"' \ nggaraggiiif i c’m—I" .. '7.) Ufli‘i’efi‘gfigy ‘ __I *w . THESIS G This is to certify that the dissertation entitled A. FERROCENYLSULFIDES: PREPARATION AND REACTIVITY AS BIDENTATE LIGANDS B. TRIS(CYCLOPENTADIENYL)ZIROCONIUM DERIVATIVES presented by 1 Beth McCulloch has been accepted towards fulfillment of the requirements for 335; H, m w \ Major professor Date a: “' 21 ~83 MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 IVISSI_I RETURNING MATERIALS: Place in book drop to LianAmgs remove this checkout from “ your record. FINES will be charged if book is returned after the date stamped below. _ . - - an, a“. . A . : ~. ‘ f _ 1:. ,al ‘- . *w ‘ l . I . I 1" a... ‘4; Inis _ :Io A. FERROCENYLSULFIDES: PREPARATION AND REACTIVITY AS BIDENTATE LIGANDS B. TRIS(CYCLOPENTADIENYL)ZIRCONIUM DERIVATIVES By Beth McCulloch A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1983 ABSTRACT A. FERROCENYLSULFIDES: PREPARATION AND REACTIVITY AS BIDENTATE LIGANDS B. TRIS(CYCLOPENTADIENYL)ZIRCONIUM DERIVATIVES By Beth McCulloch Ferrocenylsulfides, Fe(CSHuSR)2 (R = Me, iPr, iBu, iPent, Ph, Bz) have been prepared by lithiation of ferro- cene followed by reaction with the appropriate disulfide. These complexes have been characterized by use of spectro- scopic techniques such as proton and carbon NMR, ultra- violet and visible and infrared spectroscopy. The ferro- cenylsulfide derivatives readily chelate palladium and platinum halides to form the heterobimetallic complexes, Fe(C5HuSR)2MX (R = Me, iPr, iBu, Ph, BZ; M = Pd, Pt; X = 2 Cl, Br). Proton, carbon and platinum NMR spectra were ob- tained where possible and infrared, ultraviolet and visible and cyclic voltammetry data of the bimetallic complexes is presented. An X-ray crystal structure of Fe(C5HuS-1Bu)2- PdCl2 was determined. Dynamic NMR studies suggest that pyramidal sulfur inversion and bridge reversal occur in Beth McCulloch solution. Activation parameters for sulfur inversion in Fe(CSHuS-iBu)2PdX2 (X = Cl, Br) were determined as 13.88 i 0.009 and l3.A2 : 0.12 kcal/mol for the chloride and bromide complexes respectively. Variable temperature Platinum-195 NMR data indicates the presence of two di- astereoisomers that increase in relative population as the temperature is lowered. A chiral ferrocenylsulfide, (CSHS)FeCSH3[CH(CH3)N- (CH3)2][SCH2Ph], was prepared and its relevance to asym- metric hydrogenation is discussed. A series of dialkyldi- thiocarbamateferrocene derivatives, Fe(C5HuSCSNR2)2 and Fe(C5H5)(CSHuSCSNR2) (R = Me, Et, iPr) were prepared by re- action of lithioferrocene vith tetraalkylthiuram disulfides. Proton and carbon NMR, Optical and infrared data were ob- tained. Dynamic NMR studies indicate that restricted rota- tion occurs around the carbamate carbon-nitrogen bond in the methyl and ethyl derivatives. Approximate rotational free energy barriers were determined and were correlated with the "thioureide" band in the infrared. Variable temperature NMR spectra of the is0pr0pyldithiocarbamateferrocene derivatives are dominated by restricted rotation of the isopropyl-nitro- gen bond and at least two conformers are present at low tem- perature. The dialkyldithiocarbamate ferrocenes are compared to dithiocarbamate ligands and palladium complexes of the ethyl derivativeswere-investigated. The tris(cyclopentadienyl)zirconium complexes, Beth McCulloch (05H5)3er (X = Cl, nBu, Me), were prepared. The thermal stability of (05H5)3ZrBu was attributed to the three n-bonded cyclopentadienyl rings that block the beta elimin- ation pathway. Reaction of (05H5)3ZrMe with carbon mon- oxide was investigated. To My Parents and My Husband, Ralph ii ACKNOWLEDGMENTS I would like to express my appreciation to Professor Carl H. Brubaker, Jr. and to the members of the group for their help and friendship. In addition, I would like to thank Dr. T. J. Pinnavaia for many helpful discussions and Kermit Johnson for tech— nical expertise in obtaining NMR spectra. iii Chapter TABLE OF CONTENTS LIST OF TABLES. LIST OF FIGURES LIST OF ABBREVIATIONS A. FERROCENYLSULFIDES: PREPARATION AND REACTIVITY AS BIDENTATE LIGANDS I. II. INTRODUCTION. EXPERIMENTAL. General Techniques. l,l'-Bis(methylthio)ferrocene (2) l,l'—Bis(isobutylthio)ferrocene (l0). l,l'-Bis(isopropylthio)ferrocene (ll) 1,1'-Bis(phenylthio)ferrocene (l2). 1,1'-Bis(benzylthio)ferrocene (lé). 1,1'-Bis(isopentylthio)ferrocene (lg) l,l'-Bis(diphenylphosphino)ferrocene (lg) Preparation of Metal Complexes. l,l'-Bis(isobutylthio)ferrocene- palladium dichloride (IQ) 1,1'-Bis(isobutylthio)ferrocene- palladium dibromide (l1). l,l'-Bis(isobutylthio)ferrocene- platinum dichloride (lg). 1,1'-Bis(isobutylthio)ferrocene- platinum dibromide (lg) iv Page ix . xii . xvi IA IA 16 17 18 19 2O 21 21 22 22 23 23 23 Chapter Page l,l'-Bis(isopropylthio)ferrocene- palladium dichloride (20) . . . . . . . . . 2A l,l'-Bis(isopropylthio)ferrocene— palladium dibromide (2%). . . . . . . . . . 2A l,l'—Bis(isopropylthio)ferrocene- platinum dichloride (2%). . . . . . . . . . 2A l,l'-Bis(isopropylthio)ferrocene- platinum dibromide (2%) . . . . . . . . . . 25 l,l'-Bis(methylthio)ferrocenepalladium dichloride (2Q) . . . . . . . . . . . . . . 25 l ,l'-Bis(methylthio)ferrocenepalladium dibromide (25). . . . . . . . . 25 l,l'-Bis(methylthio)ferroceneplatinum dichloride (26) . . . . . . . . . . . . . . 25 1,1'—Bis(methylthio)ferroceneplatinum dibromide (27). . . . . . . . . . . . . . . 26 l,l'-Bis(benzylthio)ferrocenepalladium dichloride (2Q) . . . . . . . . . . . . . . 26 l, l' -Bis(benzylthio)ferrocenepalladium dibromide (2%). . . . . . . . . . . . . 26 l,l'-Bis(benzylthio)ferroceneplatinum dichloride (ég) . . . . . . . . . . . . . . 26 l,l'—Bis(benzylthio)ferroceneplatinum dibromide (§l). . . . . . . . . . . . . . . 27 l,l'-Bis(phenylthio)ferrocenepalladium dichloride (52) . . . . . . . . . . . . . . 27 l, l'-Bis(phenylthio)ferrocenepalladium dibromide (éé). . . . . . . . . . . . . . 27 l, l'- Bis(pheny1thio)ferroceneplatinum dichloride (QM) . . . . . . . . . . . . 28 l,l'-Bis(phenylthio)ferroceneplatinum dibromide (éé). . . . . . . . . . . . . . . 28 N, N-dimethyl-a- -ferrocenylethylamine (g6)........ ......28 Chapter Page (R)—a-[(S)—2-benzylthioferrocenylJ— ethyldimethylamine (31) . . . . . . . . . . 29 1,1'-Bis(dimethyldithiocarbamate)- ferrocene (3Q). . . . . . . . . . . . . 30 1,l'-Bis(diethyldithiocarbamate)- ferrocene (3g). . . . . . . . . . . . . . . 31 1 ,l'-Bis(diisopropyldithiocarbamate)- ferrocene (A0). . . . . . . . . . 32 Dimethyldithiocarbamateferrocene (A1) . . . 33 Diethyldithiocarbamateferrocene (A2). . . . 3A Diisopropyldithiocarbamateferrocene (A3). . 35 Reaction of Fe(C5HuSCSNEt2)2 (32) with (PhCN)2Pd012. . . . . . . . . . . . . . . . 36 Reaction of FeCp(C5HuSCSNEt2) (A2) with (PhCN)2PdCl2. . . . . . . . . . . . . . . . 36 III. RESULTS AND DISCUSSION. . . . . . . . . . . 37 A. Fe(CSHuSR)2 (R = Me, iPr, iBu, iPent, Ph, Bz) . . . . . . . . . . . . . . . . . . 37 1. Preparation . . . . . . . . . . . . . . 37 2. l H NMR. . . . . . . . . . . . . . . . . 39 3. 13c NMR . . . . . . . . . . . . . . . . 39 A. Infrared Spectra. . . . . . . . . . . . AA B. CSHSFeCSHBECH(CH3)N(CH3)2][SCH2Ph]. . . . . A6 C. Fe(C5HuSR)2MX2 (R = Me, iPr, iBu, Ph, Bz; M = Pd, Pt; X = Cl, Br) . . . . . . . . A9 1. Preparation . . . . . . . . . . . . . . A9 I 2. H NMR. . . . . . . . . . . . . . . . . 50 3. 13c NMR . . . . . . . . . . . . . . . . 55 vi Chapter Page A. Infrared Spectra. . . . . . . . . . . . 56 Ultraviolet and Visible Spectra . . . . 60 Structure . . . . . . . . . . . . . . . 63 Dynamic NMR Studies . . . . . . . . . . 7A 195Pt NMR . . . . . . . . . . . . . . . 87 \OGJNCMUW Electrochemistry. . . . . . . . . . . . 90 D. Fe(C5HuSCSNR2)2 and FeCp(C5HuSCSNR2) (R = Me, Et, iPr) . . . . . . . . . . . . . 96 1. Preparation . . . . . . . . . . . . . . 96 2. 1H NMR. . . . . . . . . . . . . . . . . 99 3.13CNMR................99 A. Ultraviolet and Visible Spectra . . . . 10A 5. Dynamic NMR Studies . . . . . . . . . . 108 6. Metal Complexes of Fe(C5HuSCSNEt2)2 and FeCp(CSHuSCSNEt2) . . . . . . . . . 116 B. TRIS(CYCLOPENTADIENYL)ZIRCONIUM DERIVATIVES I. INTRODUCTION. . . . . . . . . . . . . . . . 123 II. EXPERIMENTAL. . . . . . . . . . . . . . . . 125 General Techniques. . . . . . . . . . . . . 125 Tris(cyclopentadienyl)zirconium- chloride (AB) . . . . . . . . . . . . . . . 125 Attempted Preparation of Tris(cyclo- pentadienyl)zirconiumbutyl (A9) . . . . . . 126 Tris(cyclopentadienyl)zirconiumbutyl (A9) . 127 Tris(cyclopentadieny1)zirconium- methyl (59) . . . . . . . . . . . . . . . . 127 vii Chapter Page Photolysis of Tris(cyclopentadieny1)- zirconiummethyl under CO atomosphere. . . . 128 III. RESULTS AND DISCUSSION. . . . . . . . . . . 129 l. Cp3ZrCl . . . . . . . . . . . . . . . . 129 2. Cp3ZrR (R = nBu, Me). . . . . . . . . . 131 3. Reaction of Cp3ZrMe with CO . . . . . . 135 A. Conclusions . . . . . . . . . . . . . . 136 APPENDIX A — Photolysis of (u-Oxo)Bis(Chloro- zirconocene) . . . 138 APPENDIX B. . . . . . . . . . . . . . . . . . . . . 1A2 REFERENCES. . . . . . . . . . . . . . . . . . . . . 150 viii Table LIST OF TABLES Metal Complexes Derived from Ferro- cenylphosphine and Ferrocenylarsine Ligands . . . . . . . . . . . . 1H NMR Data for Ferrocenylsulfide Com- plexes, Fe(C5HuSR)2 R = Me, iPr, iBu, iPent, Ph, Bz. Spectra obtained in CDCl3 at room temperature 13C NMR Data for Ferrocenylsulfide Com- plexes, Fe(C5HuSR)2, R = Me, iPr, iBu, iPent, Ph, Bz. Spectra obtained in (CH2C12/D2O) solution at room tem- perature. Infrared Data for Ferrocenylsulfide Complexes, Fe(C5HuSR)2, R = Me, iPr, iBu, Ph, Bz 1H NMR and 13c NMR Data for Fe(C5HuS-iBu)2 and Metal Derivatives, PdCl2, PdBr2, PtCl2 and PtBr2 . . . . . . . . . 2 (R = Me, iPr, iBu, Ph, Bz; M = Pd, Pt; x = Cl, Br) in Region Moo—200 cm‘l. Infrared Data for Fe(C5HuSR)2MX ix Page A0 A2 A5 51 Table 6 Cont 9A 9B 10 11 12 13 1A 15 Measured as Nujol Mulls between CsBr Plates. Electronic Absorption Spectra of Ferrocenylsulfide complexes Fe(C5HuSR)2, Fe(CSHuSR)2MX (R = iBu, Ph, Me; M = 2 Pd, Pt; X = Cl, Br) at 2A°C in MeCN solution at approx. conc. 8.0 x 10—5 M. Positional Parameters and Esd's for Fe(C5HuS—iBu)2PdCl2 Anisotropic Thermal Parameters with I Esd s for Fe(C5HuS-iBu)2PdC12 IsotrOpic Thermal Parameters with Esd's for Fe(C5HuS—iBu)2PdCl2 Selected Bond Distances (A) with Esd's for Fe(C5HuS-iBu)2PdC12 Selected Bond Angles (deg) with Esd's for Fe(CBHuS-iBu)2PdC12 Dihedral Angle and Bridgehead Angle of selected [3]ferrocenophanes. Least-Squares Planes in Fe(C5HuS-iBu)2PdC12 Torsion Angles in Fe(C5HuS-iBu)2PdCl2 195Pt NMR Data for Ferrocenylsulfide- platinum Complexes, Fe(C5HuSR)PtX2, R = iBu, iPr, Ph, Bz; X = Cl, Br. Page 58 61 66 68 69 71 72 7A 75 76 Table Page 15 Cont Measurements made in CDCl3 solution l6 l7 l8 19 20 21 at room temperature . . . . . . . . . . . . 88 Cyclic Voltammetry Data for Ferro- cenylsulfide Complexes. . . . . . . . . . . 9A 1H NMR Data for Fe(C5HuSCSNR2)2 and Fe(C5H5)(C5HuSCSNR2) Complexes (R = Me, Et, iPr) . . . . . . . . . . . . . 100 130 NMR Data for Fe(C5HuSCSNR2)2 and Fe(C5H5)(C5HuSCSNR2) (R = Me, Et, iPr). Measured in CHZCl/D2O solution at ambient temperature unless otherwise shown . . . . . . . . . . . . . . 102 Electronic Absorption Spectra of Fe(C5HuSCSNR2)2 and Fe(C5H5)(C5HuSCSNR2) (R = Me, Et, iPr) at 2u°c in MeCN solution at approx. conc. 8.0 x 10"5 M. . . 106 NMR Parameters, Kinetic and Infrared 2)2 (CSHuSCSNRz) where R = Me, Et . . . . . . . 110 1 Data for Fe(C5HuSCSNR and Fe(CSH5)— H NMR Data for Cyclopentadienyl- zirconium Complexes . . . . . . . . . . . . 130 xi LIST OF FIGURES Figure Page 1 Proton decoupled (above) and gated decoupled (below) 13C NMR Spectra of Fe(CSHuS-iPr)2 . . . . . . . . . . . . . A3 2 1H NMR Spectra of Fe(C5HuS-iBu)2 and Fe(C5HuS-iBu)2PdCl2 . . . . . . . . . . 53 3 1H NMR Spectra of Fe(C5HuS-iPr)2PdC12 and Fe(C5HuS-ipr)2PtCl2 o o o o o o o o o o 5“ A Infrared Spectra of (A) Pe(C5HuSPh)2, (B) Fe(C5HuSPh)2PtC12 and (C) Fe- (CSHuSPh)2PdBr2 . . . . . . . . . . ... . . 57 5 Far Infrared Region for Fe(C5HuS-iBu)2— PdX2 and Fe(C5HuS-iBu)2PtX2 where X = C1 (above), and Br (below). . . . . . . 59 6 Structure and Numbering Scheme for Fe(C5HuS-iBu)2PdC12 . . . . . . . . . . . . 6A 7 Stereoview of Fe(C5HuS-iBu)2PdC12 . . . . . 65 8 Variable Temperature 1H NMR Spectra fOP Fe(CsHuS-1BU)2PdCl2 o o o o o o o o o o 77 9 Variable Temperature 1H NMR Spectra for Fe(C5HuS-iBu)2PtC12 . . . . . . . . . . 83 xii Figure 10 11 12 13 1A 15 16 17 18 19 20 Page Variable Temperature 13C NMR Spectra for Fe(C5HuS-iBu)2Pd012 in Region from 70-80 ppm. . . . . . . . . . . . . . . 86 Variable Temperature 195Pt NMR Spectra for Fe(C5HuS-iBu)2PtCl2 89 Variable Temperature 195Pt NMR Spectra for Fe(C5HuS—iPr)2PtC12 . . . . . . . . . . 91 Cyclic Voltammograms of (A) Fe(C5HuS-iBu)2, (B) Fe(C5HuS-iBu)2Pt012 and (C) Fe(C5HuS-1Bu)2PdCl2 o o o o o o o c o o o o o 93 1 H NMR Spectra of Fe(C5H5)(C5HuSCSNEt2) (above) and Fe(C5HuSCSNEt2)2 (below)- . . . 101 13c NMR Spectra of (A) Fe(C5HuS-1Pr)2, 2)2 and (C) Fe— . . . . . . . . . . . . A (C5H5)(C5HMSCSNEt2) 10 (B) Fe(C5HuSCSNiPr 13 C NMR Spectra Of Fe(C5H5)(C5HuSCSNMe2) (above) and Fe(C5HuSCSNEt2)2 (below). . . . 105 Ultraviolet—visible Spectra for (A) Fe(C5HuSCSNMe2)2 and (B) Fe(CSH5)- (CSHuSCSNMez) . . . . . . . . . . . . . . . 107 Variable Temperature 1H NMR Spectra of Fe(C5HuSCSNiPr2)2. . . . . . . . . . . . 111 1 Variable Temperature H NMR Spectra ' . . . . . . . . . 112 of Fe(C5H5)(C5HuSCSNiPr2) Conformers A and B of Fe(C5H5)- (CSHMSCSNiPrZ). . . . . . . . . . . . . . . 11A xiii Figure Page 21 Slow exchange 1H NMR Spectrum of Fe(C5H5)(C5HuSCSNl-Pr2) . . . . . . . . . . 11A 22 Resolution Enhanced 1H NMR Spectrum of Fe(C5H5)(CSHuSCSNiPr2) at Slow Exchange. . . . . . . . . . . . . . . . . . 115 23 Infrared Spectra of (A) Fe(CSH5)- (C5HuSCSNEt2) and (B) Pd Complex: (C) Fe(C5HuSCSNEt2)2 and (D) Pd Complex . . . . 118 2A Infrared Spectra of Fe(C5HuSCSNEt2)2 2 (below) . . . . . . . . . . . . . . . . . . 121 (above) and Fe(C5HuSCSNEt2)2PdCl 25 Infrared Spectra of Fe(C5H5)- (OSHASCSNEt2) (above) and Fe(C5H5)- (CBHuSCSNEt2)PdCl2 (below). . . . . . . . . 122 26 ESR Spectrum obtained from reaction of Cp3ZrC1 with BuLi in THF solution. . . . 133 Al 1H NMR Spectra of (Cp2ZrCl)2O during photolysis at (A) t = 0, (B) t = 75 min, (C) t = 6 h 35 min and (D) t = 16 h. . . . . . . . . . . . . . . . 1A0 A2 ESR Signal Obtained from Photolysis of (Cp2ZrCl)2O. . . . . . . . . . . . . . . 1A1 B1 1 H NMR Spectra of Fe(C5HuSPh)2 (above) and Fe(C5HuSCH2Ph)2 (below). Inset cyclopentadienyl proton region. . . . 1A2 xiv Figure Page B2 1H NMR spectrum of Fe(C5HuS—iPr)2 (above) and 13C NMR spectrum of Fe- (CSHuSCNiPr 2 in CDCl at 57°C 2) 3 (below) . . . . . . . . . . . . . . . . . . 1A3 13 0 B3 C NMR Spectra of Fe(C5HuSCH2Ph)2 proton decoupled (above) and gated decoupled (below) . . . . . . . . . . . . . 1AA BA 130 NMR Spectra of Fe(C5HuSMe)2 proton decoupled (above) and gated decoupled (below) . . . . . . . . . . . . . 1A5 1 B5 H NMR Spectra of Fe(C5HuSPh)2PtCI 2 (above) and Fe(C5HuSPh)2PtBr2 at 50°C (below). . . . . . . . . . . . . . . . 1A6 B6 195Pt NMR Spectra of Fe(C5HuPPh2)2- PtCl2 (above) and Fe(C5HuS—iBu)2PtC12 (below) . . . . . . . . . . . . . . . . . . 1A7 1 B7 H NMR Spectra of Fe(C5H5)(CSHuSCSNMe2) (above) and Fe(C5HuSCSNMe2)2 (below). . . . 1A8 1 B8 H NMR Spectra of Fe(C5H5)(C5HuSCSNiPr2) (above) and Fe(C5HuSCSNiPr2)2 (below) . . . 1A9 XV Bz Cp Et Fc dep iBu iPent iPr Me Ph CH Cl DMF MeCN THF SCE LIST OF ABBREVIATIONS benzyl cyclopentadienyl ethyl ferrocenyl 1,l—Bis(diphenylphosphino)ferrocene isobutyl iSOpentyl iSOpropyl methyl phenyl methylene chloride N,N—dimethylformamide acetonitrile tetrahydrofuran standard calomel electrode xvi I. INTRODUCTION Ferrocene, which was discovered in 1951, has proved to be a remarkably stable and unusually reactive organo- iron complex. It readily undergoes a variety of aromatic substitution reactions such as acylation, alkylation, formylation, mercuration and sulfonationl. Electrophilic substitution reactions are however limited to electrOphiles which do not oxidize the iron atom or destroy the cyclo— pentadienyl ring-metal bond. The metalation reaction com- plements electrophilic substitution in that it provides an alternate route to introducing reactive functional groups on to ferrocene. Metalation may be achieved by reaction of ferrocene with butyllithium, amylsodium or phenylsodiumZ. Ferrocene is dilithiated in over 90% yield by a mixture of butyl- lithium and tetramethylethylenediamine (TMEDA)3. h’ @UZN] Fe "‘8" Li > FE A TMEDA V © ""1 Lu W A The dimetalated species may be isolated as a pyrophoric red—orange crystalline solid where TMEDA chelates the di- lithium reagent. Use of the lithium reagent isolated as a solid, rather than the in situ slurry, leads to higher yields in the subsequent reaction with electrophiles. The advantage of the butyllithium/TMEDA metalation reagent is that only dilithioferrocene is produced whereas sodium and potassium derivatives of ferrocene result in mixtures of the mono and dimetalated species. In contrast to dilithioferrocene, preparation of mono— lithioferrocene by addition of stoichiometric amounts of butyllithium/TMEDA to ferrocene leads to a mixture of the monolithiated and dilithiated species“. Another route to lithioferrocene where alkyllithium is added to chloromer- curiferrocene produces a reactive dialkylmercury compound that forms undesirable side productsS. High yields of lithioferrocene, with no concurrent dilithiation, can however be obtained by reaction of butyllithium and bromoferro- cene6. The organic chemistry of ferrocene is extensive and literally thousands of derivatives have been prepared. One of the more recent applications of ferrocene is its use as a ligand in transition metal complexes7. Metala- tion has proved to be a useful synthetic method for the introduction of potential donors such as phosphines and arsines on to the cyclopentadienyl ring. Scheme 1 illustrates the variety of different donor substituents that may be incorporated into ferrocene. In 1970 Davison prepared ferrocenylphosphines and ferrocenyl— arsines in high yield from l,l'-di1ithioferrocene. In addition reaction of dilithioferrocene with elemental sul- fur gave 1,2,3-trithia-[3]ferrocenophane (l) which can be reduced quantitatively to l,l'-dithiolferrocene (2)8. Recently Osborne has prepared the selenium analog (3)9. An interesting class of compounds, the [lJ—ferroceno- phanes which have phosphorous, arsenic or Gp 6A elements as the bridging atoms, have been obtained from dilithio- 10,11 2 Ph; M = Si, R = Ph, C1)12, respectively. These compounds ferrocene and RPCl2, RAsCl or R2MC12 (M = Ge, R = exhibit unusual spectroscopic prOperties as the cyclo- pentadienyl rings are severely tilted towards the bridge atom. Wrighton and coworkers13 have used (l,l'-ferro— cenediyl)dichlorosilane to derivatize a number of elec- trode and silica surfaces by opening the highly reactive, strained C—Si-C bond in the ferrocenophane. The [lJ-ferrocenophanes are cleaved by alkyllithium reagents to give a ring opened ferrocenyllithium reagent. Subsequent reaction with electrophiles gives rise to ferro- cene derivatives with mixed functionality as in (A). Re- 11 reported the preparation of ring—substituted cently Cullen ferrocenophanes with phOSphorous and arsenic bridges. These are precursors to chiral ferrocenes with mixed functionality Q SH 5\ LAH Q / Q @n @va. Fe PPh PPhC'z ,9 P RZCI fl ‘ Fe ©/ éu PR, Ph Li Ismael Se ““2“ @— PPh, Q“ @“R2 \ Fe Fe Se Fe ©m g.) @M, Scheme 1 that have important applications in asymmetric synthesis. Table 1 contains metal complexes specifically derived from either ferrocenylphosphine or ferrocenylarsine ligands. DavisonlLI has prepared a variety of complexes with first row transition metal halides and 1,1'-bis(diphenylphosphino)- ferrocene, (fdpp), and concluded that fdpp is a rigid, sterically demanding ligand that reacts very much like l,2—bis(diphenylphosphino)ethane, (diphos). In a study of the reaction of ferrocenylarsines with Gp 6 metal carbonyls and Op 8 metal halides DavisonlS’l6 concluded that only l,l'-bis(dimethylarsino)ferrocene, (fdma), formed bis chelate complexes such as (fdma)2Mo(CO)3 and Pd(fdma)201+ and Pd(fdma)22+. The authors postulate that the 1,1'-bis(diphenylarsino)ferrocene ligand, (fdpa), forms only mono chelate complexes owing to the steric bulk of the phenyl groups. Ferrocenylphosphine ligands have recently been employed 18 in catalysis. Workers at Celanese reported that rhodium complexes containing 1,1'-bis(diarylphosphino)ferrocenes induce high rates and high selectivity to linear aldehydes in alpha-olefin hydroformylation. In particular the rhodium complex of l,l'-bis[bis(o,a,a-trifluoro-p-tolyl)phosphino]- ferrocene was found to catalyze the hydroformylation of l- hexene to give 100% conversion with 93.2% selectivity to hexanallg. A mechanistic study of the highly selective catalyst suggests that three phosphorous atoms are bound to each rhodium atom in a dirhodium complex. Table 1. Metal Complexes Derived from Ferrocenylphosphine and Ferrocenylarsine Ligands. Metal Complex Ref. 5223 (PhCN)2MCl2 M = Pd, Pt (dep)MCl2 l7 RhH(CO)(PPh3)3 (dep)RhH(CO)(PPh3)3 18 CuI (dep)CuI 1A M(CO)6 M = Cr,Mo,w (dep)M(CO)u v6 [(dep)V(CO)u1(NEtu) Ta(CO)6 NEtu [(dep)Ta(CO)u](NEtu) MX2 M = Co,Ni (dep)MX2 X = Cl,Br,I HgX2 X = Cl,Br,I,SCN (dep)-anX2 n = 1,2 20 Hg(CN)2 [(dep)2Hg1(BFu)2 SnXu (dep)-n SnXu n = 1.5 (Et2S)2PtCl2 (dep)Ptc12 21 dea,dea M(CO)6 M = Cr,Mo,w (dea)M(CO)u; 15 (dea)M(CO)u; (dea)2Mo(CO)3; u-(dea)-[(dea)Mo(CO)3]2 MX2 M = Pd,Pt (dea)MX2; (dea)MX2 16 x = C1,Br,I (dea)MX+; (dea)2M2+ 22 Rh(C8Hl2)C12 [Rh(dea)2](PF6) Seyferth has very recently reported the use of 1,1'- ferrocenylenephosphine oligomers in the cobalt catalyzed hydroformylation of l-hexene23. The polymeric ligands chelate to cobalt in a tridentate fashion and give hydro- formylation results comparable to triphenylphosphine ligands. Palladium phosphine complexes are known to catalyze cross-coupling reactions between organometallic compounds and alkyl and aryl halides. Kumadal7 has shown that the ferrocenylphosphine ligand in the complex, (fdpp)PdCl2, is superior to diphos for coupling sec-butylmagnesium- chloride with bromopropene, bromobenzene and bromostyrene. Another recent development in ferrocene chemistry is the use of chiral ferrocene derivatives as ligands in transition metal catalyzed asymmetric synthesis. Chiral ferrocenylphosphines are prepared by the stereoselective lithiation of a—ferrocenylethyldimethylamine followed by treatment with chlorophosphines (See Scheme 2). Rhodium complexes with chiral ferrocenylphosphine ligands such as (5) and (6) have been used as catalysts in asymmetric hydrogenation, Grignard cross-coupling and hydrosilyla— tiongu. In particular, acylaminoacids have been produced in over 90% optical purity by the asymmetric hydrogenation of a-acetaminocinnamic acids catalyzed by a rhodium com- plex of (5) as shown below. H 8 § “hr 5 ““9 NMe , . NMe, Fe 2 wBuLI Fe m 1. n-BuLi 2.nBuLi/TMEDA CIPR, icnPR, v v NNhe Nflkl Fe PR2 2 Fe PR2 2 i R = Ph 2 R : t-Bu Scheme 2 Cullen recently reported that the alkylferrocenylphosphine, (6) shown in Scheme 2, was more effective in asymmetric hydrogenation of acetamidoacrylic acids than the aryl ana- He also concluded that rhodium complexes containing ferrocenylarsine ligands were not catalytically active in asymmetric hydrogenation. Ph NHCOMe H2 NHCOMe . \ / / c = c ——————- PhCH2CH \ \ H coon é/Rh coon Kumada was able to hydrogenate carbonyl compounds with optical yields ranging from A3 - 83% by introducing a hydroxyl group on to a chiral ferrocenylphosphine ligand. Previous attempts with other classes of phosphine ligands led to less than 10% stereoselectivity26. The high asym- metric induction obtained by Kumada can be ascribed to hydrogen bonding between the carbonyl group on the substrate and the hydroxy group on the chiral ligand. These attrac- tive interactions increase the conformational rigidity of the diastereomeric transition state and consequently enhance the stereoselectivity of the reaction. Since chiral ferro- cenylphosphines can be readily modified by introducing dif- ferent functionality into the side chains27 they have distinct advantages over other classes of phosphine ligands in that they may be tailored to "fit" specific substrates. In contrast to tertiary phosphines, there has been very little interest in the chemistry of metal complexes contain— ing organic sulfides. Two recent reviews deal with the co— ordination chemistry of thioether ligands with transition metalsZB. Two structures are possible, one where a single thioether ligand chelates to the metal, (7) or alternatively where two ligands chelate to the metal, as in (8), to give a complex similar to Magnus' green salt. Traditionally metal sulfides have been used as catalysts in hydrodesulfurization29 but recently there has been much interest in the catalytic activity of transition metal sulfide complexes. 10 R 'i' '3 2' I S 8\ /CI \ /s 2’ M NI InICh] / ‘\ / ‘\ ? c' ? s.» R R R l .9. lePHJN Many metal clusters with thiolato bridges have recently been reported30. A binuclear rhodium complex [Rh2(uS-tBu)2— (POMe3)u] was reported to be catalytically active in the hydroformylation of l-hexene3l. In addition n3-ally1 metal sulfur clusters, [nB-C3H5MS(n3-C3H5M')]X where M, M' = Pd, Pt, were found to selectively hydrogenate acetylenes to 32 olefins and to isomerize olefins Rakowski-DuBois has recently reported molybdenum dimers of the type (CpMoS)2S2CR2, R = H, alkyl, which reduce acetylenes to olefins and catalyze the hydrogenation of molecules containing C = N, N = N and C = S bonds33. Very few thioether transition metal complexes have 3A however proved to be effective catalysts. James has re- ported that RhCl3(SEt2)3 hydrogenates maleic acid but as sulfur ligands are weaker n-acceptors than phosphine li- gands, Rh(l) sulfide complexes can be reduced by hydrogen to rhodium metal. James has also prepared chelating chiral sulfoxide ligands and has found that ruthenium 35. complexes are active in asymmetric hydrogenation 11 The object of this research was to develop a new class of chelating ferrocenyl ligands. The recent interest in transition metal sulfides led to the investigation of the preparation of ferrocenylsulfides. A few ferrocenylsulfide complexes are known in the 36 literature. Pauson has prepared methylthioferrocene from 37 ferrocenesulfonic acid whereas Russian workers prepared this complex from thiocyanatoferrocene. FCSO3H -> FcSOZCl -> FCSH FcSMe FcHgCl -¢-FcSCN Ferrocenylmethylsulfides have recently been prepared from 38 ferrocene and mercaptans in a one step synthesis HCHO FcH + HSR -—->- FcCH2SR HClOu These procedures are limited to the preparation of specific 39 ferrocenylsulfide complexes. Elschenbroich reported the preparation of 1,1'-bis(methylthiobenzene)chromium by di- lithiating bis(benzene)chromium followed by subsequent re- action with dimethyldisulfide. An approach similar to that used by Elschenbroich was used in this research to produce a one-pot, high yield, general synthesis of disubstituted ferrocenylsulfides. In addition the reaction of tetra- alkylthiuram disulfides with mono and dilithioferrocene 12 was examined. The preparation of a chiral ferrocenylsul- fide derivative is also outlined. A few ferrocenylsulfide metal complexes have been pre- pared from l,2,3-trithia[3]ferrocenophanes as shown in Scheme 3. Davison“O has prepared a series of 1,3—dithia- [3]ferrocenophanes with C, Si, Ge and Sn bridges from 1,1'- dithiolferrocene. Similarly Se and Te can be incorporated as bridgesg. Seyferth has recently shown that transition “1. Reac— metals readily insert into sulfur-sulfur bonds tion of (I) with Pd(Ph3)uu2 and Fe3(CO)l2u3 has resulted in unusual metal complexes with a chelating ferrocenyldi- thiolate ligand. In the last section the reactivity of the ferrocenyl- sulfide derivatives as bidentate ligands was examined. Palladium and platinum halides were complexed with the ferrocenylsulfide ligands and the physical properties of the heterobimetallic compounds were examined. (c0) ....r. 3 ?s>< @ 575:8). Scheme 3 s \ ,Me Fe / ‘Me 5 u = Si.Ge,Sn,C II. EXPERIMENTAL General Techniques Air-sensitive reagents were manipulated in a prepuri- fied argon or nitrogen atmosphere. Hexane was freshly distilled from calcium hydride. l,l'-Bis(diphenylphos- phino)ferrocene, (fdpp), was prepared according to Davison's procedure8 and (fdpp)PdCl2 was prepared following Kumada's 17 procedure (dep)PtCl2 was similarly prepared. Bis— benzonitrile) complexes, [(PhCN)2MX2], where M = Pd,Pt; X = Cl,Br, were prepared according to published pro- ceduresuS. Infrared spectra (IR) were obtained on a Perkin-Elmer A57 grating spectrOphotometer or a Perkin-Elmer 239B spectrophotometer by using Nujol or Fluorolube mulls be- tween CsBr plates. Ultraviolet and visible spectra (UV- vis) were recorded by use of a Cary 17 spectrophotometer and acetonitrile solutions. Mass spectra (MS) were obtained by means of a Finnigan A000 instrument with an Incos data system at 70 eV. Electrochemical measurements were made with a PAR 17A polargraph, coupled to a Hewlett-Packard Model 70A5A fast X-Y recorder, by cyclic voltammetry tech— niques. All melting points were determined by using a 1A 15 Thomas-Hoover capillary melting point apparatus and are uncorrected. Elemental analyses were performed by Gal- braith Laboratories, Inc., Knoxville, Tennessee. Proton NMR spectra were obtained by use of a Bruker WM-250 spectrometer at 250 MHz. Unless otherwise noted, all NMR spectra were recorded in chloroform-d solutions 1 with chemical shifts reported in parts per million down- field from a tetramethylsilane internal standard. Carbon- 13 NMR (broadband proton decoupled and gated decoupled) were obtained by use of a Bruker WM-250 spectrometer at 62.9 MHz. Carbon-l3 NMR spectra were recorded in methylene chloride with deuterium oxide as an external lock and chemical shifts, referenced to methylene chloride, are un- corrected for volume susceptibilities. A pulse width (PW) of 6 us and a relaxation delay (RD) of 2 s were parameters that were generally used. Platinum-195 NMR spectra were recorded with a Bruker WH—l80 spectrometer with a mid-range frequency probe operating at 38.7 MHz. The spectra were obtained with a pulse width (PW) of 30 us and a sweep width (SW) of A0,000. The frequency offset (01) was set at zero and the synthe- sizer frequency (SY) was varied from 70,725,000 to 70,7A0,000. NaZPtC16 was used as the reference and as all signals are upfield of the reference they are reported as negative values. The chemical shifts were calculated by using the following equations: 16 Absolute frequency = Instrument frequency — 2(SY) + 01 + cursor frequency - (SW/2) Shift(ppm) = (Abs freq(sample) - Abs freq 20(1), 3689 observed data. The data were reduced 3 A7 and refinement was by full-matrix least-squares techniques ; and the structures were solved by direct methodsu8. The final R value was 0.0A6. The final difference Fourier map showed densities ranging from +1.13 to —l.01 with no indica- tion of missing atoms. l,l'-Bis(methylthio)ferrocene (9) Ferrocene (3g, 16 mmol) was added to a solution of N,N,N',N'-tetramethy1ethylenediamine (TMEDA) (5.1 mL, 17 33 mmol) and 1,6 M n-butyllithium in hexane (20.67 mL, 33 mmol) in oxygen—free hexane (100 mL) in a 250 mL round bottom flask equipped with a side arm and serum cap, under nitrogen. The solution was stirred for 3 h then dimethyl— disulfide (2.96 mL, 33 mmol) in about 20 mL benzene was added slowly via cannula to the bright orange solution at -10°C and the solution was stirred overnight. The resulting brown solution was then filtered under nitrogen and the filtrate was evaporated to dryness. (Addition of water to destroy the excess lithio species resulted in a green then black/blue oil which could be ferricenium). Unreacted ferrocene was removed by sublimation (80"C/10'l mm) and (9) 1 was obtained as a brown oil in a 70% yield. H NMR: 5A,29 (t, AH, H ), A.2l (t, AH, H3 A) 2.30 (s, 6H, CH3); 2,5 iiQ_AME= (CH2C12/D2o): 885.21 (5, cl), 71,72 (d, J = 177.6 Hz, C3,“), 69.59 (d, J = 176.6 Hz, c2,5), 19.35 (q, J = 139.5 Hz, CH3). Mass spec: m/e (rel intensity), 278 (39, M+), 232 (85), 217 (71), 186 (20), 152 (16), 121 (2), 56 (100, Fe). l,l'-Bis(isobutylthio)ferrocene (lg) l,l'-Dilithioferrocene (36 mmol) was prepared as des— cribed above. Isobutyldisulfide (13.9 mL, 72 mmol) in A0 mL benzene was added slowly via cannula to the orange solu- tion at -10°C. After being stirred overnight at room tem- perature the solution became clear brown and 10 mL water 18 was added. The organic layer was separated, dried and evaporated to dryness. Unreacted ferrocene was removed 1 by sublimation (80° C/10_ mm) to give 10 g of (10), (75% yield). Yellow flakes can be obtained by recrystal- l lizing from hexane at —78°C, mp = 33-3A°C. H NMR: 6A.28 (5, AH, H u.20 (t, AH, H3 u), 2.50 (d, J = 6.9 Hz, 2,5): AH, CH2), 1.72 (m, J = 6.9 Hz, 2H, CH), 0.97 (d, J = 6.9 . 13 . Hz, 12H, CH3), 0 NMR. (CH2Cl2/D20), 682.76 (s, C1)’ 7u.11 (d, J = 172.9 Hz, 03 A), 70.31 (d, J = 177.5 Hz, 128.5 02 S), A6.u0 (t, J = 137.8 Hz, CH2), 28.51 (d, J Hz, CH), 21.68 ppm (q, J = 125.8 Hz, CH3); Mass spec: m/e (rel intensity) 362 (A5, M+), 27A (72), 218 (88), 152 (50), 121 (89), 56 (99), A1 (100)- 20 2 Found: C, 59.60; H, 7.21. Anal. Calcd. for Cl8H S Fe: C, 59.66; H, 7.23. l,1'-Bis(isopropylthio)ferrocene (11) ISOpropyldisulfide (17.96 mL, 113 mmol) in about 150 mL hexane was slowly added via cannula to l,l'-dilithioferro- cene (5A mmol) at -78°C. After being stirred at room tem- perature for two days 100 mL water was added to the cloudy yellow solution to give a clear brown solution. The organic layer was separated, dried and evaporated to give a brown oil. Unreacted ferrocene was removed by sublimation (80°C/ 10'1 mm). About 13 g of (11) was obtained as a brown oil (73% yield). Traces of the disulfide can be removed by 19 l crystallization from hexane at low temperature. H NMR: 5A.31 (t, AH, H ), A.2A (t, AH, H3 u), 2.86 (m, J = 6.7 2,5 Hz, 2H, CH), 1.16 (d, J = 6.7 Hz, 12H, CH3); 130 NMR: 6(CH2012/D20), 79.1A (s, cl), 76.20 (d, J = 178 Hz, C3 A), 70.96 (d, J 176 Hz, 39.6A (d, J = 1A2 Hz, CH), 02,5), 126.9 Hz, CH3); Mass spec: m/e (rel intensity), 23.A1 (0. J 33A (100, M+). 292 (19), 260 (22), 250 (26), 218 (A3), 195 (A2), 152 (38), 121 (27), 97 (35). 1,1'-Bis(phenylthio)ferrocene (12) Phenyldisulfide (18 g, 82 mmol) in about 80 mL benzene was added slowly via cannula to l,l‘-dilithioferrocene (A0 mmol) at —10°C. The solution was stirred overnight at room temperature to give a cloudy yellow solution. About 30 mL water was added and (12) which precipitated out from the benzene/hexane layer was filtered and washed with petroleum ether to remove the unreacted ferrocene. About 13 g (80% yield) of (12) was obtained and an analytically pure sample was obtained by recrystallizing from CH2C12 l to give yellow needles, mp = l72—l73°C. H NMR: 6A.A9 (t, AH, H A.AA (t, AH, H3 u), 7.08 (m, 10H, Ph); 2,5)’ £39_NM§: (CH2012/D20), 61A0.27 (s, C of Ph), 128.68 1 161 Hz, Ph), 126.38 (d, J = 161 Hz, Ph), 125.17 (d, J (d, J 161 Hz, Ph), 77.77 (s, C 76.3A (d, J = 181 Hz, 1): C3,“), 71.91 (d, J = 177.5 Hz, C2,5); Mass spec.: m/e 2O (rel intensity), A02 (100, M+), 56 (7A, Fe). Anal. Calcd. for C22H1882Fe: C, 65.67; H, A.A8; S, 15.92 Found: C, 65.69; H, A.36; S, 16.01 1,1'-Bis(benzylthio)ferrocene (13) Benzyldisulfide (27.7 g, 112 mmol) in 200 mL benzene was added slowly via cannula to l,l'-dilithioferrocene (5A mmol) at —A0°C. After being stirred at room temperature overnight the bright yellow suspension was filtered and washed with water. Unreacted ferrocene and benzyldisulfide were sublimed (80°C/10-l mm) from the sticky yellow powder. The yellow solid is recrystallized from CH2Cl2/hexane to give 10.2 g of (13) (50% yield). The filtrate was evap- orated and after sublimation an additional 9 g was iso- lated to give a total yield of 83%. 1H NMR: 57.19 (m), 7.11 (m, Ph), A.12 (t, AH, H A.10 (t, AH, H3 M), 2,5): 3.72 (s, AH, CH2); 13C NMR: (CH2C12/D2O, 28°C), 6138.8 (s, 01 of Ph), 128.87 (d, J = 16A Hz, Ph), 128.25 (d, J = 161 HZ, Ph), 126.80 (d, J = 168 Hz, Ph), 81.16 (5, Cl)’ 7A.68 (d, J 178 Hz, (23 M), 70.57 (d, J = 177 Hz, 02.5): A2.l2 (t, J 1A2 Hz, CH2); Mass spec: m/e (rel intensity), A30 (100, M+). 339 (28. MI-Cngh). 2A3 (39), 152 (6). 91 (18). Anal. Calcd. for C2uH22S2Fe: c, 66.97; H, 5.15 Found: C, 67.18; H, 5.12 21 1,1'-Bis(isopentylthio)ferrocene (1A) Isopentyldisulfide (7.A mL, 33 mmol) in about 50 mL hexane was added slowly via cannula to l,l'—dilithioferro— cene (16 mmol) at -A0°C. After being stirred at room tem- perature overnight 100 mL water was added to the bright yellow solution. The organic layer was separated, dried and evaporated to give a light brown oil. Unreacted ferro- l cene was removed by sublimation (80°C/10- mm) and traces of the disulfide were removed by recrystallization from hexane at low temperature to give A.5 g of (1A), 71% yield. 1 H NMR: 6A.26 (t, AH, H A.18 (t, AH, H3 u), 2.59 (t, 2,5)’ AH, oCH 1.A2 (m, AH, BCH2), 1.65 (m, 2H, CH), 0.86 (d, ) 2 , 6H, CH3); 13C NMR: 681.A1 (s, Cl), 77.50 (d, 03 H) 70.03 (9. 02,5), 38.33 (t, aCH2), 3A.89 (t. BCH2), 26.67 (d, CH), 21.96 (q, CH3); Mass spec: m/e (rel intensity), 390 (100, M+), 286 (16), 2A9 (18), 218 (16), 152 (19). 121 (1A), 97 (19). 1,1'-Bis(diphenylphosphino)ferrocene (15)? 13C NMR: (CH2C12/D2O), 6139.02 (d, lJCP = 10.7 Hz, s, Cl), 133.29 (d, 2JCP = 18.5 Hz, d, JCH = 161 Hz, Ph) 128.A9 (d, 3JCP = 18.5 Hz, d, JCH = 161 Hz, Ph), 128.22 (d, “JCP = 25 Hz, d, JCH = 161 Hz, Ph), 76.63 (d, 1JCP = 8 Hz, Cl), 73.68 (d, 2JCP = 1A Hz, d, JCH = 177 Hz, C3 M), 72.71 (d, JCH = 175 Hz, 02,5). 22 Preparation of Metal Complexes The complexes Fe(C5HuSR)2MX where R Me, CHMe 2 2’ CH CHMe Ph, CH Ph; M = Pd, Pt; X = Cl, Br, were pre- 2 2’ 2 pared from benzene solutions of the appropriate (PhCN)2MX2 species and a slight excess of the ferrocenylsulfide, Fe- (C5HASR)2’ in an approximate 1:1.1 molar ratio. The result— ing precipitate was filtered, washed with benzene, then petroleum ether and then recrystallized from methylene chloride/hexane by slow evaporation. 1,1'-Bis(isobutylthio)ferrocenepalladiumdichloride (16) 1 Shiny black crystals decomposed at l82-183°C. _H AME! (72°C), 65.28 (br, AH, H A.A2 (t, AH, H3 M), 2,5): 2.98 (d, J = 6.8 Hz, AH, CH 1.80 (m, J = 6.8 Hz, 2H, 2): CH), 1.01 (d, J = 6.8 Hz, 12H, CH3); 13C NMR: (CH2C12/ D20), 679.32 (s, Cl), 76.23 (d, J = 18A Hz, C3,“), 71.72 (br d, J = 178 Hz, A9.50 (t, J = 1A5.5 Hz, C82), C2,5), 26.A2 (d, J = 13A Hz, CH), 21.38 (g. J = 126.7 Hz, CH3); .15 (Nujol): 370 (sh), 365 (w), 307 (s), 290 (sh) cm'l. Anal. Calcd. for C18H26Cl2S2FePd: C, A0.06; H, A.86; C1, 13.1A; s, 11.88. Found: C, A0.1A; H, 5.00, C1, 13.35; 8, 12.01. 23 l,1'-Bis(isobutylthio)ferrocenepalladiumdibromide (17) l Shiny black needles decomposed at 185-187°C. H NMR: (23°C), 65.30 (br, AH, H A.Ao (s, AH, H3 u), 3.06 25). (br, s, AH, CH2), 1.81 (m, J = 6.7 Hz, 2H, CH), 1.01 (d, J = 6.7 Hz, 12H, CH3); 13C NMR: (CH2012/D20), 679 A8 (s, Cl), 76.36 (6, J = 176 Hz, C3,“), 71.85 (d, J = 177 Hz, 02,5), 52.85 (t, CH2), 26.89 (d, J = 129 Hz, CH), 21.39 (q, J = 126 Hz, CH3); 15 (Nujol): 365 (w), 222 (s) cm'l. Anal. Calcd. for C18H26Br282FePd: C, 3A.39; H, A.17; Br. 25.A2. Found: C, 3A.50; H, A.18; Br, 23.18. 1,1'-Bis(isobutylthio)ferroceneplatinumdichloride (18) 1H NMR: (101° Yellow needles decomposed at 218-220°C. C), 05.1A (t, AH, H2 5), A.A3 (t, AH, H3 A)’ 3.08 (br, 8, AH, CH2), 1.92 (m, 2H, CH), 1.01 (d, J = 6.5 Hz, 12H, CH3); 13 . C NMR. (CH2C12/D2O), 680.22 (Cl), 75.93 (C3 M), 73.36 , 70.81 (c A8.21 (CH2), 26.55 (CH), 21.85 (CH3); 1 (02,5) 2,5), _I_R(NUJ'01): 372 (W). 323 (s). 310 (m) cm— Anal. Calcd. for C18H26C1282FePt: C, 3A.Al; H, A.17; Cl, 11.28. Found: C, 3A.A3; H, A.20; Cl, 11.39. 1,1'—Bis(isobuty1thio)ferroceneplatinumdibromide (19) Yellow flakes decomposed at 225-227°C. 1H NMR: 65.A5 (br, 2H, H A.86 (br, 2H, H A.58 (br, 2H, H3 a). 2,5), 2,5), A.3A (br, 2H, H3 4), 2.9A (br, AH, CH2), 1.89 (br, 2H, CH), 2A 1.0A (br, d, 12H, CH3); 130 NMR: (A5°C), 680.09 (Cl), 75.32 (C3,u), 71.35 (br, C2’5), A9.21 (CH2), 26.A2 (CH), 21.5A (CH3). )5 (Nujol): 372 (w), 36A (w) cm-l. 1,1'—Bis(isopropylthio)ferrocenepalladiumdichloride (20) Brown needles decomposed at l92-193°C. 1H NMR: (50°C), 05.3 (br, AH, H A.A5 (br, 5, AH, H3 A)’ A.0A (m, J = 2,5)’ 6.8 Hz, 2H, CH), 1.23 (d, J = 6.8 Hz, 12H, CH3); 15 (Nujol): 380 (w), 360 (w), 320 (s), 305 (s) cm-l. Anal. Calcd. for C16H22C1282FePd: C, 37.56; H, A.33; Cl, 13.86. Found: C, 37.56, H. A.A0, Cl, 1A.00. l,l'-Bis(isopropylthio)ferrocenepalladiumdibromide (21) 1 Brown needles decomposed at 188-190°C. H NMR: 65.23 (br, AH. H2,5). 2H, CH), 1.2A (br, 12H, CH3); IR (Nujol): 380 (w), 360 1 A.A5 (br, AH, H3 u), A.17 (m, J = 6.7 Hz, (W), 225 (s) cm— Anal. Calcd. for C16H22Br2SZFePd: C, 32.00; H, 3.69; Br, 26.61. Found: C, 32.19; H, 3.65; Br, 26.A3. 1,1'—Bis(isopropylthio)ferroceneplatinumdichloride (22) 1 Yellow flakes decomposed at 223-225°C. H NMR: 55.51 (s, 2H, H A.72 (s, 2H, H ), A.56 (s, 2H, H3 A)’ A.28 2,5)’ 2,5 (s, 2H, H3 u), A.27 (m, J = 6.A Hz, 2H, CH), 1.37 (d, 6H, CH ). 1-12 (d. 6H. CH3); l5 (Nujol): 383 (W). 365 (w). 3 25 328 (s), 315 (s) cm-l. ll)'-Bis(isoprgpylthio)ferroceneplatinumdibromide (6%) 1 Yellow flakes decomposed at 21A-2l6°C. H NMR: 65.5A (br, 2H, ), A.75 (br, 2H, H A.60 (br, 2H, H3 A)’ H2.5 2.5): A.31 (br, 2H, H3 u), A.27 (m, J = 6.7 Hz, 2H, CH), 1.39 (br, 6H, CH3), 1.16 (br, 6H, CH3); IR (Nujol): 389 (w), 370 (w), 220 (s) cm-l. 2 2 23.19. Found: C, 28.08; H, 3.A1; Br, 23.35. Anal. Calcd. for C16H22Br S FePt: C, 27.88; H, 3.22; Br, 1,1'-Bis(methylthio)ferrocenepalladiumdichloride (66) Brown powder decomposed at 192—198°C. IR (Nujol): 355 (w). 3A0 (w). 310 (s), 290 (sh) cm'l. 1,1'-Bis(methylthio)ferrocenepalladiumdibromide (g6) Dark brown powder decomposed at 209—213°C. IR (Nujol): 3A3 (w). 205 (m) cm-l. 1,1'-Bis(methylthio)ferroceneplatinumdichloride (g6) Yellow powder decomposed at 239-2AA°C. IR (Nujol): 3A5 (w), 323 (s), 305 (s) cm‘l. 26 l,l'-Bis(methylthio)ferroceneplatinumdibromide (61) Yellow crystals decomposed at 225-230°C. IR (Nujol): 3A0 (w), 208 (s) cm—l. Anal. Calcd. for C12HluBr2SZFePt: C, 22.77; H, 2.23; Br, 25.2A. Found: C, 22.90; H, 2.15; Br, 2A.99. 1,1'-Bis(benzylthio)ferrocenepalladiumdichloride (68— l Shiny black needles decomposed at 223-225°C. H NMR: (50°C), 67.22 (m), 7.19 (m), 7.06 (m, Ph), 5.00 (br, AH, H2,5), A.3A (br, 5, AH, H3,A)’ A.32 (s, CH2); )3 (Nujol): 358 (m), 335 (w), 315 (s), 300 (s) cm'l. 2uH22C12SZFePd: C, A7.A3; H, 3.65; Cl, 11.67. Found: C, A7.66; H, 3.65; CI, 11.90. Anal. Calcd. for C 1,1'—Bis(benzylthio)ferrocenepalladiumdibromide (62) 1 Shiny black needles decomposed at 206-207°C. H NMR: 67.2A (m), 7.05 (m, Ph), A.96 (v, br, AH, H2 5), A.AO (5, AH, CH2), A.33 (br, AH, H3 A); IR (Nujol): 353 (m), 3A5 (sh), 230 (s) cm_l. Anal. Calcd. for C24H22Br2S2FePd: C, A1.38; H, 3.18; Br, 22.9A. Found: C, A1.A5; H, 3.19; Br, 22.69. l,1'—Bis(benzy1thio)ferroceneplatinumdichloride (éQ) Yellow flakes decomposed at 22A-225°C. 1H NMR: (50°C), 67.3A (s), 7.20 (m), 7.0A (m, Ph), 5.A5 (v, br, AH, H2 5), 27 A.60 (br, AH, H3 A)’ A.A3 (5, AH, CH2); )5 (Nujol): 360 (m). 350 (m), 332 (s). 312 (s) cm'l. Anal. Calcd. for C2uH22C1282FePt: C, Al.39; H, 3.18; Cl, 10.18. Found: C, A1.2l; H, 3.17; Cl, 9.90. l,l'-Bis(benzylthio)ferroceneplatinumdibromide (6%) Yellow flakes decomposed at 198-200°C. 1H NMR: 67.23 (m), 7.20 (m), 7.02 (m, Ph), 5.53 (V, br, 2H, H2 5), A.56 A.12 (br, AH, H 1 (br, s, AH, CH2), A.AA (br, 2H, H ); 2,5): 3,14 IR (Nujol): 358 (m), 3A8 (m), 215 (s) cm- Anal. Calcd. for C2uH22Br282FePt: C, 36.71; H, 2.82; Br, 20.35. Found: C, 36.67; H, 2.81; Br, 20.12. l,l'-Bis(phenylthio)ferrocenepalladiumdichloride (66) 1 Brown powder decomposed at l98—201°C. H NMR: 67.38 (m, 10H, Ph), 5.27 (v, br, AH, H A.65 (br, 8, AH, 2,5)’ H3 A); 15 (Nujol): 323 (vs), 308 (vs), 278 (s), 262 (s) cm-l. 1,l'—Bis(phenylthio)ferrocenepalladiumdibromide (6%) l Shiny black needles decomposed at 188-189°C. H NMR: 67.35 (m, 10H, Ph), 5.Al (br, t, AH, H2 5), A.6A (t, AH, H3 A): IR (Nujol): 316 (s) cm-l. Anal. Calcd. for C22H18Br2S2FePd: C, 39.52; H, 2.71; Br, 23.90. Found: C, 39.A5; H, 2.82; Br, 22.30. 28 l,1'-Bis(pheny1thio)ferroceneplatinumdichloride (£6) 1H NMR: 67.A2 Yellow needles decomposed at 210-213°C. (m). 7.38 (s, Ph), 5.31 (br, s, AH, H2 5) A.6A (br, s, AH, H3,A)3 £5 (Nujol): 350 (S), 329 (S), 317 (S), 312 (S) -1 cm . Anal. Calcd. for C22H18Cl2S2FePt: C, 39.5A; H, 2.72; C1, 10.61. Found: C, 39.7A; H, 2.79; Cl, 10.66. 1,1'-Bis(phenylthio)ferroceneplatinumdibromide (66) 1 Yellow plates decomposed at 2A5—2A7°C. H NMR: 67.38 (m, 10H, Ph), 5.30 (v br, AH, H A.6l (br, AH, H3 A); 2,5), 15 (Nujol): 3A9 (sh), 32A (w), 266 (m) cm‘l. Anal. Calcd. for C22H18Br2S2FePt: C, 3A.89; H, 2.A0; Br, 21.20. Found: C, 3A.9A; H, 2.51; Br, 21.23. N,N-dimethyl-a-ferrocenylethylamine (36) (66) was prepared and resolved using (R)-(+)tartaric A9 acid as described by Ugi The amine (66), can be re- crystallized from petroleum ether to give a yellow powder, 1H NMR: 6A.11 (m, AH, CSHu), A.08 (s, 5H, Cp), mp = 35°C. 3.55 (q, J = 6.8 Hz, 1H, CH), 2.05 (s, 6H, NMe2), 1.A2 (d, J = 6.8 Hz, 3H, CH3); 13C NMR: 685.8 (s, Cl), 68.3 (d), 67.5 (d, J 176 Hz, Cp), 66.A (d), 66.18 (d), 65.7 (d), 57.6 (d, J 1A1 Hz, CH), 39.A (q, J = 133 Hz, NMe2), 1A.8 (q, J = 128 Hz, CHMe). 29 (R)-a-[(S)—2—benzylthioferrocenyl]ethyldimethylamine (61) The amine ($6), (1.5 g, 5.8 mmol) was placed in a 200 mL flask equipped with a side arm and containing A0 mL diethylether. n—Butyllithium (A.O mL, 6.A mmol) was added slowly via syringe to the solution which was at -A0°C. The orange solution was allowed to warm to room temperature slowly and then stirred for an additional 1 h. Benzyldisulfide (1.A5 g, 5.9 mmol) dissolved in about 70 mL warm hexane was added dropwise via cannula to the orange solution at -78°C. The yellow solution was allowed to reach room temperature and then refluxed overnight. The reaction mixture was cooled and then 20 mL water was added. The organic layer was separated, dried and evaporated to give a dark oily residue. The oil was chromatographed on A1203/5% H2O, by eluting first with hexane and then with CH2C12 to give 1 0.88 g of (11), (A0% yield) as a brown oil. H NMR: 67.18 (m, 5H, Ph), A.10 (m, 3H, C A.06 (s, 5H, Cp), 3.90 5H3): (m, 2H, CH2), 2.21 (s, 6H, NMe2), 1.38 (d, J = 6.8 Hz, 3H, CHMe), 13C NMR: 6138.9 (s), 129.; (d), 128.3 (d), 126.7 (d, Ph), 79.3, 78-5, 71-6, 69-99 (d, Cp), 67.99, 67.03, 56.A (d, CHMe), A1.A5 (t, SCH2Ph), 39.95 (q, NMe2), 10.88 (q, CHMe); Mass spec: m/e (rel intensity), 379 (68, M+), 33A (98. M+-HNMe2). 2A3 (100), 121 (73). 3O 1,1'-Bis(dimethyldithiocarbamate)ferrocene (66) Ferrocene (10 g, 53.7 mmol) was added to a solution of N,N,N',N'-tetramethylethylenediamine (16.85 mL, 112 mmol) and 1,6 M n-butyllithium in hexane (68 mL, 112 mmol) in oxygen free hexane (200 mL) in a l l-round bottom flask, equipped with a side arm and serum cap, under nitrogen. The solution was stirred at least 3 h, or until bright orange, then tetramethylthiuram disulfide (26.5 g, 110 mmol) in 500 mL benzene was added slowly via cannula to the solution which was at -A0°C. The solution was allowed to reach room temperature and was stirred overnight to give a black/grey solution. Water (90 mL) was added and the solu- tion was filtered. The grey precipitate was washed with water and then chromatographed on alumina. The yellow band, eluted with CH2C12, was evaporated to dryness to yield 16 g of (i6), (70% yield). The filtrate was separated from the aqueous layer, dried and chromatographed. The first band, eluted with hexane, contained 0.95 g ferrocene whereas the second band eluted with CH2C12 contained (66) and traces of the disulfide. The thiuram disulfide is removed by washing with benzene. Total yield was 18.2 g (80% yield). Compound (66) was recrystallized from CH2C12/hexane to give 1 yellow crystals, mp = 170-173°C. H NMR: 6A.50 (t, AH, H 2,5). “.38 (t, “H, H3’u), 30MB (8, 6H, CH3), 3°u3 (S) 6H, CH3); M= (Cchlg/Dgo). 6199.7 (cs), 77.A (C2 5), 76.71 (Cl), 71.98 (C3,A)’ A5.3 (CH3), Al.7 (CH3); Mass spec: m/e 31 (rel intensity), A2A (100, M+), 328 (53), 2A0 (A2). 88 (65); IR (Fluorolube): 1A80 cm-l. Anal. Calcd. for C16H2OSuN2Fe: C, A5.28; H, A.75; N, 6.60; S, 30.22. Found: C, A5.29; H, A.81; N, 6.57; S, 30.30. 1,1'-Bis(diethyldithiocarbamate)ferrocene (62) Tetraethylthiram disulfide (32.68 g, 110 mmol) in 500 mL toluene was added slowly via cannula to 1,1'-dilithio— ferrocene (5A mmol) at -78°C. The solution was allowed to reach room temperature and was stirred overnight to give a blackish solution. Water (100 mL) was added and the solu- tion was filtered. The grey precipitate was washed with water then chromatographed on alumina. The yellow band eluted with CH2C12 gave 13 g of ($2), 50% yield. The brown filtrate, which contains ferrocene, thiuram disulfide and ($2), was chromatographed to yield A.5 g of (62), total yield 68%. Recrystallization from CHzch/hexane gave yellow crystals that decomposed at 160-163°C, melted at 181°C. 1H NMR: 6A.52 (t, AH, H2 5), A.A2 (t, AH, H3 u), 3.99 (q, 3 s J = 7.0 Hz, 2H, CH2), 3.85 (q, J = 7.0 Hz, 2H, CH2), 1.39 (t, J = 7.0 Hz, 3H, CH3), 1.26 (t, J = 7.0 Hz, 3H, CH3); 13C NMR: (CH2C12/D2O), 6197.5 (s, CS), 77.2 (d, J = 175 Hz, 76.1 (s, Cl), 71.6 (d, J = 177 Hz, C3 u), A9.A C2.5)’ (q, J = 13A Hz, CH3), A6.9 (q, J = 13A Hz, CH2), 12.5 (t, J = 126 Hz, CH3), 11.2 (t, J = 126 Hz, CH3); Mass spec: m/e 32 (rel intensity), A80 (5A, M+), 38A (51), 268 (69), 116 (100), 85 (96), 60 (72); EB (Fluorolube): 1A80 cm'l. Anal. Calcd. for C20H28N28AF8: C, A9.99; H, 5.87; N, 5.83; S, 26.69. Found: C, A9.90; H, 5.80; N, 5.75; S, 26.90. 1,1'-Bis(diisopropyldithiocarbamate)ferrocene (66) TetraiSOpropylthiuram disulfide (25 g, 71 mmol) in 250 mL toluene was added slowly via cannula to 1,1'-di- lithioferrocene (35 mmol) at -78°C. The solution was allow— ed to reach room temperature and was stirred overnight to give a black solution. Water (100 mL) was added and the solution was filtered. The green precipitate was washed with water (to remove ferricenium salts) and then chromato- graphed to give 15 g of (Ag), 79% yield. The green aqueous solution was reduced with sodium bisulfite and forms a brown solid after standing in air for 30 h. Complex (66) was recrystallized from CH C12/hexane to give yellow crys- 2 tals which decompose at 180°C and melt at 225—226°C. iH NMR: (71°C), 6A.78 (br, CH), A.A6 (t, AH, H2 5), A.38 (t, AH, H3,u), 1.A7 (d, J 6.5 Hz, 12H, CH3); 13C NMR: 6198.0 (5, CS), 77.1 (d, J = 183 Hz, C2,5), 76.1 (8, Cl), 71.3 (d, J = 178 HZ, C3,“), 53.3 (d, J = 137 HZ, CH), 19.7 (q, J = 127 Hz, CH3); Mass spec: m/e (rel intensity), 536 (8, M AA0 (8), 296 (1A), 1AA (31), 102 (100), 60 (56), A3 (92); 1 , IR (Fluorolube): 1A65 cm- Anal. Calcd. for C2uH36N2SuFe: C, 53.77; H, 6.76 Found: C, 53.77; H, 6.90. 33 Dimethyldithiocarbamateferrocene (66) Ferrocene (6.1A 2, 33 mmol) was added to a solution of TMEDA (5.1 mL, 33 mmol) and n-butyllithium (20.1 mL, 33 mmol) in 150 mL hexane and the solution was stirred for 3 h. Tetramethylthiuram disulfide (7.92 g, 33 mmol) in 120 mL benzene was added via cannula to the bright orange solution which had been cooled to -78°C. The solution was allowed to reach room temperature, was stirred overnight and then 50 mL water was added to the brownish black solu- tion. The solution was filtered and the sticky black precip- itate was chromatographed on alumina. The first band, eluted with hexane, yielded 2.9 g of ferrocene. The second yellow band which was eluted with benzene gave 300 mg of 1 an unidentified product. H NMR: 6A.28 (t), A.19 (s), A.15 (t); Mass spec: A02 (base peak). The third band, eluted with benzene, yielded A70 mg of (6%) (9% yield based on the ferrocene reacted). Numerous additional bands were eluted with benzene and methylchloroform but the products were not isolated. Compound (6%) was recrystallized from CH2Cl2/hexane to give yellow crystals which decompose l at l80-18A°C. H NMR: 6A.AA (t, AH, H A.3A (t, AH, 2,5), H, u). A.2A (s. 58. Cr). 3.51 (s. 3H. CH3). 3.A5 (s. 3H. CH3); 13C NMR: (CH2C12/D20, 28°C), 6199.9 (CS), 75.66 2,5). 7A.96 (Cl). 70.3A (C3,A)’ 69.39 (Cp), A5.2A (CH3), Al.50 (CH3); Mass spec: m/e (rel intensity), 305 (87, M+), (C 3A 2A0 (22, M+ - Cp), 217 (20), 209 <61), 121 (21) 88 (100, CSNMeZ); IR (Fluorolube): 1A75 cm—l. Anal. Calcd. for C16H2OSAN2Fe: C, 51.15; H, A.95 Found: C, 51.36; H, A.92. Dithyldithiocarbamateferrocene (6g) 6 Bromoferrocene, prepared according to Rosenblum's procedure, was dried in vacuo before use. Butyllithium (7.A7 mL, 11.8 mmol) was added slowly via syringe to a solu- tion of bromoferrocene (2.6A g, 10.0 mmol) in 70 mL dry diethylether which was cooled to -A0°C. After being stirred at room temperature for 20 min. tetraethylthiuram disulfide (3.27 g, 11.0 mmol) in A0 mL toluene was added slowly via cannula to the bright yellow solution which had been cooled to -78°C. The solution was stirred for 12-18 h at room temperature to give a brown solution. Water, 20 mL, was added and the organic layer was separated, dried and chromatographed on alumina. The first yellow band, eluted with hexane, contained bromoferrocene and a small amount of ferrocene whereas the second yellow band, eluted with methyl- ene chloride, yielded 2 g of (6%), 60% yield. Complex (6%) was recrystallized from CH2012/hexane to give yellow crystals which melt at l27.5—l28.5°C. 1H NMR: 6A.A2 (t, AH, H2,5), A.33 (t, AH, H3,u), A.22 (s, 5H, Cp), 3.96 (q, J = 7 Hz, CH2), 3.82 (q, J = 7 Hz, CH2), 1.37 (t, J = 7 Hz, CH3), 1.23 (t, J = 7 Hz, CH3); 13C NMR: (CH2C12/D20, 31°C) 35 6198.7 (CS). 75.97 (C2 5), 75.22 (C1)’ 70.37 (C3 A)’ 69-A8 (Cp), A9.61 (CH2), A7.19 (CH2), 12.56 (CH3), 11.53 (CH3); Mass spec: m/e (rel intensity), 333 (18, M+), 237 (8), 217 (6), 116 (100, CSNEt2), 88 (87), 60 (61); )5 (Fluoro- lube): 1A80 cm-l. Anal. Calcd. for C15H19N82Fe: C, 5A.06; H, 5.75; N, A.20 Found: C, 53.95; H, 5.76; N, A.29. Diisopropyldithiocarbamateferrocene (6%) Butyllithium (A.9 mL, 7.8 mmol) was slowly added via syringe to a solution of bromoferrocene (1.73 g, 6.5 mmol) in 100 mL dry diethylether. After being stirred at room temperature for 20 min, tetraisopropylthiuram disulfide (2.A1 g, 6.8 mmol) in 80 mL hexane was added slowly via cannula to the bright orange solution which had been cooled to -78°C. The solution was stirred at room temperature for 18 h to give a light brown reaction mixture. Water (30 mL) was added and then the organic layer was separated, dried and evaporated to dryness to give a brown oil which was chromatographed on alumina. The first yellow band, eluted with hexane, contained a small amount of bromoferro- cene and 3A0 mg ferrocene. The second yellow band, eluted with methylene chloride, was concentrated and then the crystalline mass was washed briefly with benzene (to remove the disulfide). Yellow needles, recrystallized from methyl- ene chloride/hexane, were obtained in 8A% yield based on the 36 bromoferrocene reacted, mp = l89-190°C (dec). 1H NMR: (59°C), 6A.A0 (t, AH, H ), A.33 (t, AH, H3 u), A20 (s, 2,5 5H, Cp) A.76 (septet, 1H, CH), 1.A8 (d, J = 6.6 Hz, 6H, CH3); 13C NMR: 6198.0 (s, CS), 76.0 (d, C2 5), 75.2 (s, Cl). 70.2 (0, C3 u). 69.A (d. Cp). 5A.1 (d. CH). 19.9 (9. CH3); Mass spec: m/e (rel intensity), 361 (100, M+), 218 (99), 217 (Al), 1AA (30). 121 (6), 102 (32). Anal. Calcd. for C H NS2Fe: C, 56.51; H, 6.A2 17 23 Found: C, 56.69; H, 6.61. Reaction of Fe(C5HuSCSNEt2)2, (39), With (PhCN)2PdCl2 A benzene solution (A0 mL) of (PhCN)2PdC12 (300 mg, 783 mmol) was added slowly to a 100 mL benzene solution of Fe(C5HuSCSNEt2)2, ($9), (361 mg, 752 mmol) and stirred over- night. The light brown precipitate was filtered, washed with benzene and dried. The black brown product was slightly soluble in methylene chloride. Reaction of Fe(CSH5)(C5HuSCSNEt2) (6%), With (PhCN)2PdCl2 An identical procedure as given above was followed. (PhCN)2PdC12 (115 mg, 0.3 mmol) was added to (6%) (100 mg, 0.3 mmol) to give a red-brown precipitate which is soluble l . H NMR. (CD3NO2), in methylene chloride and nitromethane. 63.A6 (br), 2.A0, 2.0A (br,s), 1.52 (br), 1.30 (br), 0.87 (br); IR (Fluorolube): 1550 cm-1. III. RESULTS AND DISCUSSION A. Fe(CSHuSR)2 (R = Me,_iPr,riBu, iPent, Ph, Bz) 1. Preparation Ferrocenylsulfide complexes of the type, Fe(C5HuSR)2 where R = Me, iPr, iBu, iPent, Ph, Bz have been prepared in a general, high yield, one step synthesis shown in Scheme A. The appropriate disulfide is added slowly, via cannula, Fe wBuU Fe -TMEDA TMEDA ssz <§§:§>-SR Fe ©s R = Me,d-Pr,£sBu,i-Pentyl ll: Ph,!z Scheme A 37 38 to a hexane solution of l,l'-dilithioferrocene which was cooled to -A0°C. l,l'-Dilithioferrocene is prepared in over 90% yield by reaction of stoichiometric quantities of n- butyllithium and tetramethylethylenediamine (TMEDA) with ferrocene. The metalation reagent was used as the in situ slurry even though use of the species isolated as a solid has been reported to lead to higher yields.8 l,l'-Bis(methylthio)ferrocene (Q), l,l'-bis(isopropyl- thio)ferrocene (6%) and 1,1'-bis(isopentylthio)ferrocene (66) were isolated as yellow-brown oils whereas l,l'-bis- (phenylthio)ferrocene (kg), 1,1'-bis(isobutylthio)ferro- cene (66) and 1,1'-bis(benzylthio)ferrocene (6%) were ob- tained as yellow crystals. All six derivatives are soluble in common organic solvents and are air stable in solution and the solid state. The disulfide, t-butyldisulfide, failed to react with 1,1'-dilithioferrocene under the reaction conditions em- ployed. This could be due to the steric crowding of the sulfur—sulfur bond by the bulky t-butyl groups which prevent nucleOphilic cleavage of the sulfur-sulfur bonds. The brown methyl sulfide derivative (2) turned green when water was added to destroy the excess lithio reagent. The aqueous layer was green which suggests that ferrocene or a ferrocene derivative was oxidized to the ferricenium analog. When the methyl sulfide (Q) was filtered anaero— bically the filtrate did not discolor in the presence of 39 water or oxygen. This suggests that a side product, present in the reaction mixture, is responsible for the discoloring. 2. H NMR 1 The H NMR data of the ferrocenylsulfide complexes (2)- (66) is given in Table 2. The 1H NMR spectra of these com- pounds are typical of l,l'—disubstituted ferrocenes. The resonances due to the ring protons consist of two "triplets" which is characteristic of an AA'BB' spin system where JAB and J (or JAB') are equal and smaller than the chemical 50 A'B shift between A and B These "triplets" are slightly deshielded as expected for an electron withdrawing sub- stituentl6. Deuteration studies have indicated that the low field "triplet" is assigned to the protons in the 2 and 5 positions on the ring (H2,5) whereas the high field "triplet" is assigned to the ring protons in the 3 and A position (H3,u)u3. 3. 13C NMR 130 NMR is a sensitive tool for measuring the electron density distribution on the cyclopentadienyl ring in ferro- cene. Substituents on the ring induce screening of the nuclei in two different ways, one way being due to the mag- netic anisotrOpy of the substituent and the second way due to the electronic effect of the substituent that A0 Baa.» nmp.m ooa.e oma.: EmH.A mficammomemmovoa 622.: pm:.: Emo.s mAcamemmovoa . . . . . . m m m m e m ppm 0 are H am: A cam m oma a com a A name mo mom m ovoa pam.o Ems.a oom.m oom.: pmm.: mfimozmommomemmovoa ooa.a sow.m oem.q cam.: mAmozmomemmovoa nom.m ofim.e oam.e onZmemmovoa mmo mo mmom mice a mm m m: ca ossoosoo .othwnoqemp Eoop pm maoao CH oocfimpno mppoodw .Nm .nm .pcomfi .sma .662 .62 u m mflmmemmovoa .noxofiosoo opacasnaacooossoa soc mono mzz m .m oaone H Al consists of both resonance and inductive components. The 13C NMR data for the ferrocenylsulfides (2)-(%6) is presented in Table 3. A gated decoupled experiment, performed on the isopropyl derivative (6%) simplifies the assignment of the 13C NMR spectra (shown in Figure 1). The gated decoupled spectrum confirms that the signal at 79 ppm is due to the substituted ring carbon, C as this 1, signal remains a singlet under proton coupled conditions. The resonances at 76 and 71 ppm are tentatively assigned to the nuclei at the C3,“ and C2,5 positions. Koridze51 has assigned the signals in methoxyferrocene at 61.5 and 5A.7 ppm to the C3,“ and C2,5 nuclei respectively on the basis of deuterium labelling studies. In general the C2,5 positions are primarily influenced by inductive effects and the magnetic anisotropy of the substituent whereas the C3,“ positions are more sensitive to resonance ef- fects.52’53 The assignments in the ferrocenylsulfide com- plexes (%)—(%6) are tentative and deuteration studies must be employed to make unambiguous assignments of the ring carbons. The chemical shift of the substituted ring carbon atom, Cl’ reflects the inductive and field effects of the sub- stituent and exhibits the widest range of values of any of the ring carbons. The C1 resonance of Fe(C HuSMe)2 is 5 shifted downfield by 17 ppm, relative to ferrocene (68.2 ppm), whereas Fe(CBHMSPh)2 is deshielded by 9 ppm. These A2 UNN.®NH on.mmH Umm.mmH m m z m HA.NA we.MA mc.os noo.mMH A can a ovoa omm.omH oom.wma Umm.wma m m a m mH.mz Am.OA wo.es oa.am now.mma Aha mom m ovom pea.mma owm.oma Umw.me m z m Ha.HA em.o~ ss.s~ nAm.o:H Anam m even a aw.:m m m m m a m ca.Hm Ao.om a mm.mm mo.o~ om.>e He.aw A ozmo mo mom : ovoa wc.am Hm.wm oe.ce Hm.OA HH.:A on.mw mnmozzommomemmovoa He.mm ao.mm om.ow om.os ea.ms mAmozmomermovoa mm.ma mm.mo ms.HA Hm.mm mflozmemmovoa mzo go who m.mo a.mo Ho ea ocsoosoo .opzpopoQEmp soon on coaosaon Acmo\maom:ov ca oocaoooo shoooam .saa .02 u .Nm .na .osoaa .sma m .mflmmzmmovoa .noonoEoo opacasnaacooosooa soc mono mzz omH .m canoe A3 ©S°m ©5010“, CH2CI2 . .L L L )L__ 8'0 ' 60 T 4'0 r 2.0 178 Hz 176 Hz 127 Hz 17, P12 142 E12 U i M d do T ti) T In 7 217 Figure l. Proton decoupled (above) and gated decoupled (below) 13C NMR Spectra of Fe(CSHuS-iPr)2. AA values are in contrast to methoxyferrocene where the Cl resonance is shifted downfield by over 59 ppm. A. Infrared Spectra Table A contains the infrared data for the ferrocenyl- sulphide compounds (2)-(%§). Inspection of Table A indi- cates that certain frequencies are common to all five com- pounds. These frequencies have been tentatively assigned 5A by comparison with the vibrational spectra of ferrocene and dimethylferroceneSS. The high frequency infrared bands at 3090 and 2910 cm-1 are assigned to C-H stretching frequencies. The strong absorption around 1AAO cm"l may be associated with a C-C stretching mode whereas the strong bands observed at 1160, 1 are attributed to C—H deformation modes. 1 888 and 820 cm- The strong absorption that occurs at 1020 cm- could be assigned to a ring breathing mode while the broad band in the 500—515 cm-1 region may be associated with ring-metal vibrations such as an asymmetric ring-tilt and an assym- metric ring-metal stretch. In the phenyl and benzyl analogs, (66) and (6%), the absorptions due to the phenyl group are clearly visible. Strong absorption bands located between 730 and 680 cm-1 result from out-of—plane C-H bending. In addition skeletal vibrations due to C-C ring stretching are found at 1580 cm-1. The stretching vibration assigned to the C-S linkage, A5 Table A. Infrared Data for Ferrocenylsulfide Complexes, Fe(CSHuSR)2, R = Me, iPr, iBu, Ph, Bz. R = Me iPr iBu Ph Bz 3090m 3085m 3095m 3080m 20153 2910br,s 2950br,s 2910br,s 2910m 2850m 2865m 28A5m 1065br,m 1579s 1585m 1AAOs 1A63s 1A38s 1AA58 1A20br,s 1A10s lAlAm 1A08m 1Al2m 1380s 1389s 1385m 1380m 13658 136A8 1358m 1310m 1310m 1319m 1235br,s 12A0m 1235m 1189m ll9Am ll63m 11558 116A8 11688 11618 1103w 1100w 1078m 1062m 1060w lOAOm lOAlm 10A5w 10208 10158 10188 10185 10188 10103 96Am 920m 963w 8855 8855 888s 888s 885m 863m 8785 862m 85Am 8A5m 8A5m 817br,s 810br,s 815br,s 820$ 815br,s 732s 77As 712br,s 685s 692s 6A0m 637m 620m 615m 527m 577W 515br,s 500br,s 510br,s .501m 515m A95m A7Am A78m A55m AA5m H6 which occurs in the region of 700 - 600 cm-1, is generally weak56 and was not assigned for the ferrocenylsulphide complexes. B. 9.5 £395§3ECH(CH3)N(CH3)2][SCH2Ph] 5 5 Chiral ferrocenylphosphine ligands, which have planar chirality due to a 1,2-unsymmetrically substituted cyclo— pentadienyl ring, are highly effective as ligands in transi- tion metal catalyzed asymmetric synthesis. Though few sulfide complexes have been used as ligands in catalysis, the preparation of a chiral ferrocenylsulfide complex was undertaken so that its possible application to asymmetric synthesis may be investigated. A chiral ferrocenylsulphide complex was readily pre- pared from N,N-dimethyl—a—ferrocenylethylamine, (36) (see Scheme 5). The amine (56) was prepared from ferrocene ac- cording to Ugi's procedure”8 and was resolved by using (R)-(+)tartaric acid. As illustrated by Ugi56 , the (R)- amine ($6), is stereoselectively lithiated by butyllithium to give 96% of the (R)(R) isomer, (5%). The (R)(R) deriva- tive is thought to be stabilized by the coordination of the adjacent nitrogen atom in the side chain to the lithium atom. The lithiated ferrocene derivative is then treated with the disulfide and refluxed for at least three hours. The chiral ferrocenylsulphide, ($1), is chromatographed on alumina, deactivated with 5% water, and is isolated in 40% “7 e “’2 u-BuLi : 1 If": I! 0 (R) Scheme 5 5 5 Mb Fe Li..ONm2 95 R252 '1‘ 5 e Fe SR NMe, g] (”H R = 6”,"! yield as a brown oil. The chiral ferrocenylsulfide com- pound, (él) contains two elements of chirality. The (R) configuration refers to the asymmetric carbon atom while the (8) configuration refers to the planar chirality. The presence of the asymmetric carbon atom in the amine, (Q6), gives rise to diastereotopic ring carbons. The 13C U8 NMR spectrum of the amine, (56), contains five separate signals for the ring carbons at 85.8 ppm (assigned to C1 by a gated decoupling experiment), 68.3, 66.“, 66.1“ and 65.7 ppm. This is in contrast to substituted ferrocenes without an element of chirality as one signal is attributed to the 02,5 nuclei and one to the C3,“ nuclei. In the chiral ferrocenylsulfide, (3%), the benzylic protons are diastereotopic. Their diastereotopic nature was however 1H NMR as the resonance due to the not observed in the benzylic protons is partially obscured by the cyclopenta— dienyl ring protons. Cullen57 has determined the crystal structure of [(PPFA)Rh(NBD)]PF6, where PPFA is (R,S)-a-(2-dipheny1phos- phinoferrocenyl)ethyldimethylamine; NBD is norbornadiene, and concluded that the chiral ferrocenylphosphine ligand coordinated to rhodium through both the phosphorous and nitrogen atoms. As there is much interest in ligands which have both "hard" and "soft" properties, investigation of the chelation of the chiral ferrocenylsulfide, (@1), with transition metals would be of much interest. In addition it would be interesting to study the effectiveness of the chiral ferrocenylsulfide ligand in transition metal catalyzed asymmetric synthesis. 49 c. Fe(C5EuSR)2M§ (R = Me, iPr, iBu, Ph,_Bz, M = Pd, Pt; 2 x = Cl, Br) 1. Preparation Reaction of a benzene solution of the ferrocenylsulfide compounds, (2)-(1%), with bis(benzonitrile) adducts of palladium and platinum chloride and bromide salts gave rise to the monosubstituted complexes, (lg)-(%5) (see Scheme 6). The heterobimetallic complexes are insoluble in benzene: the palladium ferrocenylsulphide complexes precipitated IR @R (PhCNhMXz ©S\ /X Fe II Fe 4 ©s / \X R <§:::>*-S\ R |W==Pd,Pt X =(N,Br R = Me,£-Pr, i-Bu Ph, 82 Scheme 6 immediately while the platinum analogs precipitated after being stirred for a few days. 50 The isobutyl and isopropyl metal complexes are soluble in methylene chloride and chloroform whereas the benzyl derivatives are only slightly soluble in these solvents. The methyl and phenyl analogs are sparingly soluble in polar solvents such as acetonitrile, nitromethane and di— methylformamide. The platinum complexes are more soluble than the palladium species, especially the phenyl derivative as 1,1'-bis(phenylthio)ferrocenepalladiumdichloride is very sparingly soluble and tends to form a suspension. The pallad- ium bromide analogs are more soluble than the chloride com- plexes. Analytically pure samples were obtained by the slow evaporation of the mixed solvent system, methylene chloride/ hexane. The isopropyl derivatives tend to trap one mole of methylene chloride in the crystal lattice as shown by the 1H NMR spectra and the elemental analysis of complex (§§)° The occluded solvent is removed by drying in vacuo at 90°C. The crystals tend to discolor once the solvent has been removed. 1 2. H NMR With the exception of the methyl metal complexes, (3%)- (gZ), the bimetallic derivatives were sufficiently soluble to obtain 1H NMR spectra. 1 H NMR data for the isobutyl metal complexes, (16)-(lg), is presented in Table 5. The ferrocenylsulfide ligand undergoes a significant 51 om: :m.am m:.om Hm.m: mm.ae mm.me mo.ow mumpmmAsmflumammovmm co: mm.am mm.sm Hm.ms wo.me ma.me mm.om afloaamlsmfi-mzmmovmm om: am.am aw.sm mw.mm mm.ae sm.se m:.ae mcmeamlzmfi-mzmmovmm 0mm mm.am m:.om om.mz ms.ae mm.ms mm.me mfioemmflsmfiumzmmovmm 03m mm.am Hm.wm 03.8: Hm.os HH.:~ ms.mm mflzmfinmzmmovmm mmo mo mmo m.mo :.mo Ho OMH . . . . . m m Izm ems :o H mm H :m m m: a ma m smug Asmfi m m ovmm OHQH Ho.H mm.H mo.m m:.: :H.m mfiopmmfismfiumzmmovmm 0mm Ho.H Hw.H oo.m 02.: mm.m mcmeamfismfiumzmmovmm owe Ho.H ow.H mm.m m:.: mm.m maoeamAsmfinmsmmovmm 03m No.0 ms.a om.m om.: mm.: mfismfiumzmmovmm ooe mac :0 mmo :.mm m.mm xmfiaeoo ma .mcmpa use mfiopm .mpmem .mfloem .mm>fipw>flcma Hmpmz use mAsmflumzmmovmm cog «use mzz oma new mzz 3H .m mfinme 52 change in the 1H NMR spectra upon complexing to platinum or palladium halides. Figure 2 indicates that the most strik- ing differenceir1the 1H NMR spectra of the complexed ligand relative to the free ligand is the large downfield shift of the resonance due to the H2,5 ring protons. This deshield- ing was originally thought to be due to a severe tilting of the cyclopentadienyl rings where the alpha protons were 59. The crystal struc- further from the shielding iron atom ture of the isobutyl palladium complex, (16), (discussed in detail in a later section) however, indicated that the cyclopentadienyl rings were tilted 2° from the plane. The large downfield shift of the alpha protons is either due to the magnetic anisotropy or the inductive effect of the metal halide. A further difference between the 1H NMR spectra of the free and complexed isobutyl ligand is the deshielding of the alkyl protons. In particular, the resonance due to the methylene protons shifts from 2.50 to 2.98 ppm as shown in Figure 2. Figure 3 illustrates the difference between the 1H NMR spectra of the isopropyl platinum and palladium chloride derivatives. The palladium complex exhibits a doublet at 1.23 ppm due to the methyl groups and a septet at u.ou ppm due to the methine protons. The broad signal at “.45 ppm and the extremely broad, low intensity signal at 5.23 ppm, which is partially obscured by the methylene chloride peak, are due to the ring protons. Both signals sharpen to give 53 ©scugum, Fe © sc H,¢H Me, 1 l —,-- .-sgf fi.7-22°c 6 2 U! .1 w -. cuprm. @s\ /.. ©s/K. CH ,cmae, ll 1 . -1. 6 5 4 1 Figure 2. H NMR Spectra of Fe(CSHuS-iBu)2 and Fe(CSHuS-iBu)2Pd012. 514 la-Pr S\\~ CH Fe PU \ / Cl 5 01,0, ‘4’! Iffrrr"frrfi* *rffi"rfi** rfi 6 5 4 3 2 1 i-Pr SI Cl F \ I e /Pt\c CHzClz I \é-Pr *r" r' '1' I " r' v'I f 6 5 4 3 2 1 1 Figure 3. H NMR Spectra of Fe(CSHuS—iPr)2Pd012 and Fe(CSHuS—iPr)2PtC12. 55 "triplets" at high temperature. The platinum analog ex- hibits a far more complex spectrum than the palladium com- pound. These differences are due to different rates of pyramidal sulfur inversion. This phenomenon is considered in detail in a later section. Both isopropyl complexes contain a peak at 5.28 ppm that is due to methylene chloride which is trapped in the crystal lattice. 1 In the H NMR spectra of the platinum complexes addi- tional coupling from the 195Pt nucleus (I = l/2, 33%) was anticipated. The methine region in Figure 3 should con- sist of a lzuzl proton resonance (in addition to coupling due to the methyl protons) where the outer signals are 195Pt 1 satellites. At 250 MHz however 195Pt- H coupling is not always observed due to the chemical shift anisotropy mechan— ism which dominates the relaxation of 195Pt at high field6o. 3. 130 NMR 13C NMR data were obtained only for the isobutyl com- plexes, (28)-(23) and the results are presented in Table 5. When the isobutyl ligand complexes to the metal halide the Cl chemical shift moves upfield by roughly 3 ppm whereas the C3,“ and 02,5 resonances move downfield by 1-2 ppm. This shift is consistent with the inductive effect of the elec— tron-withdrawing metal halide as electron density is "drawn" and 0 towards the 0 atom (upfield shift) from the C 1 2,5 3,4 nuclei (downfield shift)53. The signal due to the methylene 56 carbon moves downfield by 3-6 ppm while the resonance due to the methine carbon shifts upfield by at least 2 ppm upon complexation. The upfield shift of the methine carbon could be explained by a through space interaction rather than by a simple inductive effect. 4. Infrared Spectra The infrared spectra of the ferrocenylsulfide metal com- plexes are very similar to one another and to the free, un- complexed ligand. Figure A illustrates representative spectra of the 1,1'-bis(phenylthio)ferrocene ligand, Figure (Ma), and the corresponding platinum chloride and palladium bromide metal complexes, Figure (Ab and Ac). Close inspec— tion of the infrared spectra presented in Figure A reveals that the single absorption at 820 cm-1 in the free ligand undergoes splitting in the metal complexes. There also ap- pears to be additional splitting in the metal complexes in the 500 cm'1 region where metal-ring vibrations occur. The most striking change occurs in the low frequency region where metal-ligand vibrations are prevalent. Metal- sulfur vibrations are generally of low intensity and are found at higher frequencies than metal-halide stretches6l. Table 6 contains the infrared data for the metal chloride and bromide complexes in the region MOO—200 cm-l. Comparison of the chloride and bromide analogs, as shown in Figure 5, permits assignment of the metal-sulfur and the metal-halogen 57 .mtmoamAhamqmmovom on cam maopmmAcmmzmmovmm ADV .mflcmmzzmovom Amv no whoomom UmLMpmcH .2 opsmflm 750 20:95:23; on“ 00v 000 00. 000. 8a. 00! 8! 002 80a can a 000» can» coat 7 q q _ q q . . A a . q a . Table 6. iBu, 200 cm‘l. Plates. 58 Infrared Data for Fe(CSHuSR)2MX2 (R = Me, iPr, Ph, Bz; M = Pd, Pt; X Cl, Br) in Region A00— Measured as Nujol Mulls between CsBr Pd Pt Cl Br Cl Br Fe(C H S—iBu) MX 370sh 372w 372w 5 A 2 2 365w 365w 36uw (M-S) 307s 323s ——-- 290sh 222s 310m —-—- (M‘X) Fe(CSHuSPh)2MX2 360w 360w 372w 368w (M-S) 3283 3305 320sh 210s 318s 220m (M’X) Fe(C H SMe) MX 355w 5 u 2 2 3u0w 3u3w 3u5w 340w (M's) 3108 3233 290sh 205m 305s 208s (M‘X) Fe(C H SCH Ph) MX 358m 353m 360m 358m _ 5 u 2 2 2 335w 3u5sh 350m 3u8m (M S) 3158 3328 300s 230s 312s 215s (M‘X) Fe(C H S-iPr) MX 380w 380w 383w 389w 5 u 2 2 360w 360w 365w 370w (M’S) 320s 328s 305s 225s 315s 220s (M‘X) (PhSC H SPh)MX 323vs 3505 3A9sh _ 3 6 2 308vs 316s 329s 32uw (M S) 278s ---- 317s 262s ---- 312s 266m (M‘X) 59 .Azoaoov pm pcm .Am>onmv Ho u x when: m xpmmflsmfilmzmmovom pom coamom pmpmamcH pom .m opswfim m xodmflsmanmsmmovcm ode 60 stretching frequencies. The metal-bromide stretch occurs at a lower frequency than the metal-chloride stretch and in some cases the metal-bromide stretch is not within range of the instrumentation used. The proposed assignments are in close agreement with those for the chelated thioether complex, (PhSC3H6SPh)MX2 (M = Pd, Pt; X = Cl, Br)62, (the last entry in Table 6) in addition to other values that have been reported6l’63. The metal—chloride stretches in complexes of the type cis(Pd012L2), where L is a phosphine, are found at lower frequencies than those given in Table 66“. One hesitates to correlate the higher metal-chloride stretch in the sul- fide complexes with a weaker trans-influence of the sulfide ligands due to the possible coupling between the metal- chloride stretch and other vibrational modes within the molecule. 5. Ultraviolet and Visible Spectra Table 7 contains the ultraviolet and visible spectral data for the complexes Fe(CSHLlSR)2 and Fe(CSHuSR)2MX2 where R = iBu, Ph, Me; M = Pd, Pt; X = Cl, Br. The methyl derivatives did not completely dissolve in acetonitrile solution and consequently the extinction coefficients for these compounds are unreliable. In the electronic absorption spectrum of ferrocene the low energy bands at “#0 nm (e = 91) and 325 nm (e = M9) Table 7. 61 Electronic Absorption Spectra of Ferrocenylsul- fide Complexes Fe(CSHuSR)2, Fe(CSHuSR)2MX2 (R = iBu, Ph, Me; M = MeCN Solution at approx. conc. 8.0xlO"5 M. Pd, Pt; X = 01, Br) at 2M°c in A s max (nm) (M-lcm—l) Fe(C H S-iBu) M25a 190 5 u 2 300: 1900 352 7000 210 32000 Fe(CSHuSPh)2 M30 500 2M8 32500 220 35000 Fe(C H SMe) M30 ..... 5 u 2 210 ————— 01 A e A a max max (nm) (M-lcm'l) (nm) (M-lcm- Fe(CSHuS-iBu)2PdX2 M00 760 M03 MMO 330 2300 309 262 11600 275 5100 210 30000 210 30000 Fe(CSHuSPh)2PdX2 M00a 800 M33 500 317 1M60 2M5 1500 255 6900 210 30000 210 30000 Fe(CBHuSMe)2PdX2 M00a ----- M00 210 319 1600 309 820 25M 5700 270 1600 210 30000 210 5000 Fe(CSHuS-iBu)2PtX2 M10 MM0 M20 390 300 2200 232sh 1M800 23M 16900 210 30000 210 30000 62 Table 7. Continued. Cl Br A e A s max max (nm) (M-lcm-l) (nm) (M7lcm-l) Fe(CSHuSPh)PtX2 M20 950 “28 660 360 1220 300 2000 318 3M00 2M5 11900 278 1u300 210 30000 210 30000 Fe(CSHuSMe)2PtX2 M20 470 360 9M0 310 2800 290 1200 2U2 11800 265sh 12700 210 30000 210 30000 a . . . . Difficult to ascertain exact max1mum. b Inflection point. 63 have been attributed to spin allowed d-d transitions. In the near ultraviolet region the band at 200 nm (e = 51000) has been assigned as a ligand-to-metal charge-transfer band and the shoulders at 265 nm and 2&0 nm have been assigned as metal-to-ligand charge-transfer transitions65. The ferrocenylsulfide ligands exhibit an absorption at M25-M30 nm (e = 102) which is probably due to a d-d transi- tion. The intensity of this band is significantly increased over the corresponding band in ferrocene. In the ferro- cenylsulfide metal complexes the absorption due to the d-d transition is slightly blue shifted to MOO-M20 nm in com- parison to ferrocene. The ferrocenylsulfide palladium complexes exhibit an intense, well defined maximum around 260 nm whereas the platinum analogs have an inflection point in this region. This band could be associated with a metal-to-ligand charge- transfer band whereas the strong absorption at 210 nm is probably a ligand-to-metal charge-transfer transition. 6. Structure The structure and numbering scheme of l,l'-bis(iso— butylthio)ferrocenepalladiumdichloride, (16), is shown in Figure 6 while a stereoview is given in Figure 7. Hydrogen atoms have been omitted for clarity. The positional parameters and anisotrOpic thermal parameters are given in Table 8 and Table 9, respectively. In (16) the palladium Figure 6. Structure and Numbering Scheme for Fe(CSHuS-iBu)2— PdCl2. 65 4‘ e ....e . . .Jm': «V. .m e./<\ Hooamflsmaumsmmovoa oo soa>oohoom .8 mpswfim 66 Table 8. Positional Parameters and Esd's for Fe(CSHuS-iBu)2- PdCl . 2 Atom x Y z Pd(l) .02M86(2) .05361(3) .09M52(2) Fe(l) .03M07(3) -.00389(6) .3265M(M) 01(1) -.0M093(6) .1867(1) 0M318(8) 01(2) .1033M(7) .1636(1) 0M76(1) S(l) -.0509M(6) —.0531(1) 156M5(8) S(2) .08816(6) -.0853(1) .1M708(8) 0(1) -.0M32(2) -.01M0(M) .2609(3) 0(2) -.OMS3(2) -.0915(5) .329M(3) 0(3) -.0390(3) -.0296(6) .M017(M) C(M) -.03M0(3) .0855(5) .3816(M) 0(5) -.0366(2) .0962(5) .29MM(3) 0(6) -.12M9(2) .0076(5) .135M(M) 0(7) - 175M(3) -.0M00(5) .1873(M) 0(8) —.2299(M) .0337(7) .17M8(6) C(9) - 1887(8) .1590(7) .16M3(5) 0(11) .1028(2) —.0M01(M) .2M96(3) 0(12) .1088(2) .0711(5) .2793(M) 0(13) .1138(3) .0653(6) .3662(M) 0(1M) .1102(3) -.0M86(6) .3895(M) 0(15) .10MM(3) -.1151(5) .3182(3) 0(16) .1625(3) -.0766(6) .0997(M) C(17) .2065(3) -.16M8(5) 1278(5) 0(18) .2681(3) -.1361(6) 0899(5) 0(19) .1877(M) —.2819(7) 1013(6) 8(2) - 048(2) -.165(M) .327(3) H(3) - 036(2) -.059(M) .M52(3) H(M) - 028(2) .1M8(M) .Ml7(3) H(S) - 030(2) .163(M) .259(3) H(6A) - 130(3) -.019(5) .077(M) H(6B) - 122(2) .08M(M) .1MO(3) H(7) -.170(2) —.0M9(A) .253(M) H(12) .110(2) .139(M) .2M8(3) H(l3) .115(2) .123(M) M00(3) H(1M) .111(2) —.06M(M) .M3M(3) H(15) .101(2) -.196(M) .31M(3) H(16A) .17M(3) -.000(5) 10M(M) H(16B) .153(3) —.O95(A) .039(M) H(l7) .201(3) -.155(5) .196(M) H(8A) — 2M16 .03M2 1156 H(8B) - 2203 1119 1939 H(8C) - 26M2 0032 2093 H(9A) - 1527 — 206 1732 H(9B) - 2017 - 1618 105 Table 8. Continued. Atom X Y Z H(QC) -.2229 —.1868 .2003 H(18A) .2813 -.0606 .1092 H(18B) .26u7 -.136u .0278 H(18C) .298” -.19“5 .107 H(19A) .18uu -.2858 .OUIO H(19B) .1U75 -.2995 .128 H(19C) .2182 -.337U .1225 68 Amvs.u Amvm.u Amvm.- Amvm.m Amvm.: Amvm.m Amavo Amvm. Amvo. Amvm.- “moo.m Amvs.m Amvm.m Amavo ANVH. Amvm.- Amvm. Amvm.m Amvo.m Amva.m Aaavo Amvo.a Amos. Amos. Amvm.s Amvo.s Amos.m Asvo Amvm. Amvm.u Amva. Amvm.m Amos.m Amvw.m ono Amvm.u Amvo. Amvm. Amva.m Amos.m Amva.m Amvo Amvm.au Amva. Amos.a Amvo.m Amvm.m Amvm.m Asvo Amos. Amvm. Amos. Amvm.m Asvw.o Amvs.m Amvo Amos. Amvm. Amva.u Amvw.m Amvo.s Amvo.m Amvo Amvm. Amva. Amvm. Amvo.m Amvm.m Amva.m Aavo AsVOH.- Aavsm. Assas. Amvmm.m Amvma.m Amvom.m “mom Asvzo. Asvaa.- Asvao. Amvmm.m Amvmm.m Amvam.m Aavm Amvmm. onsm.a Amvmm.- vass.s Accom.m Asvsm.m Amvao Amvom. Amvmm.- Amvas. Acvom.m Amvmm.m Asvmo.: AHVHo Amvso.u Amvma.- Amvma. Amvma.m Amvsm.m Amvom.m AHVom Aavmo.u Afivom. Aavmo. Aavoo.m Aavmm.m Aavsm.m Aavoa mmm mam mam mmm mmm Ham soc< .m cwsohcp H pmaozo ohm h pcm H new .comccH Haoo accosoaoot 6:6 ma mm3 mmpux esp ma manmH pew .wuonplm mawcw coapompmmflp map Eopm we mums» 060:3 ms*A .w madman u.2 Ill/J Bennlllj l 3’!) la a . [/1 j q 1 ‘ Nas. :ufo, w a) \ um \\ cl 6 /\m@ ~22 10:10 78 In the slow exchange limit, which occurs at -40°C, the ring protons appear as four singlets. Enhancement of the low temperature spectrum indicates that the two low field singlets are in fact doublets and the two high field singlets are triplets. The isobutyl protons also exhibit temperature dependent NMR spectra. At the fast exchange limit, 72°C, the doub- let at 2.98 ppm is due to the methylene protons, the multip- let at 1.80 ppm is due to the methine protons and the doublet at 1.01 ppm is assigned to the methyl protons. As the tem- perature is lowered, the signals broaden with the methylene protons coalescing at around 39°C. At the slow exchange limit the methylene protons exist as four doublets. The methyl protons coalesce at around -10°C and appear as two distinct doublets at -67°C. The methine multiplet is broad at low temperature probably due to two overlapping multiplets. The temperature dependent NMR spectra could arise from two independent dynamic processes. A bridge reversal pro- cess has been observed in [3errrocenOphanes with tri— chalcogen bridges. As discussed in the section dealing with the structure, the dihedral angle between the planes Fe-S -S and S -S -S is 110.9°C in the trithiane, (l). 1 3 l 2 3 The bridging sulfur, 82, flips from one side of the Fe—Sl- S3 plane to the other side on the NMR time scale and gives rise to a fluxional molecule. This process is analogous 79 to the chair-to-chair reversal observed in six membered rings72. In the dynamic NMR spectra of the trithiane, (1), the low temperature spectrum consists of four complex signals which, upon raising the temperature, collapse to give two distorted "triplets" at high temperature. At the slow exchange limit, bridge reversal is slow with respect to the NMR time scale and this gives rise to an ABCD spectrum where the four cyclopentadienyl ring protons are aniso- chronous. As the temperature is raised, bridge reversal becomes rapid relative to the NMR time scale and the spectrum consists of two distorted "triplets" arising from an AA'BB' spin system73. The crystal structure of the isobutyl palladium com- plex, (16), indicated that the dihedral angle between the Fe-Sl—S2 plane and the S -Pd-S2 plane was 75.4°. Bridge 1 reversal could occur in this complex in much the same way it occurs in the trithiane complex. The palladium atom could flip from one side of the Fe-Sl-S2 plane to the other side as illustrated in Scheme 7. @213“ e 6%? d Scheme 7. 80 The change observed in the NMR spectra of (16), as shown in Figure 8, in the cyclopentadienyl proton region is very similar to that observed for the fluxional [3]- ferrocenophane, (1). Bridge reversal, where the palladium atom flips back and forth, could account for the temperature dependence of the NMR spectra of complex (16). The second dynamic process is the pyramidal inversion of sulfur. Inversion rates at sulfur are often too slow to be observed on the NMR time scale, as for example in the 74 sulfoxides, but recently Abel and coworkers and other groups75 have determined the barrier energies for sulfur inversion in a variety of palladium and platinum complexes. In a recent study 0rrell76 determined the inversion barriers at sulfur in compounds of the type cis[MX2L], (M = Pd, Pt; X = Cl, Br, I; L = MeS(CH2)2SMe, MeS(CH2)38Me). At low temperatures, in the absence of sulfur inversion, two diastereoisomers, meso and DL, were detected. When pyramidal inversion of sulfur is slow, the sulfur atoms are chiral and this leads to the presence of diastereo- isomers. Sulfur inversion could occur in the isobutyl palladium complex, (16), as illustrated in Scheme 8. The two sulfur atoms are centers of chirality and at low temperature, when pyramidal inversion is slow relative to the NMR time scale the methylene protons in the isobutyl groups are diastereo- topic and give rise to an ABC spin system77. At -42°C 81 4‘ ”3° é-Buz /—\ eC-C-H F—'\ 9 (j? s°é Ike ' <3; S‘fihlve \ /\) <2 "3" \ / 9-C-H. Pd I?“ "5'“! / \ Cl \Cl Scheme 8 two sets of doublets are observed for each methylene proton. Splitting pattern could be shown below where JAB is 13 Hz and J and J are 5.4 Hz and 8.8 Hz. AC BC | | | | J.. J... 1.. When pyramidal inversion is rapid, at 72°C, there is the equivalent of a planar configuration at sulfur and the methylene protons are split only by the methine proton with J = 6.8 Hz. Activation parameters for the sulfur inversion process have recently been obtained for Fe(CSHuS-iBu)2PdX2, 82 78. The AG+ X = Cl, Br, from a total line shape analysis values are 13.88 i 0.09 and 13.42 i 0.12 kcal/mol for the chloride and bromide complex respectively. The difference in the energy barriers suggest that there is a measurable halogen trans—influence on the energy process. Orrell proposes that the NMR spectral changes are due solely to sulfur inversion and that the bridge reversal of the palladium atom is fast on the NMR time scale even at -90°C. In addition Orrell prOposes that only one di— astereoisomer is present and he favors the DL isomer as ring reversal is rapid. This is consistent with the crystallographic data as the DL isomer was the struc- ture observed in the solid state. Variable temperature 195Pt NMR studies indicate that there are two diastereo- isomers in a relative population of 10:1 at -35°C. Only the major diastereoisomer however appears to be observed 1 in the H NMR spectra. Figure 9 shows the variable temperature 1 H NMR spectra for the isobutyl platinum analog, (18). The same features are present as those in the palladium complex except that higher coalescence temperatures are observed for the platinum complex. This implies that the platinum complexes have a higher energy barrier than the palladium complexes and this suggests that the Pt-S bond is stronger than the 74a 750 Pd-S bond. Abel and Cross have observed a similar trend in other platinum and palladium derivatives. 83 .maoummflzmfilmzmmovmm mom mpuooom mzz ma opSmeoQEou manwfipm> .m mpswfim 84 The only significant difference between the dynamic NMR spectra of the palladium and platinum complexes occurs in the methylene region at low temperature. In the platinum complex the four sets of doublets which are visible at -6°C appear to coalesce to give a single broad resonance at -50°C. An explanation for this behavior awaits the results of Orrell's total line shape analysis. Variable temperature 13C NMR data also suggests the presence of a dynamic process. The cyclopentadienyl ring carbon region is shown in Figure 10. At 35°C a sharp sig- nal is observed for the substituted carbon, Cl’ at 79.3 ppm and two broad signals are observed at 76.0 and 71.7 ppm for the C3 4 and C nuclei, respectively. The signal at 9 2,5 71.7 ppm coalesces at this temperature and at 9°C two separate signals are evident. The lower field signal at 76 ppm coalesces around -6°C and gradually separates into two distinct signals. At -20.8°C the carbon atoms in the alpha and beta positions, C2,5 and C3,“, are magnetically non-equivalent and give rise to four separate signals. At -54.6°C these four resonances are further split and now appear as four doublets. The signal due to the Cl carbon remains a sharp singlet with no indication of broad- ening from an overlapping signal. The additional splitting could be due to the sulfur inversion process, which is now slow on the NMR time scale, that produces diastereotopic carbon nuclei that are anisochronous. 85 Figure 10. Variable Temperature 13C NMR for Fe(CSHuS-iBu)2PdCl2 in Region from 70—80 ppm. 87 8. 195Pt NMR Table 15 contains the 195Pt NMR data for the ferro— cenylsulfide complexes Fe(CSHuSR)2PtX2, R = iBu, iPr, Ph, Bz; X = Cl, Br in chloroform-dl solutions at ambient tem- perature. The methylthioferrocene platinum complexes were too insoluble to obtain 195Pt NMR data. 195Pt (I = 1/2) has a natural abundance of 33% and has roughly the same relative sensitivity as the 13C nucleus. In Table 15 the 195Pt chemical shifts of the chloride complexes are found at 3,200 ppm whereas the bromide ana- logs are found 400 ppm further upfield. This is consistent with general trends where bromide complexes are found up- field of the chloride analogs79. As 195Pt chemical shifts are solvent and temperature dependent caution must be em- ployed when data from different sources is compared. 1,1'-Bis(diphenylphosphino)ferroceneplatinumdichloride has a signal, a triplet with J = 3,374 Hz due to coupling Pt-P to two phosphorous atoms, about 1,100 ppm upfield from the corresponding ferrocenylsulfide complexes. This is con- sistent with previously reported dataBO, shown as the last two entries in Table 15, where the phosphine compound, [(PMe3)2PtCl2], is found about 900 ppm upfield from the sulfide complex, [(SMe2)2PtCl2]. This suggests that sul- fide ligands are weaker sigma donors than the phosphine analogs. As 195Pt NMR is very sensitive to changes in the metal 88 Table 15. 195Pt NMR Data for Ferrocenylsulfideplatinum Complexes, Fe(CSHuSR)PtX R = iBu, iPr, Ph, 2’ B2; X = Cl, Br. Measurements made in CDCl3 solution at room temperature. Compound Cl Br Fe(CSHuS—iBu)2PtX2 —3,285 -3,662 ‘33353 ’33759 Fe(CSHuS-iPr)2PtX2 -3,253 -3,633 Fe(CSHuSPh)2PtX2 -3,246 -3,658 Fe(CSHuSCHZPh)2PtX2 —3,244 —3,622 Fe(CSHuPPh2)2PtCl2 —4,374t ----- . a Cls—[(SMe2)2PtX21 -3,551 '3,879 cis-[(PMe3)2PtX2]a -4,408 —4,636 aObtained from Reference 80, measured in CH2C12. coordination sphere we could eXpect to see two separate platinum resonances for the diastereoisomers obtained from slow inversion at the sulfur centers. For the isobutyl platinum compound, (18), two distinct signals, separated by 74 ppm, were observed at room temperature in a 17:1 ratio. Variable temperature measurements of this complex indicate that the relative populations of the diastereo- isomers increase as the temperature decreases. Figure 11 89 M 7 7 i-Bu / ©s\ /cn Fe pt / \CI ©s \ é-Bu VLAH9 0 }j -35°C -n°C W 22°C Figure 11. Variable Temperature 195Pt NMR Spectra for Fe(CSHuS-iBu)2PtC12. 90 shows that at —35°C two signals in an approximate 10:1 ratio were observed. A set of variable temperature data was also obtained for the isopropyl platinum compound, (22) and is presented in Figure 12. The spectrum at -10°C indicates the presence of two diastereoisomers in a 70:1 ratio. This large ratio suggests that the isomer found upfield is very much a minor diastereoisomer. Pre— gosin81 has observed two 195Pt resonances, separated by 51 ppm, for the two diastereoisomers of trans-dichloro- [(S)-N-methyl-d-methylbenzylamineJ(ethylene)p1atinum[II]. 9. Electrochemistry The electrochemistry of the complexes Fe(CSHuSR)2 2 where R = iBu, Ph, Me; M = Pd, Pt has been examined by cyclic voltammetry. All measurements and Fe(CSHuSR)2MCl were made in either DMF or CH2C12 solutions with 0.1 M [nBuuN][C10“] as a supporting electrolyte. A platinum "flag" electrode or a glassy carbon electrode was used as the working electrode and values were recorded relative to a standard calomel reference electrode (SCE). Sweep rates were varied from 100 mV/s to 200 mV/s over a poten- tial range of +2.0 to -1.5 V. Well defined, one-electron reversible redox waves were observed for the oxidation of the iron atom in the ferrocenylsulfide complexes, (9), (10), (12), and the platinum analogs in DMF and CH2C12 solutions respectively. 91 p F F 64’! / S \ Pt/c ©‘K /H\m i-Pr M A W "1°C 23°C -3§C Figure 12. Variable Temperature 195Pt NMR Spectra for Fe(CSHuS-iPr’ ) 2PtCl2 . 92 The palladium complexes, which are sparingly soluble in these solvents (especially the phenyl derivative), gave poorly defined redox waves with large peak separations. Representative cyclic voltammograms are shown in Figure 13. Examination of the electrochemical data given in Table 16 indicates that E1/2 for the ferrocene ligands becomes increasingly positive in the order S-iBu < SMe < SPh < PPh2. This is consistent with ferrocene being stabilized by electron withdrawing substituents. Upon complexation to palladium or platinum, E /2 for the ferrocenyl group in- 1 creases by +0.4 to +0.59 V. This increase could be at- tributed to a through space electrostatic interaction with the positive charge on Pd or Pt or alternatively could be viewed as Pd or Pt withdrawing electron density from the ferro- cenyl group through the sulfide bridges. The theoretical difference in peak potentials for anodic and cathodic waves (AEp) is 57 mV for a reversible one-electron process. This difference can however be affected by the cell design and the solvent. In Table 16 peak separations are as large as 350 mV. McCleverty82 has reported peak separations of up to 230 mV for the reversible oxidation of Cr(C0)5L complexes, where L is a phosphine, in CH2C12 solution. In addition Kotz83 has reported a peak separation of 310 mV for the reversible system FcPPh2 in CH2C12 solution. Moreover, under the experimental conditions employed, ferrocene, a completely 93 ISpA B C A CURRENT l l l l l l l l J l I l l 1 l.2 H 10 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.! 0 \NDLTS Figure 13. Cyclic Voltammograms of (A) Fe(CSHuS-iBu)2, (B) Fe(CSHuS-iBu)2PtC12 and (0) Fe(CSHus_1Bu)2_ PdClZ. 94 Table 16. Cyclic Voltammetry Data for Ferrocenylsulfide Complexes. Compound E%(V)a AEp(mV)b iC/iaU 815W)a Fe(C5H5)2 0.3M 250 1.00 Fe(CSHuS-iBu)2 8:37 158 8:57d Fe(CSHuSMe)2 8:35 33g 8:32d “(8585892 8:88 888 8:88.531 Fe(CSHuPPh2)2 8.2% 128 3:6ge Fe(CSHuS-iBu)2Pd012 8:86 17; 8.83 :82838 Fe(CSHuSMe)2PdCl2 0.86 270 1.20 —o.66 Fe(CSHuSPh)2PdC12 0.85 360 ---- (-1.45) Fe(CSHuPPh2)2PdC12 0.93 120 0.93 -1.20 Fe(CSHuS-iBu)2PtCl2 8.83 155 1.gé —1.6Me Fe(CSHuSMe)2PtC12 8:53 122 1:18 -1.5oe Fe(CSHuSPh)2PtCl2 1:88 165 0:87 -1.38° “(85888228812 8:88 888 1:88 -lhe aE15 values were calculated as the average of the cathodic and bAEp is the difference between the cathodic and anodic peak. 0Ratio of peak currents cathodic (ic) vs. (ia) anodic waves. dMeasured in DMF/0.1 M n- BuuNC10u at 25° at Pt electrode vs. SCE. Other measurements made in CH C1 . eMeasured in CH C12/0.l M n-BuuNClOu at 2 2 glassy carbon electrode vs. SCE. anodic peak potential. 2 95 reversible system, has a peak separation of 250 mV. These results suggest that despite the large differences in cathodic and anodic peak potentials, ferrocenylsulfide complexes can be considered to undergo a chemically revers- ible one-electron process. The sulfur complexes appear to adsorb on the platinum electrode as in some cases the platinum surface was dis— colored after use. This adsorbtion could account for the ratio of the cathodic and anodic current (ic/ia) being less than unity, particularly for the free ligands, as voltam- mograms may exhibit enhancement of the peak currents in 84 the presence of weakly adsorbed material The fact that iC/ia is close to unity suggests that the process is chem- ically reversible. In the ferrocenyl platinum and palladium complexes an irreversible wave was observed at -0.6 to -1.64 V and has been attributed to Pd3/2+ or Pt3/2+. El/2 for plat- inum is more negative than the corresponding value for palladium and this is consistent with polarographic and voltammetric data obtained for palladium and platinum 85 bis-1,2—dithiolene complexes The large negative El/2 for palladium and platinum suggest that the ferrocenyl group is a strong electron donor. Ferrocene itself has only recently been reduced electro- chemically but ferrocene derivatives with electron-with- drawing groups, such as carboxy groups, are readily 96 86 reduced The large positive E1/2 value for Fe(CSHuS-iBu)2 PdCl2, (16), suggests that the iron atom in this complex may be readily reduced. After a dimethoxyethane solution of (16) was stirred in the presence of a sodium amalgam for two days, the brown solution turned yellow and a fine metallic precipitate was observed. These results suggest that as the palladium atom was reduced the ferrocenyl- sulfide ligand was cleaved. D. Fe(CSMuSCSNR2_)_2 and Fe(C5M5)_(_C_5MuSCSNR2) (R = Me, Etz iPr) 1. Preparation Dithiocarbamates and thiuram disulfides have been used as fungicides, pesticides, vulcanization accelerators, antioxidants, flotation agents and high-pressure lubricants 87 and as drugs in medicine In particular complexes of heavy metals with thiuram disulfides are effective fungi- cides and seed disinfectants. The rich and diverse chem- istry of dithio acid and dithiolate complexes has been extensively covered in many reviews88. Tetraalkylthiuram disulfides undergo nucleophilic attack at the disulfide linkage by cyanide ions, amines 89 90 and Grignard reagents Recently Cava reported that aryllithium derivatives react with tetraisopropylthiuram disulfide to give dithiocarbamate esters that were 97 precursors to aromatic thiols. Reaction of dilithioferrocene with a series of tetra- alkylthiuram disulfides gave rise to a high yield prepara- tion of bis(dialkyldithiocarbamate)ferrocene derivatives. A solution of the tetraalkylthiuram disulfide was added slowly via cannula to a hexane slurry of dilithio- ferrocene which had been cooled to -78°C (see Scheme 9). fi ©Li fi fi @501“, Fe ~TMEDA R’NCSSCNR’ Fe :’ , a = Me, Et, i-Pr LI SfiNR, 5 Scheme 9 In contrast to the results obtained by Cava only the desired product and no thioamide derivative was observed. The thioamide species arises from competing nucleOphilic attack at the thione carbon rather than at the sulfur- sulfur bond in the tetraalkylthiuram disulfide. A series of monosubstituted dialkyldithiocarbamate- ferrocene derivatives was also prepared by reaction of tetraalkylthiuram disulfide with lithioferrocene. Dif- ficulty was encountered when lithioferrocene was prepared 98 from ferrocene and a mixture of n—butyllithium and TMEDA. Dimethyldithiocarbamateferrocene was isolated in 9% yield by using this procedure. At least six different products were observed when chromatography was employed in isolating the dimethyl derivative. One product, which was obtained from the second band on the column, has spectral data con- sistent with the binuclear complex, Fc-S-Fc, where two ferrocene moeities are linked through a sulfur atom. An alternative route where lithioferrocene is generated from bromoferrocene gave monosubstituted dialkyldithiocarbamate- ferrocene derivatives in 60—80% yields as shown in Scheme 10. 1‘: QB! QSCNRI Fe ”-8" Li + Fe 52(CSNR2)2 © a = Et, .4»: Scheme 10 On one occasion a very concentrated solution of bromoferro- cene (about 2 g in 10 mL ether) was lithiated and after addition of tetraalkylthiuram disulfide, biferrocene which was identified by NMR and mass spectrometry was obtained. 99 Consistent results were obtained when dilute solutions of bromoferrocene were used. 2. 1H NMR 1H NMR data for the dialkyldithiocarbamateferrocene derivatives, (38) - (43), is given in Table 17. The 1H NMR spectra of these complexes are very similar to the spectra obtained for the ferrocenylsulfide compounds in the previous section. Two apparent "triplets" are observed for the cyclopentadienyl ring protons that is consistent with an AA'BB' spin system. The "triplets" are slightly de- shielded as anticipated for the electron withdrawing di- thiocarbamate substituent. The 1H NMR spectra for the mono and bis(diethyldithio- carbamate)ferrocene species are shown in Figure 14 and are identical except that a singlet at 4.22 ppm is observed for the unsubstituted cyclopentadienyl ring in the mono substituted ferrocene. Two separate signals are observed for the N,N-dialkyl protons due to restricted rotation around the carbamate C-N bond. This process will be ex- amined more closely at a later stage. 3. 130 NMR The 13C NMR data for the dialkyldithiocarbamateferro- cene complexes is presented in Table 18. During acquisition 100 om:.: 58s.: mom.: omm.: uo:.: mm Amaaazmom:mmoVAmzmovog cmm.: 6mm.m osm.: omm.m mmm.: o:m.: om:.: mm Amomzmom:mmoVAmmmovma mmz.m m:m.m m:m.: o:m.: u::.: mm AmoZZmom:mmov1mmmovom os:.: Ems.: nwm.: oo:.: :s mAmcaazmom:mmovom omm.: omm.m omm.: omm.m mm:.: omm.: mm mAmomzmom:mmovog mm:.m mw:.m cwm.: oom.: mm mflmczzmom:mmovoa mmo mmo mo mmmo : mm m.mz ooe ocsoogoo .Aha: .om .6: u mv moonQEoo Ammzmom:mmov1mzmovca occ mfimmzmom:mmovom com 6666 mzz m H .NH mHDMB 101 8 ©scmst2 é T'*T*" r ' Vfi'fifr 4 3 1 s T I I I 4 3 2 1 1 Figure 14. H NMR Spectra of Fe(C5H5)(C5HuSCSNEt2) (above) and Fe(CSHuSCSNEt2)2 (below). .oom: pm UoQSmmozn .oom: um ooLSmmmzm 102 m.m: :.:m :.mo m.os c.8s m.ms o.mm: oAmamazmom:mmOVAmmmovom m.HH m.s: m : m m m w.m: m.m: m.mw :.os o.os m.ms s.mm: A umzmom m ov m ovom m.:: m : m m m m.m: :.mo m.os s.ms o.ms m.mma A ozzmom m oVA m ovoa s.m: m.mm m.:s :.ss H.8s o.wm: cmAmhmazmom:mmovom m.:: m.m: m m : m m.m: :.m: m.as m.ss H.ms m.sma A umzmom m ovmm s.:: m m : m m.m: o.ms :.ss s.os s.mm: A czzmom m ovog mzo web go memo :.mo m.mo Ho mno ocsoQEoo .szocm mmflzhmzpo mmoacs oLSpmthEop pcofinsm pm cofiusaom ONQ\Hommo CH pmHSmmoz .Asa: .o@ .62 n ma Ammzmom:meVAmmmovoa one mAmmzmom:mmovom sou sumo mzz om: .m: 6:685 103 of the 13C NMR data the parameters PW = 4 us and RD = 45 were used as the thiocarbonyl carbon has a long relaxation time, T1' The low field signal at 199 ppm is attributed to the thiocarbonyl. Figure 15 displays an expansion of the region from 70—80 ppm for Fe(CBHuS-iPr)2, Fe(CSHuSCSNiPr2)2 and Fe(C5H5)(C5HuSCSNEt2). Tentative assignments have been made for the cyclopentadienyl ring carbons. In Figure 15(A) the C resonance is found downfield at 79.14 ppm 1 whereas in Figure 15(B) and (C) the C resonance is found 1 upfield of the C signal at 76.1 and 75.2 ppm respectively. 2,5 The major difference between the spectra in Figure 15 is that the C resonance is deshielded in Figure 15(B) and 2,5 (C) whereas in Figure 15(A) the C2,5 resonance is located at 70.96 ppm. The downfield shift in the dialkyldithio- carbamateferrocene derivatives could be due to the magnetic anisotropy of the thiocarbonyl groupgl. In Figure 15(C) the peak at 69.5 ppm is due to the unsubstituted cyclopenta- dienyl ring. The 13C NMR spectra, shown in Figure 16, illustrate that the dialkyl groups exhibit two separate signals at room temperature. This phenomenon is due to the restricted rotation around the carbamate C-N bond and will be discussed in detail later. 104 c3 c2, 5 ,4 03,] cJ Gas 0 a; C25 C1 01 C1 L .___LJ LJ L___ w 0'0 7'0 8'0 7'6 7'5 7'0 A a c Figure 15. 13c NMR Spectra of (A) Fe(CSHuS-iPr)2, (B) Fe(CSHuSCSNiPr2)2 and (C) Fe(CSH5)(C5HuSCSNEt2). 4. Ultraviolet and Visible Spectra Table 19 contains the absorption spectra of the dialkyl- dithiocarbamateferrocene derivatives. The absorption maxima are identical for the corresponding mono and bis-substituted ferrocene complexes except that the extinction coefficients for the bands centered at 270 and 240 nm are significantly lower for the monosubstituted species as shown in Figure 17. Tentative assignments of the absorption maxima have been made. The band centered around 430 nm (e = 350 M'lcm-l) which corresponds to the d-d transition observed in ferrocene H SCNMQ {0) Fe 11 70 L— .i P Y fi 1' ‘7 Y Y I I 200 1.0 160 1‘0 120 1.0 I. 60 40 S s H SCNEQ I Fe 5 I ©SCNHZ Figure 16. 130 NMR Spectra of Fe(C5H5)(C5HuSCSNMe2) (above) and Fe(CSHuSCSNEt2)2 (below). 106 Table 19. Electronic Absorption Spectra of Fe(CSHuSCSNR2)2 and Fe(C5H5)(C5HuSCSNR2) (R = Me, Et, iPr) at 24°C in MeCN solution at approx. conc. 8.0 x 10"5 M. R Fe(CSHuSCSNR2)2 Fe(C5H5)(C5HuSCSNR2) A e A c max max (nm) (H'lcm‘l) (hm) (m’lcm‘l) Me M30 360 437 M10 273 17400 272 12900 2M1 24100 239sh 17000 220 30000 210 30000 Rt 435 380 435 300 276 17200 274 12800 245 25500 245 16900 223sh 29700 205sh 32M00 210 30000 iPr 437 370 “37 310 279 13500 27M 12800 2M8 22100 2M7 1M300 227 25600 205 31600 210 30000 107 30000 ‘ I> 25000 I 20000 |5000 L E (M" cm") I l0000 I 5000 200 300 400 500 600 700 Figure 17. Ultraviolet-visible Spectra for (A) Fe(C5H4- SCSNMe2)2 and (B) Fe(C5H5)(C5HuSCSNMe2). 108 at 440 nm is very slightly blue shifted compared to ferro- cene. Two well defined maxima at 270 and 240 nm are charac- teristic of dithiocarbamates and have been assigned as intra- ligand transitions92. The band at 272 - 279 nm (e m 10“) could be assigned to a n + 0* transition whereas the band at 239 - 248 nm (e m 10“) could be a n + 0* transition. The high energy band located at 210 nm (e m 30,000) is probably a ligand-to-metal charge-transfer band associated with 65 ferrocene 5. Dynamic NMR Studies Two resonance forms possible for the dialkyldithiocar- bamateferrocene complexes are shown below. 9 S 5 I! . 7! \ \ / \c_4[/ <+—+> ,C=*L\ rcs’ \R “‘5 R The second resonance form introduces a degree of double bond character into the carbon-nitrogen bond which prevents free rotation around the C-N bond. 1H NMR and 130 NMR data in Table 17 and Table 18 The respectively indicate that two separate signals are observed for the N,N-dialkyl protons for the methyl and ethyl di- thiocarbamateferrocene derivatives, (38), (39), (41) and (42) at room temperature. When the temperature is raised 109 the two N,N-dialkyl signals coalesce and as the fast exchange limit is approached they sharpen to a single peak. The protons on the cyclopentadienyl rings show no variation with temperature. The behavior of the alkyl protons is due to the restricted rotation around the carbamate C-N bond and a rough approxi- mation of the barrier to rotation about this bond has been determined. NMR parameters, rate constants and an approximate value of the barrier to rotation in compounds (38), (39), (44) and (4%) are given in Table 20. The rate constant, kc’ at the coalescence temperature, T0’ was determined from the peak S€paration, 00, at slow exchange using equation kc = 211/2//2 for coupled system593. Hdv//2 or kC = “[602 + 6J An approximate rotational free energy barrier was obtained from the Eyring equation: A01 = 2.3RT[10.3 — log(kc/Tc)]. The values of the rotational barriers lie in a fairly narrow range from 15.75 to 16.15 kcal/mol and appear to be inde- pendent of the nature of the alkyl group. 9“ has determined rotational barriers about the Holloway carbamate C-N bond in a series of N,N~dialky1dithiocarbamate esters. Activation energies of 10 to 12 kcal/mol suggested that an appreciable amount of C-N double bond character was present. Hollaway was able to correlate the C—N double bond character with the "thioureide" band between 1489 and 1498 cm”1 in the infrared region. The "thioureide" band which 110 Table 20. NMR Parameters, Kinetic and Infrared Data for Fe(CSHuSCSNR2)2 and Fe(C5H5)(C5HMSCSNR2) where R = Me, Et. k 10+ 00 j: Tc (kcal/ 1R1 (H2) (8 ) (K) mol) (cm ) Fe(CSHuSCSNMe2)2 16.48 36.61 314 16.15 1480 Fe(CSHuSCSNEt2)2 43.95 104.24 319 15.75 1480 Fe(C5H5)(C5HuSCSNMeZ) 17.70 39.32 312 16.00 1475 Fe(C5H5)(CSHuSCSNEt2) 31.74 79.40 320 15.98 1480 has been assigned to the partial double character in the carbon-nitrogen bond was observed at 1480 cm"1 in the di— a1ky1dithiocarbamateferrocene derivatives as shown in Table 20. The variable temperature 1 H NMR spectra for the isopropyl derivatives, (44) and (43), are considerably more complex and are shown in Figures 18 and 19 respectively. The NMR spectra for the mono and bisdiisopropyldithiocarbamate— ferrocene derivatives are very similar which suggests that in the disubstituted species the dithiocarbamate groups are on opposite sides and act independently of each other. Various groups have studied methyl N,N-diiSOprOpyl- dithiocarbamate, MeSCSNiPr295-98, and have concluded that 111 .NA mhmflzmomzmmovmm mo mhpomam mzz m H onsumhodsop mfinmfihm> .mH opzwfim 112 0mm. o .Amsmazmom:mmoVAmmmovom so shooodm mzz m H mLSpmthEwp oaomass> .m: ohsmam 113 hindered rotation occurs around the carbamate C-N bond and the iSOprOpyl-nitrogen bonds. Similar processes seem to occur in the ferrocene derivatives. Figures 18 and 19 in- dicate that at high temperature the 1 H NMR spectrum con- sists of a septet for the methine protons, a doublet for the isopropyl methyl groups and the signals associated with the ring protons. This spectrum is consistent with rapid rota- tion around the carbamate C-N and isopropyl-nitrogen bonds. As the temperature is lowered the signals broaden and at the slow exchange limit there are four methine septets and three methyl doublets clearly visible. The low temperature spectra can be interpreted in terms of a mixture of the two conformers A and B which exist in different relative populations (see Figure 20). Conformer A is the preferred conformer as in conformer B there is significant interaction between the isopropyl methyl groups and the bulky ferrocene. Integration of the methine septets suggest that conformer A and conformer B are present in a 2.1:1 ratio. Tentative assignments which are based on the magnetic anisotropy of the thiocarbonyl group are shown in Figure 21. The lowest field septet is assigned to HC — the proton adjacent to the thiocarbonyl. The next lowest field septet is attributed to H8 which is adjacent to the C-SFc bond. The high field septets are assigned to Hd and Hb. Assign— ment of the isopropyl methyl resonances are extremely 114 H: sued "8"” FCS\ /c‘m ch\ /’c\Hd C—N —— / / s’ MeEC’H" s’ >6?“ c ‘ "0 Me, "c " .6. .3. Figure 20. Conformers A and B of Fe(C5H5)(C5HuSCSNiPr2). H. ....J'L 6 5 4 3 2 1 Figure 21. Slow exchange 1H NMR Spectrum of Fe(C5H5)(C5Hu— SCSNiPr’2 ) . 115 tentative especially since only three of the four methyl doublets are clearly visible. Spin decoupling experiments are necessary to provide more unambiguous assignments. Close examination of the isopropyl methyl region suggests that a third conformer is present in low concentration as has been 95,97. observed in the MeSCSNiPr2 complex A resolution en— hanced spectrumcfi‘the isopropyl methyl region is shown in Figure 22 and this clearly shows the presence of the third conformer. The dynamic NMR studies suggest that in the methyl and ethyldithiocarbamateferrocene derivatives carbamate C—N bond rotation is observed whereas in the isopropyl analogs 96 iSOprOpyl-nitrogen bond rotation predominates. Sandstr8m has performed molecular mechanics calculations on MeSCSNiPr2 and has determined that the iSOprOpyl groups rotate non- synchronously. 1 Figure 22. Resolution Enhanced H NMR Spectrum of Fe(C5H5)- (CSHuSCSNiPr2) at Slow Exchange. 116 6. Metal Complexes of Fe(CSMASCSNEt2)2 and Fe(CSMS): (055“ so SNEt 2) The dialkyldithiocarbamateferrocene derivatives, (38) to (43), are very similar to the ubiquitous dithiocarbamate ligands except that the ferrocene derivatives are neutral. The dithiocarbamate ligand forms many metal complexes with many interesting properties such as a broad range of oxida- 88. The similarity between the dithiocarbamate tion states ligands and the ferrocene derivatives led to the investiga— tion of the preparation of palladium complexes of the ethyl derivatives, (34) and (42). A benzene solution of bis(benzonitrile)palladiumdichlor- ide was slowly added to a benzene solution of either complex (39) or compound (42). A red brown precipitate that formed immediately was filtered, washed with benzene and dried. The palladium complex obtained from Fe(CSHuSCSNEt2)2, (39), was slightly soluble in methylene chloride but seemed to form a metallic film on the bottom of the flask on standing. Even though high dilution reaction conditions were employed there was probably an appreciable amount of polymeric product formed as the palladium could coordinate to two dithiocarbamate 117 moeities on different ferrocene units. The palladium complex derived from Fe(C5H5)(C5HuSCSNEt2) was soluble in methylene chloride and nitromethane. At- tempts to characterize the metal complex were not very suc- cessful. The l H NMR spectrum, measured in nitromethane-d3 solution, consisted of broad, featureless peaks which were not very informative. Attempts to obtain 13C NMR data failed as over a period of time a brown film coated the walls of the NMR tube. This was unfortunate as an upfield shift of the thiocarbonyl resonance would confirm that the palladium atom complexed to the thiocarbonyl sulfur99. Infrared is helpful in characterizing the palladium complex and three specific regions are particularly helpful. As noted previously, Fe(C5H5)(C5HuSCSNEt2) displays an absorption at 1480 cm.1 which has been described as the "thioureide" vibration and is attributed to the carbamate v(C-N) mode. This region of the infrared is shown in Figure 23 for the free ligand, (42) and the palladium complex. In the palladium compound the "thioureide" vibration is found at 1550 cm'l. This shift to higher frequencies can be correlated with an increase in the double bond character of the carbamate C-N bond (structure B in Scheme 11). s\\ / Et 5\ e/Et /C -N \ <—-> /C=N\ Fc S Et Fe 5 H A .3. Scheme 11 118 ”A: M l l l l l700 l600 I500 I400 l3 I7'oo Ieloo |500 I400 I300 WAVENWBER CM" WAVENUMBER cm" Figure 23. Infrared Spectra of (A) Fe(CSH5)(C5HuSCSNEt2) and (B) Pd complex; (C) Fe(CSHuSCSNEt2)2 and (D) Pd complex. 119 The second infrared region, between 950 and 1050 cm-1, is associated with the 0(CSS) vibrations and has been used to differentiate between monodentate and bidentate dithio- carbamate ligands. The compound, Fe(C5H5)(C5HuSCSNEt2), l and 980 cm-1 100 exhibits two bands at 1005 cm- in this region. On the basis of Bonati and Ugo's work 1 the strong band at 980 cm_ can be assigned to an uncomplexed C=S stretch. In the palladium complex the presence of only one band at 1000 cm’1 in this region suggests that palladium coordinates to the thiocarbonyl group. Caution must be observed as absorbtion from the ferrocene moeity is found in this region around 1000 cm'l. Many studies have supported the use of the bands in the 950 to 1050 cm-1 region to ascer- tain if the dithiocarbamate ligand is monodentate or bi- 101,102 dentate (See Figures 24 and 25 for corresponding in— frared spectra.) The third diagnostic region, between 300 and 400 cm-1, is associated with M-S vibrations. A broad absorption at 367 cm-1 could be assigned to the Pd-S stretch whereas the two strong absorptions at 323 and 300 cm-1 are probably associated with a Pd-Cl stretch. On the basis of the infrared analysis a tentative struc- ture for the palladium complex is shown below in Scheme 12. A crystallographic study is required before the coordination mode of Fe(C5H5)(C5HuSCSNEt2) is definitely known. 120 CI ,5 Et >de :0 =10: CI ‘-s\ Et Fc Scheme 12. 121 .ASOHCQV m :ooamAmcmzmom:zmovoa obs Ao>ocev mAmomzmom:mmovoa uo ocpooom ochmhocH .:m oasmam man on“ 00' 00. 00. 000. DON. 00! 0°... 00.. 088 80 N 88 0000 000 V W q u H q q d a J u d H d _ 122 .Aoncov maooa Ampmzmomzmmozmmmovom pcm Ao>onmv Amumzmomzmmoimmmovom mo 950on UmpmpmcH .mm opszm .50 3“ 00¢ 000 00. 000. 00! 00! GOO.' 00.. 0°00 0000 0000 000” 000' 1 . a a a 4 . A q a q . . J PART B TRIS(CYCLOPENTADIENYL)ZIRCONIUM DERIVATIVES I. INTRODUCTION Atwood recently reported the crystal structure of tetra- (cyclopentadienyl)hafnium103 and indicated that the struc- ture of Cpqu contrasted sharply with the zirconium analog, Cpqulou. The hafnium complex has two o-bonded and two n—bonded cyclopentadienyl rings whereas the zirconium species has one o and three n-bonded cyclopentadienyl rings. This difference in structure prompted the investigation of tris- (cyclopentadienyl)zirconium derivatives. Despite the interest 105 in bis(cy010pentadieny1)zirconium derivatives complexes of tris(cyclopentadienyl)zirconium are relatively unknown. Russian workers106 have briefly mentioned (CSHMCH3)3ZrCl and Samuel107 has reported that reaction of Cp2ZrCl2 with Nan gave rise to (Cp2ZrC1)20. This product was due to the reaction conditions employed and the extreme hydrolytic in— stability of the tris(cyclopentadienyl)zirconium species. Recently Russian workers108 reported a vibrational analysis of Cp3MH, where M = Zr, Hf, and concluded that the three cyclopentadienyl rings are identically n-bonded. The hydride, CpBZrH, was prepared by reduction of the tetra— substituted species, Cpqu, with lithium aluminum hydride. An interesting aspect of zirconium chemistry is its application to Ziegler—Natta catalysis. Kaminsky and Sinn 123 124 have develOped a new class of Ziegler-Natta catalysts that have a high level of catalytic activity and are extremely stable in solution and the solid statelog. The catalyst consists of a bis(cyclopentadienyl)zirconium derivative com- plexed with an aluminoxane. The aluminoxanes are cyclic oligomers that are obtained by the addition of water to a trialkylaluminum species. Sinn and Kaminsky also report that tris(cyclopentadienyl)zirconium compounds are active polymerization catalysts for ethylene. Kopf110 recently reported the crystal structure of the unusual tris(cyclo- pentadienyl)zirconium complex, Cp3ZrH°A1Et3, which was ob- tained as a side product from the polymerization studies. II. EXPERIMENTAL General Techniques All operations were performed under prepurified nitro- gen or argon by using standard Schlenck techniques. Puri- fied grade nitrogen and argon, obtained from Matheson, were further deoxygenated by passing them through columns of activated BASF catalyst R 3-11 and Aquasorb (Mallinckrodt). Reagent grade solvents were used. Benzene, toluene and tetrahydrofuran were distilled from sodium/benzophenone under nitrogen. Hexane and n-pentane were refluxed over calcium hydride and were freshly distilled prior to use. Diethyl ether was distilled from lithium aluminum hydride. Electron spin resonance spectra were obtained by use of a Varian E-4 spectrometer. A Hanovia medium pressure 450 W mercury lamp with a quartz well was used as a UV light source for bench-top reactions. The glassware used for photolysis reactions was Pyrex, so the UV wavelength range was greater than 300 nm. Tris(cyclopentadienyl)zirconiumchloride (441 Sodium cyclopentadienide (3.3 g, 38 mmol) was rapidly added to a solution of zirconocenedichloride (10 g, 34 mmol) 125 126 in 300 mL THF. The solution was refluxed for 3 h and then stirred for an additional hour. The reaction may be moni- tored by 1H NMR. (Caution must be employed as 0p32r01 disproportionates to Cp2ZrCl2 and Cpqu if refluxing is prolonged.) The yellow solution was filtered and the fil- trate was reduced in volume to yield 40% Cp3ZrCl which was collected, dried and stored in a dry box. The product may be sublimed at 9090/10’l mm. 1H NMR: (00013), 06.0 (s); (C6D6), 5.67 (5); Mass spec: (rel intensity), 320 (3, M+), + 285 (2, M - 01), 255 (100, M+ _ (C5H5)), 190 (20, M+ - 2(C5H5)), 66 (13). Attempted Preparation of Tris(cyclopentadienyl)zirconium- butyl (44) Butyllithium (0.22 mL, 0.3 mmol) was slowly added via syringe to a solution of tris(cyclopentadienyl)zirconium- chloride (100 mg, 0.3 mmol) in 50 mL of THF which had been cooled to -78°C. The yellow solution was stirred at -78°C for 4 h and then was gradually allowed to reach room tem- perature. Upon reaching room temperature the solution turned black. The solvent was removed in vacuo leaving a purple black residue. 1H NMR: (C6D6), 06.0 (broad peak), 5.77 (s), 5.3 (s), 1.0 (broad peak). 127 Tris(cyclopentadienyl)zirconiumbutyl (491 Butyllithium (0.22 mL, 0.35 mmol) was slowly added via syringe to a solution of tris(cyclopentadienyl)zirconium- chloride (100 mg, 0.3 mmol) in 100 mL of toluene which had been cooled to -78°C. After being stirred for 4 h the yellow solution was allowed to reach room temperature. The solvent was removed in vacuo and the residue was ex- tracted with hexane. The hexane solution was reduced in volume and cooled to -30°C. The yellow crystals were fil- 1H NMR: (C708), 05.35 tered to give 65% yield of (49). (s, 0p), 1.6 (m), 1.2 (m), 1.0 (m); 130 NMR: (0708), 0110.6 (d, Cp), 39.7 (t), 33.47 (t), 30.67 (t), 1M.M (0); Mass spec: (rel intensity), 285 (100, Cp3Zr), 220 (18, Cp2Zr), 66 (21). Tris(cyclopentadienyl)zirconiummethyl (5Q) Methyllithium (0.28 mL, 0.34 mmol) was added slowly via syringe to a solution of tris(cycIOpentadienyl)zirconium- chloride (100 mg, 0.31 mmol) in 50 mL of diethyl ether which was cooled to -78°C. After being stirred for 3 h the solu— tion was allowed to reach room temperature. The solvent was removed in vacuo and the yellow residue was extracted with pentane. The pentane solution was concentrated and cooled to -30°C. The yellow solid was collected and stored in the 1 dry box, 60% yield of (C5H5)3ZrCH H NMR: (C6D6), 65.34 3. (s), 0.5 (s). 128 Photolysis of Tris(cyclopentadienyl)zirconiummethyl under CO Atmosphere An NMR tube was filled with a toluene-d8 solution of tris(cyclopentadienyl)zirconiummethyl in the dry box. The NMR tube was removed from the dry box and was evacuated and filled with C0 three times on a vacuum line. This solu- tion was photolyzed for 6 h in a water bath which was main- 1H NMR tained at 25°C and the reaction was monitored by spectroscopy. There was no change in the chemical shift of the methyl signal after irradiating for 6 h. III. RESULTS AND DISCUSSION 1. Qp3ZrCl Tris(cyclopentadienyl)zirconiumchloride is produced by reaction of bis(cyclopentadienyl)zirconiumdichloride with sodium cyclopentadienidei11refluxing tetrahydrofuran. An interesting feature of this preparation is that upon prolonged heating (more than three hours) the Cp3ZrCl species appears to disproportionate to Cp2ZrCl2 and Cpqu as shown in Scheme 13. CpZZrCl2 + Nan 7——’ [Cp3ZrCl] Cpqu + Cp2ZrCl2 Scheme 13 This disproportionation reaction suggests that the tris— (cyclopentadienyl)zirconiumchloride complex has limited thermal stability. Tris(cyc10pentadienyl)zirconiumchloride is extremely moisture sensitive and hydrolyzes rapidly to (u-oxo)bis- (chlorozirconocene), (CpZZrCl)20. The u-oxo dimer is readily characterized by the Zr-O-Zr stretch located at 129 130 750 to 780 cm'1 in the infrared. Table 21 shows the solvent dependence of the chemical shifts of the cyclopentadienylzirconium complexes. The chemical shift of the cyclopentadienyl ring protons moves upfield from Cp2ZrCl2 to (Cp2ZrCl)20 to Cp3ZrCl to Cp3ZrR where R = alkyl. This trend is consistent with the subse— quent removal of a deshielding chlorine atom. Table 21. 1H NMR Data for Cyclopentadienylzirconium Com- plexes. CDCl3 C6D6 C7D8 Cp2ZrCl2 6.5 6.02 (Cp2ZrCl)20 6.3 5.74 6.00 Cp3ZrCl 6.0 5.67 Cp3ZrCH2CH2CH2CH3 I:6;m 1.20m l.00m CpBZrCH3 6:26 The 1 H NMR spectrum of Cp3ZrCl consists of a singlet at 6.0 ppm in 00013 which suggests that the rings exhibit rapid fluxional behavior that is too fast to be detected on the NMR time scale. This is not surprising in view of 1 the singlet which is observed at -150°C in the H NMR 131 for Cpqulll. The tris(cyclopentadienyl)lanthanide and actinide deriva- tives have three n-bonded cyclopentadienyl rings. If Cp3ZrCl has three n-bonded cyclopentadienyl rings it is formally a twenty electron system. Hoffmann has examined the molecular orbital scheme for the hypothetical Cp3Ti+ fragment and concluded from symmetry constraints that the three cyclo- pentadienyl rings donate sixteen electrons to the metal, leaving one orbital on the metal fragment empty and avail- able for a sigma bondllz. In lanthanide and actinide chemistry the presence of f orbitals invalidate the usual electron counting rules. 2. §p3ZrR (R = nBu, Me) The compound Cp3ZrCl appears to have remarkable struc— tural similarity to Cp3UCl and Cp3ThCl. The alkyls of these actinide complexes exhibit surprising thermal stab- ility which has been attributed113 to the crowded coordina- tion sphere blocking the beta-elimination decomposition pathway. In the light of these results preparation of the analogous zirconium complexes was investigated. The alkyl complexes Cp3ZrR, where R = nBu, Me, were isolated in good yield from reaction of Cp3ZrCl and the appropriate alkyllithium reagent at low temperature. toluene Cp ZrCl + RLi 1;: CpBZrR 3 -78°c I32 Addition of small amounts of methanol to quench any un- reacted alkyllithium led to decomposition of the zirconium alkyl. The tris(cyclopentadienyl)zirconiumalkyl complexes are extremely sensitive to protic reagents such as water and methanol. When THF was used as the reaction solvent the solution turned black upon reaching room temperature. The 1H NMR spectrum (C6D6) of the black purple solid contained very broad peaks around 6 and 1 ppm which suggests the presence of a paramagnetic species. The ESR spectrum, which is shown in Figure 26, is complex and suggests the presence of two species. A possible explanation is that THF co- ordinates to the metal to give the species, Cp3Zr'THF. When a noncoordinating solvent such as toluene is used no color change is observed and the zirconium alkyl is iso- lated in good yield. There is precedence for a trivalent complex of this kind in actinide chemistry. Kanellakopulosllu has pre- pared an analogous thorium complex by reduction of Cp3ThCl with sodium napthalide in the presence of tetrahydrafuran. C10H8 Cp3ThCl + Na -;E;.-Cp3Th-THF Alternatively the black purple paramagnetic zirconium species could be an anionic zirconium III complex, Cp3ZrR-. 133 352I IOG g=L9884 Figure 26. ESR Spectrum Obtained from Reaction of Cp3ZrCl with BuLi in THF Solution. 134 115 reduced a zirconocene dialkyl complex with Lappert sodium dihydronaphthylide and obtained an ESR active species. Initially he postulated that the ESR active species was [Zr0p2(CHZCMe3)2]- but later on the basis of cyclic voltammetry data he concluded that the anionic dialkyl was unstable and the d1 complex was [ZGC(CH2CMe3)2]. The zirconium dialkyls undergo irreversible one-electron reductions at a platinum electrode at scan rates up to 1.0 v s'l. Recently Fischer116 reported that exposure of Cp3UR, where R = Me, iPr, iBu and nBu, to excess alkyllithium in THF yields saturated hydrocarbons, RH, and a solvated IIIR]. Both Fischer and Lappert species [Li(thf)X][Cp3U found that THF was necessary for formation of these an- ionic alkyl species. Stable transition metal alkyls are generally limited to complexes where the alkyl lacks a beta-hydrogen due to the facile beta elimination decomposition pathwayll7. With few exceptions, complexes of early transition metal alkyls that contain beta hydrogens are accessible only at low temperatures. For example, Cp2TiR2, where R = nBu or nPrllB, is unstable above -40°C. The complex Cp3Zr(nBu) is one of the few examples of a thermally stable early transition metal alkyl containing beta hydrogen atoms. This stability is due to the presence of three n-bonded cyclopentadienyl rings which sterically congest the 135 coordination sphere around zirconium and thus block the beta elimination decomposition pathway. The 130 NMR spectrum of Cp3Zr(nBu) was examined at low temperature in the hope of freezing out the fluxional cyclopentadienyl rings. A sharp singlet was observed for the cyc10pentadieny1 rings at -60°C. In Cp3U(iPr), Marksllg observed restricted rotation of the cyc10pentadienyl rings with the slow exchange limit at -109°C. In the analogous zirconium complex, Cp3Zr(iPr), restricted rotation of the cyclopentadienyl rings may be evident. 3. Reaction of Qp3ZrMe with C0 Many groups120 have observed a facile reaction of zirconium alkyls with carbon monoxide where 00 inserts into the metal alkyl bond to give an acyl species. Re- action of Cp3ZrMe with CO was examined at ambient tempera- ture and under photolytic conditions. The complex, Cp3ZrMe, in a toluene-d8 solution was sealed in an NMR tube under an atmosphere of CO and the 1H NMR spectrum NMR tube was protected from light. The was monitored periodically but no change was observed in the Spectrum after a few days. If an acyl species had formed the chemical shift of the methyl group would have moved downfield. Reaction of Cp3ZrMe with CO, however, may only be photo-induced as the sterically crowded coordination sphere 136 of zirconium would prevent prior coordination of CO. Photolysis could induce an excited cyclopentadienyl ring (n3 type) as Marks has prOposed for the photo-induced beta elimination of Cp3Th alkylsll3. An NMR tube containing a toluene-d8 solution of CpBZrMe under an atmosphere of CO was photolyzed for at least six hours. No change was however observed in the 1H NMR spectrum. Even though forcing conditions (high CO pressure) have not been employed, reaction of CpBZrMe with CO does not seem feasible. These results are consistent with the crowded coordination sphere of zirconium which would pre- vent prior coordination of the CO ligand. 4. Conclusions An interesting aspect of Cp3ZrR chemistry could be their thermal and photolytic decomposition. The steric- ally crowded coordination of Cp3ZrR could prevent beta elimination as a thermal decomposition pathway and the alkyl group would rather abstract hydrogen from a cyclo- pentadienyl ring to give a binuclear species, shown in Scheme 14. Beta elimination, however, may be photo-induced 3 where a cyclopentadienyl ring may be photoactivated to n coordination. Marks121 has obtained similar results in the Cp3ThR system where beta elimination is thermally in- accessible but rather photo-induced. 137 Cp Cp Cpaer —A., \Zr/Q Zr / CD//' ‘E