@ 1974
KARL DEE SMITH
ALL RIGHTS RESERVED ORGANOSCANDIUM CHEMISTRY
by
KARL DEE SMITH
A DISSERTATION
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Chemistry in the Graduate School of The University of Alabama
UNIVERSITY, AL1\BAMA
1973
--- e····
ACKNOWLEDGMENTS
The author wishes to express his deep appreciation to:
Dr. D. F. Smith and Dr. B. W. Ponder for their understanding, encouragement, and guidance throughout the course of this research. Steve Seale for his many hours spent in setting up a workable computer library to permit the completion of this work to become a reality. Merle Watson for his many services rendered in the making of the special glass apparatus needed throughout the course of this work. The computer operators, Steve Watson, Mike Webb, Bob McGwier, Al Martin, and Bill Gammon for their cooperation in efficiently running the hundreds of computer programs needed for the completion of this work. Sam Hassel, G. M. Nichols, and Harold Moore for their many services rendered in the maintenance, stockroom, and electronics fields, respectively. The secretaries for their services rendered. Segail Friedman for typing the final manuscript of this work. His wife, Becky, .and to Angela and Christopher for their confidence, encouragement, understanding, and love shown in every way.
ii TABLE OF CONTENTS
Page
ACKNOWLEDGivIENTS • • ii
LIS'l' OF TABLES . . iv
LIST OF FIGURES • vi
Chapter
I. INTRODUCTION 1
II. EXPERIMENTAL METHODS 9
Inert Atmosphere Glove Box . 9 Reagents and Solvents . . • • • 9 Preparation of Compounds . . • • • . • • • . 11 Preparation of Samples . . • . . • • • . • • 18 Computer Programs . • . . • . . . • • . 19 Instrumentation ;, . • • • . • • • 20
III. RESULTS AND DISCUSSION 22
Dicyclopentadienylscandium Chloride Dimer ...... 22 Tricyclopentadienylscandium . • . . . . • . 41 Trichlorotris(tetrahydrofuran)scandium . 63 Bis (indenyl) magnesium . . . • • . . • • 84
IV. CONCLUSIONS ...... 112
REFERENCES ..•.• 114
I iii I
...... ___ LIST OF TABLES
Table Page
1. Elemental Analysis of Scc1 . • • • • . • • 14 3 2. Elemental Analysis of Mg(C H ) • • • . • . 16 9 72 3. Elemental Analysis of Sc(C H ) . • . . . • 18 5 53 a,b 4. Final Atomic Positional Parameters for 8. Best Weighted Least-Squares Planes for [( CS HS ) 2 S cC 1] 2 • • • • • • • • • • • • 4 0 . . . a, b , .9 Fina . 1 Atomic Positiona1 Parameters for Tricyclopentadienylscandium. . . • . 48 a b 4 lo. A·niso t ropic · Tempera-uret Fae tors , (x 10 ) for Tricyclopentadienylscandium. . . . • • . • . 49 11. Ob served and Calculated Structure Factor Amplitudes for Tricyclopentadienylscandium 50 0 12. Interatomic Distances (A) and Angles (deg) for Tricyclopentadienylscandium 55 iv } ; Table Page 13. Comparison of Metal-Cyclopentadienyl Carbon Bond Distances ... •. •• 57 14. Best Weighted Least-Squares Planes for Tricyclopentadienylscandium ..•.• 58 15. Comparison of Crystal Data for Sc(C H ) 5 5 3 and Sm(C H ) .•. .•••.••• 63 5 5 3 a 16. Final Atomic Positional Parameters for ScC1 (C H O) • . . . • • • • • . • • 69 3 4 8 3 . a, b 4 17. Anisotropi� Temperature Factors (x 10 ) for ScC1 (C H O) . • . . • • . • • . • • 70 3 4 8 3 18. Observed and Calculated Structure Factors for Trichlorotris(Tetrahydrofuran)scandium • . • 71 0 19. Interatomic Distances (A) and Angles (deg) for ScC1 (C H O) . • • . . • 79 3 4 8 3 20. Best Weighted Least-Squares Planes for ScC1 (C H o) . . . • • • • . • . 84 3 4 8 3 a, b 21. Final Atomic Positional Parameters for Diindenylmagnesium 91 a b 4 22. Anisotropic Temperature Factors , (x 10 ) for Diindenylmagnesium . . • . 93 23. Observed and Calculated Structure Factor Amplitudes for Bis(indenyl)magnesium 95 0 24. Interatomic Distances (A) for Angles (deg) for Diindenylmagnesium 104 25. Best Weighted Least-Squares Planes for Diindenylmagnesium .•..••. •. 110 V i LIST OF FIGURES Figure Page 1. Molecular structure of the dicyclopenta dienylscandium dimer which lies in a general position in the unit cell •. • 33 2. Molecular structure of the dicyclopenta dienylscandium dimer which lies on a center of symmetry in the unit cell .• 35 3. Structure and unit cell packing of_ tricyclopenta dienylscandium. The atoms are displayed as the 50% probability ellipsoids for thermal motion • . • • • • • . • • • • • 52 '4. Bond distances and angles within the cyclopentadienyl groups for Sc(C H ) • . • • • • • • •••••• 60 5 5 3 5. The coordination sphere of the scandium ion with the 50% probability envelopes of the anisotropic thermal ellipsoids • .• 73 6. Molecular view of trichlorotris(tetrahydro furan)scandium with the 40% probability envelopes of the anisotropic thermal ellipsoids ...... 75 7. Structure and unit cell packing of trichlorotris (tetrahydrofuran)scandium. The atoms are displayed as the 40% probability ellipsoids for thermal motion . • • • . . . • 77 8. View looking down the Cl-Sc-0 axis displaying the configuration of the THF rings • . • • • 82 vi Figure Page 9. Illustration of magnesium(l) and its associated indenyl rings . .•••. . . . . 97 10. View of magnesium(2) and its associated indenyl rings ...... 99 11. Structure and unit cell packing of bis(indenyl)magnesium ...... 102 12. Bond distances and angles within the indenyl groups for Mg(C H ) .•.•.•.••..• 107 9 7 2 vii CHAPTER I INTRODUCTION The element scandium has been known for over one hundred years, but its coordination chemistry has been little studied. The lack of attention has been due, in part, to the difficulty of obtaining a pure source of scandium, although both the metal and oxide are now commercially available in high purity. Scandium is the first member of the 3d transition 1 2 series and has a 3d 4s ground state electronic configura tion. The +III oxidation state is the only one known. It is in many respects quite similar to yttrium and the lanthanides (1) although the di stinctly smaller radius of the scandium(III) ion affords some noteworthy difference s in chemistry. Several coordination compounds of scandium have been synthesized recently (2), although few structural character izations of scandium complexes have been reported. At the time this work was initia.ted only the structural character ization of the scandium formate complex (3), Sc(HCOO} , had 3 1 - 2 been reported. In this compound, -the scandium(III) ions are six-coordinate in a polymeric framework with formate ions acting as bridging groups. X-ray structural character izations of dicyclopentadienylscandium chloride (4), tricyclopentadienylscandium (5), and trichlorotris(tetra hydrofuran)scandium have now been carried out. In addition, the X-ray structure of tris(acetylacetonato)scandium(III) (6) has been recently reported. Other organoscandium compounds which have been characterized by means other than X-ray methods are dicyclo pentadienylscandium acetate, (c H ) ScOCOCH ; dicyclopenta 5 5 2 3 dienylscan.dium acetylacetonate, (C H ) ScAcac; (allyl) 5 5 2 dicyclopentadienyscandium, (c H ) sc(CH CH=CH ); and 5 5 2 2 2 (dicyclopentadienyl)phenylethynylscandium, (C H ) ScC=CPh 5 5 2 (7,8). Molecular weight measurements and infrared studies showed dicyclopentadienylscandium acetate to be · dimeric with bridging acetate groups. Dicyclopentadienylscandium acetylacetonate is monomeric and infrared studies showed the acetylacetonate to be bidentate. It was indicated that (allyl)dicyclopentadienylscandium was monomeric and the spin decoupled PMR spectrum confirmed the symmetrical nature of the allyl group. It was suggested that 3 (dicyclopentadienyl) phenylethynylscandium is associated to some extent with probably bridging PhC=C groups (9). Stable organoscandium compounds characterized so far are those containing anions in which unsaturation is present and TI bonding may occur between the organic anion and the scandium(III) ion. Attempts to synthesize alkyl-scandium compounds have met with limited success. There has been no confirmation (7) of the reported synthesis of Sc(Et)3-Et2o (10). The scandium-ethyl species' instability may be due to the alkene elimination reaction which is a well known method t i I of decomposition of transition metal alkyls (11). Recently, Witt and Melson (12) reported the synthesis of organoscandium I f compounds containing the trimethylsilylmethyl anion. This I anion has been used to prevent alkene elimination reactions and enable compounds containing transition metal-carbon I bonds to be isolated (11, 13, 14, 15). They isolated the two From available infrared and mass spectral evidence it was concluded that both compounds contain covalent Sc-C bonds. With the lack of unsaturation in the anion these bonds should be purely sigma in type. They propose that the com pounds are polymeric with both terminal and bridging 4 trimethylsilylmethyl anions where the terminal Sc-C bonds are 2-electron, 2-center bonds whereas the bridging bonds are weaker 2-electron, 3-center bonds. The isolation of these compounds containing Sc-C sigma bonds suggests that the instability of the scandium ethyl species is due to a facile decomposition process, e.g., ethylene elimination (16) rather than an instability inherent with scandium-carbon bonds. The structural studies of many organoscandium com plexes should determine their potential catalytic applica bility. In view of the importance of other first row transi tion elements as catalysts in industrial processes such as hydrogenation, polymerization, oligomerization, etc., the field of organoscandium chemistry and also the synthesis of species containing scandium in oxidation states other than three should receive increasing attention. A review of the influence of ligands on the catalytic activity of a transi tion metal catalyst by Olive and Olive (17) stresses the need for a large number of systematic studies to be carried out so as to deepen the understanding of transition metal catalysis and to avoid misinterpretations. 5 Soluble transition metal complexes have become extremely important as catalysts for a wide range of reac tions over the past few decades. Probably the starting point of this development was the discovery by Roelen (17) in 1938 of the reaction of olefins with carbon monoxide and hydrogen to form aldehydes which takes place on a soluble cobalt carbonyl complex. Many reactions were subsequently dis covered: the oxidation of ethylene to acetaldehyde on a / palladium complex ( "Wacker Process") (18), the specific hydrogenation of double bonds on a series of transition metal compounds (19), hydroformylation on rhodium complexes (20), the polymerization (21) and oligomerization (22} of olefins on soluble Ziegler-Natta catalysts, and the cyclooligomerization of acetylene (23} and conjugated diolefins (24) on nickel. At first the species that effected the catalysis were mostly definite complexes such as Wilkinson's (2) RhH(CO} [(C H } P] in hydroformylation and Vaska's (25) 6 5 3 3 complex IrCl(CO) [(C H } P] as a hydrogenation catalyst. 6 5 3 2 However, two fields of chemistry that also developed rapidly at the same time led homogeneous catalysis in a new and extremely interesting direction. On the one hand, transition 6 metals attracted growing interest in preparative coordination chemistry causing synthesis of many new compounds, while on the other, important advances in theoretical inorganic chemistry (particularly ligand field theory) influenced the thinking of catalysis chemists. The net result was that more attention was devoted to the effects and the signifi- cance of ligands in the transition metal complex. The ligands of a complex that was recognized as a catalyst were systematically modified to bring about specific changes both in the rate of the catalyzed reaction and in the final product in an effort to understand which physical parameters such as s teric hindrance, -orbital energies, electron density on the metal, etc., are involved. In most known cases of homogeneous catalysis on transition metal complexes, the catalytic reaction takes place between a covalently sigma-bonded ligand R {alkyl group, hydrogen) and a substrate molecule (olefin, CO) . \ coordinated to the metal M, the substrate molecule being inserted between the metal and R by a four-center reaction (concerted reaction). This is shown schematically for an olefin in the form�la: 7 R \ / I C (Lx)M complex. The catalyst may be restored to its original state by hydrogenolysis or homolysis 1 or the same process may be repeated (polymerization). Therefore, parameters such as the stability of the M-R and M-olefin bonds, the transition metal itself, and the possibility of influencing these bonds through the other ligands L are of the utmost importance. Extensive structural work and catalytic applications have been carried out with titanium. Therefore, to present a starting framework for the catalytic possibilities of . organoscandium complexes, it is logical to compare the structural information obtained thus far with that of titanium, scandium's neighbor in the periodic table. A comparison of the metal-ligand bonding and ionic radii in titanium and scandium complexes should give some insight to the similarities and differences of these substances.· The purpose of this research was to investigate some organoscandium complexes in the solid state by X-ray dif- fraction. Since no structural characterizations of organoscandium complexes have been done, it was hoped that 8 this work would form a beginning in the systematic study of organoscandium compounds. X-ray crystallography should be a valuable tool in obtaining physical measurements and structural characterization of scandium complexes to deter mine the coordination environment for the scandium ion to afford a basis upon which the nature of the scandium-carbon bond could be studied and to resolve questions of stereo chemistry, mode of bonding and stability. Possession of such information should then aid the interpretation of other physical studies of these compounds and guide the synthetic chemist in this area. CHAPTER II EXPERIMENTAL METHODS Inert Atmosphere Glove Box All preparations, transfers, and crystal mounting procedures were carried out under a nitrogen atmosphere, since all the compounds under investigation were sensitive to water and air. The glove box used was purchased from Kewannee Scientific Equipment Corporation, Adrian, Michigan. The enclosure was the Model 2C380 with the Model 2Cl982 "Kempure" recirculating gas purification system using molecular sieve and manganese (II) oxide columns. The atmosphere was tested with trimethylaluminum before use; when the atmosphere was satisfactory there was no fuming of the compound. Reagents and Solvents Technical grade magnesium turnings and. indene were obtained from Eastman Kodak Company, Rochester, New York. The indene was freshly distilled just prior to use. 9 10 Reagent grade tetrahydrofuran, toluene, benzene, and ethyl bromide were obtained from J. T. Baker Chemical Company, Phillipsburg, New Jersey, and stored over sodium wire. Analytical reagent grade ethyl ether (anhydrous) obtained from Mallinchradt Chemical Works, St. Louis, Missouri, was used without further purification. Technical grade dicyclopentadiene, purchased from J. T. Baker Chemical Company, Phillipsburg, New Jersey, was boiled·to produce the monomer just prior to use. Anhydrous scandium oxide (99.9%) was obtained as a white powder from Research Organic/Inorganic Chemical Corpora tion, Sun Valley, California and from Alfa Inorganic Ventron Corporation, Beverly, Massachusetts • .A...�hydrous scandium trifluoride (99.9%) was purchased from Alfa Inorganic Ventron Corporation, Beverly, Massachu setts. Certified A.C.S. grade ammonium chloride was obtained from Fisher Scientific Company, Chemical Manufacturing ·Division, Fair Lawn, New Jersey. • .. 11 Anhydrous scandium trichloride (99.9%) was purchased from Research Organic/Inorganic Chemical Corporation, Sun Valley, California. Preparation of Compounds Dicyclopentadienylmagnesium Dicyclopentadienylmagnesium was prepared by the method of Barber (26): Mg+ 2c H �Mg 5 6 (C5H5}2 +H2 Commercial dicyclopentadiene (B.P. 170 ° C) was placed in a flask and boiled to produce the monomer (b.p. 42 ° C}. Cyclopentadiene thus produced was mixed with nitrogen and passed through a Pyrex tube 1.25 inches o.d. which was heated electrically to 600 ° C. Excess magnesium metal turnings were supported in the furnace tube by a circle of nichrome gauze at the tube constriction. The product fell from the furnace as a white solid and was collected in a three-necked flask. The unreacted cyclopentadiene was collected in a dry ice-ethanol trap. The apparatus was initially charged and flushed with dry nitrogen. No special pretreatment of the magnesium turnings was necessary. The dicyclopentadienylmagnesium was purified by sublimation in 12 vacuo after the unreacted cyclopentadiene contaminant had dimerized. Anhydrous Scandium(III) Chloride Anhydrous scandium(III) chloride was prepared by two different methods. The first preparation followed the method of Reed (27) in which scandium oxide was reacted with ammonium chloride according to the equation: Sc 0 + 6NH c1 ➔2scc1 + 3H 0 + 6NH 2 3 4 3 2 3 Scandium oxide (0.01 mole) was mixed thoroughly with a large excess (0.12 mole) of ammonium chloride. This mixture was placed in a Schlenk tube, flushed with dry nitrogen, and heated in a furnace at approximately 200 ° C for six to eight hours. A vacuum was then applied and the temperature of the mixture raised to 300/320 ° C and held at this point until all the ammonium chloride sublimed over leaving a silvery gray residue. This procedure was not very satisfactory, perhaps because the product was contaminated with a carbonate and hydrated oxide. The second method, far superior to the first·, was the method of Stotz and Melson (28) in which anhydrous scandium trichloride was prepared from an aqueous medium 13 with hydrolysis of the scandium(III) ion prevented by the 3- formation of the Scc1 ion. One gram of scandium oxide 6 was dissolved in 28 ml of hydrochloric acid (19%HC1) by refluxing for two to three hours. The solution was allowed to cool to room temperature and 9.0 ml of concentrated (29%) ammonium hydroxide solution added with stirring. A clear solution with pH 3 was obtained. The solution was trans- ferred to a beaker which was placed on a hot plate, and the water was removed by boiling until a moist solid was obtained. The solid was dried under vacuum over P o at room tempera 4 10 ture overnight and then transferred to a constricted Schlenk subl imation apparatus made of quartz. The remaining water was removed by heating under vacuum at 150 ° C for three hours. A coarse fritted disk was then inserted in the Schlenk tube covering the constriction. The temperature was increased to 300 ° C, maintained at this temperature for four hours, and then further raised to 500 ° C for an additional thirty minutes. The ammonium chloride sublimed onto the walls of the upper portion of the sublimator. Final heating at 850 ° C resulted in a sublimation of white crystals of scandium(III} chloride onto the walls of the lower portion of the sublimator. An alysis of a sample of the resultant 14 material done by Schwarzkopf Microanalytical Laboratory, Woodside, New York, gave the results shown in Table 1. TABLE l ELEMENTAL ANALYSIS OF ScC13 Analysis Calculated for Scc13 Found Scandium 29.7% 28.6% Chlorine 70.3% 67.4% Diindenylmagnesium Diindenylmagnesium was prepared by the following reactions: Magnesium turnings (5.0g, 0.21 mole) were covered with 100 ml of sodium-dried diethyl ether in a 250 ml three-necked flask. One neck of the flask was fitted with a condenser which was in turn connected to a mercury bubbler. Of the remaining two.entrances to the flask, one was attached via a stopcock to a high purity N cylinder, and one was fitted with a 2 seal�d tygon tube for �yringe injection of ethyl bromide. The vessel was then flushed with N2 and 15 ml of ethyl L 15 bromide (0.20 mole) was slowly added with stirring. The solution was refluxed for two hours, at which time the ethyl magnesium bromide Grignard reagent was of milky-white coloration. Then with rapid N flow, the stopcock was re 2 moved from the condenser and 14 ml of freshly distilled indene (0.19 mole) and 100 ml of toluene were added. The stopcock was replaced and the reaction temperature was elevated such that the toluene solution refluxed vigorously. All diethyl ether was driven off with a slow N flow rate. 2 After two hours, the N was closed off and the solution 2 allowed to reflux for eight more hours. Solvent was then removed, the residue dried under vacuum, and the flask taken into the dry-box. The substance was transferred to a Schlenk sublimation apparatus, removed from the dry-box and thermolyzed under vacuum at 190 ° C. The crude product was resublimed to free the white crystalline diindenyl magnesium from a yellow oil contaminant. Analysis of a sample gave the results shown in Table 2. The white crystalline solid had no clear melting point. Decomposition began at approximately 170 ° C, but sublimation was accomplish ed at 190 ° C under reduced pressure with some loss of material. It was soluble in ethers, and slightly so in 16 aromatic hydrocarbons. The substance rapidly decomposed with the slightest exposure to either H2o or o2 • TABLE 2 ELEMENTAL ANALYSIS OF Mg(C9H7 ) 2 Analysis Found Magnesium 9.6% 9.8% Carbon 84.9% 85.5% Hydrogen 5.5% 5.6% Dicyclopentadienylscandium Chloride Dicyclopentadienylscandium chloride was prepared by the method of Coutts and Wailes (8). A solution of dicyclopentadienylmagnesium {3.08g) in tetrahydrofuran {50 ml) was added slowly with ice cooling to sca~dium trichloride {3.03g) in THF {50 ml). After addition was com plete the solution was warmed to 50°C for one hour, at which stage it was pale yellow in color. Solvent was removed under reduced pressure and the residue was sublimed at a tempera- -3 ture of 220°C and 10 mm Hg giving large yellow-green crystals of dicyclopentadienylscandium chloride. 17 Tricyclopentadienylscandium Tricyclopentadienylscandium was prepared by the sealed tube reaction of dicyclopentadienylmagnesium with scandium trifluoride (29). Dicyclopentadienylmagnesium (0.0032 mole) was thoroughly mixed with scandium trifluoride (0.002 mole) and placed in a bomb tube in the glove box. After sealing under vacuum, the tube was placed in a beaker of beeswax at 220°C and rotated by use of a magnetic stirring bar in the beeswax. The tumbling action served to mix the. slurry of molten dicyclopentadienylmagnesium and solid scandium trifluoride during reaction. After a reaction time of three hours the product was transferred to a Schlenk sublimation apparatus in the dry-box. The tricyclopenta dienylscandium was freed of excess dicyclopentadienyl magnesium by heating under vacuum at 100-200°C and then sublimed as straw colored needle shaped crystals from the -4 reaction residue at 220°c at 10 mm ~g. Analysis of a sample gave the results shown in Table 3. . . Trichlorotris(tetrahydrofuran)scandium In a dry-box, scandium trichloride (0.0026 mole) was dissolved in tetrahydrofuran (25 ml)· in a three-necked flask. The solution was refluxed gently for three to four 18 TABLE 3 ELEMENTAL ANALYSIS OF Sc(C5H5) 3 Analysis Found Scandium 11.4% 12.5% Carbon 81.8% 80.3% Hy,drogen 6.8% 7.3% hours at which time the solution was red in color. Solvent was then partially removed and the flask taken into the dry box. The solution was transferred to bomb tubes. Slow evaporation of the solution allowed formation of orange, plate-like crystals of trichlorotris(.tetrahydrofuran) scandium. Preparation of Samples X-ray Diffraction Crystals of dicyclopentadienylscandium chloride, tricyclopentadienylsoandium, and diindenylmagnesium were grown by slow sublimation in a sealed, evacuated tube. Crystals were mounted in 0.2 or 0.3 mm thin-walled glass capillaries with the aid of a small amount of stopcock grease. The capillaries were sealed with beeswax and then 19 taken outside the dry-box and sealed with a mini-torch. The crystals were then examined under a polarizing microscope and one giving good extinctions was affixed to a goniometer head for X-ray study. Computer Programs An IBM 360/50 computer was used to perform most calculations, but a Univac 1108 Computer was used sometimes in the final stages of structure refinement. The initi~l plotting of structures was done using a Hewlett Packard Recorder drivenby a Varian Data-6201 Computer with final plotting done using a Calcomp Plotter driven by a XDS-Sigma 7 Computer. The programs ACAC (30) and later ORABS (31) were used to reduce the raw intensities to structure factors. The program FAME (32) was used to calculate normalized structure factors and output the Wilson plot and statistically analyze for a center of symmetry. Direct methods were applied with the program MULTAN (33) which determines phases derived from E-values of FAME. ·The full-matrix, least-squares refinement was per formed using the program ORFLS (34). Calculation of Fourier, 20 difference Fourier and Patterson function maps was carried out using the program ALFF (35). The program ORFFE (36) was used to calculate interatomic distances, bond angles, principal axes of thermal motion, and the standard errors of the functions. The program HYGEN (37) was used to generate positions of hydrogen atoms from molecular geometry. The calculations of bond distances and angles were routinely done using the program JAM (38). The program BEPLA.l (39) was used for best plane calculations. The crystal structure illustrations were obtained using the program ORTEP (40). Instrumentation X-ray Diffraction A Norelco X-ray generator made by Phillips Elec tronics Company, Mount Vernon, New York, was employed in all preliminary film work. A Buerger precession camera made by the Charles Supper Company, Natick, Massachusetts, was used in preliminary examination of all crystals studied. Some preliminary film data were collected with a non-integrating Weissenberg camera also made by the Charles Supper Company. 21 Three-dimensional single-crystal X-ray diffraction data were obtained on an ENRAF-NONIUS CAD-4 diffractometer purchased from the ENRAF-NONIUS Company, Delft, Holland. Ni-filtered copper radiation was used in data collection for dicyclopentadienylscandium chloride and tricyclopenta dienylscandium. For magnesium indenide and trichlorotris (tetrahydrofuran)scandium a graphite monochromator with the (002) plane in diffracting position was used to obtain monochromatic Cu Ka radiation. CHAPTER III RESULTS AND DISCUSSION Dicyclopentadienylscandium Chloride Dimer At present the organometallic chemistry of scandium is a relatively unexplored area. Tricyclopentadienylscandium (41), triphenyl- and tri(phenylethynyl)scandium (7), and dicyclopentadienylscandium chloride and derivatives have been prepared, but no structural data have been presented. The X-ray structure analysis of the dicyclopentadienyl scandium chloride dimer gives the first view of the stereo chemistry of an organoscandium complex and a study of the nature of the scandium-carbon bond. Yellow-green rod shaped crystals of dicyclopenta dienylscandium chloride were prepared by the method of Coutts and Wailes ( 8)., and diffraction-quality crystals were grown by slow sublimation. Preliminary unit cell parameters were determined by precession (Cu Ka) photographs. Systematic absences allow the space group to be P2 /c. The 1 22 23 lattice parameters as determined from a least-squares refinement of (sin0/A) 2 values for 12 reflections are: 0 a= 13.54(1) A 0 b = 16.00(1) A 0 c = 13.40(1) A V = 2896 i 3 (3 = 93.97(5) 0 -3 The calculated density is 1.44 g cm for Z = 6. Data were taken on an Enraf-Nonius CAD-4 diffractometer•with Ni-filtered copper radiation. The crystal, a rod of dimen- sions 0.17 x 0.17 x 0.42 mm, was aligned on the diffracto meter, such that no symmetry axis was coincident with the ~ axis of the diffractometer. The diffracted intensities were collected by the w-28 scan technique with a take-off angle of 1.5°. The scan rate was variable and was determined by a fast (20°/min) prescan. Calculated speeds based on the net intensity gathered in the prescan ranged from 6 to 1° min-1 • Background counts were collected for 25% of the total scan tiine at ·each end of the scan range. For each intensity the scan width was determined by the equation· scan range= A+ B tane 24 where A= 1.0° and B = 0.5°. Aperture settings were deter- mined in a like manner with A= 4 mm and B = 4 mm. The crystal-to-source and crystal-to-detector distances were 21.6 and 20.8 cm, respectively. The lower level and upper level discrimminators of the pulse height analyzer were set to obtain a 95% window centered on the Cu Ka peak. As a check on the stability of the diffractometer and the crystal, two reflections, the (111) and the (002), were measured at 30-min intervals during data collection. No significant variation in the references intensities was noticed. The standard deviations of the intensities, o , 1 were estimated from the formula 2 2 2 2 OI = {[cN+(Tc/2TB) (Bl+B2)]+ (0.03) [cN+(Tc/2TB) (Bl+B2)] }½ where CN is the counts collected during scan time Tc and B 1 and B are background intensities, each collected during 2 the background time TB. One independent quadrant of data was measured out to 20 = 110°. A total of 1680 reflections were judged to be observed on the criterion that I>o . 1 The intensities were corrected in the usual manner for Lorentz, polarization, and absorption (31) effects -1 (µ = 85.5 cm ). 25 Fourier calculations were made with the ALFF (35) program. The full-matrix, least squares refinement was carried out using the Busing and Levy program ORFLS (34). The function w(IF I-IF 1> 2 was minimized. No corrections 0 C · were made for extinction or anomalous dispersion. Neutral atom scattering factors were taken from the compilations of Ibers (42) for Sc, Cl, c, and H. Final bond distances, angles, and errors were computed with the aid of the Busing, Martin, and Levy ORFFE (36) program. The crystal structure illustration was obtained with the program ORTEP (40). Partial structure solution was accomplished by direct methods, and an electron density map phased on the scandium and chlorine atoms yielded the positions of the remaining nonhydrogen atoms. Several cycles of least-squares refine ment with isotropic thermal parameters for all atoms produced a reliability index of R = r(IF 1-IFcl)/(EIF I>= 0.13 . 0 0 Conversion to anisotropic temperature factors, the inclusion of hydrogen atoms in calculated positions, and additional cycles of refinernent·produced a final R = 0.072 and 26 Unit weights were used at all stages of refinement, and no 2 systematic variation of w(IF I-IF 1> vs. IF I or (sin0)/A 0 C 0 was observed. The largest parameter shifts in the final cycle of refinement were less than 0.10 of their estimated standard deviations. A final difference Fourier map showed no unaccounted electron density. Atomic and thermal para- meters are given in Tables 4 and 5, respectively. Observed and calculated structure factor amplitudes are listed in Table 6. In the unit cell there are six chlorine-bridged dimers, of which four lie in general positions and two reside on a center of symmetry. Although there are two crystallographi cally different molecules, they do not differ significantly in any respect and the configuration in each case is repre- sented by Figures 1 and 2. The cyclopentadienyl rings are bonded in a penta-hapto-fashion, with the scandium-carbon 0 bond length (Table 7) ranging from 2.39 to 2.49 A, and 0 averaging 2.46 A. This value is somewhat shorter than the 0 2.49 A standard found in Sc(C H ) (5), and could reflect 5 5 3 either the somewhat greater ability of the chlorine atom to remove electron density from the scandium atom, or the more crowded environment about the scandium atom in tricyclo pentadienylscandium. 27 TABLE 4 a,b · FINAL ATOMIC POSITIONAL PARAMETERS · FRO · [ Atom x/a y/b z/c Sc(l) 0.0520(1) 0.7352(1) 0 .34"88 (2) Sc(2) 0. 2511 (1) 0.8969(1) 0.4438(2) Sc(3) 0 .4134 (1) 0.4134(1) 0.4382(2) Cl(l) .0.2030(2) 0.8118(2) 0.2842(2) Cl(2) 0.0963(2) 0.8267(2) 0.5043(2) Cl (3) 0.4202(2) 0.5729(2) 0.4594(2) C(l) -0.0336(9) 0.8097(14) ·0.2066 (13) C(2) -0.0568(11) 0.8507(8) 0.2944(15) C (3) -0.1151(10) 0.7983(10) 0.3495(11) C(4) -0.1271(7) 0.7291(8) 0.2983(11) C(5) -0.0822(9) 0.7322(10) 0.2173(11) C(6) 0.0206(10) 0.5828(7) 0.3429(13) C (7) 0.0333(10) 0.6038(8) 0.4416(12) C (8) 0.1306(14) 0.6274(8) 0.461;3(13) C(9) 0.1750(9) 0 .6237 (8). 0.3735(16) C (10) 0.1076(15) 0.5969(9) 0.2980(11) C (1.1) 0.3347(9) 0.7838(9) 0.5415 (13) C(12) 0.3852(10) 0.7913(8) 0.4572(11) C(13) 0.4298(9) 0.8685(8) 0.4602(10) C (14) 0.4092(8) 0 .9076 (7) 0.5469(9) C(15) 0.3519(10) 0.8558(10) 0.5985(9) C(l6) 0.1363(11) 1.0009(10) 0.3770(22) C(17) 0.1630(16) 1.0269(9) 0.4700(19) C (18) 0.2560(15) 1.0480(8) 0.4834(13) C (19) 0.2919(10) 1.0389(7) 0.3899(13) C(20) 0.216(15) 1.0149 (8) 0. 3302 ( 11) C (21) 0.3736(11) 0.3424(9) 0.2750(10) C(22) 0.3728(10) 0.4270(11) 0.2576(9) C(23) 0.4709(13) 0.4534(9) 0.2754(10) C (24) 0.5252(10) 0.3849(10) 0.3055(10) C{25) 0.4669(14) 0.3190(9) 0.3074(12) C (26) 0.2865(22) 0 .3125 (15) 0.4771(17) C {27) 0.3537(13) 0.3076(11) 0.5526(16) C (28) 0.3554(11) 0.3784(15) 0.6016 (11) C{29) 0.2863(18) 0.4303(9) 0.5582(19) C(30) 0.2475(9) 0.3851(18) 0.4785(17) H(l) 0.0038 0.8420 0 .1621 28 TABLE 4--Continued Atom x/a y/b z/c H(2) -0.0447 . 0. 9059 0.3211 H(3) -0.1374 0.8101 0.4166 H ( 4) -0.1640 0.6805 0.3156 H(5) -0.0742 0.6910 0.1644 H(6) -0.0375 0.5608 0.3094 · H (7) -0.0190 0.5981 0.4903 H(8) 0.1641 0.6468 0.5227 H (9) 0.2459 0.6374 0.3661 H(lO) 0.1173 0.5906 0.2251 H (11) 0.2979 0.7342 0.5585 H (12) 0.3899 0.7466 0.4094 H (13) 0.4653 0.8929 0.4055 H(l4) 0.4306 0.9616 0.5675 H(lS) 0.3254 0.8647 0.6649 H(l6) 0.0734 0.9846 0.3550 H(l 7) 0.1132 1.0286 0.5258 H (18) 0.2949 1.0637 0.5476 H(l9) 0.3600 1.0506 0.3764 H(20) 0.2236 1.0067 -o.2548 H (21) 0.3142 0.3071 0.2706 H(22) 0.3210 0.4662 0.2329 H (23) 0.4998 0.5115 0.2713 H (24) 0.5997 0.3864 0.3148 H(25) 0.4934 0.2641 0. 32 86 H(26) 0.2599 0.2782 0.4171 H(27) 0.3938 0.2545 0.5581 H(28) 0.4019 0.3849 ·0.6608 H(29) · 0.2776 0 .4 86 8 0.5907 H(30) 0 .1942 0. 4182 0.4415 a Standard deviations in parentheses refer to last digit quoted. 0 2 b Isotropic thermal parameters set at 4.0 A for all hydrogen atoms. 29 TABLE 5 · . ab ANISOTROPIC TEMPERATURE FACTORS ' (x 104) FOR [cc 5a5 ) 2scc1] 2 Atom 13 11 13 22 13 33 13 12 13 13 13 23 Sc (1) 35 (1) 30 (1) 67 (2) 2 (1) -19 (1) -5 {1) Sc(2) 41(1) 30 (1) 58 (1) 2(1) -14 (1) -4 (1) Sc (3) 36 (1) 35 (1) 55 (2) -2 (1) -21 (1) 4 (1) Cl (1) 47(1) 49 (1) 51 (2) -1 (1) -5(2) · -9 (1) Cl (2) 46(2) 47 (1) 59(2) -3 (1) -3 (2) -6 (1) Cl (3) 35 (1) 37 (1) 69(2) 5 (1) -2 8 (1) 4 (1) C (1) 43(9) 176(18) 113(16) 18(10) -15(9) 104(13) C(2) 93(13) 35(7) 193(21) -4(7) -83(13) 18(10) C (3) 73(10) 72(9) 116 (15) 32 (8) -35(9) -40(9) C (4) 24(7) 75(9) 110(14) 4(6) -9(7) 7(9) C(5) 59(10) 96 (11) 8 4 ( 14) 26 (8) -13 ( 8) -9(9) C (6) 102 ( 11) 22(6) 164 ( 17) -9 (6) -71(11) -11 (7) C(7) 94(11) 55(8) 127 (14) -7(7) 12(10) 15(9) C (8) 159(17) 38(7) 123(16) 3 ( 8) -79 (12) 12 (8) C(9) 60(9) 52 ( 8) 221(23) 15 (7) -44(12) 30 (11) C(lO) 169(17) 56 ( 8) . 98 (14) 36(10) 11 (12) -8 ( 8) C (11) 60(10) 65(9) 159(18) 0 -6°(10) 49(10) C(12) 77(10) 40(7) 135(15) 35 (7) -54(9) -39 (8) C (13) 72(9) 67 (8) 66(12) 3(7) -16(8) -10 ( 7) C (14) 61 ( 8) 31(6) 84(12) -21 ( 6) -26 ( 7) 9 ( 6) C (15) 96(11) 89(10) 37(10) 26 (8) -17(8) -2 (8) C(l6) 66 ( 11) 43(9) 419(42) -5(9) -84(18) 51 (16) C (17) 141(19) 33 ( 8) 303 (33) 9 ( 10) 126(20) 14(12) 30 TABLE 5--Continued Atom C (18) 191(19) 33 ( 7) 107(16) 8(10) -70(14) -18(8) C (19) 89 (10) 29(6) 139(16) 11(6) 14(10) 13(8) C(20) 226 (21) 43(7) 83(13) -10{11) -100(14) 8 { 8) C {21) 106(2) 73(9) 6 3 ( 12) -24(8) -11(9) -26 { 8) C (22) 107(12) 108 (11) 37(11) 5 4 ( 10) -37(9) -1 {8) (2 3) 151 (15) 60 ( 8) 65(12) . -13 (9) 7 (11) -3 (7) C(24) 95 (12) 91 ( 11) 77 (13) 33(9) 5 (9) -36(9) C(25) 172(18) 52 (8) 105 (14) 38(10) -32 (10) -19 ( 8) C (26) 249 (32) 110(16) 148(25) -129(18) 39(19) -33(15) C (27) 123(16) 68 (11) 153(23) 51(11) 40(13) 51 (11) C (28) 76(12) 135 (16) 86 (14) -31 (11) -20 (10) 13 (13) C(29) 187 (23) 38(7) 211(27) -39(10) 139(19) -18(11) C ( 30) 3 8 ( 8) 166(19) 177(24) 4(12) -26(11) 98(17) a Standard deviations in parentheses refer to last digit quoted. ~ Anisotropic thermal parameters defined by 2 2 2 exp [- Cf\ 1h +s 22k +s 33 1 +2s12hk+2s13h1+2s23kl] 31 TABLE 6 OBSERVED AND CALCULATED STRUCTURE FACTORS FOR THE DICYCLOPENTADIENYLSCANDIUM CHLORIDE DIMER 32 1'),J .... -• 1l •1., ••.~ -1 1 17.t J6.5 1 O 1!,> 1>,2 S II ll,l !•,s ',. , • l Ji.8 U,S ...... 2 1 '"·' 2 on.• "•' ·• ,. ,,., ,s., -~• 1l1J "·"ta," 1i,S"·' •l 1 •>,7 0,lu., l l JC,J l0.1 l,; 11., "·' I : "•• ll,• • " '"·' 17,) -• IJ u.c 7,2 l 1 U.• U,l • o H,6 0,1 -l 015,,lHl,O •8 l >6,t "•' 51,2 ,,., , , n,, ,,.,, 9 11 .... 11,0 -• u '"·' 11,, ., 1 !1,1 se,2 5 ., 1l,2 "·" •5,1 ,i,, -l O '"•' ,o,s -1 "v., )!., 7 11 Jl,i '"·" •9 l H,l 11,I , C11J,ll1>,0 _, 1l ,, •• 1),1 0 1, •••• ·~- 1 -• 1 Si,! S<.J 1 10 17,1 .... l I <•.7 2',D > l2 "·· 1i,; 10., 10 •• • I.! .!l.l 20,'; , 1l 2',1 1',l 20,. "·' -0 ll .-.• •l,2 l~. l '"•• 6 l2 ll,l l!,9 ; ! "1•• H,• •! l 19,1 S,J JO,• l1,S •1 1l i,.'I ll,l 1 i,.1 t11Jl{!!:! •1 l 16,9 ll,9 l2 !6.1 52,1 S•,l -~ !l. "'·' ,, •• •& J 21,l 10.6 a,, ,.,, t l 14,1 79,0 ., "••·• n,. • 1 >),'; "·' 1 1J ;,,1 "·' i 1l lJ,9 l<,7 28.1 12,2 l iiUi H,1 27,6 j ,1 i,,, "1,) 1 • l0,7 11,0 -• • 10.s 1a.2 ·• 1l "·· '7,7 ! 1l 2•.J H.I -•2 ....• 26,5 , 11.•,s.o • 1J 11.1 "·' ,8 55,0 -l 1• 21,S lS,S 1 • n.• ,u.2 1S.9 16,S 1", I l,6 o 1' .S,O ll,1 H,1 11,• ·l l 11,e 11,i a• 16,1 72,1 •SU 1',.< l1,0 i,., .,,,,, s 1• u,, 11,0 18.J 21.0 .,.s "·· H,< •••' -: 1lU Hil 2•-· 10., l9,t n,, I l !l,8 !1,• "·'07,C 10,9"·' ~ ; :::} :;:: _,. • S ll,l 30.S ;,,o 22,S a< !O,l •••• 2',S ll,5 l1 l ·••• lv.1 -11 l 18 •• 22,• 1 6 .,,, 0,5 -1 • .,., 16.8 l • E,! ••• 10l,71Gl,l -l • 10,c 15.5 l • 12,• lt,6 • l 61,t !a,1 n,o lS.S •S l i,,] 27,S .,•• ,s,s .,., ,.. , •1 l 1',C Z•,l : : :::: :i1 ...39,J , .,.~16,S 11 • "·· "·" 22.c n.~ -1 , ,~., :ii.o ,s.e 21,, -• • 17,0 11.1 < ; H.! 7'.7 ·l s,s,,e1s,,• -• g "·' 11,6 11.1 n.2 ,,_. "·" -• • 10.1 18,8 2,.2 n.1 ,c,,1,a,., -2 10 >•·• n.1 l 10 11,1 lt,5 57,S 61,1 s W , •• ~ 22 •• 22,J 2,.0 _, I l 21.0 i1.• 1 s u,2 JS,5 -• ,o ]9,6 )6,5 -7 l ll,2 2S,O 0(,0 •••• o s Jl,1 "•• 7 l 20.0 •••• Q 11 2•,7 11,l l• !l~lJ 2,...... 1 11 22.:,, .~ •• 9 l 21,! il,i -1 11 23,8 22,5 -11 l :11.• 21,1 ,.,6 17,S 11 • ,,.1 .... -2 11 2s.c 15,1 20.1 2s,, '1 l 20,l 1J,6 19,~ 16,·l -• n 2s.• 22.6 -1• J 19.6 11,7 •••• H.a -1 " n., 11.0 r • ,i.2 ••·• ;1,1 "·' 1 11 2, •• 25,5 o.s n., -1 • 11,t H.l _, 12 ,,,, ,,,, 36,J H,l 1 O 37,l ll,9 02,7 .,., -~ ll 21.2 2•.2 -• • , •• 7 ... ~ ., ... , .. J 1l 2•.9 22., n.2 u.1 -J • 91,l It,• ,o.o ,~ ••• .; n n.• n,2 , •• , ,s.s l • 8~. 1 O~.S •S 1l U,6 111.,,11.1, , "'·' ' "·" ,._., 11.• -• • ,2.8 .,•• .. . OHOHL• ••••u•• , "'·' ,s,, ,.,, 1 • JS,( JS.l 0 1 20,2 .... -s,..... 02,s 01,1 l 1 H,2 oe,J •12 2,.1 26,1 5 • ,t.• lO,O H,6 n.a g • '"·" 21,2 -2 t :il,f 21,S •1l l!,8 22,8 6 • l7,• l7,l 01,2 ,a.• 11 • n,1 ,i.o -• t J6,6 ... , 0 27,8 H.6 •1 • 1s,o l1,S -· '2,2 11., "·' o 2 n.1 2••• 16,'l 1il.1 1 • ,,., ,,., ,, . ... , -9 • :il.o H.~ ., s •••• H,1 l0,7 ll,8 1 ! •5.( .... , ... ,,,. l 7 •••• "1,o -l 7 o,,l Jg,2 -2 , ,,,.r 1n., '"·" 11,7 2 ~ 10.1 "·' l 7 ,e,1 Fig. !.--Molecular structure of the dicyclopenta dienylscandium dimer which lies in a general position in the unit cell. 34 35 Fig. 2. --Moleculat··str"u_ct6r.$ .. O·f the dicyclopenta dienylscandium dimer which ·iies ··dn a center of symmetry in the unit cell. 37 TABLE 7 0 INTERATOMIC DISTANCES (A) AND ANGLES (DEG) FOR ~C5H5)2scc1]2 Bonded Sc(l)-Cl(l) 2.585(4) C(l)-C(2) 1.40(2) Sc(l)-Cl(2) 2.583(4) C(2)-C(3) 1.40(2) Sc(2)-Cl(l) 2.580(4) C ( 3) -C ( 4) 1.31(2) Sc(2)-Cl(2) 2.559(4) C ( 4) -C ( 5) 1.28(2) Sc (3)-Cl (3) 2.568(4) C(5)-C(l) 1.41(2) Sc ( 3) -Cl ( 4) 2.565(4) C{6)-C(7) 1.36(2} Sc(4)-Cl(3} 2.565(4) C (7)-C (8) 1.38(2) Sc(4)-C1(4) 2.569(4) C(8)-C(9) 1.36(2) Sc(l)-C(l) 2.47(1} C (9) -C (10) 1.38(2) Sc(l)-C{2) 2.44(1) C(10)-C{6) 1.38(2) Sc(l)-C{3) 2.48(1) C {11) -C (12) 1.37(2) Sc(l)-C(4) 2.47(1) C ( 12 ) -c ( 13 ) 1.37(2) Sc (1) -C (5) 2.44(1) C (13) -C (14) 1.36 (2) Sc ( 1) -C ( 6) . 2.48(1) C(14)-C(l5) 1.37(2) Sc (1) -C ( 7) 2.46(1) C(15)-C(ll) 1.39 {2) Sc ( 1) -C ( 8) 2.48(1) C(l6)-C(l7) 1.34(2) Sc(l)-C(19) 2.45(1) C(l7)-C(l8) 1.30(2) Sc(l)-C(l0) 2.45(1) C(l8)-C(l9) 1.38(2) Sc (2)-C(ll) 2.46 (1) C(l9)-C(20) 1.31(2) SC ( 2) -C ( 12) 2.48(1) C ( 2 0 ) -C ( 16 ) 1.31(2) Sc (2)-C(l3) 2.46(1) C(21)-C(22) 1.38(2) Sc ( 2 ) -C ( 14 ) 2.47(1) C (22)-C (23) 1.40 (2) Sc(2)-C(l5) 2.48(1) C(23)-C(24) 1. 36 ( 2) Sc(2)-C(l6) 2.41(1) C (24) -C (25) 1. 32 (2) Sc(2)-C(17) 2.44(1) C (25) -C (21) 1.36(2) $ C ( 2 ) -C ( 18 ) 2.48(1) C(26)-C(27) 1. 32 (2) Sc(2)-C(19) 2.46 (1) C ( 2 7 ) -c ( 2 8 ) 1.31(2) Sc(2)-C(20) 2.45 (1) C ( 2 8 ) -C ( 2 9 ) 1.35(2) Sc(3)-C(21) 2.49(1) C(29)-C(30) 1.37(2) Sc(3)-C(22) 2.45(1) C(30)-C{26) 1.38(2) Sc ( 3) -C ( 2 3) . 2.45(1) Sc (3) -C (24) 2.46(1) Sc ( 3) -C (25) 2.46(1) Sc(3)-C(26) 2.44(1) Sc(3)-C(27) 2.46(1) Sc (3)-C (28) 2.44(1) 38 TABLE 7--Continued Bonded Sc(3)-C(29) 2.45(1) Sc(3)-C(30) 2.39(1) 0 Nonbonded Distances (A) C (5) -c (6) 3.19(2) C (5) -C ( 7) 3.88(2} C (5) -C (10) 3.48(2) C(2)-C(16} 3.67(2) C(4)-C(10) 3.82(2} C(4)-C(7} 3.44(2) C(4)-C(6} 3.11(2) C(3)-C(7} 3.86(2) C(3}-C(6} 3.91(2) C (11) -C (8) 3.83(2) C(ll)-C(9} 3.95(2) C(l5)-C(l8} 3.64(2) C(14)-C(17) 3.92(2} C(14}-C(18} 3.14(2) C(14}-C(19} 3.30(2) C (13) -C (18} 3.74(2) C(13)-C(19) 3.40(2) C(8}-C(29} 3.96(3) C(21)-C(26} 3.06(2) C ( 21) -C ( 2 7 ) 3.79(2) · C (21} -C (30} 3.38(3} C(22)-C(26) 3.72(3) C (22)-C (30) 3.57(2} C{25)-C(26) 3.45(3) C(25)-C(27) 3.73(3) Bond Angles Sc{l}-Cl(l}-Sc(2) 97.6(1) C(2}-C(l)-C(S) 101.1(12) Sc(l)-S1(2)-Sc(2) 98.2(1) C(l)-C(2)-C(3) 109.3(12) Cl (1) -Sc (1) -Cl-(2) 81.8(1} C(2)-C(3)-C(4) 106.5(14) Cl(l)-Sc(2}-C1(2} 82.3(1} C(3)-C(4)-C(5) 111.3(14) C1(3)-Sc(4)-C1(4) 80.4(1) Sc(3)-Cl(4)-Sc(4) 99.6(1) Sc{3)-Cl(3)-Sc(4) 99_.6(1) Cl(3)-Sc(3)-C1(4) 80.4(1) C{l)-C(6)-C(7) 109.3(12} C(l)-C(5)-C(4) 111. 8 (14) C(6)-C(7)-C(8) 107.9(14} C(15)-C(ll)-C(12) 108.3(12) C{7)-C(8)-C(9) 107.2(13} C(ll)-C(12)-C(13) 107.3(12) C { 8) -C ( 9) -C ( 10} 109.8(14} C(12)-C(13)-C(14) 108.8(12) C{9)-C(10)-C(6} 105.8(14) C(13}-C(14)-C(15} 108.2(11) C{20)-C(16)-C(17) 102.3(13} C(l4)-C(15)-C(ll} 107.4(12) C{16)-C(17}-C(l8) 113.8(17) C(25)-C(21)-C(22) 108.7(13) C{17)-C(l8)-C(l9) 104.2(14) C(21)-C(22)-C(23) 105.8(11) C(18)-C(l9)-C(20) 106.0(14) C(22)-C(23)-C(24) 107.1(13) C(19)-C(20)-C(16) 113.2(17) C(23)-C(24)-C(25) 109.7(14) C(30)-C(26)-C(27) 108.1(16) C(24)-C(25)-C(21) 108.6 (14) C(26)-C(27)-C(28) 108.8(15) C(27)-C(28)-C(29) 109.1(15) C(28)-C(29}-C(30) 103.3(1.4) C(29)-C(30)-C(26) 110.7(15) 39 0 If one takes 0.68 A as the radius (4~6} of the scandium(III} ion, a scandium-carbon bond length of 2.46-2.49 0 A in dicyclopentadienylscandium chloride agrees very well with the value predicted on the basis of the two known organosamarium structures. The average samarium-carbon bond 0 0 distance is 2.78 A in (C H ) sm (43) and 2.75 A in 5 5 3 (c H ) sm (44); the generally accepted radius of the 9 7 3 0 samarium (III} ion is 0.96 A (45). 0 The scandium-chlorine distance of 2.575 A is quite 0 long compared to that found in Scc1 (c H o} (2.413 A) (46}. 3 4 8 3 However, the structure of the latter consists of discrete molecules in which each chlorine atom is bonded to only one scandium atom. The lengthening of a bond to a bridging halide ion is quite common: in [CH AlC1 ] where there are 3 2 2 both bridging and terminal chlorine atoms, the bond lengths 0 are 2.25 and 2.05 A, respectively (47). As is shown in Table 8, the scandium atom lies on the 0 average 2.18 A out of the plane of the cyclopentadienyl groups. Within rings themselves, the bond distance and angle are normal. The packing is typical of a molecular compound: 0 the shortest nonbonded contacts are 3.1 A between carbon 40 atoms on cyclopentadienyl groups bo~ded to the same scandium atom, and the closest inter-molecular carbon-carbon approach 0 is 3.82 A. TABLE 8 BEST WEIGHTED LEAST-SQUARES PLANES FOR [cc5H5 ) 2sccl] 2 Plane Scl Ring 1 -0.7892x + 0.3790y - 0.4832z - 4.0900 = 0 Scl Ring 2 -0.2495x + 0.9472y - 0.2015z - 7.9333 0 Sc2 Ring 1 -0.7792x + 0.4130y - 0.4715z + 1.4367 = 0 Sc2 Ring 2 -0.2428x + 0.9465y - 0.2123z -13.7549 = 0 Sc3 Ring 1 0.2436x - 0.1898y - 0.95llz + 3.3665 = 0 Sc3 Ring 2 0.72llx + 0.3712y - 0.5849z - 0.6561 = 0 D Deviations of atoms from planes (A) Atom Scl Ring 1 Atom Scl Ring 2 Cl -o.oo C6 -0.01 C2 -0.00 C7 0.01 C3 o.oo ca 0.01 C4 o.oo C9 -o.oo cs -0.00 Cl0 -o.oo Scl -2.19 Scl 2.17 Atom Sc2 Ring 1 Atom Sc2 Ring 2 Cll 0.01. Cl6 -0.03 Cl2 · -0. 01 Cl7 0.02 Cl3 -0.00 Cl8 0.03 41 TABLE 8--Continued Atom Sc2 Ring 1 Atom Sc2 Ring 2 Cl4 o.oo Cl9 -o.oo Cl5 o.oo C20 ·-0.01 Sc2 2.20 Sc2 -2.17 Atom Sc3 Ring 1 At.om Sc3 Ring 2 C21 0.01 C26 -0.00 C22 -0.01 C27 -0.01 C23 o.oo C28 0.00 C24 -0.01 C29 -0.00 C25 0.01 C30 0.01 Sc3 -2.17 Sc3 2.18 Tricyclopentadienylscandium Birmingham and Wilkinson (41) first predicted the bonding in Sc(C5H5) 3 to be purely ionic on the basis of chemical reactivity and solubility measurements. More recently, Nugent, et al., (48), have determined from absorp tion and uv-excited emission spectra that the percent covalent character in the tricyclopentadienyllanthanides is not greater than 2.5%. In opposition to this view stands the work of Wong, Lee, and Lee (43) on the crystal structure of Sm(C5H5) 3 . Even though their calculation (from the observed bond lengths) of only 37% partial ionic character in the 42 samarium-carbon bonds- is at best questionable, the fact that the cyclopentadienyl rings have a definite preferred orien tation may be interpreted as structural evidence for some covalency in the metal-carbon bonds. · It has been painted out (2) tath in. view. o f t h e sma 11er ionic. . rad" ius o f Sc 3+ . 3+ ( ) . . re l ative to Sm , Sc c5H5 3 may be expected to exhibit con- siderable covalent character. The crystal structure of tricyclopentadienylscandium gives the first direct evidence for a.degree of covalency in the scandium-carbon bond. Tricyclopentadienylscandium was prepared by the sealed-tube reaction of dicyclopentadienylrnagnesium with scandium trifluoride (29). Single crystals of Sc(C5H5) 3 were gro~n by_ sublimation and sealed in thin-walled glass capillaries •. Preliminary unit cell parameters were determined by precession (Cu Ket) photographs. The crystal system is orthorhombic. Systematic absences allow the space group to be Pbcm or Pbc21 • The lattice parameters as determined from a least-squares refinement of (sin0/~) 2 values of 12 reflections are 43 0 a = 12.881(5) .A 0 b = 8.954(4) A 0 C = 9.925(4) A V = 1145 i. 3 -3 the calculated density is 1.41 g cm· for Z = 4 using the standard relation -24 D XV X 10 NM= l.66 x 10-24 N = number of molecules per unit cell M = molecular weight D = density V = volume of unit cell (~3) Data were taken on an Enraf-Nonius CAD-4 diffractometer with Ni-filtered copper radiation. The crystal, a rod of dimen sions 0.12 x 0.15 x 0.70 mm, was aligned on the diffracto meter, such that the rod axis was coincident with the axis of the diffractometer. The diffracted intensities were collected by the scan technique with a takeoff angle of 1.5°. The scan rate -1 was variable and was determined by a fast (20° min ) pre- scan. Calculated speeds based on the net intensity gathered -1 in the prescan ranged from 7 to 0.8° min • Background 44 counts were collected for 25% of the total scan time at each end of the scan range. For each intensity the .scan width was determined by the equation scan range= A+ B tane where A= 0.9° and B = 0.45°. Aperture settings were determined in a like manner with A= 3 mm and B = 3 mm. The crystal-to-source and crystal-to-detector distances were 21.6 and 20.8 cm, respectively. The lower level and upper level discriminators of the pulse height analyzer were set·· to obtain a 95% window centered on the Cu Ka peak. As a check on.the stability of the diffractometer and the crystal, one reflection, the (121), was measured at 30-min intervals during data collection. No significant variation in the reference intensity was noticed. The standard deviations of the intensities, cr, were I estimated from the formula a I = {[CN + (Tc/2TB) 2 (Bl+B2)] + (O .03) 2 [cN+Tc/2TB ,2 (Bl+B2)] 2}'· where CN is the counts collected during scan time Tc and B1 and B2 are background intensities, each collected during the background time TB. Two symmetry related octants were measured out to 20 = 100° and one octant to 20 = 150°. A 45 total of 1620 reflections was collected of which 1013 were unique and had intensities greater than background. The intensities were then corrected for Lorentz, . -1 polarization, -and. absorption ( 31) effects (µ = 53. 4 cm ) • The calculated transmission factors ranged from 0.38 to 0.51. Fourier calculations were made with the ALFF program (35). The full-matrix, least-squares refinement was carried out using the Busing and Levy program ORFLS (34). The function w ( IF F I ) 2 was minimized. No corrections were 0 1-1 C made for extinction or anomalous dispersion. Neutral atom scattering factors were taken from the compilations of Ibers (42) for Sc, C, and H. Final bond distances, angles, ahd errors were computed with the aid of the Busing, Martin, and Levy, ORFFE program (36). Crystal structure illustra tions were obtained with the program ORTEP (40). Preliminary density calculations indicated the pres- ence of four molecules of Sc(c5tt5) 3 in the unit cell. This was interpreted to mean that the scandium atoms must lie on special positions in space group Pbcm or in general positions in the acentric Pbc21 . The Patterson map clearly showed the presence of the metal atoms on or near Z = 1/4, 3/4, the 46 location of the mirror planes in Pbcm. A structure factor calculation based on the centric space group yielded an R fact?r of- 38%, but the corresponding Fourier map was complex. Many attempts at positioning cyclopentadienyl carbon atoms with both ordered and disordered models did not improve the R factor to below 33%. At this point the structure solution was sought in the acentric space group Pbc21 . Fourier maps phased on the scandium atom quickly revealed the coordinates of several carbon atoms, and several electron density maps preceded by partial least~squares refinement showed all the non-hydrogen atoms in the asymmetric unit. The final positions of the carbon atoms clearly show that the molecular grouping cannot contain the mirror plane demanded by space group Pbcm. Although the standard acentric space group is Pca21, the structure reported here is based on Pbc21 to emphasize the similarity with tricyclopentadienylsamarium. Subsequent isotropic refinement led to a discrepancy factor of Rl = [E(IF I-IF 1)/EIF ll X 100 = 9.6% 0 C 0 Anisotropic refinement lovered R1 to 7.1%. The inclusion of 0 hydrogen atom contributions at calculated positions 0.95 A from the corresponding carbon atoms followed by further 47 anisotropic refinement of all non-hydrogen atoms led to a final Rl = 4.1% and 2 R2 = [ Ew ( IF 0 1-1 F O I i2 / (wF 0 ) ] ½ x 100 = 4. 3% 2 wher~ w = 1/cr • Unobserved reflections and two reflectionsr the {200) and (111) r whi.ch appeared to suffer from extinc tion, were not included. The largest parameter shifts in the final cycle or refinement were less than 0.02 of their estimated standard deviations. A final difference Fourier 03 map showed no feature greater than 0.4 e/A. The standard deviation of an observation of unit weight was 2.39. No systematic variation of w(IF I-IF 1> 2 vs. IF I or (sin0)/A 0 C 0 was observed. The final values of the positional and t~ermal parameters are given in Tables 9 and 10, respectively. Observed and calculated structure factor amplitudes are liste4 in Table 11. The most striking feature of the structure of tri cyclopentadienyls·candium is the existence of both bridging and terminal cyclopentadienyl groups (Figure 3). Each scandium atom is thus coordinated to two c5H5 . ions in a penta-hapto- fashion and to two others through essentially 48 TABLE 9. F-INAL ATOMIC POSITIONAL PARAMETERS· a,b' FOR TRICYCLOPENTADIENYLSCANDIUM Atom x/a y/b z/c Sc. 0.2514(1.) 0.4617(1) 0 .2400 (1) . C (1.) 0. 4367 (14) o.• 4632 (7) 0.1568(7) C(2) 0.4379(5) 0.5172(9)" 0.2862(9) C(3) 0.3853(6) 0.6501(9) 0.2937(9) C (4) 0.349"5(5) 0.6809(6) 0.1616 (9) C(5) 0.3826(4) 0.5647(6) 0.0792(6) C(6) 0.1359(4) 0.5495(7) 0.4227(7) C (7) 0.0819(4) 0.4362(6) 0.3551(7) C(8) 0.0571(4) 0~4868(6) 0.2197(8) C(9) 0.0982(4) 0.6321(6) 0. 2110 (6) C(lO) 0.1446(4) 0.6697(7) 0.3362(8) C(ll) 0.2047(4) 0.2057(5) 0.5453(5) C(l2) 0.2934(3) 0.2770(5) 0.4887(6) C (13) 0.3121(4) 0.2167(5) 0.3627(5) C(l4) 0.2343(4) 0.1061(5) 0.3388(5) C(15) 0.1689 (4) 0.1025(6) 0.4466(5) H (Cl) 0. 469.3 0.3718 0.1226 H("C2) 0.4712 0.4710 0.3624 H(C3) 0.3747 0.7161 0.3719 H(C4) 0.3075 0.7646 0.1287 H(C5) 0.3700 0.5554 -0.0143 H(C6) 0.1631 0.5442 0.5160 H (C7) 0.0605 0.3409 0.3919 H (C8) 0.0236 0.4323 0.1471 H(C9) 0.0956 0.6974 0.1330 H(ClO) 0.1770 0.7669 0.3551 H (Cll) 0.1756 0.2246 0.6305 H (C12) 0.3313 0.3547 0.5331 H (Cl.3) o. 36A9 0.2478 0.2997 H(Cl4) 0.2295 0.0432 0.2620 H{Cl5) 0.1141 0.0390 0.4551 -a Standard deviations in parentheses refer to last digit quoted. b Isotropic thermal parameters for hydrogen atoms taken - as 4. O A02 . 49 TABLE 10 · a b 4 ANISOTROPIC TEMPERATURE FACTORS. ' (x 10. ) FOR TRICYCLOPENTADIENYLSCANDIUM Atom 611 e22 e33 al2 e·13 e23 Sc 23 (.1) 56 (1) 47 (1) -1(1) 4 (1) -4 (1) C (1) 25 (3) 105(7) 100(8) -4(4) 5 ( 4) 4 (7) C(2) 4 7 (4) 204 ( 14) 133(10) -50(6) -35 (5) 71(10) C (3) 62(5) 166(12) 144(9) -61 (6) 35 (5) -74(9) C (4) 56(4) 69(6) 150(9) -10(4) 25(5) -11(7) C(5) 39 (3) 117 ( 8) 73(6) -17(4) 11 (3) 4.(5) C (6) 47(4) 119 (7) 86 (6) 15 (5) 10(5) -14(9) C(7) 29 ( 3) 105(7) 89(7) 16 (4) 12 (4) 11(6) C (8) 30(2) 105(6) 65 (7) 10 (3) 5 (4) 5 (6) C(9} 53 ( 3) 90(9) 82 (7) 30(4) 11 (4) 9 (7) C(l0} 56 (3) 94 ( 8) 12 7 ( 7) 10 (4) 18 (4) -45(6} C (11) 45 ( 3} 81 (8) 46(8l -6 (4) -4 (4) 13 (7) c.(12) 37(4) 80(6) 72(6) 5 (3) -8 (3) 32 (4) C (13) 55 (3) 102(9) 50 ( 8) 8(4) 6 (5) 27 (7) C (14) 6 8 (4) 86(6) 56(5) 2 (4) -10(4) 8(5) C (15) 49 ( 3) 102(7) 49(5) -10(3) -9 (3) 14(4) a Standard deviations in parentheses refer to last digit - quoted. b Anisotr~pic the2mal par~meters defined by exp [- TABLE 11 OBSERVED AND CALCULATED STRUCTURE FACTOR AMPLITUDES FOR TRICYCLOPENTADIENYLSCANDIUM 51 U 6 12,3 12,0 J I 12,2 U,2 2 1 21,2 29,l ...... ,...... 9 6 11,1 111,1 I 21,l 21,1 ...•••11•on••••• 1 l 21,5 29,J II I 12.1 12,6 l 1 11.1 12,1 C fl 16.1 16,2 1 0 ,.s •.• 2 1 1,9 l,J 5 I 5.5 , •• • 1 n.2 u.• z C 11,6 U,1 : ;::: :::: J 1 20,1 21,1 6 1 .,•• U,4 I ti 11,1 10,9 : : ,t: ::~ 4 1 11,1 u.a 1C O J,8 l,1 5 0 11,l J,, g ! ti 'ti 1 U,8 1',1 o 1 a., 2~.1 : : t; i:i 1 1 l,C 2,5 6 O 65,7 H,1 ~= i!~i ii:! 1 15,0 15,4 2 1 n., 21., 1: 2-.0 ,,,:_:•,, 1 11,, 11.2 J 1 5,1 •• , ; ~ 2::: 2~:: l 16,• U,J 10 fl J1,l H,6 1 ,., 5,11 • 1 H,l H,6 '! :!~i :!] 5 1 ,., 6,5 11 0 11,J ! ,.s 1 ,.1 9,7 12 o ,,., n,1 ·11 im 1 1.2 1,6 & 1 ZJ,l 22,~ 16 0 21,1 22,5 i!~l l 1 B,2 l,~ I 16,I 11,J , 2 u., n.o 1 n.11 n,J 1 1 1c.a 11,1 a 2 21.s 21.2 : 1~::~ 2,:,:_',• , 1 ,.a 1.e 9 1 U,l 1 J 2 H,O 3',l :l :m ;1!1 : 1::: 1;:: 1 i ti tl 1fl 1 18,6 11,5 s 2 21.2 .iz,1 n 1 11,6 1!:; 11 1 1,6 0,6• , z u., n,1 1 2 1 2 u., 12,8 , z 11,1 n.e i i:i ,;:r 2 2 12,& U,l 2 20.1 19,9 n:.1i::'!.:. .i!:•.. 1:: I 2 H,!I J6,7 ! :ti :l:! 3 2 l,,J 6,2 9 2 10,1 10., ;! i li~i H:! 4 2 J,J 8,J 'Kl 2 fl!i,9 '5,2 : 2i:i 2;:; 5 210,710.2 ,I : 11.m ir J 2 5,1 5,1 t1 1C,l 10,1 : : 1!:~ 1::: 2 11 8,0 1,9 12 Z 31,0 J0,1 ! • 5,l •• , J 11 2,1 2.1 I 2 l,Z 1,{I 12 2 11,6 J0,7 6 I 1.0 1,1 10 2 8,ij @,ij : .. !! •• ~1;:,.!!:! 0 J 21., 20,J ~ : .n:; :::: 1 1 0 60,6 62,1 1 J ...... , 1 i i~! J l ,-,, 9,5 lil II !1,1 !1,5 ,l,:_ !l~i 1i i i:i ,. 1 J O 211,1 25,1 l 'I $,9 !l,11 1 9 H,1 :,i::..:l~_., ,; • O 16,2 11,0 • 3 25,C 2,.1 11 11 IC,5 H,J 5 0 111.6 ••• 9 5 J 1,2 '·" 5 I 21,1 H,Z Iii2::: 6 0 7,J 1,1 t l i1,9 2~,2 o,o ~:..,: ~::: il~i 1: I l 11,0 16,J •• u., 2,1.,i:_::•, 1 0 28,1 21.1 1 \ 10,1 , •• jm t ' ,., 6.~ ti :: 1::~ 12.• 9 0 16,l 15,6 1t:: ~ i;:.;lu I • 21,1 2,., 1 9 H,1 ,t: 10 O 5,6 5,1 1 • 7,5 J.~ I 9 2,9 2,9 6 15,5 2 II 1],2 1J,) 9 II 11,1 11,1 n.• : ;: 1::1 't: 11 0 22.• 21,1 10 II :illl,11 zt,11 12 0 !I,! 6,J • • U,1 U,l 11 I 1,5 ., •• I~ : :~:~ i1::.::,:.' ,i 1!:l 1l:: ~ ~ u:: U D 16,3 16,0 11.~=_:i,: 5 • 5,9 5,11 12 , n,• ~t: ;m I • 10,1 9,9 n., J 11 ,.2 10,1 1 , n.o n.• i o·, 2,.2 10., : :: 1::: 2 1 605 J,4 i : 1::~ 9 • •,J •• 5 2 ~: ~::: • 11 1,1 2,5"' 1 I 21,5 U,6 d:: 3 1 12,1 12,1 10 4 6,2 ,., 5 11 11,. U,I 2 6 u., 31.:l : ~: 1~:: 1~:~ 11 1 U,l 12,1 2 5 1::: :ti 0 5 11,J 11,5 l 6 9,1 9,11 1 1 ••••"R•ill•••••••O O Jl,9 1 5 5,5 6,0 11 6 11,6 10,t : ii i:! !:i .,,6 5 1 "·' 11,9 .; s u.1 15,11 I O , •• 9 16,1) 6 1 5,5 9,4 l 5 J,1 ••• 5 6 ,,., 15,1 2 o 21.0 21.1 1 1 6,1 5,5 I 6 U,1 11,1 : ~= u:; u:: J O 111,9 ,.,, I 1 10.J 10.1 11 5 15,2 ... , 1 I 22,6 22,J : ;: '::i .:::;·:=•=·:·:· II O IID,J 41,6 9 1 19,J H,5 6 5 12,9 12,1 a , 1s,o n,s 10 1 6,1 6,J I 5 17,1 16.t ; ,,!:.~ 5 D 11,11 1.:1,2 9 5 9,] 5,J t 6 11,Z 11,J :i~i 6 0 u.s 22,1 11 I 11,S 16,8 ..l::: ,:.:\:i:• 10 I 21,9 20,1 : ,a,a I! 0 6 ,., 5,11 l !1 i~i ti 1 2,1 1 o n,J n., u z.• !:.. 11 6 21,0 22,l 6,0 5,2 Ill 1 1,0 ,.1 1 6 4,1 1.1 I O 2 6 1C,2 10,2 D I U,9 11,6 !.• !! •..!12: ..!!:! 10 O 11.1 11,1 1 2 52,0 SJ,11 o o 11.5 n.5 : i1 Ji :i:! • 6 J,5 7,6 , • ,,., n., 2 U t.9 ,.•• U D 3,9 3,11 li Z 5,6 s.s ;m 1 D 11,9 11,0 .u ! 6 5,5 5,2 2 I 31,C 30,9 J 2 13,1 111,1 2 0 211,2 H,1 13 0 !1,0 5,7 7,2 J I 16,I 16,0 15 0 Z,1 2,2 4 2 H,S 1C,. 6 6 .1.11 l O JZ.I ll,• ...... 1 0 23.5 J••····· 2!,1 8 6 10,,'il 11,1 II I ]9,1 16,9 0 1 111,9 11,1 5 2 26,2 10 • G U,2 11•.1 2 0 21,6 H,0 2!.2 1;:i J~· 0 7 1•;2 18,J 5 I 22,l 22,2 J on.• :I: 1 21.11 JC.6 6 2 11,1 5,S I I 21,1 21,0 5 D 12,J U,11 n., 1 1 11,9 16,9 1 2 31,3 Jl,4 2 l 16,5 16,6 6 0 111,11 11,5 I O 1,5 1.1 11 1 u., u.o 1 I 1,2 6,5 5 G 911,5 54,6 II 1 ZG,l 22,0 9 2 Jl,I 31.5 12 I 11,2 .... 1 0 6,1 , •• 11.1 10 2 10.!I 10,8 t 7 11,2 U,1 I D lG,2 JD,2 6 0 1,D 6,1 s , n., o a 11.1 s.1 G 10 17,1 l1,9 1 0 11,1 11,1 I I 29,9 21,2 11 2 21,C li1,6 l 10 U,1 13,11 10 0 u,o u., 1::5 H:i ~,::::,·:::_·:::. 1 8 2,5 3,5 ~ I I 9.1 9,1 1 1 16,3 16,5 12 2 1,1 1,6 1!:: O 12 11,8 12,11 11 0 11,1 l,I 2 1&.1 2 I ••• J,9 12 0 21",1 27,11 t O 21,6 H,I I 1 22,6 21,1 n 11.s 2 12 10,1 Ul,1 9 1 J,5 l,7 111 2 J.5 1t~ 12 O 26,1 21•• 11 e JI,& Jl,1 1., ,: -... ' •• ';0 ..,.5.6 11 12 10,5 ,,., 12 0 , •• 6,11 10 I 29.0 2',4 I l 11,1 U,l 111 O 15,1 ,s,, 1::~ 1 ~ 14, 1 1], l ...... , 12 ' 6,4 , •• 11 I t,1 6,5 2 J 11.2 ...... 5!1,1 !11,9 ~ 1 0 111,1 fl!i,I 0 1 o u.t n., 13 O 11,5 19,2 12 I U,1 1',t J 1 6,5 5,11 11,1 2 0 •• , •• , 1 tu., 11.2 11 l 25,1 21,5 1i,:_l, 5 0 1!:,! 15,l 2 1 21.1 211,J 15 0 ,.o 9,5 12 1 11,6 19,6 1 0 10,5 10.• l O 15,2 17,6 1 I 31,5 H,2 U 1 5,6 5.8 5 J 11,1 .... II O J,1 ,., J I H,1 H,l 6 J 5,6 s.s 8 G 2,1 2,• 5 1 21,5 11,!I 2 I 20.6 ,,,. 111 1 12.a 11,1 9 0 19,l H,l !I O 55,1 5'1,1 3 1 39,0 Jt,2 I! 1 3,9 J,1 1 J u.o n,2 6 O 1., 1,1 6 125,921.0 I J 1,J 1,1 1 1 .lJ,9 21,0 1 1 11.1 15,J I 1 , •• ,,5 D 2 26,1 26,ll 1 0 11.1 11.1 1 2 1,2 1,,1 • J 1•.1 U,9 ;;l 2 1 u.a o.5 I 1 5,• 3,0 5 1 13,1 15,1 il:l :iJ l 1 11,3 11,1 I O 1,1 0.1• 9 1 2,1 2,J 6 110,310.5 2 2 11,J U,D 10 J 5,2 11,9 9 0 15,2 1!1,J 11 J 10,l 10,1 5 1 ,.s 5,,) 15,t 1 1 111,1 15,1 l 2 111,5 U,I 7 1 12,.i 12,7 2 1 21,1 2f,5 10 , u.2 11 2 29•• JD.6 1 4 u.1 n,6 1 11 1 9,2 ,.1 1 1 u,1 1,.2 2 1 n,2 n,s C I 1 U,8 11,U 1 1 "'·' •o.o 12 1 9 1 n.a 111.1 5 2 11,S 19,4 ~:i tL 1 2 ZC,5 20,6 4 1 11.e 11,0 u.o u.2 2 11.0 11,5 J II Jl,2 3~.1 12 1 U.J U,2 11 1 29,1 a,., t 2 Z 2 5,8 5,9 5 1 n.c 12,1 12 1 J,5 2,1 9 .a 1.11 6.4 • • ,., •• J !.. l~.,- 1: ... !;! 6 1 24,5 2',1 14 1 5,6 , •• 1 O ll,J 12,1 l 2 2C,1 19,9 15 1 l,I U 1 ICl,t 9,1 10 2 ll,2 H,O !I • Jl,2 33,0 1 1 u.2 •s., ,.o 12 2 n,1 1!1,9 • I 2.2 2.1• J O 2!,:;; 21., 5 2 15,t 15,5 • 2 11,1 15 1 n,s n,o 1 Z ••• 11,1 9 1 1a.o 29,1 u • .a 1 2 59,2 6G,5 12 2 U,9 15,9 1 4 32,1 n.1 • C 6,2 s,, to 1 E,2 6,9 1 2 24,1 24,2 5 C U,4 IIE,J 9 2 h,8 1,,2 2 2 •4,Z .,., :. 2 z,., 29,4 11 2 1,5 1.1 • II 1,2 1,9 1 J 9,6 e.2 11 1 15,2 15,6 3 2 .,••• ,.1 0 J .... 11,5 t • U,6 21,1 1 D 23.~ 21,Z U 1 6,4 6,1 J 2 24.1 2•.1 1 1 15,1 16,6 11 I , .• 1 10.1 9 0 9,6 9,D 5,0 5,l II 2 11,3 15,J II 2 10,5 10,4 11,t 11,1 U 1 6.! 6,1 !I 2 JI•• 31,0 2 l 40,5 112,1 12 • •• , 1.0 11 '16,2 16,1 U 1 11,l 11,1 5 2 l,2 6,9 3 l 1.1 1,2 U 11 111,C U,9 1li O s. 1 !,2 9. 1 ~- 5 6 2 6G,7 ,0,2 6 2 211,1 23,5 15 1 11., 20,1 1 2 21,!I 26,9 II l 21,9 21,1 1 511,118.0 1 1 5,5 5,2 u 1 6,a 1,0 1 2 12,1 12,1 5 3 i4,1 23,11 2 5 6,2 1,1 J 1 10,C 9,l I 2 !1,1 5,4 a 2 e.s 9,l 1 2 •1.9 •1,9 9 2 211,11 23,5 6 l 21,l 26,5 ! 5 11,1 10,8 • 1 1,1 10,1 :i:: :~:i 2 2 z••' n,a 9 2 7,1 l,l 1 J 6 5 12,1 12,5 5 1 5,8 4,1 10 2 26,2 25,1 10 2 11.0 ,.o ,.o 6 1 1[,6 10., l 2 11,5 11,5 11 2 12,2 11,1n.• I J ll,8 311,l 9 5 1,S 6,1 1::: 1t~ • 2 45,! •• ,5 11 2 ••• 5.6 10 5 5,9 5,11 1 1 s.2 4,Z 12 2 11,5 11,1 13 2 7,0 s•• ' J "·' 16,9 12 1 2,f ,.,. 5 ;,i 21 •• 2,.2 111 2 11,9 !1,1 111 3 21,9 22,0 11 5 11,G 11,4 1:;: 1;:i u 2 n,1 11.1 1 2 32,• ll,I 7 2 1l,6 12,9 15 2 10,2 ,o.6 12 l 6,1 1,9 12 s 2.1 2,11 I Z 12,8 12,1 1• 2 10,1 11,1 , 6 n,s n,s 2 ii 11,C 11,6 11,t 12,) 1! 2 4,3 1,1 1 J 21.• 29,t U J 1G,6 11,l 2,J 2,J 9 2 18,1 11,l 11 J U,2 111,1 2 I 10,5 12,2 l 2 u.• 2,,!1 0 J 15,1 16,1 li l 1',6 11,1 11 2 10., 11,) 10 2 •• s .,4 J 3 110,8 41,!I 0 11 J0,1 Jl,2 J 6 36,D 35,l '·" 1f,5 11 2 U,3 U,6 1 J H,1 2E,11 II 6 2,9 11,l 5 2 21,6 29,6 J,6 l,7 2 1 n., .,,, II J 26,1 25,6 2 • 2l,I U,1 6 Z 8,11 l,J 1l 2 9,9 9,8 5 J 52,1 51,l J I 9,1 9.8 5 6 zs., 29,J ,. z 5,t ~.z 3 J 5,!i 5,6 6 6 6.l 5,9 l Z 21,l U,7 ,~:! 1~:! 4 l 21,6 2J,1 I J 9,Z 9.1 II II lJ,1 32,a 15 2 1,6 1., 6 II 26,6 25,!I 1 6 11., 11.11 a 2 a,2 a.z 12., 12., 5 J 12,l 11,9 1 J 12.1 12,1 I 2 16,1 16,2 2., 2,9 1 3 !!.6 '2,2 a 1 n.1 u.a 1 • 2.1 1,1 • 6 •• , •• , 2 l 11,l 11,Z l l 9,1 8,2 I II H,1 H,1 9 6 16,5 H,!I 10 2 l,1 1,1 1 J ZC,I 9 3 19,1 20,!I ,. 2 11,!: 11,6 1::! 1i:: l J 5•,2 !••9 2~,2 11 J 2!,1 211,1 9 • !I.I 1,6 11 6 H,7 H,9 n.1 • l 11,C 10,6 • 1 20,1 n.a 1 1 5,6 1 J 1., 1,6 ,,,s 9 3 10,1 ,., U J 5,9 6,0 10 11 20.2 20.• J,t 6,J 1,1 5 3 6!i,C 11,0 J 1 u.2 u,1 Z l 1G.~ U,8 10 3 J:1,9 22,0 12 3 5,1 ,.o n • s.• 1,9 6 J 1,6 1.5• 1l 3 9,2 9,0 12 • ,.1 6,5 4 1 a,1 1,1 ] J 15.6 16,J l n•1•,-~;f ...5,S 1 l 2!,Z 25,& t1 3 12,5 12., ,. ••• , •• 1 5 115,115.9 6 l 10,4 10,6 12,1 11.0 1• 3 6,11 6,9 O 1t.8 11,Z e J S,1 5,1 n 1 0 5 21.6 20,9 6 1 ,.1 , •• 1 J 5,6 4,8 o 12 1 1a.2 11.0 15 l 12,0 12,t 2 0 ,,1 5,8 9 J 21,4 Zt,5 1 II 21,9 23,!I 1 5 1,1 9.9 1 110,610.11 I l 1,3 9,0 1~ l 11,0 10,1 1l J 3,9 11,1 I 1 3,6 Z,9 11 ) 5,6 5,9 3 o 2, 1 ~. J• J 2 II 20., 20,6 2 5 u.1 211,1 11 J ll,8 U,1 n 1c.• ,., J 5 u., 12,6 9 1 6,2 1,0 1:i l l,O 3,1 • o s,r •.1 J fl 29,1 JD,1 5 Q .,1 l,'i 12 l 6,1 5,9 0 • 54,1 s•.1 11 5 11,2 11,1 10 1 2,8 i,7 1 • n.• 111,, 1 • 21,1 2:;,,1 I II 11,C n.t & C 16,6 1,;,e 1i J 5,5 5,9 5 II ,a,3 1,9 6 5 32,1 32,2 II 1 5,4 5,0 J I 2l,3 Zl,8 13 J 20,B 21,1 2 • 41,5 ••• 2 1 2•.t 21.• 11 4 l,8 ,,1 7 0 s.z i.1 l • 2G,I 2C,6 I • u.2 1],1 I 5 21,l 21,8 a C 1 1'.l 15,5 15 3 20,( 20,1 9 5 J,5 3,9 2 I U,O 13,1 ! 4 u.2 26,7 • 4 31,0 H,1 1 • 11.2 n.2 2 1 12,1 1 4 21,7 21,7 I 4 11,11 11,1 10 5 11.• 11,1 J I 11,2 11,J 1 • 22,5 12,6 u.c :I 4 ,., 6,1 5 11 5.0 4,5 I 1 10, 1 •.J 9 • 25.11 as.2 u s n.s n.o •• ,.o 1., 9 • 21., 2.:.0 1 • n.e u,1 6 11 21,l i2,2 II 11 1',8 l'J,2 5 1 5,1 !,2 l 11 11,2 3,1 11 11 13,1 u., U 5 11,1 11,1 s I n.5 u.o 6 1 5,6 5,l 4 • 11,!I 12,2 0 6 U,1 U,4 6 I 1,1 9,11 1 5 11., 11,1 S • 6,5 5,0 8 II 21,6 21,l 1 I H,1 1f,2 l 5 11,5 11,6 l 1 5,5 ••& 10 • 11,, ;: : ,t;; 1::: I 6 1,J 5,1 f' 2 U,ij 1',9 6 • &,t 5,6 u.o I I ••2 J,I II 5 , •• 5,1 11 11 1,1 7,5 111 II 5.3 5.J 2 6 11,1 19,5 1 2 11, ~ J, l 1 ij ,.6 ~.o 9 I 13,11 U,J 5 s s.1 ,.a 12 • U,1 21,11 1 5 21,1 31,i J ' •• 2 4,1 4 :.I 15,J 15,l I • 22,1 z:i., .a 5 5,0 5,2 II 6 18,1 11,1 1 9 9,1 1,5 9 5 •.1 5,0 'I 12.• 12,1 12 I 21,G 21,11 J 5 19,1 ,.,, 9 6 1,6 1,1 2 9 1,1 e• .a 10 5 1,1 4,11 3 2 3., 2,1 1t 6,1 6.9 111 • 19,2 19,G • 5 ,., •• , J 9 1,8 1,1 1 6 u.1 21.s 4 2 12,2 12,2 15 4 •• , S.1 6 6 11.1 11,7 6 2 tG,6 10,6 U 5,4 5.1 S 5 U,3 U,I 1 6 1,2 1,3 5 9 !1,1 5,t a 6 ,.4 s.s 0 5 21,2 JC,& J 6 21., 26,4 1 U,1 •••• 1 5 12,3 12,1 1 5 11,5 11,2 I I 11,6 n,9 t 9 11,2 •• 1 ~ ~ 1::: 1::i 2 lC,9 JO,• I 5 11,1 1Zo5 1 9 7,6 1.11 • 6 s., •.• J 51,9 ~].5 2 5 31.2 37,9 • ' 8,1 9,5 2 3 10,9 11,2 l 5 1,1 1,1 9 5 JZ,2 31,6 10 6 19,9 16,0 I 9 1,9 J,I 5 6 20,1 "·' 3 l l,2 1,1 • 11.r 1,,9 11 6 ,., 5,5 J 10 10., 11.a 1 6 11,2 11.1 ! ... , ~],7 II 5 31,5 31,1 to 5 5.1 6,G • l 12,l 12,9 11 5 23,C 22,0 U 6 ,.1 ,., 9 6 19.6 n.1 6 ,., 5,9 ~ 5 11,1 11,5 •,. 1.1 a.•• • J •• , 9,J 12 5 1,7 1,1 o ., u.a 24,9 !I to U,1 15,5 2 1 t,] 6,6 7 '1,l ll,9 6 5 1'.! 16,2 II 1 J,9 1,1 0 • 'I,] ,.1 5 11.1 U 5 10.1 11,t 1 1 111,2 15,0 6 10 ••• 5,1 Z I 15,0 15,6 1 e.6 1,1 a n,1 1 21,1 • 1 5.2 •• 6 9 5 1,5 1,1 11 5 5,2 5,1 a aa., 3 • !,1 9 34,t 14,6 1 1 n,9 n,f 3 1 5,1 1,7 7 1 i,1 1,l ,.1 11 15,1 16,J 111 s n.o u,1 !•• !! ••• ,,:.J!;! I 4 11,11 U,2 2 I 1,0 ,., II 7 21,5 21,5 0 0 6.8 1,2 1 1 20.s ao.o n 5,1 s.1 11 5 •• , 9.0 1 1,1 0 5 12,l 11,5 l.i 5 !,8 5,6 J 6 31,1 31,t 5 1 10.1 10,9 2 fl 1,1 1,1 a ,,a 1J 12,9 11,l 6 1 ,,,., 19,2 J I 12,9 U,1 2 5 1C,6 11,4 ,. , •• 6,2 n 5 ,., •·• • ' 1,2 1,9 l I J,.1 0.9 U 5 i,9 l,Q ! 6 23,2 22,f f 1 5,0 1,5 • D 5,8 11,2 I I 1,1 l,J 1 U,5 U,1 9 I 12,G U,J :••• ;•R•:;! ... !:! 6 6 6,1 1,4 I 1 11,1 11,J •• 11,9 11,1 1 O 1t,O 16,l 2 13,0 12., o , n., so.a , • 2,0 a,6• l 9,1 e,4 1 6 10,1 ,o.s 1 6 .... 15,0 10 1 11,a u.o I O ••• 5,t J O ,.1 •• , I 6 8.11 1,3 11 1 J,6 3,1 10 fl 12,6 12,2 1 I 15.J 15,11 • 1,6 8.6 2 6 41,1 41,2 6,0 1 1 u,s u.a l i 12,t 13,1 9 6 11,2 10,f U 1 1!1,5 15,1 1J • 5,1 S,1 1 9 ,.o ! 5,1 5.5 11 6 10,1 ,,•• 2 9 •• J 5,1 1 2,6 1,6 6 11,2 10,1) II 6 21,9 21,9 o a za,1 zz.5 o I so.a ,1.1 12 6 J,I J.t 2 a n., 11,1 I 1 10,1 ID,11 • 9 2.1 2.1• 1 u.z 11., I 5,6 , •• ~ 6 U,I U,l 2 !,1 5,0 52 Fig. 3.--Structure and unit cell packing of tricyclopentadienylscandium. The atoms are displayed as the 50% probability ellipsoids for thermal motion. 53 cc (D ------54 · only one carbon atom. The result is a polymeric arrangement of two symmetry related chains of Sc(c5H5 ) 3 units. The average scandium-carbon bond length of the . 0 penta-hapto-cyclopentadienyl rings is 2.49 A (Table 12), and the average distance of the scandium atom from the 0 planes of the two cyclopentadienyl groups is 2.19 A. Both of these values compare favorably with the standards reported 0 for [(C5H5) 2sccl] 2 : (4) 2.48 and 2.17 A, respectively. The data in Table 13 indicate that the scandium-carbon distance fits in well with the general trend found among first-row transition metal ~-cyclopentadienyl complexes. As Stucky has pointed out (49), the only metal-carbon bond lengths which are significantly shorter than one would predict on the basis of_metallic radii are those found with iron and cobalt. For each ring the results of least-squares best-plane calculations are shown in Table 14. The fact that ring C is in an environment quite different from that of rings A and B does not affect the planarity of the group; the maximum 0 deviation in any case of 0.01 A from the plane. 55 TABLE 12 0 INTERATOMIC DISTANCES (A) AND ANGLES (DEG) FOR TRICYCLOPENTADIENYLSCANDIUM Bonded Ring A Ring B Sc-Cl 2.525(4) Sc-C6 2.473(6) Sc-C2 2.495(5) Sc-C7 2.474(4) Sc-C3 2.471(6) Sc-CS 2.521(4) Sc-C4 2.461(5) Sc-C9 2.511(5) Sc-CS 2.500 (5)' sc-Cl0 2.505(5) Ring.. C Ring C' Sc-Cll 3.847{4) Sc-Cll' 2.519(4) Sc-Ci2 3.020(5) Sc-Cl2' 3.329 (5) Sc-Cl3 2.629{4) Sc-Cl3' 4.144(5) Sc-Cl4 3.341(5) Sc-Cl4' 4.032(5) Sc-Cl5 3.961(4) Sc-Cl5' 3.i51(4) Cl-C2 1.372(9) C6-C7 1.402(7) C2-C3 1.371(9) C7-C8 1.475(7) C3-C4 1.416(9) C8-C9 1.408(6) C4-C5 1.391(7) C9-Cl0 1.419 (8) CS-Cl 1.381(7) Cl0-C6 1.382(8) Cll-Cl2 1.425(6) Cl2-Cl3 1.383(6) Cl3-Cl4 1.430(7) Cl4-Cl5 1.363(6) Cl5-Cll 1.423(6) 0 Nonbonded Distances (A) C8-Cll 3.10(1) C3-Cl0 3.13(1) C4-Cl0 3.16(1) C7-Cll 3.21(1) C2-Cl3 3.23(1) C6-Cl2 3.24(1) Cl-Cl2' 3.29 (1) C4-C9 3.30(1) C5-Cll' 3.35(1) C7-Cl2 3.35(1) C5-Cl2' 3.39(1) Cl4-Cll' 3.39(1) C5-Cl5' 3.40(1) Cl-Cl3 3.41(1) C5-Cl4 3.42(1) C6-Cll 3.43(1) 56 TABLE 12.:.-continued Bond Angles -· C4-Cl-C5 107.1(5) Cl-C2-C3 110.6(6) C2-C-3-C4 106.3(5) C3-C4-C5 107.4(5) Cl-C5-C4 108.6(5) C7-C6-Cl0 107.9(5) C6-C7-C8 109.0(4) C7-C8-C9_ 105.2(5) C8-C9-Cl0 109. 9 (5) C6-C10-C9 109.0(5) Cl2-Cll-Cl4 106.3(4) Cll-C12-Cl3 108.7(4) Cl2-C13-Cl4 107.4(4) C13-C14-C15 108.6(4) Cll-C15-Cl4 108.9(4) ~ C' is related to C by the symmetry operations (x, 1/2 - y, 1/2 + z), followed by a unit cell translation in z. 57 TABLE 13. COMPARISON OF :METAL-CYCLOPENTADIENYL CARBON BOND DISTANCES a (Sc-C)- r(Sc)- Compound M-ir-C Ref. (M-C) r(m)b Sc(C5H5 ) 3 2.49 (5) ( (C5H5 ) 2sccl] 2 2.48 (4) c5H5TiCl(ONC9H6 ) 2 2.41 o.oa 0.15 {52) {C5H5)2Ti(C6H5)2 2.31 0.18 0 .15 {53) c5H5V{CO) 4 2.28 0.20 0.28 (54) C5H5Cr {NO). 2NCO 2.20 0.28 0.34 {55) c5H5Mn(CO) 3 2.17 0.31 0. 35 {56) . Fe (C5H5 ) 2 2.04 0.44 0.36 (57) c5H5Co(CH3c2cH 3) 2co 2.07 0.41 0.37 (58) Ni(C5H5 ) 2 2.20 0.28 0. 38 (59) ~ Representative compounds have been chosen. b Metallic radii as given in L. Pauling, "The Nature of the Chemical Bond," Cornell University Press, Ithaca, N. Y., 1960, p. 403. 58 TABLE" 14 BEST WEIGHTED LEAST-SQUARES PLANES FOR TRICYCLOPENTADIENYLSCANDIUM Plane A 0.8482x + 0.4935y - 0.1927z - 6.5220 = 0 B 0.869lx - 0.3680y - 0.3307z + 1.6838 = 0 C -0.58llx + 0.7008y - 0.4137z + 2.4670 = 0 Deviation of atoms from planes (A) Atom Plane A Atom Plane B Atom Plane C Cl -o.ooa C6 0.01 Cll -o. 01 · C2 o.oo C7 -0.00 Cl.2 o.oo C3 0.00 ca -0.00 Cl3 0.00 C4 -0.00 C9 0.01 Cl4 -0.01 cs o.oo Cl0 -0.01 ClS 0.01 Sc -2.19 Sc 2.19 Sc 2.sob a The standard deviation for the distance of each carbon 0 atom from the plane is 0.01 A and for the scandium atom, 0 0.04 A. b The distance of the scandium atom from the plane of C 0 is -2.21 A. 59 Figure 4 shows the bond lengths and angles in the three cyclopentadienyl moeities. The average carbon-carbon 0 bond distance of 1.40 A is well within the expected range (45). It should be noted that the bridging c5H5 group does not differ significantly from the terminal groups with respect to either bond distances or angles and, within the group itself, no unusual variations are found. Table 11 shows that the scandium atom is bonded equally to all five carbon atoms of ring A and of ring B. On the other hand, the association with rings C and C' appears to be of a fundamentally different nature. The 0 Sc-Cl3 bond is 0.15 A·longer than the average found in A . and B, and the bond makes an angle of 73° with the plane of ring C. The Sc-Cll' bond distance is within the range of those noted for A and B, and the bond makes an angle of 61° with the plane of ring C'. A further survey of Table 11 and ~igure 3 indicates that the interaction is through only one carbon .atom. This is especially evident for Cll', where the next closest approaches to the scandium atom (C12', ClS') 0 differ by only 0.18 A. One would expect the scandium-carbon bond to make an angle of 55° with the plane of the ring if the carbon atom 60 Fig. 4.--Bond distances and angles within the cyclopentadienyl groups for Sc (C 5H5) 3. · 61 (A) 1.42 ( B) 1.41 . (C) 1.43 62 · were sp 3 hybridized (50, 51) • Unfortunately, the meaning of· the observed angles (61, 73°) _is probably obscured by the rather strict steric requirements obtained by placing four cyclopentadienyl groups about the scandium atom. It is possible that the geometry of the bridging c5a5ion is simply the result of the minimization of the potential energy of the crystal. However, the structural parameters may perhaps be more reasonably interpreted in terms of a preferential interaction between one carbon atom and the scandium atom. To the extent one wishes to view a preferred orientation as an implication of covalent bond character, this represents the first experimental evidence for an appreciable amount of covalency in an organoscandium compound. The crystal structure of tricyclopentadienylscandium also has a direct relation to the inaccurately determined ·structure of tricyclopentadienylsamarium (43) (Table 15). The only real difference in the lattice parameters is that b for the samarium compound is almost twice the value for the scandium analog. Wong, Lee, and Lee (43) state that only a few very weak reflections were found for kf2n. A careful search between layers ink for Sc(c5H5) 3 showed no such intensities. Further studies of Sm(c5H5) 3 and related 63 compounds _may reveal even closer similarities between the two substances. ... TABLE 15 COMPARISON OF CRYSTAL DATA FOR Sc(C5H5) 3 AND Sm(C5H5) 3 Sc (C 5H5.) 3 · Sm(C 5H5 ) 3 Crystal system Orthorbhombic Orthorhombic Space group Pbcm or Pbc21 Pbcm or Pbc21 0 a, A 12. 881(5) 14.23(2) 0 b, A 8.954(4) 17.40(1) 0 c, A 9.925(4) 9.73(2) 03 v, A 1145 2295 z 4 8 Space group Pbc21 Pbcm Selected Trichlorotris(tetrahydrofuran)scandium Herzob, et al. (60) have shown that anhydrous heavy metal halides form complexes with tetrahydrofuran under anaerobic conditions. All lose tetrahydrofuran quantita tively in air and react with water. More recently, Finke and Kirmse (61), have made solubility studies of Sccl3 in 64 various solvents. They also deal with the formation of addition products of Scc13 and discuss the infrared spectra of the solutions and the dry addition products. They indicate the formation of coordinate bonds of Scc13 with the oxygen~containing solvents. Trichlorotris(tetrahydrofuran)scandium was prepared by reaction of anhydrous scandium chloride with THF under anaerobic conditions (6). Single crystals of Sccl3 (c4H8o) 3 were grown by slow evaporation of solvent and sealed in thin-walled glass capillaries. Preliminary unit cell parameters were determined by precession (Cu Ka) photographs. The crystal system is monoclinic. Systematic absences allow the space group to be P2 1/c. The lattice parameters as determined from a least-squares refinement of (sin0/A) 2 values for 12 reflections are 0 a= 8.890(4) A 0 b = 12.842(6) A 0 C = 15.485(6) A V = 1767 ~ 3 e = 92.243(5) 0 -3 The calculated density is .l.38 g cm for Z = 4. Complete three-dimensional single-crystal X-ray diffraction data 65 were obtained on an Enraf-Nonius CAD-4 diffractometer controlled by a PDP8/E computer. A graphite monochromator, with-the (002) plane in diffracting position was used to obtain monochromatic Cu Ka radiation. The radiation was detected using a scintillation counter with pulse height discrimination. The crystal, a plate of dimensions 0.10 x 0.30 x 0.30 mm, was aligned on the diffractometer1 such that one.long axis was coincident with the~ axis of ·the diffrac- tometer. The diffracted intensities were collected by the scan technique with a take-off angle of 3.5°. The scan rate -1 was variable and was determined by a fast 20°(min } prescan. Calculated speeds based on the net intensity gathered in the -1 prescan ranged from 7 to 1 min .• Background counts were collected for 25% of the total scan time at each end of the scan range. For each intensity the scan width was determined by the equation scan range= A+ B tane where A= 1.0° and B = 0.46°. Aperture settings were deter- mined in a like manner with A= 4 mm and B = 4 mm. The crystal-to-source and crystal-to-detector distances were 21.6 and 20.8 cm, respectively. The lower level and upper level discriminators of the pulse height analyzer were set to obtain a 95% window centered on- the Cu Ka peak. As a check on the stability of the diffractometer and the crystal, one reflection, the (211}, was measured at 30-min intervals during data collection. No significant variation in the reference intensity was noticed. The standard deviations of the intensities, oI, were estimated from the formula where CN is the counts collected during scan time Tc amd B1 and B2 are background intensities, each collected during the background time TB. One independent quadrant of data was measured out to 20 = 114°. A total of 1227 unique reflections were collected which had intensities greater than background. The intensities were then corrected for Lorentz, -1 polarization, and absorption (32) effects (µ· = 56.6 cm ) • · Fourier calculations were made with the ALFF programs (36). The full-matrix, least-squares refinement was carried out using the Busing and Levy program ORFLS (35). The function w ( IF 1-1 F I ). 2 was minimized. No corrections were made for 0 C 67 extinction or anomalous dispersion. ~eutral atom scattering factors were taken from the compilations of Cromer and Waber (63) for Sc, Cl, o, C, and H. Final bond distances, angles, and errors were computed with the aid of the Busing, Martin, and Levy ORFFE program (37). Crystal structure illustrations were obtained with the program ORTEP (41). Preliminary density calculations indicated the pres- ence of four molecules of sccl3 (c4a8o~ in the unit cell. The structure solution was first sought using heavy atom methods. The interpretation of the Patterson map was ambiguous; no peaks verified with any certainty. The electron density map was complex and many attempts at positioning the nonhydrogen atoms yielded an R factor of no lower than 40%. At this point the structure solution was sought by direct methods using the program MULTAN (33) with three starting phases and an absolute figure of merit of 1.4271. An electron density map phased on the scandium, chlorine, and oxygen atoms yielded the positions of the remaining nonhydrogen atoms. Several cycles of least- squares refinement with isotropic thermal parameters for all atoms produced a reliability index of R = E(IF I-IF l>/CEIF I>= 0.14 0 C 0 68 Conversion to anisotropip temperature factors and additional cycles of refinement of all nonhydrogen atoms led to a final R· = 0.077 and 1 2 2 R2 = (rw(IF0 1-IF0 1> /E(wF 0 ) ]½ = 0.077 2 where w = 1/cr • Unobserved reflections were not included. No attempt was made to locate the hydrogen atoms. The atomic positions and anisotropic thermal para meters of the nonhydrogen atoms as obtained from the final least-squares cycle are given in Tables 16 and 17, respec tively. In the final cycle, no parameter shift was greater than 0.02 of the estimated standard deviation. No systematic variation of w(IF I-IF 1> 2 vs. IF I or (sin8/J) was observed. 0 C 0 Observed and calculated structure factor amplitudes are given in Table 18. The structure consists of four discrete Scc13 (c4a8o) 3 molecules within the unit cell. Figure 5 shows the coordina tion sphere of the scandium ion with the 40% probability envelopes of the anisotropic thermal ellipsoids. Figure 6 shows a similar view of the entire molecule. The unit cell packing is shown in Figure 7. Intramolecular distances and angles together with their estimated standard deviations are listed in Table 19. 69 TABLE 16 a FINAL ATOMIC POSITIONAL PARAMETERS FOR ScC1 3 (c 4a8o~ Atom x/a y/b z/c Sc 0. 7618 (2) 0.2436{2) 0.2431(1) Cl (1) 0.7680{4) 0.4028(3) 0.3252(2) Cl (2) . 0.9451(3) 0.1608(3) 0.3402(2) Cl (3) 0.5756{3) 0.3100(3) 0.1396{2) 0(1) 0.7562(9) 0.0966(6) 0.1657(5) 0(2) 0.5862(8) 0.1709(6) 0.3129(5) 0 (3) 0 .9390 (8) 0.2844(5) 0.1574(5) C (1) 0.7266(23) 0.0918(11) 0.0725(8) C (2) 0. 7572 ( 34) -0.0211(13) 0.0486(12) C (3) 0.7483(26) -0.0795(12) 0.1283(12) C (4) 0.7846(20) -0.0080(9) 0.2020(9) C(5) 0.5872(15) 0.1644(12) 0.4083(8) C(6) 0.4403(18) 0 .1193 (16) 0.4264(9) C (7) 0.3542(14) 0.1014(12) 0.3478(10) C (8) 0.4497(15) 0.1255(14) 0.2754(9) C (9) 1.0778(14) 0. 2235 (11) 0.1472(10) C (10) 1.1455(15) 0.2683(11) 0.0696(10) C (11) 1.1007(13) 0.3817(10) 0.0685(9) C (12) 0.9441(13) 0.3799(9) 0.1023(8) a Standard deviations in parentheses refer to last digit - quoted. 70 TABLE 17- ANISOTROPIC TEMPERATURE FACTORS a,b (x 104} FOR ScC13 (c4H8o} 3 Atom Sc 102 (3) 67 (2) 42(1) 1(2) 8(1) 5 (1) Cl (1} 240(6) 80(3) 64 (~) 5 ( 3) 13 {3). -16(2) Cl (2} 134(4) 104(3) 6 3 {2) 9 {3) -9 (2) 19(2) Cl (3} 160(5) 95{3) 60{2) 11 {3) -15(2) 15(2) 0(1} 236(16) 63(6) 44 (5) 6 { 8) 3(7} 3(4) 0(2} 126 (11) 115 (8) 50{4) -33(8) -5(6) 11(5) O (.3) 150(2) 72(6) 61 {5) 7 (7) 37(6) 10 (4) C (1} 645(56) 80(12) 31(8) 1 (21) -6(16) -16(7) C (2} 1083(106) 89(15) 71(12) 87(32) 5 (28) -7 (11) C (3} 673(68) · 79(13) 95 {13) 13(24) -36(24) -20(11) C (4). 465 {42) 45 (9) 6 7 {9) 12(16) 4(15) 6 (7) C(5} 221 {25) 177 (17) 36 {7) -49(18) 23 (10) 12(9} C{6) 251(31) 264(26) 54(10) -95(25) 38(14) -4(13) C(7) 148(.22) 139(15) 102(12) -30(15) 41(13) -12(11) C(8) 166(23) 222(21) 73(10) -100(19) -21(12) 24(12) C(9) 179 (22) 125 (14) 116(12) 53(15) 97(14) 40(10) C(l0) 209 (24) 107 (13) 115(12) 25 (15) 81(15) 31 (10) C {11) 180 (22) 77 (10) 83(9) 2 (12) 51 (11) 0 C(12) 183 (21) ·75 (9) 60 (8) -6(12) 21(10) 15(7) a Standard deviations in parentheses refer to last digit quoted. b Anisotropic thermal parameters defined by exp 2 2 2 [-B 11h + s22k + s331 + 2B 12hk + 2S13hl + 2B 23kl)] 71 TABLE 18 OBSERVED AND CALCULATED STRUCTURE FACTORS FOR TRICHLOROTRIS(TETRAHYDROFURAN)SCANDIUM 72 0 U u •• u.o l U.• 1••" o , o •., ,o,t }t.~ ~·-, ., I O 1.,,.1 11,'t -2 1Z 11.0 u .... •) l Ul.J lllloJ •I ' hot .ol,,J 11,1 1.,,, l • Uol til.11 10,i, J4,l •2 t ,,., 11,J 1 H,11 J'>,I -: : it: u:: •l •'-• u,, '" .: ~ ii~i '!m U,H •9,1 '" -j .:!:! .: : ft:~ it: ' i !;~i !E; u .• 1,.• , .!!J! ~J,l ""•"' 1os., n,J -: g !t: :::: :! l 11.5 111.l ,,., n.1 _: i im im t l0,1 z,._,;i 11,11 14,I .: : m~ HI! U,l l~,1 -; i g~1 mi »,I JJ.2 1 u.1 n., ,,,. 11,1 -t I 11>,1 U.l ,~., 10,1 _,' S l 11,1 •••• ' 0 , ... , , •• o I U,l U,l •I lZ,1> U,'i 0 l U,J Fig. 5.--The coordination sphere of the scandium ion with the 50% probability envelopes of the anisotropic thermal ellipsoids. ·74 75 Fig. 6.--Molecular view of trichlorotris (tetrahydrofuran)scandium with the 40% probability envelopes of the anisotropic thermal ellipsoids. 76 77 Fig. ?.--Structure and unit cell packing of trichlorotris(tetrahydrofuran)scandium. The atoms are displayed as the 40% probability ellipsoids for thermal motion. 78 79 TABLE 19- 0 INTERATOMIC DISTANCES (A) AND ANGLES (DEG} FOR Sccl 3 (c4H8O) 3 Bonded Sc-Cl(l) 2.406(4) Sc-C1(2) 2.420(4) Sc-C1(3) 2.415(4) SC-0(1) 2.236(8) Sc-O(2) 2.147(7) Sc-O(3) 2.164(7) O(1)-C{l) 1.46(1) C (1) -c ( 2} 1.52(2) C (2) -C ( 3) 1.45(2) C(3)-C(4) 1.49(2) O (l) -c (4) 1.47(1) O(2)-C(S) 1.48 (1) O(2)-C(6) 1.47(2) C (6) -C ( 7) 1.43(2) C ( 7) -C ( 8) 1.47(2) C(8)-O(2) 1.45(1) O(3)-C(9) 1.47(1) C(9)-C(l0) 1.48(2) C(l0)-C(ll) 1.51(2) C(ll)-C(l2) 1.51(2) C(12)-O(3) 1.50(1) 0 Nonbonded Distances (A) Sc-C(5) 3.21(1) Sc-C ( 8) 3.22(1) Sc-C(9) 3.24(1) Sc-C (12) 3.27(1) Sc-C(l) 3.29(1) sc-c (4) 3.30(1) Cl(l)-O(2) 3.39(1) Cl(l)-O(3) 3.42(1) Cl(l)-Cl(2) 3.49(0) Cl(l)-C1(3) 3.50(0) Cl(l)-C(S) 3.71(1) Cl(l)-C(12) 3.86(1) Cl.(2)-O(2) 3.20(1) C1(2)-O(l) 3.23(1) Cl(2)-O(3) 3.24(1) C1(2)-C(4) 3.33(1) Cl (2) -C ( 9) 3.35(1) C1(2)-C(5) 3.39(1) C1(3)-O(l) 3.19(1) C1(3)-O(2) 3.22(1) Cl(3)-O(3) 3.25(1) C1(3)-C(l) 3.29(1) Cl ( 3) -C ( 8) 3.39(2) C1(3)-C(l2) 3.47(1) O ( 1) -0 ( 3) 2.91(1) 0(1)-0(2) 2.94(1) 0 ( 1) -c ( 8) 3.29(2) O(l)-C(9) 3.31(2) 0 (2} -c (4) 3.40(2) O(3}-C(l) 3.35(2) C ( 8) -C ( 4) 3.66(2} C(9)-C(l) 3.70(2) Bond Angles O(l)-Sc-C1(2) 87.9(2) O(l)-Sc-O(2) 84.3(3) 0 ( 1) -ScO ( 3) 82.9(3) O(l)-Sc-C1(3) 86.6(2) Cl(l)-Sc-C1(2) 92.5(1) Cl (1) -Sc-Cl {3) 93.0(1) 80 TABLE 19--Continued Bond Angles Cl(l)-Sc-O(2) 96.1(2) Cl(l)-Sc-O(3) 96.7(2) Cl (3)-Sc-) (3) 90.2(2) C1(3)-Sc-O(2) 89.7(2) O(2)-Sc-Cl(2). 88.9(2) C1(2)-Sc-O(3) 89.9(2) O(1)-Sc-Cl(l) 179.5(13) O(2)-Sc-O(3) 167.2(3)· Cl(2)-Sc-Cl(3). 174.4(2) O(l)-C(l)-C(2) 104.7(12) C ( 2) -C ( 3) -C ( 4) 108.4(14) C ( 1) -C ( 2) -C ( 3) 105.6(14) C(4)-O(l)-C(l) 111.1(9) C(3)-C(4)-O(l) 104.0(11) C ( 5) -C ( 6) -C ( 7) 110.6(12) O ( 2 ) -c (5 ) -c (6 ) 104.1(11) C(7)-C(8)-O(2) 106. 2 (11) C ( 6 ) -C ( 7 ) -C ( 8 ) 108.1(11) O(3)-C(9)-C(l0) 104.4 (11) C(8)-O(2)-C(5) 110.5(9) C(l0)-C(ll)-C(l2) 103.2 (10) C(9)-C(10)-C(ll) 105.6(11) C(12)-O(c)-C(9) 109.3(8) C(ll)-C(12)-O(3) 105.1(9) 0 The average Sc-Cl distance of 2.413(4) A is quite short compared to that of the bridged dicyclopentadienyl- 0 scandium dimer (2.575(6) A), but this is not unusual (47). 0 The Sc-O average distance of 2.182 A is long compared to .. that reported by Anderson, Neuman, and Melson (6) for 0 Sc(acac) 3 (2.070 A). However, Hanson (62) reports a Sc-O 0 distance of 2.18-2.26 A for the structure of tetraaquotris- oxalatodiscandium(III) hydrate. The average carbon-carbon 0 0 distance of 1.48 A and C-O distance of 1.47 A are reasonable for single bonds. The bond angles of the ligands to the scandium ion range from 82.9° to 96.7°. It is thus clear that some dis- tortion from a regular octahedral environment is observed 81 for the coordination of the scandium ion. The configuration of the THF rings is shown clearly in Figure 8. As is shown in Table 20, the scandium atom lies 0 only 0.02 and 0.03 A·out of the plane of two of the tetra~ 0 hydrofuran groups and 0.25 A out of the plane of the third tetrahydrofuran group. The carbon and oxygen atoms deviate considerably from the plane of the rings, since tetrahydro furan is not a planar molecule. 82 Fig. 8.--View looking down the Cl-Sc-0 axis displaying the configuration. of the THF rings. 83 84 .TABLE 20- BEST WEIGHTED LEAST-SQUARES PLANES FOR . ScC13 (C 4H8o) 3 Plane Sc Ring 1 0.9937x + 0.0659y - 0.0902z - 6.4331 = 0 Sc Ring 2 0. 3935x - 0.9193y - 0.012oz + 0.0675 = 0 Sc Ring 3 -0.4903x - 0.4240y - 0.7614z + 7.4665 = 0 0 Deviations of Atoms from Planes (A) Atom Sc Ring 1 Atom Sc Ring 2 Atom Sc Ring 3 0(1) -o.oo 0(2) -0.03 0 ( 3} 0.02 C (1) -0.08 C (5) 0.01 C (9) -0.14 C(2) 0.14 C (6) 0.02 C (10) 0.21 C (3) -0.15 C(7) -0.04 C (11) -o~. 20 C(4) 0.09 C (8) 0.04 C (12) 0.11 Sc 0.02 Sc -0.25 Sc 0.02 Bis(indenyl)magnesium The properties of bis(cyclopentadienyl)magnesium have been the subject of a great many investigations since its initial preparation in 1954 (54, 65). As an intermediate in the production of other cyclopentadienyl compounds, the substance offers certain advantages over the commonly used 85 alkali metal counterparts. Mg (C 5H5) 2 may be quite readily prepared in quantitative yield from the high temperature reaction of cyclopentadiene with magnesium metal {26), and purified by sublimation. Its high solubility in hydrocarbon solvents also affords a wider range of synthetic prospects. The relation of Mg(C5H5 ) 2 to the bis(cyclopentadienyl) derivatives of the transition metals has proved to be a point of some controversy (66, 67). Although compounds of the type M(C5H5 ) 2 , {M = Mg,V,Cr,Mn,Fe,Co,Ni), are all isostructural (68, 69-}, early magnetic, spectral, and chemical investiga tions led to the conclusion that the bonding in the magnesium and manganese compounds is essentially ionic (67, 70). Sub sequent studies of the vibrational spectra of bis(cyclopenta dienyl)magnesium indicated, however, the presence of covalent ring-to-metal bonding which is weaker than that of ferrocene ( 6 6) • Compared to the role of the cyclopentadienyl group in the renaissance of organometallic chemistry, the part played by the indenyl moiety has been small indeed. Very few indenyl transition metal complexes have been reported {71, ·72, 73), and bis(indenyl)magnesium has been noteworthy in its absence. Bis(indenyl)iron exists in the solid state 86 as a sandwich compound. which exhibits the gauche configura- tion (74): Based on the geometrical ·analogy between Fe(C5H5) 2 , one might expect Mg(C9H7) 2 to be similar in structure to Fe (c9a7.> 2 • Such is not the case. We wish to report the preparation and crystal structure of bis(indenyl)magnesium, and to discuss the relation of the new compound to the well-known bis(cyclopentadienyl)magnesium. Bis(indenyl)magnesium was prepared by the thermal decomposition (190°C) of indenylmagnesium bromide in vacuum -4 ("'10 mm) . The white crysta·lline air-sensitive substance was separated from an accompanying yellow oil and purified by sublimation. The net yield of pure product was 25%. 87 Single .crystals of Mg(C 9H7 >. 2 -were also grown by sublimation and sealed in thin-walled glass capillaried. Preliminary unit cell parameters were determined by precission (Cu Ka) photographs. Final lattice parameters as determined from a ·2 least-squares refinement of (sin0/A) . values for 12. reflec- tions accurately centered on a diffractometer are 0 a= 21.496(4) A 0 b = 12.371(4) A 0 c = 10.390(4) A V = 2763 i_3 Data were taken on an ENRAF-NONIUS CAD-4 diffracto- meter with graphite crystal monochromated copper radiation. The crystal was aligned on the diffractometer such that the rod axis was coincident with the$ axis of the diffrac tometer.' The diffracted intensities were collected by the w-20 scan technique with a take-off angle of 3.5° •. The scan . -1 rate was variable and was determined by a fast (20°min ) prescan.. Calculated speeds based on the net intensity gathered int. h e prescan range d f rom 7 to 0 • 7° min. -l • Background counts were collected for 25% of the total scan time at each end of the scan range. For each intensity the scan width was determined by the equation 88 scan range= A+ B tane where A= 1.0° and B = 0.45°. Aperture settings were deter min~d in a like manner witb A= 4 mm and B = 4 mm. The crystal-to-source and crystal-to-detector distances were 21.6 and 20.8 cm·, respectively. The lower level and upper level discriminators of the pulse height analyzer·were set to obtain a 95% window centered on the Cu Ka peak. As a check on the stability of the diffractometer and the crystal, two reflections, the (112) and (410), were measured at 30-min intervals during data collection. No significant variation in the reference intensities was noted. The standard deviations of the intensities were estimated in the fashion previously described with the value of° the parameter p setat0.02. Two symmetry related octants were measured out to 28 = 120°; a total of 1112 unique observed reflections (I>3a(I))were obtained after averaging. The intensities were corrected for Lorentz and polarization effects (31), but not for absorption 1 coefficient -1 (µ = 9.30 cm ) • Fourier calculations were made with the ALFF program (35). The full-matrix, least-squares refinement was carried out using the Busing and Levy program ORFLS (34). The 89 function w(IF l~IF 1> 2 was minimized. No corrections were 0 C made for extinction or anomalous dispersion. Neutral atom scattering factors were taken from the compilations of Cromer and Waber (63) for Mg, C, and H •. Final bond distances, angles, and errors were computed with the aid of the Busing, Martin, and Levy ORFFE program (36). Crystal structure illustrations were obtained with the program ORTEP (40). Preliminary density calculations indicated the presence of eight molecules of Mg(C 9H7) 2 in the unit cell. This was interpreted to mean that there must be· two indepen dent molecules in the asymmetric unit, since the space group P2 1 21 21 has only four~fold general positions. The interpreta tiop of a sharpened Patterson map, although quiet ambiguous, led to the correct placement of both magnesium atoms. Fourier and difference Fourier maps phased on the two magnesium atom positions led to a correc::t partial model, and subsequent Fourier calculations preceded by isotropic least-squares refinement of the magnesium and carbon atom positions, allowed the location of all 38 nonhydrogen atoms in the asymmetric unit. Anisotropic refinement with unit weights led to agreement indices of Rl = E(IF I-IF 1)/EIF I= 0.08 0 C 0 90 and R (Ew(IF I-IF 1> 2 /Ew(F >2 0.092 2 = 0 C · · 0 ]½= Inclusion of hydrogen atom contributions at calculated positions, and the use of a weighting scheme1 based on the satisfaction of the criterion that ( rF 1-1 F I) 2 not vary . 0 C with either IF I or (sin8)/A produced final values of 0 R1 = 0.066 and R2 = 0.069. Unobserved reflections were not included. The largest parameter shifts in the final cycle of refinement were less than 0.20 of their estimated standard deviations. A final difference Fourier map showed no feature - 03 greater ·than 0.4e /A, the standard deviation of an observa- tion of unit weight was 1.04. The final values of the posi tional and thermal parameters are given in Tables 21 and 22, respectively. Observed and calculated structure factor amplitudes are listed in Table 23. Figures 9 and 10 show the magnesium atoms and their associated indenyl rings .. Bis_(indenyl)magnesium in the solid state exhibits magnesium atoms in two different environments and indenyl groups of a fundamentally different nature. As shown in 1 The weighting scheme is based on essentially unit weights except for a diminished contribution from the very intense reflections. 91 TABLE 21 a,b . FINAL ATOMIC POSITIONAL PARAMETERS FOR DIINDENYLMAGNESIUM Atom: x/a ' y/b z/c Mg(l) 0.4238(1) 0.8469(2) 0.5414(3) Mg(2) 0.3932(1) 0.6182(2) 0.9413(3) C (1) 0.3551(7) 0 .8116 (11) 0.3567(15) C(2) 0. 4085 (11)· 0.8671(9) 0. 3223 {13) C (3) 0.4608(8) 0.8051(11) 0.3342(13) C (4) 0.4415(6) 0.7041(10) 0. 3820 (11) C(5) 0.4745(6) 0 .6071 (13) 0.4155(15) C(6) 0.4404(11) 0.5188(11) 0.4631(19) C (7) 0.3761(12) 0.5252(14) 0.4701(18) C (8) 0.3446(6) 0.6125(13) · 0.4372(16) C(9) 0.3769(5) 0.7058(10) 0.3933(11) C (10) 0.3946(9) 0.7352(11) 0.7029(11) C(ll) 0.4436 (8) 0.7285(10) 0.7825(18) C(12) o·.4390(6) 0.7922(12) 0.8847(15) C (13) 0.3836(5) 0.8450(8) 0.8806(10) C.(14) 0.3524 (8) 0.9263(11) 0.9697(17) C (15) 0.2914(9) 0.9553(10) 0.9257(18) C(16) 0.2724(9) 0.9192(12) 0.8138(21) C (;t 7) 0.2948(7) 0.8480(13) 0.7316(16) C(l8) ·o.3535(5) 0.8176(9) 0.7700(13) C(l9) 0.4774(4) 0.5403(8) 1.048 (11) C(20) 0.4685(4) 0.4682(8) 0.9443(12) C(21) 0.5172(4) 0.4767(8) 0.8537(10) C(22) 0.5609(4) 0.5567(7) 0.9062(9) C (23) 0.6154(5) 0.5960(8) 0.8592(10) C (24) 0.6485(5) 0.6677(9) 0.9327(14) C(25) 0.6243(5) 0.7062(7) 1.0513(14) C(26) 0.5684(5) 0.6721(8) 1.0959(9) C (27) 0.5347(4) 0.5930 (7) 1.0258(10) C(28) 0.3052(6) 0.5900(8) 1.0840(12) C (29) 0.2998(4) 0.6387(8) 0.9691(14) C (30) 0.2948(5) 0.5696(9) 0.8638(12) C (31) 0.3117(5) 0.4692(8) 0.9175(11) C (32) 0.3228(5) 0.3640(12) 0.8573(12) C (33) 0.3378(6) 0.2805(9) 0.9385(18) C (34) 0.3448(6) 0.2926(9) 1.0700(15 92 TABLE 21--Continued .. Atom x/a y/b z/c C (35) 0.3373(5) 0.3909(9) 1.1307 (11) C(36) 0.3187(4) 0.4806(9) 1.0511(11) H (Cl) 0. 3102 0.8405 0.3552 H(C2) 0.4084 0.9400 0.2833 H(C3) 0.5039 0.8309 0.3076 H(C5) 0.5223 0.6053 0.4058 H (C6) 0.4628 0.4505 0.4869 H(C7) 0.3484 0.4636 0.4998 H(C8) 0.2967 0.6178 0.4498 H(ClO) 0.3928 0.6886 0.6327 H (Cll) 0.4868 0.6827 0.7653 H (C12) 0.4708 0.8026 0.9584 H(14) 0.3689 0.9572 1 .• 0577 H(C15) 0.2641 1.0044 0.9666 H(C16) 0.2289 0.9487 0.7778 8: (Cl 7) 0.2732 0. 816 4 0.6487 H (C19) 0.4480 0.5475 1.1207 H(C20) 0.4335 0.4212 0.9325 H (C21) 0.5208 0.4366 0 •. 772 3 H (C23) 0.6343 0.5735 0.7765 H(C24) 0.6890 0.7020 0. 89 32 H(C25) 0.6513 0.7548 1.1064 a·cc26 > 0.5526 0.7025 1.1767 H(C2 8) 0.3093 0.6320 1.1641 H(C29) 0.2720 0.7099 0.9540 H(C30) 0.2869 0.5846 0.7694 H(C32) 0.3213 0. 34 75 0.7637 H(C33) 0.3427 0.2051 0.9061 H (C34) 0.3584 0. 2365 1.1332 H(C35) 0.3396 0.4016 1.• 2320 a Standard deviations in parentheses refer to last digit quoted. ~ Isotropic thermal parameters for hydrogen atoms taken 02 as 5.0 A. 93 TABLE 22 ANISOTROPIC TEMPERATURE FACTORS _a,b(x 104) FOR DIINDENYLMAGNESIUM Atom all a22 a33 al2 al3 a23 Mg(l) 36 (1) 58 (3) 95(4) -4(2) 5 (2) .-6 (3) Mg(2) 24(1) 59 (3) 123 (5) 0(2) 5(2) 8 ( 3) C (1) 62(7) 86(14) 199 (27) 34(9) -40(11) -9(16) C(2) 131(13) 38(10) 148(21) -22(10) 12 (15) -26(14) C (3) 84(9) 79(14) 155(22) -54(9) 47(12) -49(17) C(4) 41(5) 101(14) .112(17) 2(7) 7 (8} -45(13) C(5) 47(6) 158(19) 231(29) 24 (10) . 1.(10) -47(22) C(6) 124(15) 58(15) 238 (33) 29 (14) 11(25) -9(20) C(7) 142 (18) 103(21} 192 (33) -63"(19) -16 (25) -49 (23) C (8) 47(6) 175(20) 205(26) -54 {11) -14 (11) -41(23) C (9) 33(5) 110(14) 124(17) -23{6) -5 (7) -37(13) C{l0) 96(10) 112(14) 75(16) -60(10) 21(9) -18(12) C (11) 82(9) 57(10) 267 (31) -5(8) 86 ( 13) 7(16) C (12) 27(5) 167(19) 291(31) -35 {8) -39 (10) 163 {21) C (13) 26(4) 55(9) 131(15) -14(5) -1 ( 7) 7(10) C(14) 83{9) 81(16) 268(35) -31(10) 66 (18) -2(20) C(lS) 99(9) 79(12) 222 (28) 19 (8) 102(15) 15(16) C(16) 97(10) 102(16) 358 (38) 4 8 (10) 140(18) 96(22) C (17) 45(6) 149(19) 278(33) -32(9) -36(12) 132(22) C(l8) 27(4) 75(11) 178 (21) -2 3 ( 6) -22 (7) 66(13) C(19) 21(4) 82(10) 135(17) 14 (5) 8 (7) -6 ( 13) C(20) 25 (3) 56(8) 172(17) 1 (4) 7 ( 8) 27(12) C (21) 22 (3) 51 (8) 126(15) 7(5) 2 (7) -4(10) C (22) 18(3) 66(9) 78 (13) 10 (4) 7 (5) 9 (9) 94 . TABLE 22--Continued Atom '\1 13 22 B33 13 12 13 13 13 23 C (23) 32(4) 62(9) 104(13) l(5) 9 (6) 0(10) C(24) 35(5) 91 (12) 172(21) -3(6) 7 (9) 15(15) C(25) 26(4) 53 (9) 206(23) .-4 ( 5) -17(9) -15(13) C(26) 30(4) 88(11) 95 (14) 10(6) -17(7) -9 (11) . C(27) 13(3) 73(9) 107(14) 10(4) 8(5) 10 (11) C (28) 44(5) 56 (9) 151(19) 6 (5) 20(8) -6(12) C(29) 24 (3) 64(10) 223 (23) 0(6) 24 (8) 30(14) C (30) 37 (4) 75(9) 140(16) 5 (5) -20 (7) 47(10) C (31) 26(4) 73(11) 116(19) -7(5) -3(7) -16(11) C(32) 36(4) 151(15) 126(16) -24 ( 7) -2 (7) -30(15) C (33) 44(5) 60 (10) 270(30) -3(6) 1 (11) . -19 (16) C (34) 46(5) 76(12) 202(24) -6 (6) 15(10) 36 ( 15) C (35) 39 ( 4) 89 (11) 120(16) -6 (6) . 8 ( 7) 8(11) C(36) 19 ( 3) 107 (13) 91(15) -3 (5) 3(6) 26(12) a Standard deviations in parentheses refer to last digit - quoted. b Anisotropic thermal parameters defined by exp [ 2 2 2 ] -13 11h + a22k + 13 331 + 213 12hk + 213 13hl + 213 23kl) 95 TABLE 23 OBSERVED AND CALCULATED STRUCTURE FACTOR AMPLITUDES FOR BIS(INDENYL)MAGNESIUM 96 J r.1 l'>, l 11,, UUHl• J•- • U•• , ' ,., .. ool t,.J 1.4 J.<; l O "8.J P.I . l •I•< f, I b.~ 11., 11.1:, ..,., " o U~.l hl.J 11.,,14,4 ''·"l',,< , ,•., ll.~ 11,) .... 5 , ... 6 II ~2.1 ~l. .. I hol 1,., .. .. H.~ IJ,1 11,0 H.3 1,.b Idol 8 0 ~2.7 ~1-" ' ' llol 4d,', S,,./ ho'• .<~. I ,~.~ ' " 1, .., 11., " ' lb.? I~,) 1.7 •• ~ lO ~ 11.7 10, Fig. 9.--Illustration of magnesium(l) and its associated inderiyl rings. 98 99 Fig. 10.--View of magnesium(2) and its associated indenyl rings. 100 101 Figures 9, 10, and 11 each magnesium atom is coordina ated to three indenyl moieties, one in a penta-hapto fas~ion and two in a less symmetric manner. The substance exists therefore in an -infinite polymeric arrangement with both bridging and terminal indenyl groups. Table 24 presents the bond length calculations upon which a detailed description of the structure can be based. The terminal group is bonded to Mg(l) at distances ranging 0 from 2.31(1) to 2.54(1) A with the larger values correspond- ing to the sterically less favorable C(4) and C(9) positions. The association with the two bridging ring systems appears to be through essentially only one carbon atom in each group: 0 0 C(lO) at 2.26(1) A and C(21) at 2.32(1) A. The extent to which the interaction is localized with these two atoms is seen with reference to the other distances in the five- membered ring fragments (Figure 12). The closest approach 0 by another atom from either ring is 2.67(1) A, greater than any approach for the penta-hapto group. The second independent magnesium atom, Mg(2), is bonded to its terminal group in a more distorted fashion than 0 is Mg(l); the lengths range from 2.26(1} to 2.60(1) A, but again the long distances correspond to the sterically less 102 Fig. 11.--Structure and unit cell packing of bis(indenyl)magnesium. 103 104 TABLE 24 ·o INTERATOMIC DISTANCES (A) AND ANGLES (DEG) FOR DIINDENYLMAGNESIUM Bonded Mg(l)-C(l) 2.46(2) Mg(l)-C(2) 2.31(1) Mg(l)-C(3) 2.35(1) Mg(l)-C(4) 2.45(1) Mg(l)-C(9) 2.54(1) C(l)-C(2) 1.38(2) C(2)-C(3) 1.37(2) C(3}-C(4) 1.41(2) C (4) -C (5) 1.44(2) C(5)-C(6) 1.41(2) C(6)-C(7) 1.29 (3) .c (7) -c (8) 1.32(2) C(8)-C(9) 1.42(2) C(9}-C(l) 1.44(2) Mg(l)-C(l0) 2.26(1) Mg (1) -C (11) 2 .• 93(2) Mg(l)-C}l2) 3.65(2) Mg(l)-C(l3) 3.63(1) Mg (1) -C (18) 2. 84 (1) C(l0)-C(ll) 1. 34 (2) C(ll)-C(l2) 1.33(2) C(12)~C(l3) 1.36(2) C(13)-C(14) 1.52(2) C(14)-C(15) 1.43(2) C(15)-C(16) 1.31(3) C(16-C(17) 1.32(2) C(17)-C(18) 1.38(2) C(l8)-C(10) 1.52(2) Mg(2)-C(19) 2.33(1) Mg(2)-C(20) 2.46(1) Mg (2) -C (21) 3.32(1) Mg(2)-C(22) 3.70(1) Mg(2)-C(27): 3.18(1) C(19)-C(20) 1.41(1) C(20)-C(21) 1.41(1) C(21)-C(22) 1.47(1) C (22)-C (23) 1.36(1) C(23)-C(24) 1.37(1) C(24)-C(25) 1.42(2) C(25)-C(26) 1.36(1) C(26)-C(27) 1.42(1) C(27)-C(19) 1.41(1) Mg ( 2 ) -C ( 2 8 ) -· 2.43(1) Mg(2)-C(29) 2.26(1) Mg(2)-C(30) 2.34(1) Mg ( 2 ) -C ( 31) 2.55(1) Mg (2) -C ( 36) 2.60(1) C(28)-C(29} · 1.38(2) C ( 2 9 ) -C ( 3 0 } 1.39(2) C(30}-C(31) 1.41(1) C(31)-C(32) 1.46(2) C ( 3 2) -C ( 3 3) 1.37(2) C (33)-C (34) 1.38(2) C(34)-C(35) 1.38(2) C(35)-C(36) 1.44(2) C(36)-C(28) 1.43(1) Mg(2)-C(ll) 2.40(1) Mg(2)-C(12) 2.44(1) 0 Nonbonded Distances (A) Mg (1) -C (5) 3.42(2) Mg (1)-C (8) 3.53(1) Mg (1) -C (17) 3.41(2) Mg(2)-C(l0) 2.87(1) Mg ( 2 ) -C ( 13 ) 2.88(1) Mg(2)-C(l8) 3.16(1) Mg(2)-C(32) 3.60(1) Mg(2)-C(35} 3.64(1) 105 TABLE 24--Continued 0 Nonbonded Distances (A} Mg(2}-C(l4} 3.92(2) C(l)-C(lO) 3.81(2) C(5)-C(10} 3.79(2) C(4)-C(10) 3.50 (2) C(6}-C(l0} 3.79(2) C(7)-C(l0) 3.57(2) C(8)-C(10) ~.33(2) C(9)-C(10) 3~26(2) C(10}-C(30} 3.41(2) C(10)-C(29) 3.76(2) C (11) -C (22) . 3.54(2) C(ll)-C(21) 3.57(2) C(ll)-C(27) 3.61(2) C(ll)-C(20) . 3.67(2) C(ll)-C(l9) 3.68(2) C(ll}-C(30) 3.85(2) C(ll)-C(29) · 3.99(2) C(12}-C(27) 3.53(2) C (12) -C (19) 3.64(2) C(l2}-C(29) 3.83(2) C(12)-C(26) 3.84(2) C(12}-C(22) 3.92(2) C(13)-C(30) 3.91(2) C(l3)-C(29) 3.38(1) C(l5}-C(29) 3.94(2) C(14)-C(29) 3.80(2) C(l7)-C(29) 3.58(2) C (16)-C (29) 3.84(2) C (18) -C (29) 3.32(2) C(17}-C(30) 3.71(2) C(l9)-C(36) 3.49(1) C(18}-C(30) 3.46(2) C(l9)-C(28) 3.77(2) C(l9)-C(35) 3.64(2) C ( 2 0) -C ( 31) 3.38(1) C(19)-C(31) 3.91(2) C(20)-C(32) 3.51(2) C(20)-C(35) 3.41(2) C(20)-C(33) 3.65(2) C(20)-C(35) 3.55(2) C(20)-C(34) 3.67(2) Bond Angles C(l)-C(2}-C(3) 112.4(10) C ( 2) -C ( 3) -C ( 4) 106.7(13) C(3)-C(4)-C(9) ·108.1(13) C (4)-C (9)-C (1) 108.4(12) C(9)-C(l)-C(2) 104.4(13) C(9)-C(4)-C(5) 119. 0 (14) C(4}-C(5)-C(6) 118.4(14) C(5)-C(6)-C(7) 119. 7 (16) C(6)-C(7)-C(8) 122.9(17) C ( 7 ) -C ( 8 ) -C ( 9 ) 119.8(16) C ( 8) -C ( 9 ) -C ( 4) 120.1(14) C(18)-C(10)-C(ll) 1_02. 4 (11) C(10)-C(ll)-C(l2) 113.5(15) C{ll)-C(l2)-C(13) 108.9(14) C(l2)-C(13)-C(l8) 108.9(13) C(l3)-C(18)-C(10) 106.1(12) C(18)-C(l3)-C(l4) 117.9(13) C(l3)-C(l4)-C(15) 112.0(15) C(14)-C(15)-C(l6) 118.9(16) C(l5)-C(16)-C(17) 133. 5 (21) C(16)-C(l7)-C(18) 109.3(18) C(l7)-C(18)-c(l3) 127.8(16) C(27)-C(l9)-C(20) 106.4(10) C(l9)-C(20)-C(21) 111.2(9) C(20)-C(21)-C(22) 106. 0 .( 9) C{21)-C(22)-C(27) 106.3(9) C(22)-C(27)-C(19) 109.9(10) C(27)-C(22)-C{23) 122.5 (10) C(22)-C{23)-C(24) 118.6(11) C(23)-C(24)-C(25) 120.8(11) 106 TABLE 24--Continued Bond Angles C(24}-C(25}-C(26} 121.0(11} · C(25)-C(26)-C(27) 119.6(11) C(26}-C(27}-C(22} 117.3(9) C(36)-C(28)-C(29) 104. 9 (11) C(28}-C(29}-C(30} 113.2(9) C(29)-C(30)-C(31). 104.5(10) C(30}-C(31)-C(36} 109.3(11} C(31)-C(36)-C(28) 108.1(11) C(36}-C(31)-C(32} . 119. 7(12) C(31)-C(32)-C(33) 116.3(12) .c ( 3 2) -c ( 3 3) -c ( 3 4) 123.5(13) C(33)~C(34)-C(35) 122.3(13) C(34)-C(35}-C(36) 116.7(12) C(35)-C(36)-C(31) 121.3(12) 107 Fig. 12.--Bond distances and angles within the indenyl groups for Mg(C9H7) 2 ...... 0 00 109 favorable C(31) and C(36) positions. In this situation the bridging groups are coordinated through two carbon atoms 0 2.33(1) to 2~46(1) A. This is only the second single-crystal study of a n-c5a5 group being coordinated to the magnesium. Stucky 0 has an average Mg-n-C distance of 2.55 A, whereas in bis (indenyl) magnesium an average Mg-n-C distance of only 0 2.43 A is found (Table '24). The normal magnesium-carbon single bond is about 0 2.18 A (76, 77). Thus, it is seen that the bridge bonds are longer than normal single bonds, but shorter than then-bonds 0 (2.43 A). We view the bonding as being either essentially ionic with some directional {covalent) character or weak covalent bonds, such that the lattice,effects (packing) dominates. Such is not the case with diindenyliron where the strong covalent bonds cannot be broken even for more desirable packing. For each ring the results of least-squares best- plane calculations are shown in Table 25. The maximum 0 deviation in any case is 0.04 A from the plane indicating 110 planarity of the groups. Figure 12 shows the bond lengths I and angles in the four indenyl moieties. The average carbon carbon bopd distance is well within the expected range (45). It should be noted that the bridging indenyl groups do not differ significantly from the terminal group with respect to either bond distance of angles and, within the group itself, no unusual variations are found. TABLE 25 BE.ST WEIGHTED LEAST-SQUARES PLANES FOR DIINDENYLMAGNESIUM Plane Mgl-Ring 1 -0.1018x - 0.3210y - 0.9416z + 7.4862 = 0 Mgl-Ring 1 -0.0989x - 0.3299y - 0.9397z + 7.4587 = 0 Mgl-Ring 2 0.4484x + 0.7440y - 0.4953z - 6.6593 = 0 Mgl-Ring 2 0.428sx + 0.7558y - 0.4952z - 6.8951 = 0 Mg2-Ring 3 -0.4900x + 0.7232y - 0.4868z + 5.5100 = 0 Mg2-Ring 3 -0.4814x + 0.7317y - 0.4825z + 5.3167 = 0 Mg2-Ring 4 0.9619x + 0.2442y - 0.1232z - 6.6971 -- 0 f.ig2-Ring 4 0.9642x + 0.2322y - 0.1283z - 6.5799 = 0 0 Deviations of Atoms from Planes (A) Atom Mgl-Ring 1 Atom Mgl-Ring 2 Cl -0.00 ClO -0.00 C2 -0.00 Cll -0.00 C3 0.01 Cl2 0.01 C4 -0 .01 Cl3 -0.01 111 TABLE 25--Continued Atom Mgl-Ring 1 Atom Mgl-Ring 2 I C9 0.01 Cl8 0.01 I Mgl -2 .10 Mgl 2.14 Cl -0.02 ClO -o.oo C2 -0.01 Cll -0.02 C3 0.02 Cl2 o.oo C4 0.00 Cl3 0.01 cs 0.01 Cl4 0.04 C6 o.oo ClS 0.02 C7 -0.03 Cl6 -0.04 ca -o.oo Cl7 o· .02 C9 0.02 Cl8 -0.02 Mgl -2.10 Mgl 2.14 Atom. Mg2-Ring 3 Atom Mg2-Ring 4 Cl9 0.01 C28 0.01 C20 -0.01 C29 -0.01 C21 0.01 C30 0.01 C22 -o.oo C31 -0.01 C27 -0.01 C36 -o.oo Mg2 2.14 Mg2 2.10 Cl9 0.01 C28 -0.00 C20 -0.03 C29 -0.03 C21 -0.00 C30 0.02 C22 0 .• 01 C31 0.01 C23 · o .01 · C32 0.01 C24 0.03 C33 0.01 C25 -0.03 C34 -0.02 C26 -0.02 C35 -0.02 C27 0.02 C36 0.03 Mg2 2.12 Mg2 2.09 CHAPTER IV CONCLUSIONS From our studies of the preparation and properties of organoscandium compounds we have presented evidence for: 1. The solid state existence of two types of scandium-carbon bonds, one of the classic ~-description and one which could be viewed as a in character. The normal Sc--C ~-bond is 2.48 A, and the a-bond lengths are closely similar. 2. The existence of some degree of directional covalent bonding in the series Sc, Yt, La Lu. 3. The use of Mg(C9H7) 2 as an intermediate in the preparation of new organoscandium compounds. 4. The "normal" values of Sc-Cl and Sc-0 bond l ength s and ..ionic . rad' ius o f Sc 3+ The structural studies described herein represent the first detailed characterization of organometallic 112 113 scandium complexes. We expect that based on this work as a foundation, the field of scandium chemistry will grow and perhaps even develop in the fashion of titanium chemisty. REFERENCES 1. T. Moeller, D. F. Martin, L. c. Thompson, R. Ferrus, G. R. Feistel, and W. J. Randall, Chem. Rev., 65, 1 (1965) • 2. G. A. Melson, and R. W. Stotz, Coard~ Chem. Rev., J_, 133(1971). 3. M. K. Guseinova, A. S. Antsyshkina, and M.A. Porai Koshits, Zh. Strukt •. Khim.,~, 1040(1968); J. Struct. Chern. (USSR) , 2_, 926 (1968) • 4. K. D. Smith, and J. L. Atwood, Chem. Comm., 593(1972). 5. J. L. Atwood, and K. D. Smith, J. Amer. Chem. Soc., 95, 1488 (1973). 6. T. J. Anderson, M.A. Neuman, and G. A. Melson, Inorg. Chem., 12(4), 927(1973·). 7. F. A. Hart, A.G. Massey, and M. s. Saran, J. Organometal. Chem., 21, 147(1970). 8. R. s. P. Coutts,and P. C. Wailes, J. Organometal. Chem., 25, 117(1970). 9. E. A. Jeffery, and T. Mol:-e, J. Organometal. Chem., 11, 393 (1968). 10. V. M. Pletz, Cornpt. Rend. (USSR), 20, 27(1938). 11. G. Vagupsky, W. Mowat, A. Shortland, and G. Wilkinson, Chem. Comm., 1369(1970). 12. J. A. Witt, and G. A. Melson, Amer. Chem. Soc. Nat. Meeting, New York, N. Y., Abstr., No. 81 (1972). 13. F. Hug, W. Mowat, A. Shortland, A. C. Skapski, and J. Wilkinson, Chem. Comm., 1079(1971). 114 115 14. F. Hug., w. Mowat, A. c. Skaps·ki, and G. Wilkinson, Chem. Comm., 1477 (1971). · 15. M. R. Collier, M. F. Lappert, and M. M. Trudock, J. Organometal. Chem., 25 C36 (1970). 16. P. s. Braterman, and R. J. Cross, J. Chem. Soc. (Dalton), 657(1972). 17 •. O. Roelen, Angew. Chem., 60, 62 (1948) • 18. J. Smidt, w. Hafner, R. Jira, R. Sieber, J. Sedlmeier, and A. Sabel, Angew. Chem., 72, 93(1962); Angew. Chem. internat. Edit, 1, 80 (1962). 19. L. E. Lyons, and L. E. Rennick, Ind •. Eng. Chem. Prod. Res. Develop.,~, 2(1970). 20. D. Evans, J. A. Osborn, and G. Wilkinso~, J. Chem. Soc. (A), 3133(1968). 21. G. -Henrici-Olive and S. Olive, Advan. Polymer Sci., §_, 421(1969). 22. H. Bestian, and K. Clauss, Angew. Chem., 75, 1068(_ 1963); Angew. Chem. internat. Edit, 2, 704(1963). 23. W. Reppe, O. Schlichting, K. Klager, and T. Toepel, Liebigs Ann. Chem., 560, 1(1948). 24. G. Wilke, B. Bogdanovic, P. Hardt, P. Heimbach, w. Keim, M. Kroner, w. Oberkirch, K. Tanaka, E. Steirrucke, D. Walter, and H. Zimmermann, Angew. Chem. l!!_, 157(1966); Angew Chem. internat. Edit., ,?., 151 (1966) • 25. L. Vaska, and J. w. Diluzio, J. Amer. Chem. Soc.,~, 2784 (1961). 26. W. A. Barber, Inorg. Syn.,~, 11(1960). 116 27. J.B. Reed, B. s. Hopkins, and L. F. Audrieth, Inorg. Syn., l, 28(1939). 28. R. W. Stotz, and G. A.· Melson, I:riorg. Chem., 11(7), 1720(197i). 29. A. F. Reid, and P. C. Wailes, Inorg. Chem., 1(7), 1213(1966). 30. L. J. Guggenberger, and C. Prewitt, "Program ACAC," E. I. du Pont de Nemours and Co., Willmington, Delaware. 31. D. J. Wehe, W.R. Busing, and H. A~ Levy, "ORABS, A Fortran Program for Calculating Single-Crystal Absor;ption Corrections," Report ORNL-TM-229 Oak Ridge National Laboratory, Oak Ridge, Tenn., 1962. 32. B. K. Dewar, A. L. Stone, and E. B. Fleischer, "Program FAME," private communication, 1966. 33. P. Main, and M. M. Woolfson, "MULTAN, A Computer Programme for the Automatic Solution of Crystal Structures," Department of Physics, University of York, York, England, and G. Germain, Laboratoire de Chimie Physique, Universite de Louvain, 39 Schapenstraat, Leuven, Belgium, 1971. 34. w. R. Busing, K. o. Martin, and H. A. Levy, "ORFLS, A Fortran Crystallographic Least~~quares Program," Report ORNL-TW ... 305, Oak Ridge, Teri.:n:., 1962. 35. c. R. Hubbard, C.·O. Quicksall, and R. A. Jacobson, "Program ALFF," USAEC Report IS, 2625, Ames . Laboratory, _Iowa State University, Ames, Iowa, 1971. 36. w. R. Busing, K. o. Martin, and H. A. Levy, "ORFFE, A Fortran Crystallographic Function and Error Program," Report ORNL-TM-206, Oak Ridge National Laboratory, Oak Ridge, Tenn., 1964. 37. F. K. Ross, "Program HYGEN," private communication, 1967. ! 117 38. P. K. Hon-, "Program JAM," private communication; 1964. 39. P. K. Hon, "Program BEPLAl," private communication, 1964. 40. C. K. Johnson, "ORTEP, A Fortran Thermal-Ellipsoid ·Plot Program for Crystal Structure Illustrations," Report ORNL-3794, Revised, Oak Ridge National Laboratory, Oak Ridge, Tenn., p. 70, 1965. 41. J.M. Birmingham, and G. Wilkinson, J. Amer. Chem. soc • , 7 8 , 4 2 ( 19 5 6 ) • 42. J. A. Ibers, "International Tables for X-ray Crystallography," Vol. III, Kynoch Press, Birmingham, England, 1962, pp. 202, 203. 43. c. H. Wong, T. Y. Lee, and Y. T. Lee, Acta Cryst., Sect. B, 25, 2580 (1969) •. 44. J. L. Atwood, J. H. Burns, and P. G. Laubereau, J. ,~er. Chem. Soc., 95, 1830 (1973}. 4 5. F • A. Cot ton, · and G. Wilkinson, ·"Advanced Inorganic Chemistry," 2nd ed, Wiley-Interscience, New York, N. Y., 1966. 46. J. L. Atwood, and K. D. Smith, unpublished results. 47. G. Allegra, G. Perego, and A. Immirzi, Makromol. Chen., 61, 69(1963). 48. L. J. Nugent, P. G. Laubereau, G. K. _Werner, and K. L. Vander Sluis, J. Organometal. Chem., 23, 487(1970). 49. C. Johnson, J. Toney, and G.D. Stucky, J. Organometal. Chem., 40, C11(1972). 50. V. I. Kulishov, E. M. Brainina, N. G. Bokiy, and Yu. T. Struchkov, Chem. Commun., 475(1970). 51. J. L. Calderon, F. A. Cotton, and P. Legzdins, J. Amer. Chem. Soc., 91, 2528(1969). 118 52. J. D. Matthews, and A.G. Swallow, Chern. Commun., 882(1969). 53. v ..Kocrnan, J. c. Rucklidge, R. J. O'Brien, and W. Santo, ibid., 1340(1971). 54. J.B. Wilford, A. Whitla, and H. M. Powell, J. Organornetal. Chern.,~, 495(1967). 55. M.A. Busch, G. A. Sim, G. R. Knox, M. Ahmad, and C. G. Robertson, Chern. Commun., 74(1969). 56. A. F. Berndt~ and R. E. Marsh, Acta Crystallogr., 18, 118 (1963) • 57. J. D. Dunitz, L. E. Orgel, and A. Rich, ibid., ~r 373 (1956). 58. L. F. Dahl, and D. J. Smith, J. Arner. Chem. Soc., 83, 752 (1961). 59. L. Hedberg, and K. Hedberg, J. Chern. Phys., ,?2, 1228(1970). 60. s. Herzog, K. Gustav, E. Krueger, H. Obenender, and R. Schuster, z. Chem., I(ll), 428(1963). 61. A. Finke, and E. M. Kirrnse, z. Chern., 5(5), 193(1965) (Ger.). 62. Hansson, Acta Chern. Scand., ~(4), 1337(1972). 63. D. T. Cromer, and J. T. Waber, Acta Cryst., 18, 104, 1965. 64. F. A. Cotton, and G. Wilkinson, Chem. and Inc., 307 (1954). 65. E. 0. Fischer, and w. Hafner, Z Na turf ors chg. , 9b, 417 (1954). 66. E. R. Lippincott, J. Xavier, and D. Steele, J. Amer. Chem. Soc., 83, 2262(1961). 119 67. F. A. Cotton, and L. T. Reynolds, J. Arner. Chem. Soc., ~, 269 (1958). 68. J. D. Dunitz, L. E. Orgel, and A. Rich, Acta Cryst., i, 373 (1956). 69. Von E. Weiss, and E. o. Fischer, Z. Anorg. Chem., 278, 219(1955). 70. G. Wilkinson, F. A. Cotton, and J.M. Birmingham, J. Inorg. and Nucl. Chem.,~, 95(1956). 71. E. O. Fischer, and D. Seus, Z Naturforschg., ~b, 694 (1953). 72. E. O. Fischer, D. Seus, R~ .Jira, ibid., 8b, ·692 (1953). 73. P. L. Pauson, and G. Wilkinson, J. Amer. Chem. Soc., 76, 2024(1954). 74. J. Trotter, Acta Cryst., 11, 355(1958). 75. c. Johnson, J. Toney, and G.D. Stucky, J. Organometal. Chem., 40, Cl.1(1972). 76. J. Toney, and G.D. Stucky, Chem. Commun., 1168(1967). 77. L. G. Guggenberger, and R. E. Rundle, J. Amer. Chem. Soc., 2.Q.(20), 5375 (1968). .~ 4 ,., n., l,/ 11,•l, io.l 1,.1 l';,f Jh.', z~.~ t 1.~ '" ll,7 1'.-J 5 t,J a.o e.l 17 •• 1 ·,.~ " " II.I 1,1 11.~ J].,l 6 U ~.(' t., 11.1 lo.ij l~, I l~ Iii ~o<> 3,l UU .. I.• l .. U* .. " ,., lf,,1 I~.! :; tl ill1 1 I~ Is,•; \'lo\ " ' " \(J.] . lJ.5 ll.2 IJ ;; 1<,f 74.4 I l'l•~ \d.4 " ,., ,., 14.J '" 1,1 ,., II ll,'• 14.I o' ?,'..', !",.~ " .;1.1 ,~. r '" 2 ll U.I U •., la.z 1~.~ 11.• ,, \'l.L lJ.9 ' " '"·" 1 .... 1 ' If.~ /b.~ " c. , I}.~ IJ.2 " ... ,, ) 11 , ••, ~ l''• 4 '1. ' .. ' , " It.~ H.l II ,n.,, 2J,I l •.l I~. 1 l.l • .; 1r,.~ P.1 1~ • .1 1,.,; Doij I •'•' ,,.,. ,.,., 1•;.1 ,l "It 11 11 u.l ~.~ " " 7,l' ll O.J ;.1 !?,• 74.o .:J.} n ~- f ~- 5 ' "' 11.4 ,~ .. l '•" l•J.,. 2 i< 11.1 J,:.\ It; 1.,.1 )"•'• ,.... '" 11 z,i.e ;o.i 11 It" I Ll ~. ~ ~.~ ,~. ·' ~~-·' n.~ ll r,.J II,·) ~ 3 U ti,/ 13d J lJ Jd ••" ?~., ,. .,.1 UUOL• lh .. ou n.~ l'lo) IU.I i.::]:!! / JI ,., lv.1 J ~--· ')],. l O lbd ~3.1 ...... ,. 7 ...... I H,.l It.) ho• lf.l If.• ~~.1 ,~. \ :,.4 2~.4 2 0 , ... •;.:i 'J '> l 7l.4 l"'J.l " ,., .. 1., 41.l .,.I If.I )uo• n:1 H:~ l O 4d, d >4o4 t! l ~ IJ,.7 !'..:.,, " I l<-1 IJ.S 't J lv'i•~ <. H.~ J.). I 1~. ·, 11','> 11.1 I~-" IJ. 1 2 J H." J4 0 J " 5 ~ I~.• I 'lou , J ~4. I ~< .•; ?o,l ,," 11.J llos ' I J.I I••~ 6 •'.! <,]..; 41.'J 4<,.7 :,.,; !·,.! ,,1.• ' 11.1 11,l 7 3•,.1 ~, .. , J,.1 \eoO l2. I 1~.l ! ~"-" 2'1. • l•,f 77.• " :iii I~.~ II.~ ;!~i a o 15,> II.I 11.1 ,:1.~ ' " q 'lh.3 H,~ 2~., n., !1.4 /', •• ' Ji'.! 11., \LI <) ~.,; 1,.1 ,.1 1, •., .•.i.1 1·1.J ll [) ~. I 3,'J n.1 , .... , 11.J 11.·, ' " I I.~ ;~:~ i;~i It ) ~•.' ~•,.-l 1,,.4 •. ., :: :m :i~I l~ 4.1 l,• " 1,,.,, 1,., " b« ll-' 1-,.,. !', .. n.J 71.• " " 1:,,1, h.l ! l):~ I!:~ 16 ~ ,.., •• , :i .., .. .. ~ lo; ~~ •·• ~- 4 ~-~ ,,,. / ~''• I ,, Iv.~ 14., H,I II.I 22 J. ~ 1, <> :.1.1 .. 1 •., llJ.J [.'I.S ' 6,,. 0 I "" I.~ ~~-~ ,,,,.s ••• ~ 1~. J l"•' l'>oJ 1,1 I,: H:! I t 11,.1 Uld \].I i~.1 <7.L :t~ " , , 1, • ., ll,