 —1 l 3 12 2 -3 -4 4 3 —1 IE! 2.68 3.37 2.48 2.56 a'+ is an origin reflection Symbola Phase (deg.) 180 62 least~squares refinement were based upon the cell being centrosymmetric, i.e., the space group is :1. This conclusion arose from the statistics obtained from FAME (Table 20). A structure factor calculation based on the determined atomic coordinates and isotropic thermal parameters were used to obtain a Fourier map from which some of the carbon positions were located. It was necessary to repeat this process twice more before all non—hydrogen atom positions were defined. By using these initial coordinates and isotropic thermal parameters of 2.5A2 for all atoms, four cycles of full matrix least— squares refinement of coordinates and thermal parameters lowered R1 from 0.467 to 0.096. At this time a difference Fourier map was calculated and the positions of all the ring hydrogen atoms were determined. To avoid errors in coordinates, they were placed 0.97A (68) from their bonded carbon atoms along the line of the appropriate angle bisector. The "acid" hydrogen was placed between 01(1) and 01(3). It was thought that the hydrogen might be bridging between these oxygen atoms because there were short contacts between them (”2.5A) and the Sc—O bond distances were longer than those for the other oxygen atoms. Further refinement of all coordinates and isotropic thermal parameters continued to a converged R = 0.064. At this time the acid hydrogen had moved “1A. A difference Fourier map at final convergence (Figure 10) further suggested the new hydrogen position between 01(1) and 01(4) across the center of inversion (see the hydrogen bonding discussion). Further refinement of coordinates and anisotropic thermal parameters for all non—hydrogen atoms and isotropic thermal parameters for the hydrogen atoms was conducted in two parts. There appeared to be two mutually perpendicular 63 Table 20. Statistical Analysis of [E] Values from FAME (<[E|2> rescaled to 1.00) lheoretical Values Calculated Values Centric Acentric Average Magnitude of E's 0.798 0.886 0.806 <|E2 - 1|> 0.968 0.736 0.950 Percentage of [El '.s>l 32.00 37.00 32.49 Percentage of |E|'s>2 5.00 1.80 3.99 Percentage of [El's>3 0.33 0.01 0.26 64 .NIH .sla .xIH as mm was NN means; 5" engine cu commas“ mum wmumcwpuooo oflaoflw $3.0 n m um qma mo moauoom .AUao as w Qvao macaw cowbno cooBuwp pwumooa .N .aowuflmoa A5 cmmouvbi MGHBocw Qua Howuoom wonmuowwflm .os masses Figure 10 . 66 quasi—mirror planes in the molecule, and atoms in each quasi—mirror plane were refined separately (due to core storage limitations of the computer for full-matrix refinement). To check on correlations between parameters, the tropolone ligands were refined in pairs not contained in a quasi—mirror plane for one cycle and did not correlate significantly. Consequently, the final cycles were not overlapped and the final reliability factor was 0.027 while RW = 0.029 (w = 1). The final refinement obviously justified the choice of Pi as the space group. The final observed and calculated structure factor amplitudes are listed in Table 21. No parameter shift in the final two cycles was greater than 0.08 of its estimated standard deviation. A final difference Fourier map showed no peaks greater than 0.2e_A3. Description of the Structure and Discussion Hydrogentetrakis(tropolonato)scandium(III) crystallizes from aqueous methanol solution in the space group 2:. The final atomic coordinates and thermal parameters for scandium, oxygen and carbon atoms are listed in Table 22. Those for the hydrogen atoms are listed in Table 23. The compound exists as a hydrogen-bonded dimer located about the center of inversion of the cell. A stereoscopic view of one half of the dimer, excluding ring hydrogen atoms, is shown in Figure 11. The shortest non—bonding separations are listed in Table 24. These are classified as intramolecular (within the dimer) or intermolecular (between dimers). For the intramolecular separations, the shortest 0...O distances (2.49A) involves the acid hydrogen (3) of the hydrogen bond, with the shortest separation not involving \l 0 1 X l.\ S e d U t .1 1.. m. r 0 t C a F e r u t C u r t S d m e S b 0 1m a d e t a 1 U C l a C l a n .1 F Table 21. a .. u— an n "It! um y... u. L run in. Table 21 . (Continued) . . .I-III.‘ .. I In...“ I_. In...“ . I. rmvnl I I IIIIIIIII .. I mun. I. I I . II-IIIII - I "III":- - I. man-u - I "In I... ~ I "I "I . I I _ . . ‘- I " . _ 2 . I. I I I .. . I . . . .I I I l I . .. I. . . . . I I - .I I .4 - — I I I I I l - I I - I I II I. In - I I 9 ‘ 5- I l - I I - I r I a II - — I e I - - - I -— I I 7‘ . . _ . . - . I I r I I . . . . - . I I I .. II . . I - I I I. . I . . I. I . _ . . . . I I - I I . I I - . I I _ I - l I - | \ - ' " 3 .I o - - I I I \ I I - ‘ 0 I I I . . . . - I I I 4 - - - I I ‘9 . . . . . I I. >4 . . I I . I I - I a I II I - - - I I - u- I u a . I I I I . . . I I III . I I . . . - . I I . I I I I I I I: . . I I . I I I I u . .. .. I I. I I . - I I . I I I. _ - - - I I h I u . . . _ . . . I . I I u . . . _ I I . . I I I ll . I I I I I - I I . I I III I . . I I l I I - I I I I v: - — I I - I I - - I I a M . I I . . - . . I I II -_ I . . . I II n - . . I I . I I I I I I n . _ , _ _ . . I .I I I I v v I I M - - I I - - I l I I i . . . . . . . I I I I . . I . . . — - I . I I 4 - I l - I I ' e S . . I . .. I I . I I I I . . . I I I - . . I I I . . I I I I . - . I II . . _ . , . I I I I I I — . I I I I . I I I I I I . _ . I . I I - -. .. . — II . I I I I. Table 223. Atom Sc 01(1)b 02(1) c1(1) c2(1) c3(1) 04(1) C5(1) ca(1) C7(1) 01(2) 02(2) c1(2) C2(2) C3(2) 04(2) 05(2) 06(2) C7(2) 01(3) 02(3) Cl(3) C2(3) Final Atomic Coordinates 0.32858(5)a O O O O O O O O O O O —O. —0. —0. -0. O O O O O O x/a .307l(2) .l9l4(2) .216l(2) .l906(3) .1016(3) .0139(3) .0069(3) .0538(3) .1531(3) .1389(2) .1841(2) .0427(3) 0784(3) 1904(4) 2163(4) 1371(4) .0131(3) .0710(3) .4299(2) .4703(2) .4947(3) .5352(3) 69 O O O O O O O O O O O O O O O O O O O O O O y/b .62313(4) .5869(1) .8086(2) .6843(2) .6597(3) .7458(3) .8797(3) .9623(3) .9336(3) .8112(2) .5736(2) .7091(2) .6087(2) .5739(3) .5942(3) .6548(3) .7146(3) .7301(3) .6852(2) .4098(2) .5152(2) .3265(2) .1895(3) z/c 0.66803(4) 0.4923(1) 0.5363(1) 0.4092(2) 0.3092(2) 0.2102(2) 0.1844(2) 0.2544(2) 0.3621(2) 0.4377(2) 0.6861(2) 0.7953(1) 0.7610(2) 0.7743(3) 0.8507(3) 0.9391(3) 0.9675(3) 0.9180(2) 0.8271(2) 0.6853(1) 0.8256(1) 0.7810(2) 0.7997(3) 70 Table 22 a . (Continued) Atom x/a y/b z/c 03(3) 0.6035(4) 0.0821(3) 0.8953(3) 04(3) 0.6520(4) 0.0792(3) 1.0004(3) 05(3) 0.6461(3) 0.1860(3) 1.0337(3) C6(3) 0.5909(3) 0.3195(3) 0.9744(2) c7(3) 0.5200(3) 0.3876(2) 0.8636(2) 01(4) 0.5342(2) 0.6160(2) 0.5693(1) 02(4) 0.4013(2) 0.7742(2) 0.6878(1) 01(2) 0.5890(3) 0.7043(2) 0.5625(2) 02(4) 0.7036(3) 0.7073(3) 0.4912(2) 03(4) 0.7779(3) 0.7900(3) 0.4729(3) 04(4) 0.7586(4) 0.8874(3) 0.5230(3) 05(4) 0.6591(3) 0.9276(3) 0.6078(3) 06(4) 0.5521(3) 0.8875(3) 0.6556(3) C7(4) 0.5103(3) 0.7906(2) 0.6371(2) aNumbers in parentheses refer to the estimated standard deviations (esd) of the last decimal place. bThe numbering system given refers first to the atom of a given ring, then to the ring, g5}, 01(1) is the first oxygen on ring 1. 01 is bonded to Cl and 02 is bonded to C7. 71 Aqumoo.OI Amvwaoo.OI Aflvmmoo.OI Amvwmoo.OI ANVnHOO.OI ANvHNoo.OI Amvmooo.OI ANVmHoo.OI Aqumoo.OI Amvwmoo.OI Amvhaoo.OI flavomoo.OI Aavmfioo.OI Amvomfioo.OI AnvNHoo.o Aquooo.o ANvoNoo.o Amvmmoo.o Amvwooo.o Awesooo.o AmVHHoo.OI Aqumoo.OI ANVHHOO.OI ANvoooo.o Amvmooo.o Amvoooo.OI Afivmooo.OI Aqvaoooo.o mam Amvwcoo.OI Amvmwoo.OI vawsoo.OI Amvmooo.oa ANvflmoo.OI Amvnmoo.OI Amvmaoo.OI Amvamoo.OI “mesmoo.OI Amvmmoo.OI ANVqNOO.OI ANvaoo.OI ANvaoo.OI Amvmflmoo.OI Nam Amewoo.o Amvmmoo.o ANvoooo.o Amvwmoo.o Amvnqoo.o Amvomoo.o Amvoooo.o fiNvaoo.o ANmeoo.o Amvmmoo.o Amvmqoo.o Amvaoo.o Aavoqoo.o Aqvommoo.o mmm m “6voHHO.o AmVoooo.o Amvomoo.o ANVQHHo.o Anvsmoo.o Amvomoo.o AmVoooo.o Aqvwmoo.o Amvsmoo.o Amvmooo.o Amvwmoo.o Amvmmoo.o Amvomoo.o Amvomqoo.o NNm Amvmaao.o A6V6woo.o Amvflmoo.o Anvqfifio.o Amenoo.o Aqvwoao.o AqvaHo.o AQVmHHo.o AquHHo.o Amvmwoo.o Amvsmoo.o AmVOHHo.o ANvaoo.o onquoo.o Ham muwuoamumm HmEuwse camouuomfln< Hmafim AmvNo ANVHo vamo vaao havno AHVoo Afivmo Aavqo Avao AHvNo AHvHo Aavmo AHvHo 6m Ecu< .LNN mHQmH 72 Aavmmoo.OI “Neomoo.OI ANVaNoo.OI Amvmfioo.OI Amvoaoo.o AmVnooo.OI vawfioo.OI Amvflmoo.OI havomoo.OI Aavqfioo.OI ANVOHOO.OI Auvmmoo.OI Amvamoo.OI Anvnfloo.ou Amvomoo.OI mmm AHVmHoo.o Amvmooo.o Awesooo.on Amvaoo.OI Amvmmoo.OI Amvmmoo.OI AmeNoo.OI ANVnooo.o Amvmaoo.OI ANVwHoo.OI Amvaooo.o Amvofioo.o Aqumoo.o Amemoo.o AmVsaoo.o mam Amvomoo.OI AmeNoo.OI AmeHoo.OI AmvhNoo.OI Amvoqoo.OI Amquoo.OI Amvoqoo.OI Amvomoo.OI Amvamoo.OI Amvnmoo.oI Amvmaoo.OI Ameqoo.OI Aqvmqoo.OI Asemqoo.OI Aqvfinoo.OI mam AHVoooo.o ANvoqoo.o Auvqmoo.o Amvomoo.o Amvmsoo.o AMVOCHo.o Amvmooc.o Amvmmoo.o ANvHooo.o Aavoqoo.o Amvfiqoo.o Amvnmoo.o Amvmooo.o Amvoooo.o Aqvmoao.o mmm Amvnmoo.o Amenoo.o Aweomoo.o Asemoao.o Aqvmwoo.o Ameooo.o Amvhooo.o AmvNooo.o Amvoooo.o ANmeoo.o AmVemoo.o AmVHOHo.o Aquwao.o Aqvnmao.o Aqvamfio.o NNm ANanoo.o AmVnooo.o Aqvoaoo.o Aqvmmoo.o Amvomao.o Amvwoao.o Amquao.o Amvhsoo.o AmvaHo.o AmeNHo.o Aqvmnoo.o Aqvmmfio.o Amvomao.o Amvwmoo.o Amvooao.o Ham Avmsdfiusoov Aqvflo Amvno Amvoo Ameo Aquo Ameo Amvuu Amvao Amv~o Amvao Asto Amvoo ”memo Aquo Auvmo sou< .QNN oHQNB 73 Hfiaxmmmm + azmamm + Mfimamm + Nammm + NMNNm + NfiaavaHaxw "Show m£u «0 who muwuoamumm Hafiuonu 0:90 Amvomoo.OI Amemoo.OI AmVHmoo.OI Aqvqnoo.OI Aquooo.0I vamqoo.OI ANVwHoo.OI Aavmmoo.OI Amvmooo.o Amvomoo.o Amvmsoo.o Aqvmmoo.o Amemoo.o Amszoo.o Amvmooo.OI vamuoo.o mam ANv6NOO.OI Amvmmoo.OI Aqvomoo.OI Aqvmoao.OI Aqvwnoo.OI Amvnsoo.OI Amvnuoo.OI Aquqoo.OI Nam Amvssoo.o Amvswoo.o AqvaHo.o Aqvwmfio.o Amvmoao.o Anvmhoo.o ANvmsoo.o Amvmooo.o mmm AmVnmoo.o Amvmwoo.o A4VOHHo.o Aqvmmfio.o AsvaHo.o Amvnwoo.o Amvomoo.o Amvmsoo.o NNm Amvosoo.o Aqvhaao.o Amvmqfio.o Amvmsao.o Aqufifio.o Aqvfiaoo.o Amvmhoo.o Aqumoo.o Ham Acmnafiuaoov .LNN Aqvno Aqvoo Aqvmo Aqvqo Asvmo Aqvmo Asvau Aqvmo 8% magma 74 Table 23. Final Hydrogen Atom Coordinates and Isotropic Thermal Parameters Atom x/a y/b z/c B, R2 Ha 0.595(4) 0.536(4) 0.542(3) 7 30(3) 112(1)b 0.240(3) 0.570(2) 0.309(2) 2.41(52) H3(1) 0.101(3) 0.704(3) 0.149(2) 3.63(60) 04(1) —0.040(3) 0.922(3) 0.110(2) 3.78(62) H5(1) ~0.068(3) 1.051(3) 0.225(2) 3.22(59) H6(1) 0.027(3) 1.009(3) 0.391(2) 2.82(55) H2(2) -0.081(3) 0.533(3) 0.715(3) 4.87(74) H3(2) —0.264(4) 0.564(3) 0.838(3) 5.38(78) H4(2) —0.303(4) 0.661(3) 0.979(3) 5.49(79) H5(2) -o.162(4) 0.748(3) 1.034(3) 6.19(88) H6(2) 0.030(3) 0.774(3) 0.954(2) 3.83(64) H2(3) 0.507(3) 0.170(3) 0.732(2) 3.85(62) H3(3) 0.613(3) -0.004(3) 0.981(3) 5.99(82) H4(3) 0.687(3) -0.006(3) 1.061(3) 5.46(77) H5(3) 0.677(3) 0.169(3) 1.108(2) 3.39(61) H6(3) 0.601(3) 0.374(3) 1.018(2) 3.96(65) H2(4) 0.737(3) 0.645(2) 0.449(2) 2.43(52) H3(4) 0.852(3) 0.771(3) 0.419(2) 3.88(65) H4(4) 0.817(3) 0.934(3) 0.499(3) 5.07(75) 75 Table 23. (Continued) Atom x/a y/b z/c B, 3.2 H5(4) 0.665(3) 1.000(3) 0.634(2) 4.49(67) H6(4) 0.494(3) 0.927(3) 0.713(3) 4.42(69) aHydrogen attached to 01(4) bFirst number refers to the carbon atom to which the hydrogen is attached, the second to the ring. 76 . R15) HSL‘W GLUME TRY Figure 11. Stereoscopic View of half of the HSc(trop)4 dimer, 25% probability envelopes, ring hydrogens omitted for clarity. Figure 12. Interpenetrating trapezoids and hydrogen bonding in HSc(trop)4. 77 Table 24. Shortest Non-bonding Separations Intramoleculara Type H — H Hl(1) — H 01(3) - H 01(3) - 01(4) C2(1) - 01(4) Cl(l) — Cl(3) 01(1) - 01(4)b Intermolecular Type Distance 3 H2(l) - H5(2) 2.31 H1(1) — 111(2) 2.43 Hl(3) - H4(4) 2.48 aAtoms are from opposite halves of the dimer. b Hydrogen H is located between these atoms. Distance 2 2.73 2.24 2.62 3.13 Atoms related by (X, y, 2‘1) (-X, 1_Y9 l-Z) (X, Y‘1 9 Z) 78 H being 3.13; (01(3)—01(4)). (The shortest O...O distance within one HSc(trop)4 unit is the ligand "bite" distance, the average value being 2.505;). For separations involving other atoms, no unusually short distances are found. The closest intermolecular contacts involve the tropolonate ring hydrogen atoms, the shortest distance being 2.13A, (H2(l)-H5(2)). Coordination Polyhedron Each of the scandium(III) ions in the dimer is surrounded by four bidentate ligands. Interatomic distances and angles within the coordination polyhedron are listed in Table 25. The average Sc—O distance calculated from bond lengths involving oxygen atoms of ligands 2 and 3, which are not part of the hydrogen bonding system is 2.18; from which an ionic radius for eight—coordinated scandium(III) of approximately 0.78; is estimated. This value is similar to that obtained from the Sc~0 distance in ScPO4 (22), Na5[Sc(C0 ].2H20 (21), 3)4 Sc2(C0 .6H20 (24). For a coordination number of eight, there 3)3 are two important idealized coordination polyhedra, the dodecahedron and the square antiprism, although the polyhedron found in a structure is often distorted from one of these types (69, 70). The intermediate polyhedron between these two extremes is the bicapped trigonal prism (42, 71). If the method of Lippard and Russ (70) is employed to distinguish between the two extremes, several parameters must be defined. The determination of the best trapezoidal planes for a dodecahedron and the calculation of the angle of intersection between these planes is necessary. Table 26 and Figure 12 identify the planes T1 and T2. The angle between them, is 89.5°. For an a , T1T2 79 idealized dodecahedron, this value should be 90° whereas for a square antiprism this value should be 77.4°. The values 9A and SB are close to those expected for dodecahedral stereochemistry. A/‘\A 0,. Table 25 lists the O...0 separations for the polyhedron edges and Figure 13 shows the polyhedron with edges 3, b, m, and g_(as defined by Hoard and Silverton (69) for a dodecahedron. Table 27 lists the normalized values for these edges (the length of the edge divided by the average Sc—O distance). The bidentate ligands span equivalent E edges of the polyhedron and thus the dodecahedron has approximate D2d symmetry. This mode of chelation is quite common for bidentate ligands (72, 73), although in Sc2(C2 2H20 the three oxalato 04)3. ligands span two m_edges and one §_edge of the dodecahedron (62) and in Cs[Y(hfac)4], hfac_ = hexafluoroacetonate ion, the diketonate ligands span g_edges resulting in D2 symmetry (74). In an idealized D2d dodecahedron, there may be a difference in the bond lengths between the A and B type Sc~0 bonds as defined by Hoard and Silverton (69). 0 Table 25. Interatomic Distances (A) and Angles 80 Coordination Polyhedrona Sc - 01(1) Sc — 02(1) Sc - 01(2) Sc - 02(2) Sc — 01(3) Sc - 02(3) Sc - 01(4) Sc — 02(4) Average 01(1) - 02(1) 01(2) — 02(2) 01(3) — 02(3) 01(4) — 02(4) Average 01(1) - Sc — 02(1) Sc — 01(1) - 01(1) Sc — 02(1) - 07(1) 01(2) ~ Sc - 02(2) Sc - 01(2) — Cl(2) N N N N N N N N N N N N N N Distances .314(2) .209(2) .180(2) .l64(2) .l78(2) .183(2) .260(2) .228(2) .215(5)b .516(2)c .496(2)c .507(2)C .500(2)C .505(5)b Angles 67.53(6) 119.9(1) 123.6(1) 70.50(7) 120.2(2) 01(1) — 01(2) 01(1) - 01(3) 01(1) - 01(4) 02(1) - 01(2) 02(1) — 01(4) 02(1) — 02(2) 02(1) — 02(4) 01(2) - 01(3) 01(2) - 02(3) 02(2) - 02(3) 02(2) - 02(4) 01(3) - 01(4) 02(3) — 01(4) 02(3) - 02(4) H - 01(4) H...Ol(l) 01(3) - Sc — 02(3) Sc — 01(3) — 01(3) Sc - 02(3) - 07(3) 01(4) - Sc - 02(4) Sc — 01(4) - Cl(4) (deg) within the N .747(3) N .518(3) N .715(3) N .995(3) U.) .l92(3) N .983(3) N .736(3) N .726(4) LA) .577(4) N .767(4) N .768(4) N .920(4) la.) .059(4) N .691(4) 1.00(4) I—l ..49(4) 69.84(7) 120.2(1) 120.3(1) 67.69(6) 121.8(1) 81 Table 25. (Continued) Sc - 02(2) - C7(2) 120.9(1) Sc - 02(4) - C7(4) 122.9(1) 01(1) - H — 01(4) 175.9(33) H — Ol(l) - Sc 137.5(21) H — 01(4) — Sc 119.0(2) aErrors referred to last significant digit are in parentheses. bErrors for averages are computed using the method of small sample statistics: See W. Blaedel and V. Meloche, "Elementary Quantitative Analysis", Row, Peterson and Co., Evanston, 111., 1957, p. 557. CLigand "bite". 82 Table 26. Least Squares Planesa Trapezoids Plane a b C d Atom Dev. from Plane 2 T1 0.8337 0.3754 -0.4050 —3.4044 01(1)b 0.154(2) 02(1)b —0.109(2) 01(3)b —0.190(2) 02(3)b 0.125(2) Sc -0.l33(l) T2 -0.5291 0.3826 —0.7574 2.9261 01(2)b 0.141(2) 02(2)b —0.182(2) 01(4)b -0.105(2) 02(4)b 0.175(2) Sc -0.118(1) Tropolonato Rings6 3 b c d Ring 1 0.7828 0.4542 —o.4253 —4.0678 Ring 2 —0.5389 0.5289 —0.6557 0.4985 Ring 3 0.9040 0.1692 -0.3926 -1.7281 Ring 4 -0.6187 0.2501 -0.7448 4.7944 Deviations from Planes A Atom Ring 1 Ring 2 Ring 3 Ring 4 01 0.174(2) -0.089(2) -0 012(2) —0.077(2) 02 —0.026(2) 0.026(2) -0.o44(2) 0.110(2) Table 26. Atom C1b 02b 03b 04b 05b 06b C7b Sc aGeneral equation for planes: (Continued) Ring 1 0.002(3) 0.000(3) 0.000(3) 0.001(3) -0.006(3) 0.008(3) —0.005(3) -0.227(1) 83 Ring 2 -0.024(3) 0.015(4) 0.023(5) -0.018(5) -0.018(4) —0.015(4) 0.009(3) -0.l94(1) O O O 0. O O O O Ring 3 .004(2) .011(3) .005(3) 016(3) .002(3) .018(3) .010(3) .428(1) ax + by + cz + d = 0. Ring 4 -0.022(3) —0.010(3) 0.024(4) 0.023(4) —0.o34(4) —0.014(3) 0.035(3) 0.153(1) O X, y, and z are coordinates (A) in an orthogonal system defined by_h X_g, 2, b Atoms defining planes. * E 3 respectively. 84 .q Amouuvumm CH cow. AHHHVnBflvcmom 9.3 How Gouvmnhflom COHumcflwHooo .MH mhnwwm ’ w. / ‘ 85 .nH ounwwb 86 Table 27. Shape Parametersa for HSc(trop)4 Parameterb HSc(trop)4 Dodecahedron a 1.19 1.17 m° 1.13 1.17 b 1.45 1.49 g 1.26 1.24 0A, deg. 36.4 35.2 GB, deg. 74.7 73.5 0T T d 89.5 90.0 162 dT 0.005 0.0 l dT 8 0.007 0.0 2 aSee references 17 and 18 for definition of shape parameters. bPolyhedron edges normalized by dividing by the average Sc-O bond distance (2.215;). Cm is the normalized ligand "bite" distance. d§T T2 is the angle between the best trapezoidal planes T1 and T2. edT and dT are mean displacements of ligand atoms from best trapezoidal planes T and T2. 1 87 However, for HSc(trop)4, no significant differences are apparent. Another method used to describe eight-coordinate polyhedra was first proposed by Porai—Koshits and Aslanov (71) and expanded by Muetterties and Guggenberger (50). This again is based upon dihedral angles of the edges in the polyhedron as in six—coordination. The non-planarity of the trapezoids distorts the polyhedron from D2d symmetry through the bicapped trigonal prism of C2v symmetry until the limit of a D4d square antiprism is reached. The pertinent structural parameters are listed in Table 28. The dihedral angles, 5', are defined as the angle made by the normals of the planes adjacent to the edges defined by double lines in Figure 14. From this comparison, HSc(trop)4 does not belong with the dodecahedral class nor with the bicapped trigonal prism class, but is intermediate. It is preferred, however, to consider the coordination polyhedron to be distorted D2d dodecahedron as the C2V bicapped trigonal prism should be more regular in shape than a dodecahedron. Hydrogen Bonding and Ligands In the crystal, the two polyhedra are held together about the center of inversion by two almost linear hydrogen atoms attached to the 01(4) oxygen atom (O—H bond length is 1.00(3)A) are hydrogen bonded to the 01(1) oxygen atoms across the center. The hydrogen bonded 0...H distance is 1.49(3)A and the 0-H...O angle is 175.9°. The atoms, Sc, 01(1) and H and the related atoms across the center form a chair-shaped arrangement with 01(4), H and 01(1) atoms almost planar. The scandium atoms are located such that they make an angle of 132° with this plane. ' w 88 Table 28. Dihedral Angles for HSc(_trop)4 and Ideal Polyhedra (deg.) Complex 6' Anglesa 9'sb Ideal Dodecahedron 29.5, 29.5 0.0 29.5, 29.5 Ideal Bicapped 0.0, 21.8 14.1 Trigonal Prism 48.2, 48.2 Ideal Square 0.0, 0.0 24.5 Antiprism 52.4, 52.4 HSc(trop)4 13.2, 28.7 9.7 42.4, 42.9 aFor definition of 6, see text. b9 is defined as the dihedral angle between the dotted and dashed triangles for the trapezoidal atoms, BAAB. "w ll" . > O .QH Ghawflrfl. .mocHH manner up woumofivnfi madman anuwmsfiv man mdfinfimow mmmwo coupon H m \ e.~m .e.~m N.we .N.we m.m~ .m.m~ . . n e n . . 8.8 .o o o.o . .m.a~ .o.o .s .m.m~ n mm W 9 fl 8 I .\//V __\ \ /% .. .1. “mm 90 During the refinement of the strucutre, the hydrogen was placed initially between 01(1) and 01(3) of one polyhedron, because of the short 0...0 separation between these atoms (2.518;). However, after refinement, the hydrogen was located in the position described above, with H...01(3) across the center equal to 2.62(4)A. The position of 01(3) is probably influenced by the hydrogen, but the hydrogen bond is not bifurcated as in tropolone itself (64). Any effect that the hydrogen produces on ring 3 is of much smaller magnitude than that produced in rings 1 and 4. Further information concerning the hydrogen bonding is obtained by considering the nature of the ligands coordinated to the scandium(III) ion. Table 29 records the interatomic distances and angles for these ligands, The C7H5 rings are nearly planar with only small deviations of the carbon atoms from the planes defined by C1—C7. Table 26 lists the deviations and equations of the planes. However, 01 and 02 for the four rings show a complex pattern of distortions from the ring planes. Rings 2 and 3 show much smaller deviations than do rings 1 and 4, with a total out of plane distance being 0.20; for ring 1, 0.11A for ring 2, 0.03; for ring 3 and 0.19; for ring 4. The twisting of rings 1 and 4 may be related to the nature of these ligands. The bond lengths and angles within the four rings all show systematic variations (see Table 29). Rings 2 and 3 conform to patterns observed in Sc(trop)3 and Al(trop)3 (42). These complexes each contain coordinated tropolonato ligands with the C-C bond lengths decreasing as one proceeds from the carbon atoms attached to the oxygens to the C4 ring atom. Rings 1 and 4 do not exhibit this 0 Table 29 . Interatomic Distances (A) and Angles (deg.) for the Atoms Ol—Cl 02-C7 Cl-C7 C1-C2 C2—C3 C3-C4 C4—CS C5-C6 C6—C7 C2—H2 C3—H3 C4—H4 C5—H5 C6-H6 Ol—Cl-C2 Ol-Cl-C7 C7-C1—C2 Cl—CZ-CB Cl-CZ-HZ H2—CZ—C3 C2—C3-C4 91 Tropolonato Ligandsa Ring 1 1.315(3) 1.267(3) 1.459(3) 1.381(3) 1.395(4) 1.368(4) 1.389(4) 1.370(4) 1.424(4) 0.94(2) 0.99(3) 0.96(3) 0.92(3) 0.96(3) 120.2(2) 112.1(2) 127.7(2) 130.5(2) 114.5(14) 115.1(14) 129.9(3) Ring 2 Distances 1.275(3) 1.276(3) 1.467(4) 1.413(4) 1.366(4) 1.384(5) 1.369(5) 1.383(4) 1.400(4) 0.97(3) 0.97(3) 0.94(3) 0.95(3) 0.98(3) Angles 120.2(2) 114.3(2) 125.7(2) 130.9(3) 111.5(18) 117.5(18) 129.4(3) Ring 3 1.274(3) 1.281(3) 1.468(3) 1.403(4) 1.381(4) 1.380(4) 1.379(4) 1.373(4) 1.407(4) 1.00(3) 0.97(3) 0.98(3) O.9l(3) 0.97(3) 120.0(2) 114.1(2) 125.9(2) 130.5(3) 112.3(15) 117.2(15) 129.7(3) Ring 4 1.324(3) 1.257(3) 1.465(3) 1.373(3) 1.398(4) 1.360(4) 1.394(4) 1.355(4) 1.425(4) 0.93(2) 0.93(3) 0.93(3) 1.00(3) 0.95(3) 120.6(2) 111.0(2) 128.5(2) 129.5(3) 115.4(15) 115.1(15) 129.5(3) Table 29. Atoms C2-C3—H3 H3-C3-C4 c3-c4-cs C3-C4-H4 H4—C4—CS C4-C5-C6 C4—C5-H5 H5-C5—C6 C5—C6-C7 C5-C6—H6 H6—C6—C7 C6-C7—C1 C6—C7-02 02—C7-C1 aErrors referred to last significant digit are in parentheses. (Continued) Ring 1 115.5(15) 114.6(15) 126.7(13) 117.6(16) 115.7(16) 130.0(3) 116.8(16) 113.1(16) 131.3(3) 114.5(15) 114.1(15) 123.9(2) 119.9(2) 116.2(2) Ring 2 Angles 114.2(19) 116.4(18) 127.5(3) 115.0(20) 117.3(19) 130.1(3) 119.3(20) 110.4(20) 130.4(3) 117.5(16) 112.0(16) 125.8(2) 120.4(2) 113.8(2) Ring 3 115.4(19) 114.7(19) 127.4(3) 117.2(18) 115.2(18) 130.1(3) 117.0(17) 112.8(17) 130.2(3) 113.5(16) 116.2(16) 126.2(2) 120.5(2) 113.3(2) Ring 4 112.1(17) 118.4(17) 127.7(3) 118.0(19) 114.2(19) 129.7(3) 114.9(16) 115.3(16) 130.9(3) 117.6(17) 111.5(17) 123.9(2) 119.8(2) 116.3(2) 93 pattern. The bonding pattern in ring 4 is quite similar to that found in tropolone (64) where a more localized positioning of double bonds than in tropolonato ions is expected. The placement of the hydrogen atom on 01(4) is consistent with an increased Cl-Ol bond length and the corresponding decrease in C7—02 bond length (1.324 gs, 1.2572, respectively). Ring 4 is thus more of a "tropolone" ligand than a "tropolonato" ligand. Ring 1 also is affected by the hydrogen atom placement since the C—C bond length pattern is similar to that observed in ring 4, and C—0 distances for ring 1 are unequal at 1.315 and 1.267;. Thus, the hydrogen influences electronic distributions with rings 1 and 4, such that they appear similar to tropolone molecules. However, the difference Fourier map (Figure 10) does not indicate any "averaging" of the hydrogen position. The hydrogen bonding in HSc(trop)4 also affects the Sc—O bond lengths (2.314, 2.209, 2.260 and 2.2284). The differences are considerably greater than the errors expected in the observed values. However, although the hydrogen bond has weakened the bond between the scandium and the oxygens of ligands l and 4, these ligands are still considered to be coordinated in a bidentate manner. The carbon-hydrogen distances in the ligands average to 0.95;, very close to the optimal C—H distances as discussed by Churchill (68). These distances are shorter than the true C—H distances (expected to be ~1.10:4) because the scattering factor curve for hydrogen devised by Stewart, g£_a1. (34), assumes a spherical shape for the hydrogen atom instead of some polarized shape. The average ligand "bite" distance (01...02, Table 25), 2.505;. falls within the narrow range observed for other compounds containing chelating tropolonato 94 ligands (42). The average Ol—Sc—02 angle is 69.9, smaller than 73.8° found in Sc(trop)3. Thus, the change in effective ionic radius of the scandium(III) ion on increasing coordination number 6 to 8 results in a closing of the O—Sc-O angle rather than an increase in the ligand bite distance. The thermal amplitudes of the ligand groups may be interpreted in terms of rigid body motions. The root mean square (rms) deviations of the observed gij from that for a rigid body corrected for the total number of degrees of freedom [EU 2/(n ~ s)]1/2 (U = 8 /22a *a * ’ —ij —ij ij —i—i’ * * °_1 31 and gj are the two reciprocal lattices in A , Agij — yij(found) — O Hij(rigid body)), of the four ligands range from 0.0016 to 0.0019A2 and 0(gij) ” 0.0020A2 (75). The analysis of the librational motion of the ligands, although the motions are independent, indicates each has an angular movement of approximately equal magnitude. CHAPTER 6 DISCUSSION OF EIGHT COORDINATION AND FURTHER WORK Hydrogentetrakis(tropolonato)scandium(III) has been shown to exist as a hydrogen bonded dimer with the coordination environment intermediate between a D2d dodecahedron and a C4V bicapped trigonal— prism. Thus, the prediction of Muetterties and Wright (14) that the tropolone system may be used to form complexes of high coordina— tion numbers has been confirmed for the scandium(III) ion. Blight and Kepert (76) have discussed the effect of bidentate ligands on the stereochemistry of eight—coordination. For the M(bidentate) case, these authors conclude that as the normalized ligand bite distance increases, there is a change in preferred stereochemistry from the D2d dodecahedron to the D2 square antiprism. For HSc(trop)4 the normalized ligand bite is 1.13 (0"'Obite/M—Oavg)’ which is in the region where an intermediate stereochemistry may arise. Presumably, the steric requirements of the tropolonate ligands are an important factor in the determination of the stereochemistry adopted. It is interesting to speculate at this point, the stereochemistry of the species, Cs(trop)4, where hydrogen bonding should not be of concern. It would also be educational to speculate about the stereochemistry + + that would result for the system M[Sc(hfac)4], M = Na+, K , Rb , or Cs+, reported by Gurevich, g£_§1. (ll, 77). If the ligand bite 95 96 distance obtained by Bennet, 33 31. (74), in the complex Cs[Y(hfac)4].2H20 is assumed (2.77;) and an eight-coordinate Sc—O distance of 2.20;, a normalized ligand bite of approximately 1.25 is calculated. Blight and Kepert (76) suggested that there are three stereochemistries possible for values of this magnitude, 315., the D2 square antiprism, the D4 square antiprism, and the D2 dodecahedron. Values for a few complexes are listed in Table 30 together with the stereochemistry adopted for the complex. There is little difference between these stereochemistries and it is impossible to predict which one will be preferred. It should be anticipated, however, that the stereochemistry for the [Sc(hfac)4]- ion are presumed to be something other than the distorted D polyhedron found for HSc(trop)4 2d/C4V The determination of the crystal and molecular structures of CsSc(tr0p)4 and Cs[Sc(hfac)4] would verify the comments just presented. Attempts have been made to determine these structures without success. For various reasons, CsSc(trop)4 crystals are twinned and are poor X—ray scatterers. Attempts to recrystallize this compound from solvent mixtures has only resulted in the isolation of Sc(trop)3 or poor quality CsSc(trop)4 crystals. The problem of recrystallization has been noted by Hoard (79), also. The structure determination of Cs[Sc(hfac)4] has been attempted with difficulties arising. First, the crystals are not well formed, with a broad mosaic spread usually being observed. Second, a reasonable crystal was found, but the space group determination was initially incorrect. This was discovered after a data set had been collected based upon assignment to an orthorhombic space group, which later was found to not satisfy the diffraction pattern. Further study revealed the space group to be 97 Table 30. Normalized Ligand Bite Distances for Several Eight-Coordinate Complexes Complex N.L.B. HSc(trop)4 1.13 Sc2(C204)3.6H20 1.18 Cs[Y(hfac)4] 1.19 MISc(hfac)4] 1.26 I‘Tb(dpm)4 1.29 Stereochemistry D2d’Dodecahedron D2d'Dodecahedron D 2 Dodecahedron D 2 Sq. Antiprism D 4 Sq. Antiprism D Dodecahedron 2d D2 Sq. Antiprism Reference Chapt. 5 62 74 78 98 consistent with the monoclinic space group,.P21/c. The lattice parameters are listed in Table 31. The crystal was lost, unfortunately, before a complete data set could be collected. (Monoclinic space groups require a quadrant of data to be collected.) Attempts to recrystallize more of the material have met with difficulty. The only reasonable solvent mixture the author has found is ethanol/water. However, hexafluoroacetylacetone dissociates if the concentration of water exceeds 40%(V/V). Table 31. Crystal Data Molecular Formula Molecular Weight Crystal Habit Crystal Size Crystal Color 0, cm"1 Space Group Systematic Absences 99 Cs[Sc(C5H02F6)4] 1006.07 Plate Undetermined Clear 15.44 le/n, monoclinic E+£=23+1 050, _E = Zn_+ 1 8.195(2) 20.871(6) 18.408(5) 90.15 3148.5 2.013 2.052 4 8Measured by flotation in chloroform/bromoform. APPENDICES APPENDIX I ESCA Spectrum in the O ls Region for Htrop 100 HfrOp 0 1s KE 716.4 ev BE 532.0 " 2.7oV R I 715 T ' 7 719 Figure Al. ESCA spectrum in the 0 1s region for Htrop. APPEND IX I I ESCA Spectrum in the O ls Region for Sc(trop)3 101 Sc (trop) 3 0 1s KE 716.4 eV BE 532.0 7 v 715 719 Figure A2. ESCA spectrum in the 0 15 region for Sc(trop)3. APPENDIX III ESCA Spectrum in the O ls Region for HSc(trop)4 102 HS 9 C( l'op)4 We? 01s KE 716.4ev BE 532.0 Figure A3. ESCA spectrum in the O ls region for HSc(trop)4. y APPENDIX IV ESCA Spectrum in the O ls Region for Cs[Sc(trop)4] 103 Cs[$c(trop)4] KE 0 Is 716.4 eV BE 532.0 1W Figure A4 ESCA spectrum in the 0 ls region for Cs[Sc(trop)4]. APPENDIX V Analytical Data for the Compounds Studied Table A1. Compound Sc(trop)3 HSc(trop)4 Sc(acac)3 104 Analytical Data for the Compounds Studied Theoretical Found ZSC ZC 2H ZSc ZC ZH 11.01 61.77 3.70 10.69 60.51 3.99 8.49 63.40 3.99 8.71 62.88 4.18 Mass spectral data indicate the parent ion peak at 342 M/e. Theoretical value is 342 M/e. REFERENCES REFERENCES In Text 1. L. F. Nilson, Chem. Ber., 12, 554(1879); Comptes Rend., 88, 645 (1879). 2. D. I. Mendeleev, J. Russ. Phys. Chem., 1(1879). 3. R. C. Vickery, "The Chemistry of Yttrium and Scandium," International Series of Monographs on Inorganic Chemistgy, Vol. II, Oxford: Pergammon Press, 1960, and references therein. 4. L. Pauling, "The Nature of the Chemical Bond," 3rd. ed., Ithaca: Cornell University Press, 1960. 5. T. Moeller, D. F. Martin, L. C. Thompson, R. Ferrus, G. R. Feistel and W. J. Randall, Chem. Rev., 62, 1(1965). 6. F. A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry: A Comprehensive Text," 3rd. ed., New York: Interscience, 1972. 7. R. W. Stotz and G. A. Melson, Inorg. Chem., 11, 1720(1972). 8. E. Hansson, Acta Chem. Scand., 21, 2827(1973). 9. L. R. Melby, N. J. Ross, E. Abramson and J. C. Catis, J. Amer. Chem. Soc., 86; 5117(1964). 10. H. Bauer, J. Blanc and D. L. Ross, J. Amer. Chem. Soc., 86, 5125(1964). 11. M. Z. Gurevich, B. D. Stepin and V. V. Zelentsov, Zh. Neorg. Khim., 15, 890(1970). 12. M. Z. Gurevich and T. M. Sas, Zh. Neorg. Khim., 16, 93(1971). 13. L. I. Kononenko, I. G. Orlova and N. A. Poluekov, Ukr. Khim. Zh., 32’ 627(1966). 105 14' 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 106 E. L. Muetterties and C. M. Wright, J. Amer. Chem. Soc., 8%, 4706 (1965). C. Horovitz, K. Gschneidner, Jr., C. A. Melson, D. H. Youngblood and H. H. Schock, "Scandium: Occurence, Chemistry, Physics, Metallurgy, Biology and Technology," London: Academic Press, to be published, 1974. G. A. Melson and R. W. Stotz, Coord. Chem. Rev., 7 133(1971). “1’ K. H. Jack, Acta ngstallogr., TR, 780(1957). W. Klemm and E. Krose, Z. Anorg. Chem., 253, 218(1947). K. Schubert and A. Seitz, Z. Anorg. Chem., 256, 226(1948). H. E. Swanson, R. K. Fugat and G. Ugrinic, National Bureau of Standards Report, Project 0907—41—4449, Report Number 2290, 1952. T. A. Zhdanova, A. A. Voronkov, L. N. Komissarova and Yu A. Pyatenko, Sov. Phys.—Dok1., 16) 70(1971). R. C. L. Mooney, Acta CEystallogr., 9, 677(1956). W. O. Milligan and L. W. Vernon, J. Phys. Chem., 56; 145(1952). E. Hansson, Acta Chem. Scand., 26, 1337(1972). H. Schwarz, Z. Anorg. All. Chem., 323] 44(1963). M. K. Guseinova, A. S. Antsyshkina and M. A. Porai—Koshits, Zh. Strukt. Khim., 92 1040(1968). E. C. Lingafelter and R. L. Braun, J. Amer. Chem. Soc., 88, 2951 (1966). B. Morosin and H. Montgomery, Acta ngstallogr., Sect. B, 25} 1354 (1966). G. T. Morgan and H. W. Moss, J. Chem. Soc., 189(1914). V. I. Ivanov and O. M. Petrukin, Sov. Phys.—Crystallogr., 15, 492 (1970). R. B. Roof, Acta ngstallogr., 9, 781(1956). J. Iball and C. H. Morgan, Acta ngstallogr., 22; 239(1967). D. T. Cromer and J. T. Waber, Acta Crystallogr., 18] 104(1965). R. F. Stewart, E. R. Davidson and W. T. Simpson, J. Chem. Phys., 42} 3172(1965). 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. Chem. , g, 1770(1968). 107 E. W. Hughes, J Amer. Chem. Soc., 6%, 1737(1941). F. A. Cotton and J. S. Wood, Inorg. Chem., 3, 245(1964). B. Morosin, Acta Crystallogr., 22, 315(1967). M. J. Bennett, F. A. Cotton, P. Legzdins and S. J. Lippard, Inorg. K. Dymock and G. J. Palenik, Acta Crystallogr. , Sect. B, 30, 1364 (1974). B. Morosin, Acta Crystallogr., 19, 131(1965). Al(trop) crystalizes in the space group CZ/c with §_= 10.855, b_= 10.214, 3 = 17.710 A and B = 112.71° (42). E. L. Muetterties and L. J. Guggenberger, J. Amer. Chem. Soc., 94, 8046(1972). L. J. Guggenberger, personal communication. M. Renninger, Z. Physik, 106,, 141(1937). =l/o2 (F), 0(F)9 = O. 05F for F >4Fmin and 0(F) = 0.20Fmin for min’ Fmin F<4F M90 see reference 35. T. A. Hamor and D. J. Watkin, Chem. Commun., 440(1969). R. A. D. Wentworth, Coord. Chem. Rev., 9; 171(1972—1973). D. L. Kepert, Inorg. Chem., 11, 1561(1972). E. I. Steifel and G. F. Brown, Inorg. Chem., 11) 434(1972). E. L. Muetterties and L. J. Guggenberger, J. Amer. Chem. Soc., 96, 1748(1974). B. Morosin and H. Montgomery, Acta Crystallogr., Sect. B, 25) 1354 (1969). J. P. Fackler, Jr. and A. Avdeef, Inorg. Chem., 13, 1864(1974). ‘ A. Avdeef, J. A. Costamagna and J. P. Fackler, Jr., Inorg. Chem., 1}} 1854(1974). R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect. B, 25, 925(1969). R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect. B, %2’ 1046(1970). 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 73. 74. 75. 108 K. D. Smith and J. L. Atwood, Chem. Commun., 593(1972). J. L. Atwood and K. D. Smith, J. Amer. Chem. Soc., 9;; 1488(1973). J. L. Atwood and K. D. Smith, J. Chem. Soc., Dalton Trans., 921 (1974). S. S. Eaton, G. R. Eaton, R. H. Holm and E. L. Muetterties, J. Amer. Chem. Soc., 2;, 1116(1973). T. J. Cardwell and R. J. Magee, Anal. Chim. Acta, 36, 180(1966) and references therin. L. Pokras, M. Kilpatrick and P. M. Bernays, J. Amer. Chem. Soc., 75, 1254(1953). E. Hansson, Acta Chem. Scand., 26) 1337(1972). The X—ray photo electron spectra were provided by DR. D. M. Hercules of the University of Georgia. H. Shimanouchi and Y. Sasada, Acta Crzstallogr., Sect. B, 22] 81(1973). A. D. Hamer, D. G. Tisley and R. A. walton, J. Inorg. Nucl. Chem., 36, 1771(1974). R. B. K. Dewar and A. L. Stone, University of Chicago. The version of these programs is dated 1964. G. Germain, P. Main and M. M. Wbolfson, Acta Crzstallogr., Sect. B, 26) 274(1970). M. R. Churchill, Inorg. Chem., 1%} 1213(1973). J. L. Hoard and J. V. Silverton, Inorg. Chem., %, 235(1963). S. J. Lippard and B. J. Russ, Inorg. Chem., 7, 1686(1968). M. A. Porai-Koshits and L. A. Aslanov, Zh. Strukt. Khim., 13, 266 (1972). E. L. Muetterties and C. M. Wright, Quart. Rev., Chem. Soc., glv 109(1967). S. J. Lippard, Prog. Inorg. Chem., 8, 109(1967). M. J. Bennett, F. A. Cotton, P. Legzdins and S. J. Lippard, Inorg. Chem., 7, 1770(1968). V. Schomaker and K. N. Trueblood, Acta Crzstallogr., Sect. B, 24, 63(1968); program TLS modified for Mass Eeighted Displacement Coordinates by M. A. Neuman, American Crystallographic Society Meeting, Ames, Iowa, 1971. 109 76. D. G. Blight and D. L. Kepert, Inorg. Chem., 11, 1556(1972). 77. M. Z. Gurevich, B. D. Stepin, L. N. Komissarova, N. E. Labedeva and T. M. Sas, Zh. Neorg. Khim., 16, 93(1971). 78. T. J. Pinnavaia, G. Podolsky and P. W. Codding, Chem. Commun., 242 (1973). 79. J. L. Hoard, personal communication. General "International Tables for X—Ray Crystallography," Birmingham: Kynoch Press, Vol. I, 1969, Vol. II, 1972 and Vol. III, 1968. F. R. Ahmed, S. R. Hall and C. P. Huber, "Crystallographic Computing," Copenhagen: Munksgaard, 1970. M. J. Burger, "Crystal Structure Analysis," New York: John Wiley and Sons, Inc., 1960. G. H. Stout and L. H. Jensen, "X-Ray Structure Determination," New York: Macmillan Company, 1968. Computer Programs W. R. Busing, K. Martin and H. Levy, "OR FLS", a Fortran Crystallographic Least— squares Program, ORNL—TM—305, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1962; modified at the University of Wisconsin. W. R. Busing, K. Martin and H. Levy, "OR FFE", a Fortran Crystallographic Function and Error Program, ORNL-TM—306, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1964. C. K. Johnson, "0R TEP", a Fortran Thermal-Ellipsoid Plot Program for Crystal Structure Illustrations, 2nd Revision, ORNL-3794, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1970. R. B. K. Dewar and A. L. Stone, "FAME", "MAGIC", "LINK" and "SYMPL", a dircet methods phase determination package, 1964. P. Main, M. Woolfson and G. Germain, "MULTAN", a direct methods phase determination program, 1971. M. A. Neuman, "CONNIE", a Fourier summation program. M. A. Neuman, "FOROl", a structure factor calculation program. M. A. Neuman, "PLANE", a best plane calculation program. 110 V. Schomaker and K. N. Trueblood, "TLS", a Rigid—Body thermal analysis program, see footnote 75. J. M. Stewart, T. A. Kundell and J. C. Baldwin, "FOURR", a Fourier summation program for all space groups, from X-RAY70, University of Maryland, 1970. \ ARIES J mi 20 N \ \ 082 NIVERSITY i r; \ E 3 \ AT 1L \ \ 3 \ \ MICHIGAN ST