.2... . .. 1.3....) . . n 1.... If; v. .1. 5.7 2.47/1 ..¢ Z :r...../.i’../... 3...... .43... .1... g: .a . .62... .V _. ..r.. . ....M...... . .1..r £4.40.” a. .3:: .r .32....2 . 2......3... 33...... .3... .... .1; . . 2.._r....... .......,....«....wnuw... .<..I r). r57... .4. 1.... ....Z...,.4.............. ... .4.» 2.04.3... .7... . .5... l: I.... .... with. I MER :s‘io 3".- AN VEL BESS 'FQR PRO ... . .~ . . .. ... ‘v..a.l.\....pu»-u:.. T .7 w .. ., . w . .. w . . . . if... .3..fi...a!.¥.u.r ....\ _ . L, .., ,. .13. 5.31. c!!!» gavz- .. ......u_.u.......:...... 3K}... :2. 3: .. .. . : w.” 25.... . _ ... .. , . . .. . . .. .N..\e.fi.fi....£..fi :. . . .... .( . A - J . . .. flaw... MAHPII .5... .3}... .. . i in hi: This is to certify that the thesis entitled PART I. A NOVEL STEREOCHEMICAL REARRANGEMENT PROCESS FOR AN ISOMER OF TRIMETHYLSILICON ACETYLACETONATE, A SILYL ENOL ETHER PART II. DYNAMIC STEREOCHEMICAL PROPERTIES OF SOME CYCLOPENTADIENYLZIRCONIUM B-DIKETONATE COMPLEXES presented by Jerry J. Howe has been accepted towards fulfillment of the requirements for Ph .0. degree in ChemTStY'y My professor Date July 20, 1971 0-7639 ABSTRACT PART I A NOVEL STEREOCHEMICAL REARRANGEMENT PROCESS FOR AN ISOMER OF TRIMETHYLSILICON ACETYLACETONATE, A SILYL ENOL ETHER By Jerry J. Howe Trimethylsilicon acetylacetonate has been prepared from trimethylchlorosilane and acetylacetone as described in the literature.1 Nuclear magnetic resonance studies of trimethylsilicon acetylacetonate show that the compound possesses an open-chain enol ether structure which gives rise to configurations in which the uncoordinated carbonyl oxygen atom is positioned either gi§_or tran§_to the siloxy group. In chlorobenzene at room temperature the equilibrium value of the [gj§J/[trgnsj ratio is equal to 0.33. The gig isomer undergoes a rapid, intramolecular rearrangement process at room temperature which interchanges the nonequivalent allylic and acetyl methyl groups on the acetylacetonate ligand. First-order rate constants for the stereochemical rearrangement of the cj§_isomer in chlorobenzene solution were determined by proton nmr line-broadening techniques over the temperature range -36.2 to 38.4°. The values of the Arrhenius activation energy and frequency factor are l3.8 :_0.5 kcal/mol and exp(13.05 i 0.54), respectively. The tran§_isomer of trimethylsilicon acetylacetonate is stereochemically rigid at temperatures even as high as 120°. The difference in lability is Jerry J. Howe attributed to the ease of formation of a five-coordinated silicon intermediate or transition state. Comparison of gi§;(CH3)3Si(acac) with the relative rearrangement rates of the gi§;R(CH3)ZSi(acac) series (R = nyC4H9, C2H5, CH2=CH, CF3CH2CH2, and C6H5) plus gi§y(C6H6)2(CH3)Si(acac) supports a mechanism involving a pentacoordinate silicon intermediate or transition state. PART II DYNAMIC STEREOCHEMICAL PROPERTIES OF SOME CYCLOPENTADIENYLZIRCONIUM B-DIKETONATE COMPLEXES The eight-coordinate complex (n-C5H5)Zr(dpm)2Cl, where dpm = dipivaloylmethanate, exists in solution on the basis of an octahedron in which the n-CSH5 ring occupies one stereochemical position, the chlorine atom is positioned gj§_to the ring, and the oxygen atoms of the diketonate ligands occupy the remaining coordination sites. At elevated temperatures, the compound undergoes a kinetically first-order stereochemical rearrangement process which interchanges the nonequivalent environments of the diketonate ligands. The rate of ligand interchange is determined by nmr line-broadening techniques. Since the rate is found to be comparable to the rates of stereochemical rearrangement for analogous acetylacetonate complexes, the steric requirements of the terminal groups on the diketonate ligands are not an important factor in the activation process. An ionic complex, [(n—C5H5)Zr(dpm)2][SbCl6], has been prepared. The cation possesses a configuration based on a trigonal bipyramid in which the n-CSH5 ring occupies an axial vertex and the oxygen atoms Jerry J. Howe of the diketonate ligands are positioned at the remaining coordination sites. At room temperature, the nonequivalent diketonate ligands are rapidly interchanged on the nmr time scale. I§j§flB-diketonato)-n-cyclopentadienylzirconium complexes, which possess a stereochemistry based on a pentagonal bipyramid in which the center of the cyclopentadienyl ring occupies an axial vertex, undergo two distinct types of stereochemical rearrangement processes. The faster process (process I) interchanges the nonequivalent environments of the terminal groups on the equatorial diketonate ligands, and the slower process (process II) exchanges the equatorial ligands with the unique ligand spanning an equatorial-axial edge. An intramolecular mechanism operates in process I and first-order rate constants have been determined by nmr line-broadening techniques. In the (n-C5H5)Zr(dik)3 series, the lability increases in the order dpm < hfac < acac. Process II has been studied for the hexafluoro- acetylacetonate and acetylacetonate derivatives, (n-C5H5)Zr(hfac)3 and (n-C5H5)Zr(acac)3, with the latter complex being more labile. Plausible mechanisms are discussed for both processes. l. R. West, J. Amer. Chem. Soc., go, 3246 (l958). PART I A NOVEL STEREOCHEMICAL REARRANGEMENT PROCESS FOR AN ISOMER OF TRIMETHYLSILICON ACETYLACETONATE, A SILYL ENOL ETHER PART II DYNAMIC STEREOCHEMICAL PROPERTIES OF SOME CYCLOPENTADIENYLZIRCONIUM B-DIKETONATE COMPLEXES By Jerry JG Howe A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry T971 To Jennifer ii ACKNOWLEDGMENT I would like to extend my appreciation to Dr. Thomas J. Pinnavaia for his interest, guidance, probing discussions, and patience during this investigation. I also wish to thank the other members of my guidance committee, Dr. Carl H. Brubaker, Jr., Dr. Alexander I. P0pov, and Dr. Max T. Rogers for their interest in this research. Special thanks goes to Dr. Carl H. Brubaker, Jr., who served as my second reader, twice. I am deeply grateful to my wife, Jennifer, for her endless encouragement and constant love which have made these years as a graduate student very enjoyable. I wish to thank my parents, Mr. and Mrs. Roger F. Howe of Napakoneta, Ohio, for their continual encouragement during my educational pursuits. II. III. II. TABLE OF CONTENTS PART I INTRODUCTION .......................... l EXPERIMENTAL .......................... 7 A. Reagents and Solvents .............. . . . . . 7 B. Synthesis of 2-Trimethylsiloxy-Z-pentene-4—one . ...... 7 C. Analytical Data ...................... 8 D. Index of Refraction .................... 8 E. Molecular Weight Determination . . . . . . ......... 8 F. Vapor Phase Chromatography . . . . . . . .......... 9 G. Infrared Spectra . . .................... 9 H. Nuclear Magnetic Resonance Spectra . . . . ......... 9 I. Preparation of Solutions .................. 10 J. Computer Computations . . . ................ 10 RESULTS AND DISCUSSION . . . . ................. 12 A. Preparative Chemistry ................... l2 8. Characterization of Trimethylsilicon Acetylacetonate . . . .l2 C. Kinetic Study of Trimethylsilicon Acetylacetonate ..... 27 PART II INTRODUCTION . . . .............. . . . . ..... 4l EXPERIMENTAL .......................... 47 A. Reagents and Solvents ................... 47 B. General Synthetic Techniques ................ 49 C. Preparation of Compounds . . . . . . . . .......... 49 l. Chlorobis(2,2,6,6-tetramethyl-3,5-heptanedionato)- cyclopentadienylzirconium ............ . . . .49 2. Tris(l,l,l,5,5,5-hexafluoro-2,4-pentanedionato)- cyclopentadienylzirconium ................ 50 3. Tris(l,l,l-trifluoro-2,4-pentanedionato)cyclo- pentadienylzirconium .................. 5l 4. Tris(l,l,l-trifluoro-5,5-dimethyl-2,4-hexanedionato)- cyclOpentadienylzirconium ............... 5l a. From zirconocene dichloride in triethylamine . . ...5l b. From zirconocene dichloride in neat pivaloyltri- fluoroacetone ................... 52 iv 5. Tris(2,2,6,6-tetramethyl—3,5-heptanedionato)cyclopenta- dienylzirconium ................... 52 a. From tetrakiscyclOpentadienylzirconium ...... 52 b. From zirconocene dichloride in triethylamine . . . 53 6. Tris(2,4-pentanedionato)cyclopentadienylzirconium . . 53 a. From zirconocene dichloride in triethylamine . . . 53 b. From tetrakiscyclopentadienylzirconium ..... 54 7. Tris(l,3-diphenyl-l,3-propanedionato)cyclopenta- dienylzirconium ................... 54 8. Tris(5,5-dimethyl-2,4-hexanedionato)cyclopenta- dienylzirconium ................... 55 a. From zirconocene dichloride in triethylamine . . 55 b. From tetrakiscyclopentadienylzirconium ..... 55 9. Tris(l—phenyl-l,3-butanedionato)cyclopentadienyl- zirconium ...................... 56 lO. Bis(2,2,6,6-tetramethyl-3,5-heptanedionato)cyclopenta- dienylzirconium hexachloroantimonate ......... 56 D. Analytical Data ................... . . 57 E. Molecular Weight Determinations . . . . ......... 57 F. Conductance Measurements ................. 57 G. Melting Point Determinations ............... 58 H. Electron Paramagnetic Resonance Spectra ......... 58 I. Nuclear Magnetic Resonance Spectra . . .......... 58 J. Infrared Spectra . . . ................ . 58 K. Preparation of Solutions for Nmr Studies ....... . . 59 L. Determination of Mean Lifetimes . . . . ......... 59 RESULTS AND DISCUSSION ................. 6l A. Reactions for the Preparation of (n-C5H5)Zr(dpm)3Cl and (n-CsHs)Zr(dik)3 Complexes ................ 6l B. Constitution of (n-CsHs)Zr(dpm)2Cl and (n-C5H5)Zr(dik)3 Complexes in Solution .................. 63 C. Dynamic Stereochemical Pr0perties of (n-C5H5)Zr(dpm)2Cl . 64 D. Preparation and Pr0perties of (n-C5H5)Zr(dpm)2Cl-SbCl5 . . 82 E. Characterization of the (n-C5H5)Zr(dik)3 Complexes . . . . 9l l. Nmr Spectra .......... . . . . ....... 9l 2. Process I ...................... 98 3. Process II ................... . . . l23 BIBLIOGRAPHY ......................... l33 APPENDIX A. B. C. Nmr Chemical Shifts in Benzene_om do FE ooF\m w.- cowpmcpcmocoo “Asz oov om.mm pm corps—cm mcm~cmnocopco cw Aumomvwmmflmrov do Echomam LE: :ouoca .P mczmwd thm 7:1 Q. «.1524 I/g/rl 263 .w wo..m _O..m Ohm mud 0%.? WiJ O_u.1_n:2< /7 . r 7 0N u.1_n=2< 15 é—ZOHz—————> J1; 385° ________}+_______> Figure 2. Temperature dependence of the acetylacetonate methyl proton resonance lines of (CH3)3Si(acac); concentration 11.4 g/100 ml of chlorobenzene. Figure 3. 16 Spin-spin decoupling of trap§;(CH3)3Si(acac). A. The spectrum at 100.0 MHz. B. Irradiation of the sample at the -CH= frequency. C. Irradiation of the sample at the =CCH3 frequency. 0. Irradiation of the sample at the COCH3 frequency. The concentration of the solution is 11.4 g/100 ml of chlorobenzene. II II I I I 18 field (pf, Figure 2). The magnitude of the coupling constants for both the gi§_and t:gp§_isomers is in very good agreement with those previously reported for certain B—dicarbonyls.18 Chemical shifts in carbon tetrachloride, benzene, chloroform, and methylene chloride, and as the neat liquid for the acetylacetonate and Si-CH3 protons of (CH3)3Si(acac) are collected in Table I; shifts for (C6H5)(CH3)ZSi(acac) and (C6H5)2(CH3)Si(acac) in carbon tetrachloride are included for comparison. No significant concentration dependence was observed for (CH3)3Si(acac) in carbon tetrachloride over the range 3.0-20.0 g/100 ml of solvent. Since some concentration dependence was found in benzene, the Shifts in this solvent were extrapolated to infinite dilution. The concentration dependence of the chemical shifts in chloroform and methylene chloride was not investigated. In each traps isomer the magnitude of the coupling between the —CH= and COCH3 protons is m 0.6 Hz, and the allylic coupling constant is 0.4 Hz or less. It is to be noted that the relative chemical shifts for the COCH3 and =CCH3 protons of the trgp§_isomer are not in agreement with the empirical ”ene-one" rule of Anteunis and Schamp19 for assigning chemical Shifts of similar types of protons in B-diketone enol ethers. The “ene-one” rule States that for organic compounds of the type 0L 1 l RC=C—COR the "ene” protons (R-C=) resonate at lower field than the same protons in the "one" position (COR). However, at least three pieces of evidence can be cited in support of the assignments made here.(l) Replacement of alkyl groups on silicon by phenyl groups leads to 0.09-0.14 ppm upfield shifts for the =CCH protOnS, whereas the COCH3 and —CH= protons 3 19 .m mocwcmewmw .uwzcwp pmmzu .cowpstu mpwcwch op nmpmpoamcpxm mew mcmNCmn aw mpwwcm __mnwswhn .umpo: wmwzcmzpo mmmpc: pcm>Fom 60 FE oo_\m op mw cowpmcpcmucoo moov mw mczpmcmgamp ”mm:_m> Asaav 6 mm umpgogmc mpcwzm _F_om ecaoasou cmsomw m.u casemw mecca m.mmpm:0pmom_>umu< coow_wmo:mmcowcb-mcmcu use umflm LOI memo pewzm Pmuwemgu copocm .H m_QMH 20 Show little or no change in chemical shift. An examination of molecular models of the trans phenylsilyl derivatives reveals that reasonable configurations are possible in which the =CCH protons are within the 3 diamagnetic cone of a phenyl group, but that configurations which can lead to upfield shifts for the COCH3 protons without also appreciably influencing the chemical shifts of the -CH= protons are unlikely. (2) The upfield shifts for the Si-CH3 protons of both sis: and trsps: (CH3)3Si(acac) in benzene solution show that these protons experience the diamagnetic anisotropy of the benzene ring to a greater extent than the internal TMS reference. Apparently, a stereOSpecific solvent-solute association results from the interaction of the n-electrons on benzene and the Siloxy group. Indeed, similar stereospecific interactions between benzene and a variety of other types of solute molecules are well known.20 Such an interaction should be expected to lead to upfield shifts for the =CCH3 protons, which is indeed the result observed for the trsfls_isomer. (3) For (CH3)3Si(acac) in dichloromethane solution at -40°, where the COCH3 and =CCH3 resonances of the pjs_isomer are well resolved, the =CCH3 and COCH3 protons of the Ersps_isomer are deshielded by 0.12 ppm and shielded by 0.04 ppm, respectively, relative to the analogous protons of the pis_isomer. Deshielding of the =CCH3 protons in the Ersps_isomer is expected because of the paramagnetic anisotropic 21 Finally, it might be mentioned effect of the adjacent COCH3 group. that for (CH3)3Si(acac) at -40°, the magnitude of the allylic coupling is slightly greater in the sis_isomer (m 0.5 Hz) than in the EEQDE. isomer (< 0.3 Hz). Under the same conditions, the long-range coupling between the -CH= and COCH3 protons is smaller in the pjs_isomer (m 0.0 Hz) than is the trans isomer (m 0.5 Hz). The relative magnitudes of the allylic systems, yig,, that cisoid coupling is slightly larger than 21 transoid coupling.22 The coupling constants alone, however, would not constitute a reliable basis for the chemical shift assignments, because no relationship exists between allylic coupling constants and the stereochemistry of related a,B-unsaturated esters.23 Several unsuccessful attempts to separate the sjs_and srsps_isomers of (CH3)3Si(acac) by gas chromatography (sf, Experimental Section) or by vacuum distillation at 68° through a Spinning band column suggest that equilibrium is established readily between the two isomers. Facile isomerization is further supported by the fact that a freshly distilled sample contained the same ratio of isomers as a sample that had aged 6 months at room temperature. Also, after 12 hr, at 120° the pis:to- spsps_ratio (0.38 :_0.04) was equal, within the 95% confidence level estimated error, to the ratio observed at room temperature (0.34 :_0.04). Both the sample aged at room temperature and the sample heated at 120° should be at equilibrium because the g1s_methyl enol ether of acetylacetone is known to undergo conversion to "pure" sgsps_isomer within 8 days at ambient temperature and within 3 hr. at 100°.12 Thus, solutions of the compound in chlorobenzene were assumed to be at equilibrium after one week at room temperature, and the sis:to-t§sps_ ratio, Shown in Table II, was determined by planimetric integration of -CH= nmr lines. Other triorganosilicon acetylacetonates are included for comparison. With the exception of the phenylsilyl derivatives, the equilibrium amount of sis_isomer increases with increasing electron- withdrawing ability of the substituents on silicon. This relationship between the sisrto-spsps_ratios and the polarity of the silicon substituents suggests that a long-range electrostatic interaction may exist between silicon and the dangling carbonyl oxygen atom in the sis_ isomers. Such an interaction would also account, in part, for the 22 Table II. Equilibrium Ratio of sis:to-trans Enol Ether Isomers for Triorganosilicon Acetylacetonates.a Compound [cisl/[transl (p—C4H9)(CH3)ZSi(acac) 0.28b : 0.02c (C2H5)(CH3)ZSi(acaC) 0.29 :_0.02 (CH3)3Si(acac) 0.34 :_0.04 (CF3CH2CH2)(CH3)ZSi(acac) 0.39 i 0.03 (CH2=CH)(CH3)ZSi(acaC) 0.38 i 0.02 (C6H5)(CH3)ZSi(acac) 0.31 :_0.02 (C6H5)2(CH3)Si(acac) 0.25 i 0.02 aIn chlorobenzene solution at room temperature; concentration is 0.60 m, Compounds other than (CH3)38i(acac) from Reference 8. bAll values are averages of five Spectral copies. CErrors are estimated at the 95% confidence level. 23 enhanced stability of these pjs;triorganosilicon acetylacetonates relative to the sis_form of the methyl enol ether of acetylacetone. The sisrto- sssps_ratios for (C5H5)(CH3)25i(acac) and (C5H5)2(CH3)Si(acac) are lower than expected on the basis of inductive effects, but steric factors and ligand + metal pn-dn bonding could weaken the long-range silicon-oxygen interaction in the gis_isomers of these derivatives.8 It is noteworthy that a ground-state stereochemistry based on a trigonal bipyramid in which acetylacetonate acts as a bidentate ligand and spans an axial-equatorial edge cannot be assigned in place of the gis_isomer. A trigonal-bipyramidal structure of this type should give rise to two Si-CH3 lines in the absence of exchange. However, only one sharp Si-CH3 resonance line is observed for the labile isomer at -53.7°, which is the result predicted for the sis_enol ether structure. The ability of the sis_isomer to undergo rapid rearrangement which leads to environmental averaging of nonequivalent acetylacetonate methyl groups is attributed to the facile formation of a five-coordinated silicon intermediate. Five-coordinated silicon intermediates have also been postulated by Pike6b and Ainsworth.13 Two possible mechanisms for the rearrangement process are illustrated in Schemes I and II, Figure 4. In Scheme 1, the five-coordinate silicon intermediate possesses either a trigonal bipyramidal or a square pyramidal geometry in which both Si-O bonds are equivalent. This equivalence dictates equal probability for the rupture of either Si-O bond therefore permitting the environmental interchange of the two methyl acetylacetonate groups. Scheme 11 also utilizes a trigonal bipyramidal intermediate; however, the acetylacetonate ligand now spans an axial-equatorial edge. This is in accordance with two "preference" rules for trigonal bipyramidal Aomomvwmmfimzuvnmmm cw maaocm Pagpms mumcopmumpxumum pampm>wacmcoc ecu mo mmcmcuxm mcp com pcsooom v—zoz gown: mam—cocoa: .e mczmym nIU nIU m A U I ZOC© Io mocmvcmamu mczpwcwaewp mg» mzozm o m>gzu .>@ .mpcmcanoo mucmcommc mcp :wmzpmn cowpmcwamm zocmzcmgc .u m>c=u ”mucmcomwc fizgpme nmmmcm>6165wp 659 do spew: chF .m m>czo ”mocmCOmmL mzooo exp 40 spew: mCWF .< w>g=o “mcmNcmnoco_cu :w AomomvwmmAmIovumwm Io mcopocq fixgyme mumcopwomrxpoom wzw Low mcwpmamcwa wamgm-wcw_ LE: 40 mucmucmamc oczpmgmaewh .m manure 29 VTI‘T'TIVIIrTIIIImITITI¢l—r1rl] .m mcamwd 8 L h i 8. on on 0' ON 0 ON- 0?. 00. 411 4 _ 41 _ 41 _ . d . d 41 _ . _ 4 _ . _ . l1 1 L od 9] .9, (300 <13.29 >12.6 (CF3CH2CH2)(CH3)2Si(acac) ~12. 621 :_118 12.2 aIn chlorobenzene solution; 0.60 9, total concentration of R3Si(acac) is bTemperature at which k was determined. cBasis for estimates of error are described in Reference 8. dExtrapolated value from Arrhenius plot. eResults other than those for (CH3)3Si(acac) are from Reference 8. 39 Similar rearrangement phenomena with other diketonate or pseudo- 8’13’35 Unfortunately, those diketonate ligands have been observed. studies are not detailed enough to permit a quantitative comparison to the sis;(CH3)3Si(acac) results. Collins8 reports that the 1,1,1,5,5,5—hexaf1u0ro-2,4-pentanedione (hfac) and 2,2,6,6-tetramethy1-3,5-heptanedione (dpm) derivatives of (CH3)3Si(dik) exist solely as the spsps_and gjs_isomers, respectively. He rationalizes the sis:t:gps_distribution by the relative long-range interaction of the nonbonded carbonyl oxygen atoms in the two ligands. The relative negative charges on the free carbonyl Oxygen atoms, based on inductive effects, would be hfac < acac < dpm. Hence, if only the gis_isomers are capable of interaction between the silicon and nonbonded carbonyl oxygen atoms, the sjs;§rpps_distributions are rational. Finally, Collins states that even at temperatures as low as -95°C, gisICH3)3Si(dpm) undergoes a very fast stereochemical rearrangement process analogous to that of sisflCH3)3Si(acac). Ainsworth's data13 for several ketene methyl trimethylsilyl acetals suggest stereochemical rearrangements, but they are insufficient to permit even a qualitative comparison with the present work. Shvo35 found a free energy of activation (15.4 :_0.5 kcal/mol) for the analogous process in the triacetylmethane derivative. This value is to be compared with an Arrhenius activation energy of 13.8 :_O.5 kcal/mol for the similar process in sisfiCH3)3Si(acac). The parameters Ea and AGi are related through AH1 by the equations 80* = AHI - TAS _ 1 38 and Ea — AH + RT. 40 Therefore, for the rearrangement process of sis;(CH3)3Si(acac), Ea * AG¢ since A5256 is essentially zero (-0.8 :_2.5 eu) and RT at 25° is very small (0.6 kcal/mol). The Taft constant for the acetyl group (0* = 0.87439) indicates that the acetyl moiety functions as an electron-withdrawing substituent relative to the hydrogen atom. Therefore, the sis ketonic oxygen atom of the triacetylmethanate derivative would be less nucleophilic than the analogous oxygen atom of the acetylacetonate compound, and would account for the higher activation energy of the triacetylmethanate enol ether if bond—making is an important Step in the activation process. However, the experimental technique and the basis for the :0.5 kcal/mol uncertainty are inadequately described for the triacetylmethanate complex, and the error inherrent in such a determination of AG* fr0m a Single experimental temperature prohibit placing any significance on the absolute difference between the activation energies. In conclusion, all data support the mechanisms proposed in this work and suggest that a bond—making process is an important step in the stereochemical rearrangement of these comp0unds. I. INTRODUCTION The history of B-diketonate coordination can be traced back to 1894 with the preparation of Be(C5H702)2.40 Since then the growth of this area of organometallic chemistry has been phenomenal. Compounds containing acetylacetone and other B-diketones as ligands are known for almost every metallic or semi-metallic element. Activity in this area continues to be very brisk, as verified by the appearance of six review article52’40-44 in the last six years. The organometallic chemistry of cyclopentadiene began shortly after the arrival of the twentieth century with the preparation of potassium cyclopentadienide from cyclopentadiene and potassium metal.45 Although other group IA and IIA cyclopentadienides were prepared subsequently, the current interest in metal derivatives of cyclopentadiene stems from 46 and the independent discovery of ferrocene by Kealy and Pauson, Miller, Tebboth and Tremaine47 in 1951. Since the synthesis and characterization of ferrocene, cyclopentadiene derivatives have been prepared for almost every transition and post—transition metal. Interest has been so keen that individual investigators have found it nearly impossible to review the entire field, but the reviews of several authors48—5] do provide a good overview of the area. It was inevitable that the areas of B-diketonate and cyclopentadienyl organometallic chemistry would soon meet. In 1956 Thomas52 prepared a chromium compound of the formula C5H5Cr(05H702)Br. The compound was 41 42 obtained in 3% yield by the reaction of chr0mium(III) acetylacetonate and cyclopentadienylmagnesium bromide. Thomas suggested that an entire class of compounds of this type should exist. However, not until 1961 was another example of a cyclopentadienylmetal B-diketonate reported. Brainina and her coworkers53 prepared chlorobis(2,4—pentane- dionato)cyclopentadienylzirconium, (n-CSH5)Zr(acac)2C1, by the reaction of zirconocene dichloride and the neat diketone and, also, by the metathesis reaction of Zr(acac)2C12 and Na(C5H5) in tetrahydrofuran. Additional means of preparation which were discovered later by Brainina54’55 involve a redistribution reaction between (n—C5H5)2Zr612 and Zr(acac)4, and the reaction of [(n-C5H5)ZrC12]20 or (n-C5H5)Zr(OC2H5)2C1 and the neat diketone. The reaction of (n-C5H5)22rC12 and H(acac) is preferred as the experimental technique is simple and the product is obtained in high yield. All of Brainina's synthetic methods, however, have been used56’58’61’65 to prepare a variety of cyclopentadienylzirconium B-diketonate complexes of the general type (n-C5H4R)Zr(dik)2X, where R may be H, CH3, or tgssfc4H9, and X = Cl, Br, or OCOCH3. Similar hafnium derivatives have also been prepared by analogous procedures.58’61 More recently, Newton62 has modified Brainina's original method for the preparation of (n-C5H5)Zr(acac)201 by carrying out the reaction of (n-C5H5)2Zr612 and H(acac) in the presence of triethylamine. The addition of a base apparently represents a significant improvement as the method is applicable to a wider variety of diketonate derivatives including tropolonate. Newton has also succeeded in preparing the first cyclopentadienyltitanium derivative, (n-C5H5)Ti(acac)2Cl. A hexafluoroacetylacetonate complex, (n—CSH5)Ti(hfac)2Cl, has been reported,63 but no preparative details were given. 43 Brainina has also discovered two additional types of cyclopenta— dienylzirconium B-diketonates which have the general empirical formulas [(n—CSH5)Zr(dik)2]20 and (n-C5H5)Zr(dik)3. The zirconoxanes, which contain a Zr-O-Zr linkage, were prepared by hydrolysis of (n-C5H5)Zr(dik)2X complexes in the presence of triethylamine and ethanol. The (n-CSH5)Zr(dik)3 complexes were obtained by the reacti0n of the highly 66 reactive compound (C5H5)4Zr and the free diketone. Some fluorinated B-diketonate analogues of the latter type of complex have been recently reported by Graham t al.,63 presumably by a method different from that used by Brainina, but the experimental procedure has not yet been described. Three transition metals have been found to form cationic complexes of the type [(n-CSH5)2M(dik)]+, where M = Ti, V, or Mo.66’67 These complex cations are formed by the reaction of the metallocene dichloride and free diketone in aqueous solution and are isolated as salts of large uninegative anions, such as C104', BF4', or PF6'. A neutral complex with squarate, (n-C5H5)2Ti(C4O4), has also been isolated.68 Since the initiation of this study, the structures of two cyclopentadienyl metal B-diketonates, 113,, (n—C5H5)Zr(acac)2Cl and (n-C5H5)Zr(hfac)3, have been determined by X—ray diffraction methods.59’69’64 The stereochemistry of (n-C H )Zr(acac) Cl is best described in terms 5 5 2 of a D dodecahedron in which the (w-CSHS) occupies a triangular face, 2d the chlorine is positioned at a vertex sis to the cyclopentadienyl group, and the donor oxygen atoms of the bidentate acac ligands occupy the remaining coordination position. A simpler description of the molecule can be formulated in terms of an octahedral coordination polyhedron in which the center of the 65H5 ring is at a vertex sis to the chlorine atom. In (n-CSH5)Zr(hfac)3 the center of the C5H5 44 group is at an axial vertex of a pentagonal bipyramid. Although it appears acceptable from a stereochemical viewpoint to consider the 05H5 group as occupying a Single coordination position of higher coordination number polyhedrons, it is likely that three metal orbitals are used in bonding to the C5H5 group. Thus it is preferable to regard (n-CsH5)Zr(acac)2C1 and (n-C5H5)Zr(dik)3 as being eight- and nine-coordinate complexes, respectively. No definitive structural data are available for other cyclopentadienyl metal n—diketonates, but + (n-C5H5)Cr(acac)Br and the [(n-C5H5)2M(dik)] complexes are believed to possess stereochemistries in which the center of the 65H5 groups are at tetrahedral vertices.52’66’67 Recent nmr studies in this laboratory70’71 have shown that in solution (n—C5H5)Zr(acac)2C1 adopts the same type of stereochemistry that is found in the solid state. Moreover, the complex undergoes a rapid Stereochemical rearrangement which interchanges the nonequivalent acac ligands. Similar properties were observed for the hafnium and bromine analogues, (n-C5H5)Hf(acac)2C1 and (n—CSH5)Zr(acac)2Br. Graham, §t_gl,,63 subsequently showed that (n-C5H5)Zr(hfac)3 is also nonrigid but that the molecule undergoes two discrete rearrangement processes. The faster process (process 1) results in the interconversion of nonequivalent environments of the CF3 terminal groups on the two diketonate ligands which span the equatorial edges of the pentagonal bipyramid. In the second process (process II), the unique diketonate ligand which spans an axial-equatorial edge undergoes exchange with the two equatorial ligands. The cyclopentadienyl ring plays an important role in influencing the stereochemical lability of these higher coordination number compounds. They represent the first examples of 45 higher coordination number metal chelates for which dynamic stereo— chemical information can be obtained. The objectives of this work were (1) to investigate the stereochemical lability of a new (n—C5H5)Zr(dik)2X complex with dipivaloylmethanate and (2) to Study quantitatively the kinetics and possible mechanisms for the rearrangement processes of type I and II for a series of (n—C5H5)Zr(dik)3 complexes. There is reason to suspect that the steric requirement of the ssstfc4H9 gr0ups in (n—CSH5)Zr(dpm)201, may cause the stereochemistry of the molecule in solution to be much more complex than that observed for the acac analogues. Butler70 has pointed out that a D dodecahedral coordination polyhedron can lead 2d to as many as 12 geometric isomers with C1 symmetry for a complex of the general type (w-C5H5)M(dik)2X. Although (n-C5H5)Zr(acac)2Cl exhibits only one C] isomer in solution, (w—C5H5)Ti(hfac)201 is reported to exhibit two such isomers. Clearly, the gis;octahedral formalism is adequate for describing the stereochemistry of the former complex but not of the latter. It was hoped that (n—S-H5)Zr(dpm)201 would exhibit two or more C1 isomers in solution and that an investigation of their stereochemical labilities would provide mechanistic information. The second objective of the work involved preparing several new (n-C5H5)Zr(dik)3 derivatives and, also, a new complex cation of the type [(n-CSH5)Zr(dik)2]+. Diketonate abbreviations used throughout this section of the thesis are summarized for ready reference in Table V. 46 Table V. Abbreviations and Formulas for Some B-Diketonate Ligands. Abbreviation Ligand Formula acac acetylacetonate [H3CCOCHCOCH3J‘ bzac benzoylacetonate [H3CCOCHCOC6H5]' bzbz dibenzoylmethanate [H5C6COCHCOC6HSJ' dpm dipivaloylmethanate [H9C4COCHC0C4H9]' pvac pivaloylacetonate [H3CCOCHCOC4H9J' tfac trifluoroacetylacetonate [F3CCOCHCOCH3]' hfac hexafluoroacetylacetonate [F3CCOCHCOCF3J' pvtf pivaloyltrifluoroacetonate [H964C0CHC0CF3]' dik any diketonate [R'COCHCOR]' II. EXPERIMENTAL A. Reagents and Solvents Zirconocene dichloride was purchased from Aldrich Chemical Co., Inc., and used without further purification. Acetylacetone was purchased from Matheson, Coleman & Bell and was fractionally distilled before use (bp l36—l40°). Trifluoroacetylacetone was purchased from Columbia Organic Chemical Co. and was fractionally distilled before use (bp 107°). Benzoylacetone and dibenzoylmethane (Eastman Organic) were used without further purification. Hexafluoro- acetylacetone, dipivaloylmethane pivaloylacetone, and pivaloyltrifluoro- acetone were prepared by slight modification of previously described procedures.72'75 Tetrakiscyclopentadienylzirconium was prepared according to the method described by Brainina and coworkers.76’77 Since the compound does not melt sharply, its purity was checked by chemical analysis, a molecular weight determination, and nmr spectroscopy. Apsl, Calcd for (C5H4)4Zr: C, 68.32; H, 5.73; mol wt, 352. Found: C, 67.70; H, 5.50; mol wt (C6H6), 350. Nmr (saturated solution in CDC13 at 40°) 6 5.83 (lit.77 6 5.75). Antimony pentachloride (Matheson, Coleman & Bell) was distilled under reduced pressure through a 30 cm column packed with glass helices. The pressure was adjusted to 30 torr by controlling a flow of dry argon through the system. The fraction boiling at 90-91° was collected. 47 48 The antimony pentachloride was subsequently manipulated in an argon or dry nitrogen atmosphere and stored in the dark. Chlorobis(2,4-pentanedionato)cyclopentadienylzirconium and bromobis— (2,4-pentanedionato)cyclopentadienylzirconium were obtained from E. D. Butler and A. L. Lott respectively. Tetrakis(2,2,6,6—tetramethy1- 3,5-heptanedionato)zirconium was obtained from J. T. Woodard. All were used without further purification. All organic solvents used in the synthesis and/or studies of the compounds were dried over suitable dessicants. Benzene (Mallinckrodt), hexane (Matheson, Coleman & Bell), diethyl ether (Fisher), diisopropylether (Fisher), di-pfbutylether (Fisher), gyxylene (Fisher),toluene (Fisher), and tetrahydrofuran (Matheson, Coleman & Bell) were distilled from lithium aluminum hydride. Dichloromethane (Matheson, Coleman & Bell), acetonitrile (Matheson, Coleman & Bell), and tetrachloroethylene (Matheson, Coleman & Bell) were distilled from calcium hydride. Carbon disulfide (Fisher) and triethylamine (Eastman Organic) were dried over phosphorus pentoxide and anhydrous magnesium sulfate respectively. Acetone (Fisher Certified, A.C.S.) was distilled from over Drierite. Fisher Certified A.C.S. nitrobenzene was purified by distilling it ip_ygspg_from over phosphorus pentoxide. The distilled nitrobenzene was allowed to stand over Linde Type 5A molecular sieves for two days and then redistilled ip_ysspp_from the molecular sieves. Only one fraction of nitrobenzene was collected, bp 50-53° (gs, 0.5 torr). Only benzene-d6 (Diaprep) and chloroform-d1 (Diaprep) were used without further purification. 49 B. General Synthetic Techniques The compounds used in this study are sensitive to atmospheric moisture. Therefore, all preparative reactions and manipulations of the products were conducted under a dry argon atmosphere [(n—C5H5)Zr(dik)3 compounds] or under nitrogen [(n-C5H5)Zr(dik)2X compounds]. As an extra precaution against hydrolysis in the preparation and manipulation of the (n-C5H5)Zr(dik)3 compounds, solvents were deaerated by dispersing argon through the liquid for several minutes. All glassware was dried at 175°, cooled in a calcium sulfate desiccator whenever possible, and flushed with argon before use. Filtrations were carried out in Sintered glass Buchner funnels equipped with ground glass joints. Ground glass joints were usually greased with silicone grease. The compounds were generally more stable towards hydrolysis by atmospheric moisture as solids than in solution. Among the (n-C5H5)Zr(dik)3 compounds, only the hexafluoroacetylacetonate derivative could be safely stored in a screw-cap vial fitted with a Teflon cap liner. All other (n-C5H5)Zr(dik)3 complexes were stored under vacuum. In general, the (n-C5H5)Zr(dik)2X complexes were less sensitive to atmospheric moisture than were the (n-C H )Zr(dik)3 complexes, and they 5 5 were stored in screw-cap vials in a calcium sulfate desiccator. C. Preparation of Compounds 1. Chlorobis(2,2.6,6—tetramethy1-3,5—heptanedionato)cyclopentadienyl— zirc0nium. A mixture of dipivaloylmethane (4.97 g, 27.0 nmol) and triethylamine (1.88 ml, 13.4 mmol) in 90 m1 acetonitrile was added dropwise to a suspension of (w—C5H5)2ZrCl2 (3.92 g, 13.4 mmol) in 100 m1 acetonitrile. The mixture was stirred at ambient temperature 50 for 39 hr. The white insoluble solid (identified as (n-C5H5)Zr(dpm)3 by its nmr spectrum) was separated by filtration washed with four 30—ml portions of acetonitrile, the acetonitrile washings were combined with the original mother liquor, and the solvent was removed under vacuum. The resultant white solid was suspended in 100 ml benzene, and the insoluble portion (identified as (C H ) 2 5 3 Spectrum) was removed by filtration. The product was isolated from N'HC1 by its 1r the filtrate by vacuum distillation of the benzene at room temperature. The white compound was recrystallized from 200 ml acetonitrile at —78° and was dried ip_ys£p9_at room temperature; mp l75—l77°. The yield was 5.76 g (76.8%). A311. Calcd for (C5H5)Zr(C]]H]902)2Cl: C, 58.09; H, 7.76; Cl, 6.35; mol wt, 558. Found: C, 57.93; H, 7.64; Cl, 6.37; mol wt (C6H6), 579; A (1.84 X 10'3 M_Solution in C6H5N02 at 25°), 0.151 1 2 -1 ohm' cm mol 2. Tris(l,1,l,5,5,5—hexafluoro-2,4-pentanedionato)cyclopentadienyl— zirconium. A mixture of (n—C5H5)2ZrC12 (10.0 g, 34.2 mmol) and hexafluoroacetylacetone (25 ml, 170 nmol) was heated at reflux temperature until all of the (w—C5H5)22rC12 dissolved (ss. 11 hr). The excess hexafluoroacetylacetone was removed by vacuum distillation at room temperature to obtain a yellow mass. The mass was dissolved in a minimum amount of dichloromethane at room temperature and the solution was cooled in a Dry Ice—acetone bath. The resulting yellow crystals were separated from the cold solution by filtration and dried ifl_vssgg at room temperature; mp 92-95°. The yield was 26.3 g (98.8%). The compound was recrystallized from dibutyl ether; mp 98-100°. Apgl, Calcd for (C5H5)Zr(C5H02F6)3: C, 30.88; H, 1.11; F, 43.95; mol wt, 777. Found: C, 31.01; H, 1.04; F, 44.16; mol wt (C6H6), 51 790; A (1.45 X 10'3 M_solution in C6H5N02 at 25°), 0.014 ohm-1 cm2 mol—1. 3. Tris(l,1,l-trifluoro-2,4-pentanedionato)cyclopentadieny1zirconium. A mixture of (n—C5H5)ZZrC12 (2.50 g, 8.56 mmol) and trifluoroacetyl— acetone (20 ml, 160 nmol) was heated just below reflux temperature for 1 hr. The reactiOn mixture was cooled to room temperature and 50 m1 of diethyl ether was added. The green-brown solution was evaporated to dryness at room temperature under reduced pressure. The residue gave a white crystalline product upon recrystallization from benzene- hexane and subsequent vacuum sublimation at 70°; mp 87-90°. The yield was 2.25 g (42.7%). Apgl. Calcd for (C5H5)Zr(C5H402F3)3: C, 39.03; H, 2.78; F, 27.78; Zr, 14.82; mol wt, 615. Found: C, 38.89; H, 2.76; F, 27.60; Zr, 14.83; mol wt (C6H6), 630; A (1.79 x 10'3 M_solution in C6H5N02 at 25°), 0.028 ohm"1 cm2 mo1‘1. 4. Tris(l,1,l—trifluoro-5,5—dimethyl-2,4-hexanedionato)zirconium. a. From zirconocene dichloride in triethylamine. Pivaloyltri— fluoroacetone (4.52 g, 23.1 mmol) was added to a suspension of (n-C5H5)22r012 in 45 m1 triethylamine, and the mixture was stirred at room temperature for 10 hr. After the white precipitate [(C2H5)3N-HC1] had been separated from the mixture by filtration, the volume of the solution was reduced to ss, 10 ml, and the solution was cooled in a Dry Ice-acetone bath. The resultant white crystalline solid was collected by filtration and dried ip_vsgps_at room temperature; mp 115-118°. The yield was 2.97 g (52.0%). Apgl, Calcd for (C5H5)Zr(C8H]0F302)3: C, 46.96; H, 4.76; Zr, 12.30; mol wt, 742. Found: C, 47.04; H, 4.84; Zr, 12.39; 52 mol wt (C6H6), 752; A (7.81 x 10‘3 M_Solution in C6H5N02 at 25°), <0.01o ohm‘1 cm2 mol-1. b. From zirconocene in neat pivaloyltrifluoroacetone. A mixture of (n—CSH5)22rC12 (0.64 g, 2.19 mmol) and pivaloyltrifluoroacetone (5.00 ml, 28.8 mmol) was heated at 100° for 6 hr. The excess pivaloyltrifluoroacetone was removed by vacuum distillation at room temperature, yielding a brown mass. Two vacuum Sublimations at 75° produced a white crystalline solid, mp 102-106°. The yield was 1.52 g (99.3%). The product was identified as (n-C5H5)Zr(C8H]0F302)3 by its nmr spectrum in benzene. Since a few minor, unidentifiable impurities were also observed in the nmr Spectrum, this product was not used for any physical measurements. 5. Tris(2,2,6,6-tetramethyl-3,5—hgptanedionato)cyclopentadienylzirconium. a. From tetrakiscyclopentadienylzirc0nium. Dipivaloylmethane (7.40 g, 40.2 mmol) was added to a solution of (C5H5)4 13.4 mmol) in 150 ml of dichloromethane, and the mixture was heated at ZY‘ (4.71 g: reflux temperature for 20 hr. After the solvent was removed by vacuum distillation at room temperature, the residue was slurried in 150 ml of boiling hexane and the solution was filtered. The nmr spectrum of the hexane-insoluble fraction of the residue (1.48 g) was identical with that of (C5H5)4Zr. The volume of the hexane filtrate was reduced under reduced pressure, giving 4.46 g of off-white crystals. The yield based on the amount of (C5H5)4Zr that had reacted is 62.4%. A second recrystallization of the product from acetone gave colorless crystals. The crystals were dried is 19E92.at 80° for 4 hr; mp 3 230° dec. Apsl. Calcd for (65H5)Zr(C]]H1902)3: C, 64.64; H, 8.85; Zr, 13.17; mol wt, 706. Found: C. 64.86; H, 9.01; Zr, 12.92; mol wt 53 (C6H6), 720; A (1.25 x 10'3 M solution in C6H5N02 at 25°), 0.059 ohm'1 cm2 mol']. 6. From zirconocene dichloride in triethylamine. Dipivaloyl- methane (5.67 g 30.8 mmol) was added to a suspension of (n-C5H5)22rC12 (3.01 g, 10.3 mmol) in 35 ml triethylamine, and the mixture was stirred for 5 hr at ambient temperature. The insoluble white precipitate [(C2H5)3N-HC1] was removed by filtration, and the filtrate was cooled in a Dry Ice—acetone bath. The resultant white crystals (2.42 g) were separated from the cold solution by filtration and dried jp_yssps at room temperature for several hours; mp 238-240° dec. The remainder of the solvent was removed under reduced pressure, yielding an additional 0.59 g of the product. A third fraction of the compound (0.98 g) was extracted from the (C2H5)3N-HC1 precipitate with anhydrous diethyl ether. The total yield was 3.99 g (55.2%). All three fractions of the product were identified by nmr as (n-C5H5)Zr(C11H1902)3 free of proton-containing impurities. 6. Tris(2,4-pentanedionato)cyclopentadienylzirconium. a. From zirconocene dichloride in triethylamine. A suspension of (n-C5H5)22rCl (5.00 g, 17.2 mmol) and acetylacetone (5.25 ml, 2 51.3 mmol) in 50 m1 triethylamine was stirred at room temperature for 6 hr. The white precipitate of (C2H5)3N'HC1 was separated from the solution by filtration, and the filtrate was cooled to -25° overnight. The resultant white, crystalline solid was collected by filtration and dried ip_!sspp_at room temperature for several hours; mp 135-137°. The yield was 4.65 g (59.7%). Cooling the mother liquor in a Dry Ice-acetone bath resulted in no further crystallization. 54 mg. Calcd for (65H5)Zr(CSH702)3: c, 52.96; H, 5.78; Zr, 20.11; mol wt, 454. Found: C, 53.11; H, 5.80; Zr, 20.44; mol wt (C6H6), 448; A (2.65 x 10‘3 p solution in C6H5N02 at 25°), <0.029 ohm'1 cm2 mo1". b. From tetrakiscyclopentadienylzirconium. A mixture of (C5H5)4Zr (6.09 g, 17.4 mmol) and acetylacetone (5.39 ml, 55.3 mmol) in 100 ml benzene was stirred at ambient temperature for 2 hr. The solvent was removed by vacuum distillation at 50°. The resultant brown oil was dissolved in ss, 75 ml of hot hexane, and the solution was cooled to —25°. A heterogeneous precipitate formed over the period of several days. The majority of the precipitate was an amorphous brown solid, but it was covered with spherulites of off—white crystals. The cold mother liquor was decanted from the heterogeneous mass, and the heterogeneous product was dried ip_vssp9_at room temperature overnight. The crystalline spherulites were manually separated from the parent mass; mp l33—136° dec. The yield was pg, 1.0 9 (pg, 13%). The product was identified as (n-C5H5)Zr(C5H702)3 free of proton-containing impurities by its nmr spectrum in benzene. Attempts to recrystallize the compound from hexane were unsuccessful. 7. Tris(l,3—dipheflyl—l,3—prppanedionato)cyc1opentadigpylzirconium. A mixture of (C5H5)4Zr (0.94 g, 2.68 mmol) and dibenzoylmethane (1.79 g. 8.00 mmol) in 125 ml benzene was heated at reflux temperature for 1 hr. The reaction mixture was cooled to room temperature, and the solvent was removed by vacuum distillation. The resultant yellow solid was recrystallized from a 1:1 benzenezhexane solution and dried ip_vs£pp_at room temperature; mp 189—191° (lit.66 186-187°). The yield was 1.21 g (54.8%). Assl. Calcd for (C5H5)Zr(C15H1302)3: C, 72.70; H, 4.63; mol wt, 55 826. Found: C, 72.82; H, 4.70; mol wt (C6H6), 840; A (0.95 X 10'3 M solution in C6H5N02 at 25°), 0.063 ohm" cm2 mo1-1. 8. Tris(5,5-dimethy1-2,4-hexanedionato)cyclopentadienylzirconium. a. From zirconocene dichloride in triethylamine. A mixture of (n—65H5)22r612 (2.00 g, 6.84 mmol) and pivaloylacetone (2.92 g, 20.5 mmol) in 25 ml triethylamine was stirred at room temperature for 5 hr. The (C2H5)3N'HC1 was separated by filtration, and the solvent was removed by vacuum distillation at ambient temperature. The resultant white solid was recrystallized from acetone at —78° (Dry Ice—acetone bath), separated from the cold solution by filtration, and dried ip_vsspg at room temperature overnight; mp 149—152°. The yield was 2.96 g (74.4%). Apsl. Calcd for (C5H5)Zr(C8H1302)3: C, 60.07; H, 7.65; Zr, 15.73; mol wt, 580. Found: C, 59.88; H, 7.77; Zr, 15.95; mol wt (C6H6), 557; A (1.93 x 10‘3 M_solution in C6H5N02 at 25°),<0.040 ohm‘1 cm2 mo1'1. D. From tetrakiscyclopentadienylzirconium. Tetrakiscyclo- pentadienylzirconium (1.35 g, 3.84 mmol) was added to a solution of pivaloylacetone (1.62 g, 11.5 mmol) in 50 ml benzene, and the mixture was Stirred at ambient temperature for 16 hr. The solvent was removed by vacuum distillation at room temperature. The reSultant white solid was recrystallized from 20 ml hot hexane, separated by filtration, and dried is ypspg_at room temperature for several hours; mp 150-153°. The yield was 0.80 g (35.9%). The compound was identified as (n-C5H5)Zr(C8H1302)3 free of proton—containing impurities from its nmr spectrum in benzene. 56 9. Tris(l:phenyl-l,3-butanedionato)cyclopentadienylzirconium. A suspension of (n-05H5)22rC12 (2.24 g, 7.67 mmol) and benzoylacetone (3.73 g, 23.0 mmol) in 40 ml triethylamine was stirred at room temperature for 3.5 hr. The white precipitate was separated from the solution by filtration and washed with three 25—ml portions of benzene. The volume of the combined benzene washings was reduced to gs, 25 ml by vacuum distillation at ambient temperature. Hexane (gg, 50 ml) was added to commence precipitation, and the solution was cooled in a Dry Ice-acetone bath. The resulting white powder was separated from the cold mixture by filtration and dried ip_ygggg_at room temperature; mp l44—l47° (lit.66 141—142°). The amine mother liquor was removed by vacuum distillation at room temperature, but no additional product was obtained. Apgl, Calcd for (C5H5)Zr(C]OH1002)3: C, 65.70; H, 5.04; Zr, 14.26; mol wt, 640. Found: C, 65.61; H, 5.24; Zr, 14.15; mol wt (C6H6), 624; A (2.31 X 10'3 M_solution in C6H5N02 at 25°), 0.052 ohm'] cm2 mol'I. 10. Bis(2,2,6.6—tetramethy1—3,5-heptanedionato)cyclopentadienyl— zirconium hexachloroantimonate. A solution of SbCl 1.34 g, 4.48 mmol) 5 ( in 30 ml acetonitrile was added dropwise to a suspensiOn of (n—C5H5)Zr(C]]H1902)2Cl (2.50 g, 4.48 mmol) 1n 20 m1 aceton1tr1le at 0 , and the mixture was stirred at 0° for 0.5 hr. The volume of the solution was reduced to gs, 30 ml by vacuum distillation at 0°, and the solution was cooled to -25° for 12 hr. The resultant yellow crystalline solid was separated from the cold solution by filtration and dried jp_vacuo for 3 hr at Dry Ice temperature. The yellow crystals slowly decompose at room temperature to give a black oil. The product was insoluble 57 in benzene but was readily soluble in polar solvents such as acetonitrile, acetone, dichloromethane, and nitrobenzene. Aggl, Calcd for (C5H5)Zr(C]1H1902)25bCl6: C, 37.83; H, 5.06; (:1, 24.81. Found: c, 39.06; H, 5.20; (:1, 25.07; A (2.81 x10'3 g in C6H5N02 at 25°), 18.6 ohm'1 cm2 mol']. Ir and nmr data (gf, Results and Discussion, Section 0) indicate the presence of CH3CN; hence, calcd for (C5H5)Zr(C]1H1902)25bC16(CH3CN): C, 38.77; H, 5.16; C1, 23.68. D. Analytical Data The microanalyses of all compounds were performed by Galbraith Laboratories, Inc., Knoxville, Tennessee. E. Molecular Weight Determinations The molecular weights of the compounds were determined cryoscopically in dry benzene. Recrystallized benzil was used as the calibrating solute in the determination of the molal freezing point depression constant of benzene (5.38° mfI). Freezing point depressions were measured in the concentration range 0.030 to 0.057 m_with a Beckman differential thermometer graduated at 0.01° intervals. Temperature readings, however, were estimated to :0.001° with the aid of a magnifying thermometer reader. F. Conductance Measurements Molar conductivities were measured with a Wayne Kerr Model 8221 universal bridge and a FreaS—type solution cell with bright platimum electrodes. The cell constant was determined to be 0.194 cm'] at 25° by using a standard KCl solution. Purified nitrobenzene was used for molar conductance measurements. The conductivity cell was also used for conductometric titrations. 58 G. Melting Point Determinations Melting points were determined in sealed glass capillaries with a Thomas-Hoover 6406-H capillary melting point apparatus. H. Electron Paramagnetic Resonance Spectra The absence of paramagnetic impurities in the compounds were verified with a Varian E/4 spectrometer. Spectra were taken at room temperature and 77°K in sealed quartz-glass tubing. High gain settings and a 1500 to 5500 gauss scan range produced no evidence of any paramagnetic species in the samples. 1. Nuclear Magnetic Resonance Spectra Nuclear magnetic resonance spectra were recorded on a Varian 1H spectra) or A56/6OD analytical spectrometer operated at 60.0 ( 56.4 MHz (19F spectra). The probe temperature was controlled to within :0.5° by a Varian V-6040 variable-temperature controller. The methanol and ethylene glycol nmr thermometers described by Van Geet14 were used to determine the probe temperature. Magnetic field-sweep widths were calibrated by the audiofrequency side—band technique. All spectra were recorded at low radio frequency field strengths in order to avoid saturation effects. Occassionally, a Varian A/60 or a Varian T/60 analytical spectrometer were used to obtain routine spectra. J. Infrared Spectra Infrared spectra over the range 4000 to 250 cm‘1 were recorded on a Perkin-Elmer 457 grating infrared spectrophotometer. Spectra were obtained as nujol or fluorolube mulls between CsI plates. The 2851, 1603 and 907 cm'] peaks of polystyrene were used as reference frequencies. 59 K. Preparation of Solutions for Nmr Studies All solutions used in the nmr studies were prepared in an argon atmosphere. DiisOpropyl ether and g;xylene were used as solvents for (n-C5H5)Zr(hfac)3 and (n-C5H5)Zr(dpm)3, respectively. In the selection of a solvent for studying the exchange of terminal CF3 groups on the equatorial ligands of (A-C5H5)Zr(hfac)3, the following solvents were rejected, because of the low solubility of the complex or extensive overlapping of the CF3 resonances at 56.4 MHz: dichloromethane, chloroform, toluene, hexane, acetone, mfdimethoxybenzene, and 1:1 mixtures of mfdimethoxybenzene with diphenylmethane, diethyl malonate, benzaldehyde, anisole, and chlorobenzene. Unfavorable ng4H9 proton chemical shift differences were observed for (n-C5H5)Zr(dpm)3 in carbon tetrachloride, chlorobenzene, dichloromethane, and 1,2-dich10roethane. The choice of solvents for the study of (n-C5H5)Zr(acac)3 was dictated by the liquid range and nmr chemical shift values of the solvent. Carbon disulfide was selected for variable temperature studies below -100°. For studies of the complex above ambient temperature, none of the above solvents was suitable, but tetrachloroethylene proved to be useful. Benzene was chosen as the solvent for (n-C5H5)Zr(dpm)2Cl because it was used in an earlier study of similar compounds.71 L. Determination of Mean Lifetimes Each of the stereochemical rearrangement processes studied in the present work results in the interchange of nuclei between two nonequivalent sites. Rates of exchange were determined by comparing the experimentally observed nmr spectra with spectra calculated for various mean lifetimes, 16 T. The Gutowsky-Holm equation, as expanded by Rogers and Woodbrey,15 was used to calculate the theoretical spectra. Each calculated spectrum 60 consisted of 100-250 coordinate points spaced at maximum intervals of 0.15 Hz. The calculations were carried out on a CDC 3600, CDC 6500, or IBM 1130 computer. In some cases the populations of the two sites and/or the values of the transverse relaxation times of the nuclei in the two environments are unequal. In the region of exchange below the coalescence temperature, the observed and calculated spectra were compared with respect to the widths at half-maximum amplitude of the two resonance components and/or the parameter, r, which is defined as the ratio of the maximum to central minimum intensities. Whenever possible, the best value of T was taken to be the average value obtained by fitting each line width and r. Above the coalescence temperature, values of I were determined by fitting the width of the time-averaged resonance line. All line widths and values of r were obtained by averaging the results of at least three Spectral copies. III. RESULTS AND DISCUSSION A. Reactions for the Preparation of (n-C5H5)Zr(dpm)2Cl and (n-C5H5)Zr(dik)3 Complexes The tris-hexafluoroacetylacetonate,-trifluoroacetylacetonate, and -pivaloy1trifluoroacetonate cyclopentadienylzirconium complexes were prepared by the reaction of zirconocene dichloride and the neat B-diketone under anhydrous conditions (eq. 1). Earlier work53’7O has (n—C5H5)2ZrC12 + 3H(hfac) + (n-C5H5)Zr(hfac)3 + 2HC1 + C5H6 (1) Shown that reactions of zirconocene dihalides and nonfluorinated diketones under similar conditions yield products of the type (n—C5H5)Zr(dik)2X. Apparently, the substitution of the second chlorine atom by a nonfluorinated diketone is slow, especially in the absence of a base. An appreciable difference in the relative ease of chlorine substitution by fluorinated and nonfluorinated diketones is also found in the substitution reactions of ZrC14. The reaction of ZrC14 with 78 79 hexafluoroacetylacetone or trifluoroacetylacetone, for exnnple, readily affords the tetrakis (B—diketonato)zirconium complex, 69,80 whereas the reaction with acetylacetone and certain other nonfluorinated diketones80’81 gives mainly the chlorotris(diketonato)metal complex. Although replacement of chlorine in Zr(acac)3Cl by acetylacetonate 82 can be achieved in the presence of water, the hydrolyses of (n-c5H5)Zr(acac)2C1 and analogous B-diketonates afford stable dizirconoxanes of the type [(n-CSH )Zr(dik)2]20.56’83’84 5 61 62 The only preparative pathway previously reported for (n-C5H5)Zr(dik)3 complexes containing nonfluorinated diketonate ligands involves the reaction of (C5H5)4Zr and the free diketone84 (eq. 2). In the present (C H 5 5)4Zr + 3H(dpm) + (n-C H )Zr(dpm) + 3C H (2) 5 5 3 5 6 study, analogous reactions with the appr0priate diketones gave (n-C5H5)Zr(dik)3 complexes where dik = dpm, pvac, bzbz, and acac. An easier method which produces a higher yield of products was discovered during the course of this study. The replacement of both chlorine atoms and one cyclopentadienyl group of (n—C5H5)22rCl2 by nonfluorinated diketones readily occurs under ambient conditions in a triethylamine medium (eq. 3). The acetylacetonate, dipivaloylmethanate, benzoylacetonate, (C2H5)3N (n-CSHS)2 2 + 3H(acac) + 2(C2H5)3N-————-————+ (3) (n-C5H5)Zr(acac)3 + 2(C2H5)3N-HC1 + C5H6 ZrCl and pivaloylacetonate derivatives have been prepared successfully in this manner. The method is also applicable for at least one fluorinated derivative, (w—C5H5)Zr(pvtf)3, and it may be applicable for the preparation of all (n-C5H5)Zr(dik)3 complexes. The preparation of (n-C5H5)Zr(dpm)2C1 by utilizing a stoichiometric amount of base is an adaptation of the work by Newton62 (eq. 4). This (n-C5H5)22rCl + 2H(dpm) + (C2H5)3N + 2 (n-C5H5)Zr(dpm)2Cl + (C2H5)3N'HC1 + C5H6 (4) method does not use excess ligand and is most advantageous with the more expensive diketones. It has the disadvantage of also producing (n-C5H5)Zr(dpm)3 as a by-product. At 1:1 molar amounts of (n-CSH ZrCl 5)2 and (C2H5)3N, the major product is (n-C5H5)Zr(dpm)2C1 (76.8% yield). 2 63 However, comparable amounts of (n-C5H5)Zr(dpm)3 and (n-C5H5)Zr(dpm)2Cl are obtained when the (n-C5H5)22r012 to (C2H5)3N molar ratio is increased to 1:2. In a triethylamine medium, the reaction of (n-C H ZrCl 5 5)2 2 with H(dpm) gives only (n-C5H5)Zr(dpm)3 as a product regardless of the molar ratio of (n-C5H5)22r012 to H(dpm). B. Constitution of (n—C H )Zr(dpm) 5 5 C1 and (n—C H5)Zr(dik)3 Complexes 2 5 in Solution. The molecular weight and conductivity data for each compound are presented in the Experimental Section. (n-C5H5)Zr(dmp)2Cl and all of the (n—C5H5)Zr(dik)3 complexes are monomeric in benzene solution and very weak electrolytes in nitrobenzene solution. Their molar conductances range from <0 010 to 0.060 ohm—1 cmz mol—1 at a concentration of gs, 10‘3 M, At this concentration a typical 1:1 electrolyte, such as [Ti(acac)3][Sb016], exhibits a molar conductance slightly greater than 1 cm2 mol-1.87 20 ohm— The ir spectra of the compounds show that the carbonyl stretching vibrations of the diketonate ligands occur in the region 1650 to 1500 cm_]. The vibrations are shifted to lower energy by gs, 75 cm'] from their respective absorptions as ”free” ketone carbonyl groups in open chain silyl enol ethers.8 Thus bidentate coordination of the diketonate ligands is verified. For each compound a weak, sharp band is observed near 3,100 ch] which is characteristic of the C-H stretching vibration a n-bonded (35H5 group.70 The appearance of a Single, sharp CSHS resonance at gs, T 3.50 in the nmr spectra of the compounds is also indicative of a n-bonded cyclopentadienyl group.90 These data indicate that the constitution of (W-C5H5)Zr(dpm)2Cl and the (N—C5H5)Zr(dik)3 complexes in solution is analogous to that found in the solid state for 59 64 (n-C5H5)Zr(acac)2Cl and (n-C5H5)Zr(hfac)3. 64 C. Dynamic Stereochemical Properties of (n-C5H5)Zr(dpm)2Cl. (A-C5H5)Zr(dpm)2 temperature to deduce its stereochemistry by nmr spectroscopy. The C1 is sufficiently stereochemically rigid at room proton nmr spectrum of the compound in benzene solution is presented in Figure 8. The resonance line at T 3.53 is attributed to the protons on the cyclopentadienyl ring. The remaining six lines are due to the —CH= and tggng4H9 protons on the dipivaloylmethanate ligands. The two -CH= lines at t 4.16 and 4.24 have equal relative intensities. Four tgsg7C4H9 lines of equal relative intensity occur in the region T 8.85 to 9.06. The —CH= and gsgth4H9 proton resonances result from nonequivalent environments for these groups in the (w—C5H5)Zr(dpm)2Cl molecule. The existence of a single, sharp cyclopentadienyl resonance is attributed to rapid rotation of the ring about the metal-ring axis. The simplest configuration which places each of the —CH= and tggt7C4H9 groups in nonequivalent environments is based on an octahedron with the center of the C5H5 ring at one stereochemical position and the chlorine atom at a position gis to the ring (see Figure 9I). A gssss configuration (gf, Figure 911), which would have apparent C2v symmetry in the presence of rapid rotation of the C5H5 ring, cannot be present in an appreciable amount, as judged from the relative intensity data. Thus the dipivaloylmethanate complex has the same stereochemical configuration as the other known (n—65H5)Zr(acac)2X complexes.57’58’8] The gisfoctahedral formalism used to describe the gross stereochemical features of the complex is an approximation of the 02d dodecahedral description given for (n-C5H5)Zr(acac)2Cl in the solid state.59’60 In this higher polyhedron (Figure 10) it is assumed that the ring occupies a triangular face, an AAB (1,2,3) face for example, in forming three n bonds to zirconium. The two dipivaloylmethanate ligands would 65 .p:m>_0m me _E oo_\m o.o_ .:owumLu:oo:oo MANIz oov com um :owp:_0m m:m~:mn :2 FQNAEanvLNAmImQIEV do EzLuomam LE: :ouOL: .w mLsmw: 3%: L; 00.0 3o.w mw.m mwd «Ni» 3.3 mm.m _ _ _ L _ _ _ L/ m6 u I??? N.m u 1.52:. m6 n :Eré. 66 ._quE Fvamzmuuo mFQEwm m :o :wmmn FUNAEQUVLNAmIm01tV L0: m:o_me:mww:oo mcmLp ucm.mflm mFDImmoa .m mLonu .0 amt 3m.” 67 (8) Figure 10. The D2d dodecahedron. Alphabetic edge and vertex notation is from J. L. Hoard and J. V. Silverton, Inorg. Chem., 2, 235 (1963). 68 span a g-g (6,7 and 4,8) pair of edges, and the chlorine atom would occupy vertex 8 (5). Seven additional isomers are possible if the C5H5 ring is capable of bonding through the A88 triangular faces. However, since only one C1 isomer is detectable, the octahedral formalism is adequate to describe the stereochemical features of (n-C5H5)Zr(dpm)2Cl. The zirconium complexes, (w—C5H5)Zr(acac)2X and (n-C5H5)Zr(dpm)2Cl, are in direct contrast to a related titanium compound (w-C5H5)Ti(hfac)2C1, which exists as two C1 isomers in solution. Thus, one must conclude that steric factors due to the terminal groups of the diketonate ligands are not a critical stereochemical factor. The temperature dependence of the nmr spectrum of (n-CSH5)Zr(dpm)2Cl (Figure 11) is very similar to that observed for (w-C5H5)Zr(acac)2Cl7O’71 7] As the temperature of a benzene solution of and (n—C5H5)Zr(acac)28r. (n-CSH5)Zr(dpm)2Cl is increased above gs, 60°, the two -CH= resonances broaden and coalesce into a single line which becomes very narrow by gs. 130°. The four tggng4H9 lines broaden and coalesce into a single sharp line, but the C5H5 resonance remains sharp over the entire temperature range. The temperature dependence of the -CH= and gsgg7C4H9 resonances indicates that the molecule undergoes a stereochemical rearrangement process which interchanges the two nonequivalent dipivaloyl— methanate ligands. Analogous rearrangements have been observed for the (n—C5H5)Zr(acac)2X molecules. The mean lifetimes TA and TB for the nonequivalent dipivaloylmethanate ligands were determined by comparing the experimental -CH= nmr line shapes with line shapes calculated for various trial values of T, where T = tA/2 = tB/2. Curve A in Figure 12 indicates that the frequency separation between the -CH= resonances in the region of slow exchange, 60, is temperature-dependent, presumably, because of 69 . Figure 11. Temperature dependence of the -CH= and tert-C4H9 proton nmr lines for (n—C5H5)Zr(dpm)2C1 in benzene. Concentration is 11.0 g/100 ml of solvent. 71 temperature-dependent solvation effects. Therefore, values of 6v in the region of exchange were determined by linear least-squares extrapolation of the temperature dependence of 6v in the slow-exchange region. The widths of the two -CH= lines are equal within experimental error in the region of Slow exchange, thus the apparent transverse relaxation times, T2A and T28 were determined by estimating a small temperature dependence, (9.1 :_5.5) X 10‘4 Hz/°C, for the line width in the region of slow exchange as shown by the dashed line connecting curves B and C in Figure 12. A similar temperature dependence, (33.4 :_5.5) X 10'4 Hz/°C, for the -CH= lines of Zr(dpm)4 is observed over the temperature range 42.0 to 104 5°. Values of the nmr line-shape parameters and the calculated values of T for the nonequivalent dipivaloylmethanate ligands of (n—C5H5)Zr(dpm)2Cl in benzene are given in Table VI. Shown in Table VII are the Arrhenius activation energy, Ea’ the frequency factor, A, the activation entropy, ASi, the activation enthalpy, AHI, and the extrapolated value of the first—order rate constant at 25°, k250. The Arrhenius activation energy and the frequency factor were determined from linear least-squares plot of log k gs, l/T (Figure 13) where k = (2r)']. The activation entropy and activation enthalpy were determined from the Eyring plot of log (k/T) gs, l/T. The activation parameters are rather insensitive to choices of 6v, T2A’ and TZB' The uncertainty in the linear least-squares temperature dependence of 0v propagates a 0.4 kcal/mol error in Ea and a 0.30 error in the value of log A. A reasonable uncertainty of 0.07 Hz (17% error) in estimating 12 values generates 1.1 kcal/mol and 0.68 errors in Ea and log A, respectively. The mean lifetimes of the —CH= protons are independent of the concentration of (n—C5H5)Zr(dpm)2C1 over the range 0.195 to 0.393 M, Figure 12. 72 Temperature dependence of the proton nmr line-shape parameters for the -CH= protons on the dipivaloylmethanate ligands of (n—C5H5)Zr(dpm)2Cl in benzene. Curve A is the frequency separation between the resonance components below coalescence. Curve 8 is the mean width at half—maxumm amplitude of the resonances below coalescence; Curve 0, the width of the time—averaged resonance above coalescence. The significance of the extrapolated (dashed) lines is described in the text. 73 .2 2%: (111) A8 no “LOIM arm 74 1 Table VI. H Nmr Line-Shape Parameters and Kinetic Data for the Interchange of the Nonequivalent Dipivaloylmethanate Ligands on (w-C5H5)Zr(dpm)2C1.a Temp, sz 60c rd Line Width,e 102 T, °C sec Hz Hz sec 69.4 0.777 4.74 .86 37 2 74.1 0.769 4.62 5.71 1.10 23.3 75.3 0.767 4.59 4.92 1.26 20.6 77.7 0.763 4.53 4.01 1.40 18.0 79.6 0.760 4.48 3.67 1.53 16.8 82.1 0.756 4.42 2.33 2.04 12.7 83.6 0.753 4.38 2.11 2.33 11.8 85.3 0.750 4.34 1.86 5.41 9.42 87.2 0.747 4.29 1.40 5.25 7.98 89.7 0.744 4.23 1.15 4.90 6.79 91.7 0.740 4.18 1.00 4.22 5.45 94.5 0.736 4.11 4.04 5.15 96.0 0.734 4.07 3.47 4.52 97.5 0.731 4.03 2.95 3.97 100.0 0.728 3.97 2.46 3.47 102.1 0.724 3.92 2.13 3.07 105.5 0.719 3.83 1.64 2.39 109.1 0.714 3.74 1.26 1.77 aIn benzene; concentration is 0.20 M, bTransverse relaxation time for the -CH= resonances. CFrequency separation between the -CH= . d . . resonances in absence of exchange. Ratio of the max1mum to central 75 Table VI. (Continued) minimum intensities. eWidths in the range 69.4-83.6° refer to the mean line widths of the two -CH= lines below coalescence; in the range 85.3-91.7°. The values are for the total widths of the overlapping resonance lines; above 91.7°,the widths are for the time-averaged line above coalescence. .Fn 88:8L888m E8L88 ._8>8P 88:8:88:88 888 8:8 #8 nmu8E8888 8L8 8LOLL8 F_< .88:F8> :888_888prm 8L8 omm 88 88:888:oo mme LmuLo meww __<8 76 8 8.8 H 8.8 8; H :8 88; H :1: 8; H 8.88 8.2 x _.8 8.8 8:883 18:88-: 8.8 .+. 8.8 8.8 H 8.88 88; H 88.: 8.8 1+. :8 8-2 x 8.: 888 88:88: i8: 88-: 8.8 H 8.8 8; .I... 8.88 8: 1+. :4: 8.8 H L88 8.2 x 8.8 88 88:88: N8; 88-: L; H 8.8- 8.8 H 8.2 :88 H 88.2 88.8 H 8.2 8.2 x 8.2 88:83 A 188-: :8 .8m8 _oE\_88x .818 < mo_ _8E\_88L .88 F.888 .ommx xm_8Eou .m:8~:mm :8 mmxwpano xmflanvLNAmImo-eV :8 88:8Lo ":81 p:m_8>8:cm:oz mo 8m:8:8xm Low 8L888E8L88 8888:8x .HH> anmE 77 ' 1 I ' 1,4 1. q 1.2 - d 1.0 - «I ll .8 .8 - .. D O J .6 '- q .4 P -1 .2 1. 1 l l l J 2.6 2.7 2.8 2.9 3 x It) Figure 13. Log k gs, l/T plot for (n-C H 5 5)Zr(dpm)2Cl in benzene solution. 78 Also, in the presence of an equimolar amount of H(dpm), no exchange is observed between the coordinated and free ligand at 116°. Therefore, the rearrangement process is first order, and the mechanism does not involve total dissociation of the diketonate ligands. Other workers in this laboratory71 have determined the activation parameters for similar first—order, intramolecular, stereochemical rearrangement processes of some acetylacetonate complexes (see Table VI). The difference in Ea for (n-C5H5)Zr(acac)2C1 and (W-C5H5)Zr(dpm)2C1 is_ greater than the experimental uncertainty expressed at the 95% confidence level. Only a systematic propagation of errors in 6V and T2 values would eliminate the differences in the activation parameters. Furthermore, the systematic errors would have to be in opposite directions for the two compounds. For example, only if the values of T2 for (n—C H )Zr(dpm) C1 5 5 were all too large by 20% and those for (n-C5H5)Zr(acac)201 were all too small by an equivalent amount, would the Arrhenius activation 2 energies become equal. This seems very unlikely since the technique used in both studies is identical. Therefore, the difference in the Arrhenius activation energies is undoubtedly real. - It is possible that (n-C5H5)Zr(dpm)2+, formed by the dissociation of the chlorine atom, is an intermediate in the rearrangement process for the (n-C5H5)Zr(dpm)2C1 molecule since a salt which contains that cation has been isolated in the solid state (gf, Section D). A conductometric titration of (n-C5H5)Zr(dpm)2Cl with SbCl5 in nitrobenzene solution (See Figure 14) indicates that such a cationic species can exist in solution. A similar titration curve was also obtained for 86 (w-C5H5)Zr(acac)2C1. The nmr spectrum of a mixture of (n-C5H5)Zr(dpm)2Cl and SbCl in dichloromethane exhibits a dependency on the concentration 5 of SbClS. In the absence of SbC15, the nmr spectrum consists of one 79 T l T T E l l T F l 140- -1 - -1 II ’ vi;;"“2()" a'i; " :1: Q. .. L?) . : 100 .. L1J g) '1 5 80 - :1 D 2: -A O L) 60 .8 40 -‘ 20 '- 1 1 J u 1 4 1 1 1 1 O 0.4 0.8 1.2 1.6 2.0 MOLES SbClS PER MOLE COMPLEX Figure 14. Conductometric titration of 1.69 X 10'2 M_(n-C5H5)Zr(dpm)2 with 0.161 MSbCl5 in dichloromethane at room temperature. C1 80 sharp w-CSH5 line, two sharp -CH= lines, and four sharp gsgg;C4H9 lines at 0.4°. The gggst4H9 peaks have become slightly broader in a 1:75 molar amount of SbCl5 to complex. Increasing the relative concentration of SbCl5 produces progressively broader gsgg;C4H9 and -CH= peaks until the peaks coalesce into a very broad peak at gs, 1 to 1 molar ratio of SbCl5 to complex. 7] have disfavored the formation of a Pinnavaia and Lott bis(8-diketonato)cyclopentadienylzirconium cation as an intermediate in the stereochemical rearrangement process of the acetylacetonate analogues Since (n-C5H5)Zr(acac)2Cl and (n-C5H5)Zr(acac)28r exhibit a very small difference in their activation energies (0.3 :_2.7 kcal/mol). This conclusion can be tested by determining the rate of halide exchange between the bromide and chloride complexes. The cyclopentadienyl chemical shifts in (n-C5H5)Zr(acac)2Cl and (n-C5H5)Zr(acac)28r differ by 1.8 Hz in a benzene solution in which the concentration of each compound is 0.17 M, Exchange of the halides between the two compounds also interchanges the cyclopentadienyl environments, and the rate of halide exchange can be determined by applying nmr line-broadening techniques to the cyclopentadienyl resonances. A fundamental assumption is that the n-C5H5 ligands are incapable of dissociation and subsequent interchange of their environments. An equimolar mixture of (n—C5H5)Zr(acac)2Cl (0.17 M) and (n—C5H5)Zr(acac)28r (0.17 M) in benzene exhibits two sharp n-CSH5 lines, four -CH= lines, and five resolvable acetylacetonate CH3 lines at room temperature. The relative intensities of the acetylacetonate CH3 lines indicates that some of their chemical shifts are nearly coincidental. When a fresh mixture of the bromide and chloride complexes is heated to 80°, a temperature at which the intramolecular rearrangement of 81 (n-C5H5)Zr(acac)2Cl and (n-C5H5)Zr(acac)28r can be observed by the line- broadening of the -CH= resonances, the two cyclopentadienyl lines remain sharp, indicating that the halogen exchange is slow. However, it should be mentioned that as the mixture ages at 80°, the cyclopentadienyl lines gradually broaden and eventually coalesce into a single sharp peak. After gs, 75 hours at 80°, only one cyclopentadienyl line with a peak width of gs, 2.0 Hz at half-maximum amplitude is observed. After a moderate length of time at 80° (gs, 8 hr), partial collapse of the two C5H5 lines is observed. If the same solution is allowed to age at room temperature for several days and then reheated at 80°, the mean lifetimes of the cyclopentadienyl protons are initially equivalent to those for a freshly-prepared mixture. However, after prolonged heating at 80° (gs, 200 hr), the broadening of the two C5H5 lines is no longer reversible in the above manner. A new, unassigned nmr peak in the acetylacetonate methyl region and a distinct yellow coloration of the sample were observed after the prolonged heating period. Apparently, the line- broadening phenomenon is catalyzed by a disproportionation product and a decomposition product of the complexes at elevated temperatures. The disproportionation reaction must be reversible, but the decomposition reaction apparently is not. Even though an appropriate nmr signal is not observed, it is possible that (n-C5H5)22rCl2 is a disproportionation product which catalyzes the coalescence of the C5H5 lines. The cyclopentadienyl lines immediately coalesce at 80° in a freshly-prepared mixture of (n-C5H5)Zr(acac)2Cl (0.16 mmol), (A—C H )Zr(acac)28r 5 5 (0.20 mmol), and (A-C H5)ZZrC12 (0.01 mmol). Thus, regardless of the 5 origin of the line-broadening phenomenon in a mixture of the chloride and bromide analogues, the initial rate of exchange is at least ten times lower than the rate of the stereochemical rearrangement process. More 82 importantly, the rate of stereochemical rearrangement is independent of the rate of catalyzed "halide" exchange since the mean lifetimes of the acetylacetonate methyl protons in a equimolar mixture of the bromide and chloride complexes do not change after 200 hr of heating at 80°. Therefore it can be concluded that the dissociation of a halide ion to form a (A-C5H5)Zr(acac)2+ intermediate is not an important step in the rearrangement process of the (n-C5H5)Zr(acac)2X complexes. A plausible intramolecular mechanism involving formation of an eight- coordinate symmetric intermediate, in which the halogen is position sgsss to the n-CSH5 ring, has been proposed for the acetylacetonate complexes.7] It has been suggested that the formation of the symmetric intermediate may occur gis_the rupture of a Zr-O bond. Random Zr-0 bond rupture to form the intermediate would account for the fact that random terminal R group interchange accompanies the interchange of the diketonate ligands. It is also possible for the molecules to form the symmetric intermediate gjs_a twisting motion which does not involve bond rupture. It should be noted, however, that whether the mechanism involves bond rupture or a twisting motion, the steric requirements of the terminal groups on the diketonated ligands do not drastically affect the activation energy for the rearrangement process. 0. Preparation and Properties of (n-C5H5)Zr(dpm)2C1-SbCl5 The reaction of equimolar quantities of SbCl5 and (n-C5H5)Zr(dpm)2C1 at 0° gives a yellow crystalline compound which slowly decomposes into a black oily substance at room temperature. The chemical analysis of the compound suggests its formulation as a 1:1 adduct of the reactants. Admittedly the analytical data are not as good as hoped for, but the disagreement between the calculated and theoretical elemental analyses 83 is probably due to the unstable nature of the compound and the presence of an impurity (see below). The compound has a molar conductivity at 25° of 18.6 ohm“1 cm2 mol-1 3 (2.81 X 10' M_in C6H5N02), and hence must be appreciably dissociated in solution. The relatively high conductivity cannot be attributed directly to either of the reactants. Antimony pentachloride and (n—C5H5)Zr(dpm)2Cl exhibit molar conductivities of 0.799 (17.0 x 10'3 and 0.151 ohm‘I cm2 moi‘1 (1.84 x 10'3 M) M) in nitrobenzene, respectively. A typical 1:1 electrolyte, [Ti(acac)3][SbCl6], has a molar conductivity 1 1 2 87 at 25° of 19.5 ohm' cm2 mol- in a 1.0 X 10- M solution in nitrobenzene. Thus, the formulation of the compound as a 1:1 electrolyte, L(n-C5H5)Zr(dpm)2][SbCl6], seems reasonable. The solubility of the compound is also consistent with an ionic substance. It is insoluble in the nonpolar solvent benzene, but it is readily soluble in such polar solvents as dichloromethane, acetone, and acetonitrile. The carbonyl stretching bands in the ir spectrum appear between 1660 l and 1500 chl, a shift to lower energy of gs, 50 cm" from the uncoordinated B—diketone, and indicate bidentate coordination of the dipivaloyl- ] due to the C-H methanate ligands. The weak band at gs, 3100 cm- vibrations of the cyclopentadienyl group is absent, but its absence is inconsequential in view of the nmr spectrum of the compound (see below). A band at gs, 342 cm.1 typically occurs for SbCl6- anions,89 and indeed a strong absorption is observed at 345 cm'1 in this case. However, a strong band of similar shape occurs at 325 cm-1 in the parent complex, (n-C5H5)Zr(dpm)2Cl, and hence the 342 cm'] absorption cannot be unambiguously attributed to the SbC16' anion. 84 The nmr chemical shifts offer further support for the [(n-C5H5)Zr(dpm)2] [SbC16] formulation. The nmr spectra for (n-C5H5)Zr(dpm)2C1 and [(n-C5H5)Zr(dpm)2][SbCl6] are compared in Figure 15. The shift of the n-C5H5, -CH=, and EEEEfC4H9 resonances to lower field strength is consistent with the formation of a cationic cyclopentadienylmetal-8- diketonate complex. Fay and Serpone88 report 0.25 to 0.54 ppm and 0.54 to 1.37 ppm shifts to lower fields for the —CH= and CH3 chemical shifts for several acetylacetonate cationic complexes relative to analogous neutral species. Due to the inductive effect of the EEEEfC4H9 groups, the electric field effects would be less for a dipivaloylmethanate complex cation and would account for the observed smaller changes in chemical shifts. Shifts to higher field strength would be expected for anionic complexes.88 The chemical shift of T 3.53 for the C5H5 groups suggests that the ligand is n-bonded to zirconium. The chemical shift of a time-averaged o-CSH5 peak is usually gs, 1.0 ppm downfield relative to a n-CSH5 peak.90 The n-CSH5 chemical shifts for two compounds whose structure is known, (n-C5H5)Zr(acac)2C1 and (n-C5H5)Zr(hfac)3, are t 3.69 and 3.54 in dichloromethane respectively. In addition, the C5H5 chemical shifts in (n-C5H5)Zr(dpm)3 and (n-C5H5)Zr(dpm)2 occur at T 3.99 and 3.69, respectively. The temperature-independence of C1, the cyclopentadienyl resonance (Figure 17) also suggests, in part, a "‘C5H5 ligand. Although, the chemical analysis, molar conductivity, solubility, and ir and nmr spectral data are consistent with the suggested ionic formulation, it should be noted that the data could be interpreted in terms of the formulation [Zr(dpm)2Cl][(o-C5H5)SbC15]. The expected 1.0 ppm downfield shift for a G‘C5H5 chemical shift could be countered by an almost equivalent upfield shift due to the negative charge on the 85 .:888 p:8>_88 8:8 +8 88pw_—8888 8:88-88I8 m:w::w:8 Lo um_ 8L8 : .8 I 8-8 ”838__88 88 8:8 _8NAEIIVLNA8I88-IV 8:8 I8_888IINA8888LNA8I88-888 x: :8 :pwz :8xL8E 8:88: 8:E .Eaa 80.0- .mxeuumuwm “Egg 88.8- .nI8- m:88 8_.8- :88388: 888w:8 F88IE8:8 :88E 8:8 :2 888:8L888I8 8:E .60: 88 8:8>_88 FE oo_\m o.o_ .fl8EQDmH mNAEQUVLNAmImu-evu .m .608 :8 p:8>_88 FE oo_\m o.o_ .EQNAEQUVLNAmImo-Iv .< .8:8:88E8Lo_:8_8 :w flo—unmgflmAEavaNAmIm818VQ 8:8 FUNAEQUVLNAmIm8188 mo 8L88888 LE: :OpOLQ 8:8 8o :088L88Eou .m— 8LImII 86 mi» -~—_ m2:- _~._—..- .“M .8: 8L=88I 288x. HIUI mImUlz. 87 .8_ 8L88II 8: 88:8I88888 >_8:ow>8L8 :88: 8>8: 888:8:888L L8:8o 8:» .:888 u:8>_88 8:8 88 888IEF8888 8:88-88w8 m:w::w88 Lo UMP L8:pw8 8L8 =x= U888:mvm8v 8:88: 8:E .:owp:_88 8:8 88 _E 58.0 op AFOEE mm.ov Z88I8 I: 88.8 :8 :828I888 8I8 L8888 EIL888I8 .8 .888>_88 _E 88_\8 8.8_ 8: :8I88L8888888 moo: 88 8:8:88E8Lo_:8w: :I moronmgflNAEavaNAmzmu-evQ 80 Asz 088 E3L88888 LE: :ouoL: .< .8_ 8L=mr8 88 my; 288:8 288:8 .8_ 8LIIII x 89 ion, and the C5H5 of the ligand. However, [(n-C5H5)Zr(dpm)2][SbC16] is favored since singlet at -55° could be due to the fluctional character SbClG' is a stable anion. Furthermore, (o-C5H5)SbC15' is not known, and attempts to prepare Zr(dik)2Cl+ from Zr(dik)2Cl2 have been unsuccessful.93 The presence of a CH3CN impurity is clearly indicated by ir and nmr spectroscopy. Two ir bands occur at 2285 and 2315 cm'1 where the CEN stretching frequencies for acetonitrile are expected. Furthermore, the nmr peat at T 7.91 is due to acetonitrile. Addition of CH3CN to a solution of [(n-C5H5)Zr(dpm)2][SbCl6] increases the relative intensity of that peak (Figure 16). The shift to a higher field of the chemical shift upon addition of CH3CN is presumably due to a mean decrease in the interaction of acetonitrile with the complex. At temperatures as low as -75°, one sharp CH3CN peak is observed in the mixture. This indicates that if CH3CN is coordinated to the cation, it exchanges rapidly with free CH3CN. The temperature dependence of the nmr Spectrum of [(n-C5H5)Zr(dpm)2] [SbC16] in dichloromethane is shown in Figure 17. The spectrum at -55° can most simply be interpreted on the basis of a trigonal bipyramidal coordination polyhedron in which the n-C5H5 ring occupies an axial vertex and the oxygen atoms of the diketonate ligands occupy the remaining coordination sites. As the temperature is increased above -55°, the two -CH= lines of relative intensity 1:1 and the three gggg7C4Hg lines of relative intensities 1:2:1 broaden and coalesce into sharp singlets. Cooling below -55° results in some broadening again, presumably due to solvent viscosity effects. At all temperatures, only one sharp cyclopentadienyl peak is observed. 9O .888I88EOLOFI888 :8 I8L888LENA8888LN oxen-u u Jm£< Am Imu-evg Io E8L88888 LE: :OHOLQ 8:8 88 88:88:8888 8L388L88E8E .NP 8LsmII 28828 0— JES< ":81 8:88-: nnl n—l 0— H 4monm mocmCOme vmmmcm>m-me?p mgp do gpvwz .o w>czu ”mucmummeoo zo~mn .x_m>wpomamwc .mpcmcoaeou mUCMCOme n_mww-£mvg new -20_ ecu to wczp?_qam E:E.me-e_mg pm mgpuwz .0 new m mm>czo ”mucmumm_moo zo_mn mpcmcoagou mUCGCOMwL mcp cowzpmn cowumcmamm xucmzcmcm .< w>cso “cmgpm anOLQOWwwu c? mAumwcchAmImu-ev do mucmm?_ FwwLOpmscw ms“ co masocm mam _mcwecmp ms“ toe mcwmemLma mamcm-m:w_ LE: ¢m_ do wocwncmawu wczpmcmasmh .NN mczm?m 1(33 .NN ma=m_a Anya. h. a V n a N 3 m m m H 0 o. “a 8 n SxH: Z ( v— o. m— l04 the lines, and hence the transverse relaxation times, are equal in the region of slow exchange. In a manner exactly analogous to that depicted in Figure 22, values of do and the transverse relaxation times were determined from extrapolations of the appropriate lines. The two low-field CH3 lines of (n-C5H5)Zr(acac)3 overlap in the region of slow exchange, and their accurate line widths are impossible to determine. However, in the region above the coalescence temperature for process I, the line-widths of the CH3 resonances of the unique ligand are equal. Therefore, it was assumed that T2A and T2B for the CH3 protons of the equatorial ligands were also equal. Values for T2A and T28 were determined by extrapolating the observed temperature- dependence of the CH3 line-widths into the region of exchange. The extrapolation was aided by three data points in the region of exchange, -8l.3, -73.7, and -58.9°, for which T2A and T28 were variables in total line shape analyses of the spectra. Finally, a temperature dependence was not observed for 6v in the region of slow exchange, 1,3,, below -94.9°, and therefore, a constant value for 6v was extrapolated into the region of exchange. Values of the nmr line—shape parameters and the calculated values of T for the terminal groups on the equatorial ligands of (n-C5H5)Zr(hfac)3 in diisopropyl ether, (n—C5H5)Zr(dpm)3 in gfxylene, and (n-CSH5)Zr(acac)3 in carbon disulfide are given in Tables VIII, IX, and X, respectively. Shown in the Table XI are the Arrhenius activation energy, Ea’ the frequency factor, A, the activation entropy, AS¢, the activation enthalpy, AH , and the extrapolated value of the first-order rate constant at 25°, k250, for the compounds. Values of Ea and A were determined from the slope and intercepts of graphs of log k vs, l/T, where k = (2T)-]. These graphs are shown in Figure 23. Graphs of log (k/T) vs, l/T were used to determine AH and AS . .mocmommFmou m>onm mucmcommg nwmmcm>mnmewp mgu com mew mcpuwz esp .oo.m~- m>onm mmczpmcmaewp mucmummpmou any 3o_mn mpcmcoaeou wocmcommc vpmwmuzmw; ucm Ume$-30F msp do muapw_qsm Ezewxme-$_mc pm mcpuwz on» op Loewe om.m_- op m.mm- mace; esp c? mcpuez mcwum .mmcmgoxm to mucmmnm c? mpcmcoaeoo mucmc0mmc mgp :mmzumn cowpmcmamm xucmscmcm .mucmcommg mug u—mww-;m?; Low wave cowpwxMch mmgm>mcmceo .mocmcommc 105 u mum upmwwuzoF coy we_p :owpmxm—mc wmcm>mcmchn as N_.o mw cowpmcpcoucou mgmcaw Paaocaomwwv :Hm ¢¢¢.o Fm._ No.¢_ nmm.o me.o o._m mmw.o mn.F mm.¢F mwm.o NvN.o ¢.mm mm._ m_.m mp.m_ NmN.o mmm.o 0.0— mm.~ mm.m mm.mp mmN.o wmm.o w.¢— m~.N am.m Nm.m_ mmm.o omm.o “.mh o¢.m mw.m _¢.m_ mmN.o vmm.o m.op mm.m mm.¢ ¢¢.mP mmm.o wmm.o N.m mo.m om.¢ mo.m_ me.o _mm.o _.¢ F.No mo.¢.om.¢ Nm.oP oom.o m_m.o o.mpi m.mm Nn.m._m.¢ mm.o_ Nom.o N_N.o ¢.NF- m.mm N_.m,_m.m m¢.o_ Nom.o N_N.o m.m—- m.mm ow.m._¢.m em.m_ mom.o mrm.o m._m- m¢_ m_.m.mn.m No.o_ mom.o ¢_N.o N.¢N- «mm Nn._.mm.m mx.m_ mpm.o p_m.o m.NN- mmm em._.mm.m mn.m_ Nfim.o FPN.o m.mm- mow mm._.mo.m mm.o_ m_m.o mom.o m.Nm- C. a a? awscwcoz mo mmcmgucmch mgp Low mums uwumcwx new mcmpmsmcma mamcm-m:w4 L52 8 .HHH> mFQmH @— 106 Table IX. 1H Nmr Line-Shape Parameters and Kinetic Data for the Interchange of Nonequivalent tert-C4H9 Groups on the Equatorial Ligands of (n-C5H5)Zr(dpm)3.a Temp, °c 86,sz rc Line widths, sz 102 T, sec 65.2 11 60 1.16 27.0e 70.1 11.54 1.61 15.2 74.0 11.48 2.16 10.5 74.5 11.60 9.83 2.26 9.83 79.5 11.43 5.72 3.08 7.09 79.8 11.42 5.29 3.13 6.88 83.3 11.37 3.06 4.42 5.04 84.3 11.36 2.66 4.66 4.73 87.1 11.31 1.68 3.48 aIn gyxylene at a concentration of 0.2l M, bFrequency separation between the resonance components in absence of exchange. CRatio of the maximum to center minimum intensities. dAverage width of the resonance components below the coalescence temperature. eThe value of T2 used in the calculation of all values of r was 0.590 sec. l0? Table X. 1H Nmr Line-Shape Parameters and Kinetic Data for the Interchange of Nonequivalent CH3 Groups on the Equatorial Ligands of (n-CSH5)Zr(acac)3.a Temp, °C secb rC Line widths,d Hz 102 T, sec -85.8 .24l 3.25 9.27 -85.6 .24l 4.46 2.98 l0.2 -8l.3 .249 2.56 4.l7 6.02 -80.3 .259 2.46 4.27 6.l2 -76.0 .277 l.ll 2.88 -75.0 .277 l.22 3.22 -74.9 .277 l.057 2.69 -73.9 .277 l0.40 2.08 -72.l .289 9.85 l.96 -72.0 .289 9.70 l.93 -69.7 .300 8.47 l.70 -67.l .3l2 4.80 l.00 -64.8 .328 4.24 0.887 -6 .6 .328 4.45 0.904 -62.0 .346 3.40 0.694 —58.9 .364 2.52 0.473 —58.4 .370 2.38 0.439 -54.4 .393 l.8l 0.29l aIn carbon disulfide; concentration is 0.22 M, bTransverse relaxation time for the CH3 resonances. CRatio of the maximum to central minimum intensities. dLine widths in the range -85.8 to -80.3° refer to the widths at half-maximum amplitude of the high- field resonance component below the coalescence temperature; at -73.9° and above, the widths are for the time-averaged resonance above coalescence. eThe value of 60 used in the calculation of all values of T was 10.40 HZ. 108 as 5N.o m5 cowpmcpcmucoo mmcmFAXLm :Hm am m_.o m5 cowumcpcmocoo mcmcpm F>QOLQOm558 CH8 .Fm>m_ mucmuwwcoo 5mm 855 pm umpmewpmm mew mcoccm 55 Uwpmroamfiwxw wLm omN pm mvcmgmcou mum.» LwULO pm.» .6? _._.wzcwcoz mo mmcwzucmvcH 855 L05 mcmmemme :o_pm>5pu< .Hx mpnmh Figure 23. 109 Arrhenius plots for the exchange of nonequivalent terminal groups on the equatorial ligands of (n-C5H5)Zr(dpm)3 in gfxylene (line A), (n-C5H5)Zr(hfac)3 in diisoprOpyl ether (line B), and (n—C5H5)Zr(acac)3 in carbon disulfide (line C). log k 110 I r I I I 300 F. 2.5 . 2.0 r- C B 1.5 1- C 1.01- A .5 b 1 L l I J 3.0 3.5 4.0 4.5 5.0 1 3 T x 10 Figure 23. 111 The activation parameters for the exchange process are not appreciably dependent on the estimated temperature dependence for the relaxation times or 6v. For example, and estimated error of 0.35 Hz in 60 for (n-C5H5)Zr(acac)3 generates errors of only 0.l kcal/mol and 0.13 in E6 and log A, respectively. An error of 0.15 Hz in calculating the T2 values of (w-C5H5)Zr(acac)3 produces errors of 0.5 kcal/mol in Ea and 0.52 in log A. These errors are within or nearly within the 95% confidence level estimates of error obtained by the above procedures. A comparison of the activation parameters for (n-C5H5)Zr(hfac)3, (n-05H5)Zr(dpm)3 and (n-C5H5)Zr(acac)3 shows that the differences in stereochemical rigidity of the molecules is due primarily to differences in their activation energies. Solvent effects cannot account for the difference in activation energies. Rate constants for (n-C5H5)Zr(dpm)3 in diisopropyl ether at five temperatures in the region 69.5 to 80.l°, as well as in toluene at four temperatures in the region 60.4 to 74.6°, gave estimated Arrhenius activation parameters which are equal within experimental error to the values found in Qfxylene: E 22.l :_l.l a kcal/mol, log A = l4.68 :_0.68 in diisoprOpyl ether; Ea 22.5 :_l.3 kcal/mol, log A = 14.86 :_0.81 in toluene. Likewise, the activation parameters for (n-05H5)Zr(acac)3 seem solvent-independent within the 95% confidence level of estimated error. Rate constants for four temperatures in the range -87.3 to -80.90 gave Arrhenius activation parameters in a lzl mixture of methylene chloride and trichloroethylene which fall within the limits of error expressed for the complex in carbon disulfide: Ea = 11.1 :_4.8 kcal/mol, log A = 13.84 i 5.32. The mean lifetimes of the terminal groups on the equatorial ligands of the three complexes are independent of concentration over the range 0.12-0.32 M, and also, are not affected by the presence of equal molar amounts of free ligand. Therefore, the rearrangement processes l12 are indeed first order, and an intramolecular mechanism operates. Figure 24 shows four simple intramolecular mechanisms which would lead to the interchange of nonequivalent environments for the terminal groups on the equatorial ligands of a (n—C5H5)Zr(dik)3 complex in which the diketonate ligands are symmetric. Mechanism A, which involves a sliding motion of the unique ligand, leads not only to the interchange of environments for the terminal groups on the equatorial ligands but, also, to the exchange of nonequivalent environments for the terminal groups on the unique ligand. Since coalescence of the nmr lines for the terminal groups on the unique ligand in these complexes does not accompany the collapse of the lines for the terminal groups on the equatorial ligands, mechanism A can be eliminated. Mechanism B involves the bond rupture of a Zr-O bond on the unique ligand. The Zr—O bond of RS’C, adjacent to R4’B, ruptures on the front side of the pentagonal bipyramid and re-forms by attacking the back side of the octahedral intermediate. Terminal group RS’C is now adjacent to R3 and R1 thereby exchanging the A and B environments on the equatorial ligands while retaining the identity of R52C and R5,D of the unique ligand. In mechanism C, a Zr-O bond of an equatorial ligand ruptures on one equatorial edge of the pentagonal bipyramid and re-forms along a different equatorial edge. Again RA and RB environments of the equatorial ligands are interchanged without interchanging the RC and RD environments of the unique ligands. The fourth mechanism (0) invokes a diagonal twist motion92 of the equatorial ligands. This motion also results in the desired effects. The relative merits of these last three mechanisms, the unique ligand bond rupture, the equatorial ligand bond rupture, and the 113 .-q .5I ucm mm masocm 56:5ELop mc5:_mpcou UcmmWF 555Lopczcm 855 L05 mwmmwooca mzomOchm 5pm: zmcmcm c5 wumcmcwmwu wee a wen u mamwcmgomz .oncm I I858 50 mucchocw>cm asp >55ucmu5 ncw 5man mpqwgomgmazm 8:5 .waQEou mfianvamImu1rv m 50 muccmWF 5mwgoumscw on» :o azocm FQCTEpr mg“ 505 mpcmEcoc5>cw pcm_m>vscmcoc 5o :o5mcm>coocwpc5 mcg cow mgmwcmgumg ImpsomFOEmcpcw 8585mmom .5N wczmwm ll4 115 equatorial ligand diagonal twist mechanisms, can be evaluated, in part, by the investigation of a (n-C5H5)Zr(dik)3 derivative which contains asymmetric B-diketonate ligands. Because of the asymmetry of the B-diketonate ligands, (n-CSH5)Zr(dik)3 may give rise to the six geometric isomer shown in Figure 25. In the absence of any exchange processes, an equilibrium mixture of all six isomers should give six €5H5, fourteen -CH=, fourteen R, and fourteen R' nmr lines. The proton nmr Spectrum of the trifluoroacetylacetonate derivative, (n-C5H5)Zr(tfac)3, consists at room temperature of two sets of 1:2 methyl doublets, four —CH= lines, and two C5H5 resonances. The spectrum has been interpreted previously63 in terms of a stereochemical rearrangement process which averages the terminal groups on the equatorial ligands in the set of three isomers in which one of the distinguishable terminal groups on the unique ligand is axial (1+ +II+ +III) and in the other set in which the distinguishable terminal groups is equatorial (IV+ +V+ +VI). However, the existence of all six isomers in solution has not yet been demonstrated. Elder64 has found that the bond between zirconium and the axial oxygen atom in (n-C5H5)Zr(hfac)3 is substantially shorter, and presumably stronger, than the five equatorial Zr-O bonds in the molecule. One might expect, therefore, that only one set of three (n-05H5)Zr(tfac)3 isomers (g,g,, I, II and III) may exist, because of the potentially different donor properties of the two carbonyl oxygen atoms on the asymmetric trifluoroacetylacetonate ligand. In the presence of only one set of isomers, a bond-rupture mechanism involving the bond to the equatorial oxygen on the unique ligand (mechanism 8) would interchange the two isomer with apparent CS symmetry (§,g,, I+—+II) and interconvert the environments of the nonequivalent "equatorial" terminal methyl groups on the isomer with C] symmetry (III). 116 vcm m .vcmm55 mmeOme581m uvamEEzmm mcp co masocm 5ocwecwp w_nmzm5:mc5pm5u asp pcwmwcamc .m .mucmm55 mpmcoumxvn1m owcpmeexmm mcw>mc mfixwuVLNAmImU1ev Lo» mLmEOW5 owcpmeomm 8585mmoa .mm 5.5555 117 .mm 5.5555 H a v. 118 Thus the expected number of proton nmr lines in the region of fast exchange would be in agreement with that observed. 0n the other hand, if all six isomers were present, then either rupturing of a bond to oxygen on the equatorial ligands to give a lower coordination number intermediate (mechanism C) or a digonal twisting motion of the equatorial ligands (mechanism 0) would interchange the three isomers in each set and would give rise to the observed number of proton nmr lines in the fast-exchange limit. Mechanism 8 in the presence of all six isomers would give rise to twice the number of observed nmr lines at room temperature. The temperature dependence of the nmr spectra of (11-C5H5)Zr(tfac)3 in dichloromethane is illustrated in Figure 26. In contrast to the room-temperature spectrum, the proton spectrum in the range -90 to -75°, where the exchange process is slow, contains three C5H5 lines, five broad -CH= resonances, and seven broad CH3 lines. The widths of the -CH= and CH3 lines lie in the range l.5-4.0 Hz. The 19F spectrum at -80° showed at least seven rather sharp lines. Although the number of nmr lines supports the presence of only one set of three isomers, the relative intensities of the CH3 resonances cannot be explained on this basis. It must be assumed that more than three isomers, probably all six isomers, are indeed present, and that the 1H and 19F chemical shiftfsfor many of the isomers are nearly coincident. This assumption is also supported, in part, by the broadness of the CH3 and -CH= lines. The pivaloyltrifluoroacetonate derivative, (n-C5H5)Zr(pvtf)3, is stereochemically less labile than the trifluoroacetylacetonate complex, hence the region of slow exchange for the former complex is the more accessible of the two. At -48.l° (Figure 27), nine CF3 lines are resolved in the 19F spectrum, and three C5H5 and six tert-C4H9 lines 119 c... .pcm>50m 5E oo_\m o.m m5 corpmcpcmucou .wcmcmeOLOFIUWU mAumwpchAmsz1ev Low mmc55 LE: couocq mzmu1e use qu mg“ 50 mucwucmawu mcsumcmqemh .mw wcamwm 120 :U .85 5.5555 001 on: a: mu mznu 1 t :5 j“ 5 3 121 11" c5115 -cB-I; 144119 (:Fa Figure 27. 1H and 19F nmr spectrum of G1-C5H5)Zr(pvtf)3 at -48.l°. Concentration is l6.0 g/100 ml dichloromethane. '—‘_— 122 are observed in the 1H nmr spectrum. Integration of the CF3 peaks does not reveal even a single set of peaks in the l:2 ratio expected for an isomer. Again, it must be concluded that more than three isomers are present, and that the chemical shifts for many of the isomers are nearly coincident. It should be noted that the benzoylacetonate and pivaloylacetonate derivatives are more labile than (n-C5H5)Zr(tfac)3 and (n-C5H5)Zr(pvtf)3. Either their regions of slow exchange are not readily accessible or, if they are accessible, the nmr resonances are broad and poorly resolved. The nmr data for (n-C5H5)Zr(tfac)3 and (n-C5H5)Zr(pvtf)3 are believed to be consistent with either a bond-rupture or a digonal—twist mechanism involviNg the equatorial ligands (8,9,, either mechanism C or D), but not mechanism B as the sole pathway. Digonal-twist mechanisms have been regarded as unlikely processes in the isomerization of octahedral complexes,92 but in rearrangements of higher coordination number metal complexes, such twists could be favored and cannot be ruled out. It may be argued that since the bulkiness of the terminal groups increases in the order CH3 < CF3 < tertyC4H9, the corresponding increase in activation energies for the complexes would be more consistent with a twist mechanism than a bond—rupture mechanism. However, it is possible that steric effects could predominate in a rearrangement subsequent to the initial rupture of a Zr-O bond, and therefore the activation parameters would not reflect the relative strengths of the Zr-O bonds. Thus, neither mechanism can yet be eliminated. It should be noted that among the dipivaloylmethanate complexes, the stereochemical lability increases with decreasing coordination number of zirconium in the following order: (11-C5H5)Zr(dpm)3 < 123 (n-C5H5)Zr(dpm)2Cl < [(n-C5H5)Zr(dpm)2]+. A decrease in coordination number is expected to cause an increase in the Arrhenius energy of activation in which bond-rupture is an important part of the activation process. On the other hand, just the Opposite result is expected for a twisting mechanism. 3. Process 11. The exchange of the unique B-diketonate ligand with the two equivalent equatorial ligands on (n-C5H5)Zr(hfac)3 and (w-C5H5)Zr(acac)3 was examined in diiSOprOpyl ether and tetrachloroethylene solutions, respectively, by observing the temperature dependence of the -CH= proton resonances. A similar study was not attempted for (n-C5H5)Zr(dpm)3 because of the rapid thermal decomposition of the complex at temperatures above 100°. In the study of process II for (n-C5H5)Zr(hfac)3 it was found that the frequency separation of the ~CH= lines decreases linearly with temperature in the region of slow exchange from -2 to 50° (Curve A, Figure 28). Curves B and C in Figure 28 indicate that the line widths are unequal and temperature-dependent over the same temperature range. Therefore, values of 60, T2A’ and T for the -CH= protons at 28 temperatures in the region of exchange were determined by extrapolation into the region of exchange as shown by the dashed lines in Figure 28. Similar temperature dependences were observed for the -CH= lines of (n-C5H5)Zr(acac)3, and values of 60, T2A’ and T28 were obtained by analogous extrapolations. The nmr line-shape parameters and values of r for (n-C5H5)Zr(hfac)3 and (n-C5H5)Zr(acac)3 are given in Tables XII and XIII, respectively. Since the population ratio of -CH= protons in the two nonequivalent sites is l:2, the parameter T is related to the mean lifetimes for the 124 .px0p 0:0 :5 0005L0000 m5 m0055 A005000V 00005000pr0 050 50 000005550050 0:5 .00000005000 0>000 00:0:000L 0000L0>010E5p 0:0 00 003555080 Esewx0515_0g 00 £005: .0 0>L30 MN 505mc0005 0>55050L 50 00:0:000L 0:0 50 00:0550E0 53E5x051050g 00 £0052 .0 0>L30 m5 5050:0005 0>50050L 00 0000:000L 0;» 50 000055050 Eze5x051050g 00 I005; .m 0>L:0 M00000005000 30_00 0000000500 00:0:000L 0:0 00 00500L000m x000300L5 .< 0>L30 ”L0Ip0 5500L000550 c5 m5000IVpNAmIm01IV 50 mucmmTF 000:000I501m 0:0 :0 mcopoLa “:01 050 L05 mL0p050L00 0005010055 LE: 0:0 50 0000000000 0L300L00E05 .wN 0L5550 125 (2 H) #9 HO HlOlM 3N1? I40 120 100 80 60 40 20 1(°c) 28. Ziguv‘e .00000005000 0>000 0055 0000L0>010E55 005 L05 0L0 005053 005 .om.mw 0>000 M00000005000 30500 00055 0505510050 000 050551305 005 05 L050L om.0m15.00 0000L 005 05 0050535 .00555000505 0005050 50L5000 05 Eze5xms 005 50 055000 .000000x0 50 0000000 05 m00000000L n00- 005 0003500 00550L000m 5000:00L50 .00000000L ":01 0505510050 005 L05 0055 00550x050L 0mL0>m00L50 .00000000L ":01 050551305 005 L05 0055 00550x050L 0mL0>m00L5 n2 Nm.o m5 00550L5000000 mL0050 5500L000550 050 126 n m_.5 mo.m op.“ 0mm.o m¢m.o m.©o~ mm._ mm.m m5.m 0mm.o wvm.o ~.moF mm.— 50.0 mm.m mmm.o mqm.o 0.005 mm.m m©.¢ 5N.m mmm.o omm.o o.mm N¢.¢ m¢.v om.n Nmm.o omm.o m.mm ww.m mm.m om.m mm.m Nmm.o FmN.o m.mm ww.m mm.m mm.¢ nm.n 5mm.o mmm.o m.mm N.NF om.m.mN.N o¢.m o¢.n 5mm.o mmm.o N.Nw $.05 mm.m.mo.m om.m 00.5 omm.o va.o ¢.wm m.mN mm.N.wm.— w¢.m omm.o mmm.o 0.05 000 .5 05 N0 .005053 0050 L NI .>0 000 0N5 000 05 0000055000 0 .000_0> 00505000L5x0 0L0 600 50 050050000 050L L00L0 50L55 55<0 0 00 00.0 05 00550L5000000 .00050 50000000550 000 5.0.5 5.0 0.5.5 5.00 50.0.5 50.05 0.5.5 5.50 0-05 x _.50 050000VLNL0I0010V 0.0.5 0.0_ 0.0.5 0.00 00.5.5 00.05 00.5.5 5.00 0-05 x 50.5 0005500L000I00150 00 500\5000 500\_000 F1000 . . .0 . 00 5 00 0 00 6 x0 000 5 5I I L 0 0 I 5 0 m 00005500L05 I 010V 00 0000050 0 .m50000vLNAmImu10V 000 505L050000 000 00050: 005 50 000000xm L05 0L050E0L00 0550050 .>Hx 05005 132 scattered to permit making any conclusions about the rate law. A separate experiment, conducted over a 48 hr period, indicated a large time-dependence for the rate of intermolecular exchange between free diketone and the coordinated ligands. This would help explain the inconsistency of the earlier data. Also, it suggests that the intermolecular exchange is catalyzed by disprOportionation products and/or decomposition products of the complex. Alternatively, a series of competing reactions may occur in the presence of the free diketone. It can be stated, however, that the rate of exchange between equimolar amounts of free diketone and complex in a freshly prepared solution is initially about ten times slower than the rate of process II at the same temperature. It is possible that the exchange of free and coordinated ligand occurs by a mechanism similar to that for halogen exchange in the (n-C5H5)Zr(acac)2X complexes since a great time-dependency is observed in both cases. Previous studies have shown that ligand exchange between zirconium B-diketonates and free diketones can occur vis_first-order as well as second-order paths.9] In the case of (n-C5H5)Zr(dik)3 complexes, the formation of a square— pyramidal (11-C5H5)Zr(dik)2+ intermediate in the rate-limiting step should lead to comparable rates for process II and the exchange of free and coordinated ligand, providing that the transfer of the enolic proton between the free ligand and its conjugate base in the latter process is fast. Thus it would seem that the formation of a (11-C5H5)Zr(dik)2+ intermediate is not an important pathway in process II. Furthermore, the magnitude of the activation energy favors an intramolecular mechanism for process II rather than a complete ligand-dissociation mechanism. BIBLIOGRAPHY 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. BIBLIOGRAPHY R. West, J. Amer. Chem. Soc., 88, 3246 (1958). R. M. Pike, Coord. Chem. Rev., 8, 163 (1967). a) w. Dilthey, Ber. Deut. Chem. Gesell., 36, 926 (1903); b) ibid., ss, 1595 (1903);—T_"E“d"‘36—3207‘(T1 1 ., _, 903T. — E. L. Muetterties and C. M. Wright, J. Amer. Chem. Soc., 82, 21 (1965). 0. w. Thompson, Inorg. Chem., 8, 2015 (1969). a) R. M. Pike and R. R. Luongo, J. Amer. Chem. Soc., 82, 1403 (1965); b) ibid., 88, 2972 (1966). J. J. Howe and T. J. Pinnavaia, J. Amer. Chem. Soc., 81, 5378 (1969). w. T. Collins, M. S. Thesis, Michigan State University, East Lansing, Michigan, 1970. w. H. Knoth, Ph.D. Thesis, the Pennsylvania State University, University Park, Pennsylvania, 1954. sf, L. H. Sommer, ”Stereochemistry, Mechanism and Silicon,” McGraw Hill Inc., N.Y., N.Y., 1965, p. 14. S. K. Dahr, V. Doron, and S. Kirschner, J. Amer. Chem. Soc., 88, 753 (1958). R. E. Herter, Chem. Ind. (London), 1397 (1963). B. Eistert, F. Arndit, L. Loewe, and E. Ayca, Chem. Ber., 88, 156 (1951). Y. Kuo, F. Chen, and C. Ainsworth, Chem. Comm., 137 (1971). a) A. L. Van Geet, Anal. Chem., 88, 679 (1970); b) ibid., 88, 2227 (1968). M. T. Rogers and J. C. Woodbrey, J. Phys. Chem., 88, 540 (1962). H. S. Gutowsky and C. H. Holm, J. Chem. Phys., 88, 1228 (1956). U . Gibson, Coord. Chem. Rev., 8, 225 (1969). J. L. Burdett, Ph.D. Thesis, Michigan State University, East Lansing, Michigan, 1963. M. Anteunis and N. Schamp, Bull. Soc. Chim. Belges, Z8, 330 (1967). 133 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 134 J. W. Emsley, J. Feeney, and L. H. Sutcliffe, "High Resolution Nuclear Magnetic Resonance Spectroscopy," Vol. 2, Pergamon Press, New York, N. Y., 1966. PP. 841-857. L. M. Jackman and R. H. Wiley, J. Chem. Soc., 2886 (1960). S. Sternhell, Rev. Pure Appl. Chem., 22, 15 (1964). R. R. Fraser and D. E. McGreer, Can. J. Chem., 88, 505 (1961). T. J. Pinnavaia, J. M. Sebeson, II, and D.A. Case, Inorg, Chem., 8, 644 (1969). R. C. Fay and R. N. Lowry, ibid., 8, 1512 (1967). D. P. Craig and C. Zauli, J. Chem. Phys., 82, 609 (1962). A. A. Lavigne, J. Tancrede, R. M. Pike, and C. T. Tabit, J. Organomet. Chem., 28, 57 (1968). E. L. Muetterties, W. Mahler, and R. Schmutzler, Inorg, Chem., 2, . 613 (1963). ’ E. A. Dennis and F. H. Westheimer, J. Amer. Chem. Soc., 88, 3432 (1966). F. H. Westheimer, Ace. of Chem. Res., 2, 70 (1968). R. S. Berry, J. Chem. Phys., 82, 933 (1960). F. Ramirez, A. S. Gulati, and C. P. Smith, J. Amer. Chem. Soc., 88, 6283 (1967). F. Ramirez, M. Nagabhushanam, and C. P. Smith, Tetrahedron, 1785 (1968). F. Ramirez, Acc. of Chem. Res., 2, 168 (1968). H. Shanan-Atidi and Y. Shvo, Tetrahedron Lett., 603 (1971). J. Chatt and A. A. Williams, J. Chem. Soc., 4403 (1954). C. Eaborn, ”Organosilicon Compounds“, Butterworths Scientific Publications, London, 1960, p. 112. W. J. Moore, ”Physical Chemistry”, Prentice-Hall, Inc., Englewood Cliffs, N.J., 1962, p. 298. H. H. Jaffe, Chem. Rev., 88, 191 (1953). A. Combes, C. R. H. Acad. Sci., 119, 1222 (1 94). J. P. Fackler, Jr., Prog. Inorg. Chem., 2, 361 (1966). J F. Bonati, Orggnomet. Chem. Rev., 2, 379 (1966). D. P. Graddon, Coord. Chem. Rev., 2, 1 (1969). 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 135 D. Gibson, ibid., 2, 225 (1969). J. Thiele, Ber. Deut. Chem. Gesell., 82, 68 (1901). _1 . J. Kealy and P. L. Pauson, Nature, 168, 1039 (1951). S. A. Miller, J. A. Tebboth, and J. F. Tremaine, J. Chem. Soc., 888 (1952). G. Wilkinson and F. A. Cotton, Prog. Inorg. Chem., 2, 1 (1959). P. L. Pauson in I'Organometallic Chemistry”, H. Zeiss, Ed., Reinhold Publishing Corp., New York, N. Y., 1960. Pp. 346-380. H. P. Fritz, Adv. Organomet. Chem., 2, 240 (1964). W. F. Little, Survey Prog. in Chem., 2, 133 (1963). J. C. Thomas, Chem. Ind., 1388 (1956). R. Kh. Freidlina, E. M. Brainina, and A. N. Nesmeyanov, Proc. Acad. Sci. USSR, 138, 628 (1961); Dokl. Akad. Nauk SSSR, 138, 1369 (1961). E. M. Brainina and R. Kh. Freidlina, Bull. Acad. Sci. USSR, Div. Chem. Sci., 1331 (1964); Izv. Akad. Nauk SSSR, Ser. Khim., 1421 (1964). E. M. Brainina, R. Kh. Freidlina, and A. N. Nesmeyanov, Proc. Acad. Sci. USSR, 154, 143 (1964); Dokl. Akad. Nauk SSSR, 154, 'ITFI—(T964)T__— ____ E. M. Brainina and R. Kh. Freidlina, Bull. Acad. Sci. USSR, Div. Chem. Sci., 756 (1963); Izv. Akad. Nauk SSSR, Ser. Khim., 835 (1963). T. J. Pinnavaia, J. J. Howe, and E. 0. Butler, J. Amer. Chem. Soc., 88, 5288 (1968). . M. Kh. Minacheva, E. M. Brainina, and L. A. Federov, Bull. Acad. Sci. USSR, Div. Chem. Sci., 1007 L(1969); Izv. Akad. Nauk SSSR, Ser Khim., 1104 (1969). J. J. Stezowski and H. A. Eick, J. Amer. Chem. Soc., 82, 2890 (1969). V. S. Sundarikov, N. G. Bokii, V. I. Kulishov, and U.T. Struchkov, Zh. Strukt. Khim., 28, 941 (1969). M. Kh. Minacheva, E. M. Brainina, and R. Kh. Freidlina, Proc. Acad. Sci. USSR, 173, 282 (1967); Dokl. Akad. Nauk SSSR, 173, 581 (1967). W. E. Newton, Ph.D. Thesis, University of London, London England, 1969. M. Elder, J. G. Evans, and W. A. Graham, J. Amer. Chem. Soc., 82, 1245 (1969). 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 136 M. Elder, Inorg. Chem., 8, 2103 (1969). E. M. Brainina and R. Kh. Freidlina, Bull. Acad. Sci. USSR, Div. Chem. Sci., 756 (1963); Izv. Akad. Nauk SSSR, Ser. Khim., 835 (1963). G. Doyle and R. S. Tobias, Inorg. Chem., 8, 1111 (1967); ibid., 2, 2479 (1968). . Gore, M. L. H. Green, M. G.Harris, W. E. Lindsell, and H. Shaw, . Chem. Soc. (A), 1981 (1969). Cal-1'1 . Doyle and R. S. Tobias, Inorg. Chem., 2, 2484 (1968). G T. J. Pinnavaia and R. C. Fay, ibid., 2, 502 (1968). E. 0. Butler, M. S. Thesis, Michigan State University, East Lansing, Michigan, 1969. T. J. Pinnavaia and A. L. Lott, II, Inorg. Chem., 10, 1388 (1971). H. Gilman, R. G. Jones, E. Bindschadler, D. Blume, G. Karmas, . G. A. Martin, Jr., J. F. Nobis, J. R. Thutle, H. L. Yale, and J. A. Yoeman, J. Amer. Chem. Soc., 28, 2790 (1956). J. T. Adams and C. R. Hauser, ibid., 88, 1220 (1944). .1. 6. Reid and M. Calvin, ibid., _z_2_, 2948 (1950). T. Shigematsu, M. Matsui, and K. Utsunomiya, Bull. Chem. Soc. Japan, 22, 1278 (1969). E. M. Brainina, M. Kh. Minacheva, and R. Kh. Freidlina, Bull. Acad. Sci. USSR, Div. Chem. Sci., 1839 (1965); Izv. Akad. Nauk SSSR, Ser. Khim., 1877 (1965). E. M. Brainina and G. G. Dvorgantseva, Bull. Acad. Sci. USSR, Div. Chem. Sci., 427 (1967); Izv. Akad. Nauk SSSR, Ser. Khim., 442 (1967). S. C. Chattoraj, C. T. Lynch, and K. S. Mazdiyasni, Inorg. Chem., 2, 2501 (1968). M. L. Morris, R. W. Moshier, and R. E. Sievers, Inorg. Svn., 8, 50 (1967). G. T. Morgan and A. R. Bowen, J. Chem. Soc., 228, 1252 (1924). L. Wolf and C. Troltzsch, J. Prakt. Chem., 22, 78 (1962). E. M. Brainina and R. Kh. Freidlina, Bull. Acad. Sci. USSR, Div. Chem. Sci., 1489 (1961); Izv. Akad. Nauk SSSR, Ser. Khim., 1595 11961). R. Kh. Freidlina, E. M. Brainina, La. A. Petrashkevich, and M. Kh. Minacheva, Bull. Acad. Sci. USSR, Div. Chem. Sci., 1338 (1966); Izv. Akad.7Nauk SSSR, Ser. Khim., 1396 (1966). 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 137 E. M. Brainina, E. I. Mortikova, L. A. Petrashkevich, and R. Kh. Freidlina, Proc. Acad. Sci. USSR, 169, 681 (1966); Dokl. Akad. Nauk SSSR, 169, 335 (1966). R. C. Fay and R. N. Lowry, Inorg. Chem., 8, 1512 (1967). A. L. Lott, II, University of Oklahoma, Norman, Oklahoma, personal communication, 1970. T. J. Pinnavaia and R. C. Fay, Inorg, Chem., 2, 502 (1968). R. C. Fay and N. Serpone, J. Amer. Chem. Soc., 88, 5701 (1968). L. Atkinson and P. Day, J. Chem. Soc. (A), 2423 (1969). N. J. Bennett, Jr., F. A. Cotton, A. Davison, J. W. Faller, S. J. Lippard, and S. M. Morehouse, J. Amer. Chem. Soc., 88, 4371 (1966). A. C. Adams and E. M. Larson, Inorg. Chem., 8, 814 (1966). M. Gielen, Bull. Soc. Chim. Belges, 28, 351 (1969). 7 M. Cox, J. Lewis, and R. S. Nyholm, J. Chem. Soc., 6113 (1964). T. J. Pinnavaia, W. T. Collins, and J. J. Howe, J. Amer. Chem. Soc., 82, 4544 (1970). J. J. Howe and T. J. Pinnavaia, ibid., 82, 7342 (1970). APPENDIX A. Nmr Chemical Shifts in Benzene—d and Chloroform-d1 6 Table XV presents 1H chemical shifts in benzene-d6 and chloroform-d1 for the B-diketones which were used as ligands in this study. Table XVI presents 1H and 19F chemical shifts for the complexes which were studied in this investigation. 138 139 1 Table XV. H Chemical Shifts for B-Diketones in Benzene-d6 and Chloroform—d1.a Ligand Solvent C6H6 -CH= CH3 28:2;C4H9 H(acac)b C6D6 -5.00 -l.62 CDC13 -5.50 -2.05 H(dpm) C6D6 -5.71 —1.06 CDC13 -5 72 -1.16 H(hfac) C606 -5.67 , c01213 —6.36 H(bzbz) C606 -7.85, —7.13° -6.61 CDC13 -7.97, -7.47° -6.80 H(tfac) C6D6 -5.30 -1.34 CDC13 -5.91 -2.20 H(pvac)b 6606 -5.37 -1.70 -1.03 coc13 -5.64 —2.08 -l.l6 H(pvtf) c606 -5 85 —0.81 CDC13 -6.06 -1.22 H(bzac) 6606 -7.78, -7.13C —5.82 -1.75 CDC13 -7.90, -7.45C —6.17 —2.18 aShifts are given in ppm relative to TMS as an internal standard (1% by volume). The probe temperature is 42° and all concentrations are 10.0 g/100 ml solvent. bEnol form. CC6H6 resonances are broad multiplets. J.4l() Table xvx. ‘H and ‘9r Chemical Shifts for Some Zirconium Complexes in Benzene-d6 and Chloroform-d].a Complex Solvent C6H6b 65H5 -CH= CH3 5gv£7c4H9 CF3C (n-C5H5)ZrC12 c606° -5.91 CDC13: —6.48 (C5H5)4Zr (:606 -5.57 coc13 -5.83 (n-C5H5)Zr(acac)3 C606 -6.48 —5.31. —5.16 —1.92, -l.80. -1.5o CDC13 -6.11 .5.40. -5.22 -1.97. -1.89. -1.63 (n-C5H5)Zr(dpm)3 C606 -6.34 -5.86. —5.66 —1.33. -1.32. -1.12, -o.93 CDC13 -5.94 -5.61. -s.43 -1.16, -1.14. -o.97. -o.71 ("-C5H5)Zr(hfac)3 C6D6 -5.89 -6 10, -6.07 +0.22. +0.29. +0.60 CDC13 -6.42 —6.25. -6 12 +0.06, «0.27, +0.57 (n-C5H5)Zr(bzbz)3 C606 -8.13. -7.12 ~6.72 -6.89, -6.81 coc13 -8.00, -7 35 -6.48 —6.83, -6.78 (n-C5H5)Zr(tfac)3 C606 -6.23. -6.20 -5.72. -5.64. -1 73, -1.61, -0.86. -0.39. —5.60, -s.57 -l.58. -1.31 -0.35. -0.23 CDC13 —6.25. —6.22 -5.98. -5.81. -2.18. —2.11. -1.04. -o.62. -5.60. -5.57. -2 09. -1.80 -0.56. -o.39 (i-C5H5)Zr(pvac)3 6606 —6.46. -6.42 -5.59. -5.55. -1.97, -1.86. -1.28, .1.20. -s.42. .5.40 -1.83, .1.50 -1.18, —o.93 CDC13 -6.11. -6 09 -5.51. -5.43. -1.98, -1.92. -1.19. -1.08. -5.35. -s.29 ~1.88, -1 59 -1.o7. -0.80 (n—C5H5)Zr(pvtf)3 C606 -6 20 -6.05. -5.99. —1.10. -0.98. -1.11. -0.80. -5.97. -5.93 -o.74 -0.61 601:13 —6,19 -5,98, -5.96, —1.24. -1.14. —0.86. -0.77. ~5.85, -5.81 -0.86 -o.53 (n-C5H5)lr(bzac)3 C606 -8.07. -7.20 -6.63. -6 62 -6.16. -6.01. -2.15. -1.95. -5.90. -5.88 -1.93. -1.51 CDC13 —7.94. -7.40 -6.32 —6.19. -6.02. -1.38. -1.28. -5.88 -1.21. 50.87 (u-C5H5)Zr(dpm)2C1 C606 -6 47 -5.84, -5.76 -1.15. -1.12. -1.01, ~O.94 CDC13 -6.33 -5.81, -5.79 -1.19. -1.o3. 99 -0. aShifts are given in ppm relative to TMS as an internal standard (1% by volume). b C6H6 shifts are given in ppm relative to the free diketone as an internal standard (1% by volume) 7.4 g/100 ml solvent. The probe temperature is 42° and c 19 all concentrations are 10.0 g/100 ml solvent unless otherwise noted. resonances are multiplets. F chemical dConcentration is eConcentration is 6.7 g/100 ml solvent. fSaturated solution. 8. Linear Least-Squares Analysis, Computer Program This least-squares Fortran program, written by Charles Sokol for a CDC 3600 computer, was used for linear extrapolations of nmr line—shape parameters and for determining the best straight lines for the Arrhenius and Eyring plots of kinetic data. The program calculates the slope and intercept for a straight line along with their standard deviations. The data cards are prepared as explained by the comment cards in the program itself. 141 142 xqcuzc¢51335u520<$mzc mz+x 400200.xm.x5.0050 2450.xm.x5.8005Ie.fi.>0q::c:3 520 010600 mum $0500805<:ac0 L5.:3$20 05503 0m30535.08.m n Hac35.w 5003322 01o¢uc022 Lc_0525:_.L0000020.L000055201.L000vqcmzc zc~uzwz.: x051..0.x0:z.<:z.mx.5x 00mg N520330 Sqqcczo 35 143 C20 5:2_5zcu 00. oo 05 cc L0.m50.*w:252 ac m:000.>.5 m~u.*u»amoam525*.\.m.m_m.*m:255 ac m:41*.w.m~m.*nmac4m0.¢1¢.hdzacu c5 525>mohm.5auqm5z<.cgm>mapm.wQCOvLo5.Ho. m5503 LLxm.5.m550m.x5_.m5.* 055.LHVx.5VLo.5cV 05503 5$>*.xm_.*x*.x55.0qwm232-525208.*I#0 HfiZaCu x 00.500 0550? 42:m xzzm1>x2200*.xm.c.m5m.*uml<20w m 0.\.xm.c.n5m.$ucmam 002C55ac<39m mhmizv8.xm.3.m55.002\cwLxs:m0Icmx::ve.xm.c.m55.*n:0a0:cmszL>m qqzchpacsam xz:m01>xz:m0 *.\.o.m_0.802\5>5:m xzbwvi>xz:m*.xm.c.m50.0uz\c005 042c5503>22m08.xm.c.m50.002\0m5523m00.xm.0.mfim.0n2\0»z:w.02300548 805x::v5mm.5<0m55ez 002C550c .c.m50.0ucmL>23m00.xm.c.w5u.0nw Adchpaccmszzm.*.xm.o.mflu.*u>::v x23m0.xm.o.m50.*u>x?:m $.\.c.r~m.*u2v>§:5 qucwpacm00xmoc.mHu.*ucvx53m$.xm.c.m~u.*H>23m$.xm.c.m5m.*uxérv#.* *quzacu cm C. Nmr Line-Shape Analysis, Computer Program This program has evolved into its present status through several minor modifications in a program originally written by John M. Sebeson, II. It is used to calculate nmr line shapes by use of the Rogers-Woodbrey modification of the Gutowski-Holm equation.16 The program indexed here has been used with a CD 3600 computer, but minor changes will permit it to be used with CDC 6500 and IBM 1130 computers. The input for the program is as follows: Card #1 01 1-80 Card #2 col 1-10 col 1-20 21-30 31-40 41—50 Card #3 col 1-10 11-20 21-30 31—40 41—50 51—60 Identification of the run separation in absence of exchange of the two resonance components, in Hz T2A in seconds where A is the low-field peak TZB in seconds where B is the high-field peak Pa’ decimal fractional p0pu1ation of the low-field peak Pb, decimal fractional p0pu1ation of the high—field peak N1, number of tau values (right justified) N2, number of frequency values/spectrum increment in tau, in seconds increment in omega (frequency), in Hz lower limit of tau, in seconds lower limit omega (frequency) in Hz Repeating this sequence of three card may be done for another data set. After the last card of the last data set, there should be two blank cards. 144 145 40225500 5.3.0.>x2:u.cv>§:m.30x33m.>5:v.x;:vLcm.50005.03 5m..*u 0250 FIGHkaU LC\cmw0051:wnCLU>0C5J + c\cvmnl 5 Lmicv\zcvmucvu 0\Cm012m1cw>r:muzomm 5 rimvyzrmno 0 N48000m0 fl 2>x23m1>x23000 0\1.._HZM1._ .55. O\:Hq .105 C\Uuz>xz:v. m. m00>23mnm ,5. K050505030”: 5. 553.50.053.20“... 5x00x22ml3mXEIM$fv\A>x2:m*xftvl>?:u$cvy§:mv”Pauzuhfq Am**x23m13mx23w$cv\A>EJI$X?:UI»X§ZV*TVH0&C.m % inc 5504052070 3 m0855v>+3fyzrmncu>firm J Apv>+>22mn>zrv L50>0050x.>x2:mn>xrsm 5 «#0550x.cmxztvucvxz:u 0 050x+x3300x2:v ACIZHRV 51.22.10. 51 000>.L50x 05.000 0000 5.2. H5 5 CC Liqcflv 5123cu 55 L_.c.._u5.050500:50055.000 £003 x 0.005.00_ 020 05 .FauuamvzH mI5 05 Hauam520.m:50 5ICH<35U ¢I5 2: 052560 050: 00 20x::2 015 m5 2 cn>xz:m 0 cucv>22m 5 :ncmx5:m 5 cu>2:m 0 :nxsrv 5050 5420C0 5 7 00.000 C003 00 .2100 050: y200x 023 ml 00::10 000:5 .?:3 5000 mr5 uc QI<1 20c :C» zwrp .> a: u:0<> 0:5 100 :m C5 .5 722:0:u 920 x u: 0:40> 5:» arm 05 C5 5 mzs:0co c250: 525:0 0052055303x8 Iccm 3C0 C1 2015.00 C5 5 022:0:0 025m: 231 0:5 >0552025 20:0:m 0000 055: 020000 015.m :5 5 m2220co 25 mHZHCQ 0qhzmz5quax0 0c 3805:: 0:5 0>01 c0:CIm £009 050: .0050 m15 .wz:a m0055032 05002u00u c» 309: ml 200 :qaccza 0515.0525c1 0052025101x0 @0300 a: ocm CZHZquzCU 0254 HIOH<15m 4 1:0 510umm525 224 00000 0:5 0: m205505>m3 ca