University of Southern California Dissertations and Theses

A.The anions of C(60) and its pyrrolidine derivatives B.The search for hydridocobaloximes

(USC Thesis Other)

A.The anions of C(60) and its pyrrolidine derivatives B.The search for hydridocobaloximes

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly fiom the original o r copy submitted. Thus, some thesis and dissertation copies are in typewriter free, iiW ule others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard m argins, and improper alignment can adversety afifrct reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, b%inning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photognq>hed in one exposure and is included in reduced form at the back o f the book. Photographs included in the original manuscript have been reproduced xerographicalfy in this copy. Ifigher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Ifowell Information Company 300 North Zed> Road, Ann Aibor MI 48106-1346 USA 313/761-4700 800/521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A. THE ANIONS OF C eo AND ITS PYRROLIDINE DERIVATIVES B. THE SEARCH FOR HYDRIDOCOBALOXIMES by Yongping Sun A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment o f the Requirements for the Degree DOCTOR OF PHILOSOPHY (Chemistry) August 1997 Copyright 1997 Yongping Sun Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMJ Number: 9816073 UMI Microform 9816073 Copyright 1998, by UMI Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeh Road Ann Arbor, MI 48103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 90007 This dissertatioTir written by Y ongping Sun under the direction of h..^..... Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY Dean of Graduate Studies Date DISSERTATION COMMITTEE itrperson cnm Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. To the memory of my father u Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgments First o f ail, I would like to thank my advisor. Professor Christopher Reed, for his encouragement, guidance and support. Ifis creativity, ambition and enthusiasm have been a good example for me to emulate as I pursue my career in chemistry. I am greatly indebted to Professor Gordon Nfiskelly, for his advice, encouragement and his belief in me, especially when I first came here. Ifis help made the difiScult time much easier. My special thanks also go to Professor Peter Boyd, the “third professor” fi-om New Zealand. From the enlightening discussion with him, in person or via e-mail, I have learned a lot about EPR and NMR spectroscopy. I feel fortunate that I have met Bob Bolskar when I switched to the project of fiiUerides. He introduced me to the world of EPR and offered me plentiful help in handling the intricate spectrometer in addition to many other things. PCs careful proof reading for this dissertation is also gratefully acknowledged. I appreciate Dr. Zuowei Xie for his fiuitful discussion and insightful suggestion. From him, I have learned “tricks” of crystal growing and synthetic chemistry. I am grateful to Professor Robert Bau for teaching and helping me to solve the crystal structures. The efforts o f Dr. Tatiana Drovetskaya for solving those difScult fuUerene-related crystal structures, which are included or not included here, are gratefully acknowledged. ut Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I would like to thank Timothy La for his continuous help and beneficial discussion. I have had a great pleasure to work with and learn fi*om Raj Mathur, Dr. Dan Evans, Mary Drabnis, Dr. Than Fackler and Dr. Anna Larsen. I truly appreciate their help and fiiendship. I would like to thank Dr. Dan Evans, for proof-reading the cobaloximes part o f this dissertation and for stimulating discussion. Many people have provided invaluable help to me and I am appreciative to all of them: Roy Lu, for collecting the X-ray data; Dr. David Joilie, for always being there to help fix the problems o f the EPR spectrometer, Jeff Clites, for obtaining the SQUID data on the extremely air-sensitive Cso^ samples; Antao Chen, for help with NIR spectra on the quickly degrading C eo*' solutions; Dr. Allen Benesi, for collecting the solid state NMR data; Jim Merritt, for preparing excellent glassware; Allan Kershaw, for assistance with NMR experiments. I will certainly cherish the memorable experience working days and nights here in OCW and Stabler Hall while struggling with C e o * ' and the “hydride”. I taught general chemistry lab here at USC for three and a half years and I would like to thank all of my students, for the joy and confidence they brought to me. I would like to thank all the teachers and fellows I had met at Peking University, my alma mater “Beida”, for introducing me to my professional career I would like to express my gratitude to my teacher Ms Yaoqing Zheng, for her care, encouragement and for the inspiring conversation about life and career via long distance telephone. IV Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I am greatly indebted to my mother, sister and brother. Without their support and help, I would not be able to have made through the difBcult years in college and graduate school. I deeply appreciate their continuous understanding and encouragement for my efforts. I am most grateful to my father, for teaching me to be an honest, responsible, generous, solid and diligent person; for bringing me the courage to face the challenge and the endurance to survive under the difficult situations. K s love and belief in me are the first driving force for me to succeed. I hereby dedicate this dissertation wholeheartedly to the memory of him. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents Page Dedication ii Acknowledgments iii List of Abbreviations xiii List o f Figures xv List of Tables xxi List of Schemes xxii Abstract xxiii Part A. The Anions of C eo and Its Pyrrolidine Derivatives Chapter I. Introduction to the Fullerene and Fullerides I I 1. Overview of the Fullerene I 1.1.1. Discovery and Production 1 1.1.2. Geometry and Structure 2 1.1.3. General Chemical Properties 5 1.2. Fullerides 6 1.2.1. The Electronegative Characteristics o f C o o 6 1.2.2. Fullerides in Condensed Media 8 1.2.3. Potential Applications o f the Fullerides 10 1.2.4. Isolation and Studies o f Discrete Fullerides 1 1 References 12 VI Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter H. n. 1. NMR Characterization o f C e o Fullerides 19 n .l.I. Introduction 19 H I .2. Results and Discussion 21 n. 1.2.1. ‘ ^CNMRof[Na(crown)(THF)2]„C 6o(n= 1, 2, 3)at 21 Ambient Temperature n . 1.2.1. *^C NMR o f [Na(crown)(THF)2]nC 6o (n = 1, 2, 3) in 26 Solution at Variable Temperatures n 1.2.3. Supplement: the Related Background Knowledge 30 of NMR of Paramagnetic Species n.1.3. Conclusion 32 n 2. Reativities o f Ceo^' with Alkyl Halides 33 n.2.1. Introduction 33 n.2.2. Results and Discussion 36 n.2.3. Conclusion 39 n.3. Experimental 39 References and Notes 41 Chapter HI Synthesis and Isolation o f Discrete Ceo^ Salts 46 H I.l. Introduction 46 HI.1.1 Ceo^ and Its Electronic Structure 46 HI. 1.2. Attempts of Preparing C f i o * * * 47 HI. 1.3. Cryptand and Crown Ether 49 HI 2 Results and Discussion 51 vu Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. m .2.1. Synthesis and Isolation 51 in.2.2. Instability and Reactivity 53 in.2.3. Characterization 57 NIR 57 NMR 60 EPR 63 SQUID 70 m .3 Conclusion 70 in.4. Experimental 71 in.4.1. Physical Measurements 71 in.4.2. Solvent Purification 72 in.4.3. Synthesis and Isolation 73 n i.4 .4 . Sample Sealing 74 References 74 Chapter IV. Fullerides of Pyrrolidine-Fimctionalized C eo 77 IV 1 Introduction 77 IV.2. Results and Discussion 81 IV.2.1. Synthesis of Neutral Starting Materials 81 rv.2.1.1. Compound 1 and Ceo(CH2)2NCPh3 and Applications 81 rV.2.1.2. 1,3-Dimer 4 and 1,4-Dimer 5 82 IV .2.1.3. Structural Fluxionality o f the Fullerene Dimers 87 VUl Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV.2.1.4. Other Prato-Type C « > Derivatives 90 IV.2 2. Monoanions ofFulleropyrrolidines, l and 2' 92 Synthesis 92 Vis-NIR 93 NMR 95 EPR 99 IV.2.3. Dianion o f V-Methylfulleropyrrolidine, I^' 104 Synthesis and NMR 104 Vis-NIR 105 EPR 105 X-ray Structure 107 rv.2.4. Anions of 1,3-Dimer and 1,4-Dimer 111 Sample Preparation 111 NIR 113 EPR o f Monoanions 117 EPR o f Dianions 121 rv . Conclusion 126 IV. Experimental 127 1 128 2, CgoCCHzhNCPhs 128 [Cp2Co1[lT 129 IX Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [Na(crown)(TiT^2l[l-] 130 [Cp2Co1[2-] 130 [Na(crownXTHF)2l2[l^T 131 X-ray Structure o f I^ * 131 1,3-Dimer, 4 132 1,4-Dimer, 5 133 6 133 7 134 EPR Sample of 4 135 EPR Sample of 4 ^ * 135 erences and Notes 135 Part B. Chapter V. The Search for Hydridocobaloximes 141 V.l. Introduction 141 V2. Results and Discussion 144 V.2.1. Starting Materials: Synthesis and Reactivities 144 V.2.2. Reinvestigation of “Hydridocobaloximes”: Synthesis and 147 Characterization o f Co(U) V .2.2.1. Reduction of ClCo(DH)2PBu3 with NaBH4 147 V.2.2.2. Characterization o f “Hydridocobaloxime” 148 Supplement: The Evans Method for Determining Magnetic 157 Moment o f Paramagnetic Species in Solution Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V.2.2.3. Is the Reported “Hydride” a Co(II) Species? 159 V.2.2.4. Synthesis o f Co(DH)2PBu3 ' . the “Hydride” is Co(II)! 162 V.2.2.5. Comments on Schrauzer’s and Espenson's Works 166 V.2.2.6. Synthesis and Characterization of 169 Co(DHXDBF2> 2PBu3 and Co(DBp2 )2PBu3 V.2.3. Attempts of the Synthesis o f Hydrides by Protonating 174 Co(I) Species V.2.3.1. Synthesis and Isolation of the Co(J) Species 175 V.2.3.2. The NMR of BF2-Bridged Cobaloximes; the 180 Unusual B-F Coupling V.2.3.3. Reactions of Co(I) with Acids 183 V.2.3.4. Are Hydridocobaloximes Accessible? 184 V.3. Conclusion 189 V.4. Experimental 190 Co(DH)(DH2)Cl2 191 C1Co(DH )2PBu3 191 CICo(DBF2)2PBu3 192 C1Co(DH)(DBF2)PBu3 193 ClCo(DH)2Py 193 ClCo(DBF2> 2Py 194 [Co(DH)2(PBu3)2][Co(DH)2Cl2] 194 “HC o(DH)2PBu3 ” , Co(DH)2PBu3 by NaBH* 195 Magnetic Moment Measurement by the Evans Mehtod 195 XI Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Molecular Weight Determinatioa by Solution Vapor Pressure 196 Equilibrium Method Spectroscopic Titration of ClCo(DH)2PBu3 with CpzCo 197 C o(D ï^PB u 3 : Generated by CpzCo 197 Co(DH)(DBF2>2PBu3 198 Co(DBF2)2PBu3 198 Co(DH)2PBu3 * 199 Co(DH)(DBF2)PBu3 199 References and Notes 200 XU Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Abbreviations BDE bond dissociation energy C222 cryptand[2.2.2] Cp cyciopentadienide CsHs* Cp* pentamethyicyclopentadienide CaCCHs):' crown dibenzo-18-crown-6 crypt cryptand[2.2.2] CV cyclic voitammetry DMF A^-dimethylformaniide HCON(CH3 ) 2 DMSO dimethyisulfoxide (CH3)2 SO DPPH a, o ' -diphenyl->3-picrylhydrazyl EPR electron paramagnetic resonance eq. equivalent, equivalents Fc ferrocene CpzFe HMPA hexamethylphosphoramide [(CH3> 2N]3P(0 ) HOMO the highest occupied molecular orbital Im imidazole IR infrared LB line broadening U q. liquid LUMO the lowest unoccupied molecular orbital X lll Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NIR NMR NOE MW ODCB For PPN* Py SCE SQUID TDAE THF TMS TPP TpTP UV vis vs. near infrared nuclear magnetic resonance nuclear overhauser effect molecular weight odichlorobenzene l.Z-C^H^Clz porphyrinate bis(triphenylphosphine)iminium [(C6Hs)3P=]2N* pyridine saturated calomel electrode superconducting quantum interference device tetrakis(dimethylamino)ethylene tetrahydrofuran tetramethylsilane (CH3)<Si tetraphenylporphyrinate tetra-p-tolylporphyrinate ultraviolet visible versus XIV Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Figures Figure Page 1 .1 The fullerene Cso- 1 1.2 Ceo structure. 3 1.3 The frontier Hückel molecular orbitals o f Ceo. 7 1.4 Reduction o f Ceo in CHaCN/toIuene at - 1 0 "C using (a) cyclic 7 voitammetry and (b) differential pulse voitammetry. H. 1 ‘ ^C NMR spectrum o f [Na(crown)(THF)2^ 3 [C eo^‘] in DMSO-de (the 23 peak at 39.5 ppm is from the solvent, and the rest are from crown ether and TïDO- n.2 The temperature dependence of the "C isotropic shift of 28 [Na(crown)(THF)2i[CeoT in T H F ^ n.3 The temperature dependence of the C isotropic shift of 28 [Na(crown)(THF)2%C 6o * ] in propionitrile. n.4 The temperature dependence of the C isotropic shift of 29 [Na(crown)(THF)2 " ^ 3[C6o^l in propionitrile. n.S The temperature dependence of the "C isotropic shift of 29 [Na(crown)(THF)2^ 2[C6o^l in propionitrile. n.6 H NMR spectrum o f the product o f the reaction o f Ceo^' with 37 CH2I2 (in toluene-i/g, the peak at 2.09 ppm is from the solvent). m i A mixture o f Ceo^' and Cgo^ in DMSO. 48 m .2 Dibenzo-18-crown-6 and cryptand[2.2.2]. 50 m .3 Vis-NIR spectrum of C«>^ in DMSO. 58 m .4 Vis-NIR spectrum o f [Na(crypt)^4[C6o ^ in DMSO/Na: (a) before 59 subtraction; (b) the absorbance of DMSO/Na; (c) after subtraction o f the absorbance o f DMSO/Na. XV Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ni-5 " c NMR spectrum o f [Na(crypt)']4[C6o ^ in DMSO-64 (the peaks 61 at 67.9, 66.9 and 52.1 ppm are from cryptand). in .6 Confirmation o f the formula of [Na(crypt)^4 [C6 o^ by "C NMR 61 integration (in DMSO-d^, NOE-suppressed, pulse delay 15 s, LB = 1.5). in. 7 Microwave power dependence of the EPR spectra of 65 [Na(crypt)"^4[C6o‘ *l in DMSO (at 4 K, modulation 0.25 G). in. 8 Nficrowave power dependence of the EPR spectra of 66 [Na(crypt)")3 [C6o ^ * ] in DMSO (at 4 K, modulation 0.25 G). m 9 Temperature dependence o f the EPR spectra o f [Na(crypt)']4[C6o^ 67 in DMSO (modulation 0.25 G, microwave power 200 jiW). m 10 Temperature dependence o f the EPR spectra o f [Na(crypt)"^3[C6o^'] 68 in DMSO (modulation 0.25 G, microwave power 200 pW). m i l A typical EPR spectrum o f powder [Na(crypt)^4 [Cgo^ salt (4 K, 69 modulation 0.5 G, microwave power 0.2 mW, gain 8 x 10\ frequency 9.4388 GHz). IV. 1 JV-methylfulleropyrrolidine 1. 78 rV.2 H NMR spectrum o f 4 in CS2/CDCI3 (the peak at 7.24 ppm is from 85 the solvent, and that at 7.29 ppm is from impurity). rV.3 NMR spectrum o f 4 in CS2 (acetone-d^ as the insert, not shown 85 is the C =0 peak at 168.3 ppm). IV.4 ‘H NMR spectrum o f 5 in CS2 (acetone-d^ as the insert). 86 rv.5 NMR spectrum o f 5 in CS2 (acetone<4 as the insert). 86 rv 6 The resonance structures o f amides. 88 rV.7 Vis-NIR spectra of [Cp2C o^ [1 * ] and [Cp2Co"l [C eo"] in PhCN. 94 IV 8 ‘h NMR spectrum o f [Cp2Co^ [1 * ] in DMSO-<4 (the peak at 2.49 97 ppm is from the solvent). XVI Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV.9 " c NMR spectrum o f [CpzCo^ [1 * ] in DMSO-<4 (the peak at 85.1 97 ppm is from CpzCo , and those at 110-133 ppm are from a trace of PhCN). IV. 10 ‘H NMR spectrum o f [CpiC o^P” ] in PhNOa-^s (those peaks at 98 7.50, 7.67 and 8.11 ppm are from the solvent). IV. 11 EPR spectra o f [Cp2 C o ^ [l’ ] in pyridine under different microwave 100 powers (4 K, modulation 0.25 G). IV. 12 EPR spectra of [CpaCo^[l'] in PhCN under different microwave 100 powers (157 K, modulation 0.25 G). T V . 13 EPR spectra of [CpaCo^[l'] in PhCN at different temperatures 102 (modulation 0.25 G, microwave power 20 pW). IV. 14 Schematic representation o f the electronic states o f C^o' and 1 101 IV. 15 EPR spectra of [CpaCo^fl*] in three different solvents (4 K, 102 modulation 0.25 G, microwave power 20 pW). rv. 16 EPR spectra o f 1 * in DMSO (4 K; microwave power 20 pW; 103 modulation for Na(crown)^ 0.0125 G, for CpaCo" 0.25 G). IV 17 Vis-NIR spectrum o f [Na(crown)(THF)a^2[1^3 in DMSO. 103 IV. 18 EPR spectra of [Na(crown)(THF)2^ 2[l^T in pyridine (4 K, 106 modulation 2 G). rv. 19 Dual disorder; the two orientations of in the single crystal of 109 [PPN^2 [1 ^'] toluene. IV.20 The unit cell o f [PPhTlzCl^l'toluene (for clarity all H atoms are 110 omitted). IV.21 Vis-NIR spectra o f neutral 4 and the progressive reduction of 4 with 114 cobaltocene in PhCN. IV.22 Vis-NIR spectra o f neutral 5 and the progressive reduction of 5 with 115 cobaltocene in PhCN. IV 23 EPR spectra o f [Cp2Co^[4"] in PhCN under different microwave 118 powers (4 K, modulation 0.05 G). XVU Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. rv.24 EPR spectra of [Cp2Co^[5*] in PhCN under different microwave 118 powers (4 K, modulation 0.025 G). IV 25 EPR spectra o f [Cp2Co1[4*] in PhCN at different temperatures 119 (modulation 0.05 G, microwave power 20 pW). rv.26 EPR spectra o f [Cp2Co"][5^ in PhCN at different temperatures 119 (modulation 0.025 G, microwave power 20 pW). rv.27 EPR spectra of [Cp2Co^[4*] in three different solvents (4 K; 120 modulation in PhCN 0.05 G, in pyridine and DMSO 0.25 G; microwave power 20 pW). IV.28 EPR spectra of [CpzCo'][S'] in three different solvents (4 K; 120 modulation in PhCN and pyridine 0.025 G, in DMSO 0.25 G; microwave power 20 pW). rv.29 EPR spectra o f [Cp2Co^2[4^T in PhCN under different microwave 122 powers (42 K, modulation 0.25 G). IV.30 EPR spectra of [Cp2Co^2[4^’ ] in PhCN at different temperatures 123 (modulation 0.25 G, microwave power 200 pW). IV 3 1 EPR spectra of a mixture of 4 ^ and 4 ^' prepared by 4 and 2.7 124 equivalents o f Cp*2Co in PhCN (modulation 0.25 G; microwave power 200 pW; temperature: top: 4 K, bottom: 106 K). IV. 3 2 EPR spectra o f [Cp2Co^2[5^T in PhCN at different temperatures 125 (modulation 0.025 G, microwave power 200 pW). rv.33 Vertically expanded EPR spectrum of [Cp2Co^2[5^T in PhCN (188 125 K, modulation 0.025 G, microwave power 2 mW). V .l The structures of cobaloximes. 142 V.2 The mono- and bis-Bp2 bridged cobaloximes. 142 V.3 NMR ofCo(DH)2PBu3 (generated by NaBH4> in MeOH-d!» (the 149 missing bridging H signal is presumably due to the deprotonation or the proton exchange with the solvent). XVIU Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V.4 UV-vis spectrum o f Co(DH)2PBu3* (generated by NaBH*) in 149 MeOH. V.5 UV-vis spectra of “HCo(DH)2PBu3” in MeOH at two different 1 51 concentrations. V.6 UV-vis spectra o f “HCo(DH)2PBu3” in MeOH at different 151 temperatures. V.7 UV-vis spectra of “HCo(DH)2PBu3” in hexane at different 152 temperatures. V.8 Variable temperature ‘ H NMR spectra of “HCo(DH)2PBu3” in 154 toluene-e/g (the peak at 2.09 ppm is from the solvent). V.9 Variable temperature ^ P NMR spectra o f “HCo(DH)2PBu3” in 155 toluene-f/g (from bottom to top: temperature, number of scans: RT, 53840; -45°C, 6008; -60°C, 8200; -65®C, 2056; -75°C, 800; -85°C, 600). V.IO Magnetic moments o f “HCo(DH)2PBu3” in acetone-(4 measured by 156 the Evans method at different temperatures. V. 11 Spectroscopic titration of ClCo(DH)2pBu3 with CP2C0 in methanol 164 (C1Co(DH)2PBu3 2.0 X 1 0 * ^ M, CP2C0 2.0 x 10'^ M). V 12 ‘H NMR spectrum of Co(DH)2PBu3 in toluene-^g at room 164 temperature (the peak at 2.09 ppm is from the solvent). V. 13 UV-vis spectrum o f a mixture o f Co(DH)2PBu3* and Co(DH)2PBu3 167 in MeOH. V. 14 UV-vis spectra o f Co(DH)(DBp2)2PBu3 in toluene at different 170 temperatures. V. 15 UV-vis spectra o f Co(DH)(DBF2)2PBu3 in Et20 at different 170 temperatures. V. 16 Variable temperature NMR spectra of Co(DH)(DBp2)2PBu3 in 172 toluene-(4 (from bottom to top: RT, -10°C, -20°C, -35°C, -43°C, -50®C, -90®C). XIX Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V. 17 H NMR spectrum o f the reduction product of ClCo(DH)2PBu3 by 178 1% Na(Hg) in CD3CN (not shown is the peak of the bridging H at 19.6 ppm). V. IS ‘ H NMR spectrum o f [(CH3> 4N l[Co(DH )2PBu3T in CD3CN (not 178 shown is the (CH3> 4N^ peak at 3.09 ppm). V. 19 NMR spectrum ofClCo(DBFz)2PBu3 in CDCI3. 182 V.20 NMR spectrum o f [Cp2Co"][Co(DBF2)2PBu3T in CD3CN. 182 XX Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables Table Page n. 1 Some typical "C chemical shifts of Ceo"' in solid state at room 20 temperature H 2 NMR results o f discrete fiiUerides in DMSO-de at room 23 temperature n.3 The solvent search for NMR experiments at variable temperatures 26 n.4 TLC results fiar the reaction mixture o f and PhzCCk (solvent: 38 toluene/hexane 1:4) n. 5 NMR chemical shifts o f [Na(crown)(THF)2lnC 6o " * at variable 41 temperatures m 1 Stability constant (log^) in MeOH at 25°C 50 m 2 The purification o f the solvents used or tested for C eo '* ' solubility 73 rv. 1 H chemical shifts of 1 and 2 and their monoanions 96 IV.2 Crystal data o f [PPN^2 [l^ l’ toluene 132 V. 1 Molecular weight of “HCo(DH)2PBu3” determined by the method 157 of vapor pressure equilibrium o f solutions (solvent: acetone) V.2 The NMR results of some BF2-bridged cobaloxime complexes 180 V.3 Magnetic moments of “HCo(DH)2PBu3” at different temperature 196 in acetone-<4 XXI Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Schemes Scheme Page m i The equilibrium between C e o ^ * and C e o '* * - 49 m .2 Two sub-equilibria in C e o * " synthesis. 49 IV. 1 The cleavage of the trityi group o f 2. 81 IV.2 The synthesis of 1,3-dimer 3 and 1,4-dimer 5. 84 r v 3 The synthesis of 1,3-dimer 6. 90 IV 4 The synthesis of Prato-type derivative 7. 91 V .l The equilibrium between monomeric and dimeric Co(DH)zPBu3 160 V.2 The reactivities of hydride and Co(II) with hydroxide. 166 V.3 Thermodynamic cycle of self decomposition of hydridocobaloximes. 185 V.4 The equilibrium for Co-R dissociation in alkylcobaloximes. 187 X X ll Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Abstract The fUUerides Ceo', C e o ^ * and Cao^' have been characterized by NMR spectroscopy in solution at room temperature as well as at variable temperatures. The Curie plots (S vs. 1/T) for the "C chemical shifts reveal that within the range of experimental temperatures, C e o * and Ceo^' obey Curie's law, although non-zero intercepts are observed. These results are consistent with isolated spin states in these systems. For Ceo^', the chemical shift exhibits an inverse temperature dependence, suggesting the singlet ground state with a low lying triplet exciting state. The reactivity o f C e o ^ * with alkyl cfrhalides has been studied. The competition between intra- and intermolecular nucleophilic reactions in the second step has been examined. A discrete salt of C e o * * * with Na(crypt)* as the countercation was synthesized and isolated in an analytically pure form for the first time. It is exceedingly sensitive to oxidation. It shows properties similar to those o f C e o ^ * : paramagnetism, low intensity of EPR signals, unusual temperature dependence o f its magnetic susceptibility. Like the C f io ^ * ion, these suggest a singlet ground state and a thermally accessible triplet state. The mono- and dianion of A'-methylpyrrolidine-Ceo have been investigated systematically. The electronic structure of the fullerene ball is essentially retained despite functionalization, but the LUMO energy level is slightly increased. In the monoanion, electron spin density is distributed over the buckyball with only slight xxiii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. delocalization onto the pyrrolidine ring. A pair of covalently-linked fUUeropyrrolidine dimers were synthesized. By studying these dimers at various stages o f reduction, it is shown that the intramolecular ball-to-ball interactions are weak. The existence o f the so-called hydridocobaloximes has been examined. The reported "HCo(DH)zPBu3" was found to be a mixture of Co(DH)2PBu3 and Co(DH)2PBu3 It is concluded that hydridocobaloximes are synthetically inaccessible. Co(DH)2PBu3 was shown to be essentially monomeric at room temperature, but dimerization occurs via the formation of a Co-Co bond as the temperature decreases. Systems in which one or both o f the bridging hydrogen atoms are replaced by B F % groups have been studied in a parallel manner. Unusual B-F NMR couplings were observed in diamagnetic BFz-bridged cobaloxime species. XMV Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter I. Introduction to the Fullerene and Fuilerides 1.1. Overview of the Fullerene L l.l. Discovery and Production C g o is a spherical hollow cluster o f 60 carbon atoms. Its structure had been conceived in imaginations in 1970 in an effort to find three-dimensional “super- aromatic” re-system.^ Also prior to its discovery, several molecular orbital calculations^'^ had been applied on this imaginary molecule. Figure LI The fullerene C eo. In 1984, large carbon clusters C m (n = 30-190) were observed upon the laser vaporization of graphite.^ It was noted that in the mass spectra, those of 60 and 70 carbons were always present among all the peaks. In 1985, Kroto, Smalley, Curl and their co-workers found that under specific clustering conditions, C eo and C70 exhibited unusually high intensities in mass spectra, and in some cases C eo signal could even Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dominate the whole spectra.® Based on these observations, Ceo was believed to be a species with unusual stability and its structure was proposed to be like a soccer ball. It was named Buckminsterfullerene after the architect Buckminster Fuller, who is renowned for his geodesic dome structures. In this way, Ceo was identified for the first time. Five years later in 1990, Kratschmer and Huffinan achieved the second breakthrough o f fullerene research by producing Ceo in macroscopic amounts.’ * It was generated through resistive heating of graphite rods under a helium atmosphere and its identity was primarily verified by the four distinct IR bands predicted by theoretical calculations.^ Kroto et al.‘® successfully isolated Ceo fi’ om soot extract via column chromatography. The development o f modified production," separation and purification methods" immediately fi^Uowed and Ceo soon became commercially available and its chemistry flourishes within a very short time period. In 1996, Kroto, Smalley, and Curl were awarded the Nobel prize for their contributions in fullerene identification. L1.2. Geometry and Structure Ceo possesses the highest symmetry (/« point group) a non-single-atom molecule can have. It has a truncated icosahedron geometry with 60 vertices. 12 out of the 32 6 ces are pentagonal and 20 are hexagonal. Each pentagon is surrounded by five Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hexagons, while each hexagon is surrounded, alternatively, by three pentagons and three hexagons (Figure 1.2). According to Euler's theorem,'”’ ’^ ^ polyhedral Q, should contain 12 pentagons and any number o f hexagons, m, such that m = (n-2 0 )/2 . For Ceo, u = 60, therefore it should have 20 hexagons (m = 20). In a polyhedral Cn, the existence o f adjacent pentagons will be energetically unfavorable due to the increased o-bonding ring strain or the formation of pentalene-type 8 7 c-electron system. In all the fullerene structures, the pentagons are isolated by hexagons. That is the so-called “isolated pentagon rule”. 1 3 '^ '* C fio is the smallest closed network that can have all pentagons non-adjacent. As a result, it is remarkably stable and dominates all other fullerenes in quantity during the clustering process. Figure L2 C eo stmcture. The structure o f Ceo has been investigated with several methods, including 3c NMR, neutron diffraction. X-ray powder or single crystal diffraction, etc. It was found that even in solid state the spherical balls rotate rapidly and isotropically at room Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. temperature/^ In solid state, the distance between two nearest neighbor buckybails (center-to-center) is determined to be 1 0 A * and within each ball, the atom-to-atom diameter is 7.1 A . In buckminsterfullerene, all 60 carbon atoms are equivalent since they all are at the comer o f two hexagons and one pentagon. Connecting the 60 atoms with identical environments, however, two types of C-C bonds exist: the 6 : 6 bonds, those at the fusion o f two six-membered rings and the 6:5 bonds, those at the juncture of a six- and a five-membered ring. The thirty 6 : 6 bonds are considered as C=C double bonds, since the length is 1.39 A, while the sixty 6:5 bonds are considered as single C-C bonds with a length o f is 1.43 A.^^ The long/short bond alternation in the six-membered rings suggests that the double bonds in Ceo are somewhat localized, unlike in benzene ring, all the K electrons are delocalized and the C-C bond lengths are equalized. Fullerenes are considered as the third allotrope o f carbon. The other two, graphite and diamond, are actually infinite network solids. On the surface of them, there exist inevitably dangling bonds due to the unsatisfied carbon valences. In graphite, all the a bonds are planar and the p -7 C orbitals are perpendicular to the a bond skeleton. The introduction of five-membered rings in C eo makes the sheet of carbon curl up and close in on itself with all the dangling bonds tied up. In this sense, the C eo surface is perhaps more stable. The hybridization of fullerene carbon falls between sp^ (graphite) and sp^ (diamond), since a rehybridization will occur when the a bonds at the conjugated carbon atoms deviate fi'om planarity." Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Li.3. General Chemical Properties For crystalline Ceo, the standard heat o f formation (A^f^ was determined to be 9.1 kcal/mol per C atom.^^ Compared to 0 for graphite and 0.4 kcal/mol per C atom for diamond, C eo is obviously thermodynamically unfavorable. Therefore, the formation o f Ceo should be a kinetic process.™ In the early stage of its research, Ceo was expected to be a species with evident aromatic characteristics,'”’ ’ ^ ^ of which the Birch reduction^ is one example. However, due to the lack o f attached hydrogen atoms, no aromatic reactions related to these H * s (e.g., aromatic electrophilic substitution) could occur. As our knowledge accumulates, it is found that chemically C g o behaves more like an electron-deficient alkene rather than like a "super-aromatic" molecule.^ The fullerene chemistry could be summarized in the two following category.™ ( 1 ) alkene-like reactivities Osmylation, 7 ^-transition metal complexation, halogénation, cycloadditions ([4+2], [2+2], [6+2]™, [8+2]™), radical addition, hydrogenation, epoxidation are some typical chemical reactions of C g o . As we know, alkenes show all the similar reactivities. O f these reactions, cycloaddition reactions are among the most successful and versatile to functionalize C eo- (2 ) electron-deficient reactivities C eo undergoes nucleophilic reactions not only with anionic species such as organolithium, Grignard reagents, hydroxide and cyanide,™ but also with neutral bases. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. such as amines. Ceo can be reduced to form the so-called fuilerides. It also forms molecular complexes with electron-donor species. All these reveal the electron- deficient character of C eo 1.2. Fuilerides L2.1. The Electronegative Characteristics of C eo To explore the electronic structures of Ceo, quite a few theoretical calculations^*^’ ^ * have been perfisrmed, fi'om which similar molecular orbital schemes were obtained. The most fi-equently cited one is by Haddon et al.^ using Hückel molecular orbital (HMO) theory (Figure 1.3). Our interests focus mostly on the fi-ontier orbitals, i.e., HOMO (A u ), LUMO (/lo) and LUMO + 1 (Ag) The low-lying LUMO and LUMO + 1 levels (/lu at -0.1 3 9 y 0 , A g at -0.382^ suggest that the electron afOnity of C o o will be high and under suitable conditions C o o can possibly accept up to 12 electrons. The mixing o f low-lying carbon 2s atomic orbitals into the 7C-orbital and the presence of 12 five-membered rings are considered the two factors to account for these. The electron afiSnity (EA ) o f C oo was determined to be 2.6-2 8 eV by the UPS measurement (ultraviolet photoelectron spectra) o f C o o * .^ ® This value is in a good agreement with the calculated result (2.4 eV).^^ Accordingly, the mono- and dianion of Coo have been observed in the gas phase/^ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LUMO ‘lu H O M O Figure L3 The frontier Hûckel molecular orbitals o f Ceo/' <0 10 uA « b - 3 . 0 - 2.0 - 1 .0 Ootwtlal fVolts vm re/re* I Figure L4 Reduction o f C eo in CHsCN/toluene at -10"C using (a) cyclic voltammetry and (b) dififerential pulse voltammetry.^^* Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cyclic voltammetry (CV) is a powerful tool to investigate stepwise redox processes in solution. This method was applied to study the fullerene reduction shortly after the preparation o f C g o in macroscopic quantities.^^ The observation of increasingly charged C e o " * (n = 1 to 5) in solution immediately followed.^* Carrying out CV experiments at low temperature in a mixed-solvent system will expand the potential window and by doing this, two independent groups successfully detected the sixth reduction wave of C eo .^ ^ These reversible waves (Figure 1.4) illustrate that six electrons can add facilely to the C eo buclqfball. L2.2. Fuilerides in Condensed Media As shown in Figure 1.4, all the reduction waves of C eo are well-separated, indicating that Ceo can be reduced to form distinct fuilerides, the anions of Ceo- Extensive studies have been focused on fuilerides and the synthesis and investigation are among the most successful parts o f fullerene chemistry. In condensed media, fuilerides have been synthesized and studied by various methods, most of which are briefly summarized below. ( 1 ) doped with alkali, alkaline earth or rare earth metals;^^ The M xCeo systems prepared in this way, whether stoichiometric or non- stoichiometric, are all in the solid state. For most o f alkali metal fuilerides, x is no greater than 6 and electrons fill into the triply-degenerate LUMO o f C eo - In stoichiometric systems, M nC eo"’ , n = 1, 2, 3, 4, 6 and M could be Li, Na, K, Rb, Cs. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Alkaline earth metals (Mg, Ca, Sr, Ba)^^ and rare earth metals (Yb,^* Sm^^ can be also intercalated into C eo to form fuilerides, e.g., CasCeo,'” BaeCeo/^ S reC eo-'*^ The ternary fuilerides with two different types o f metal cations are also known.^^’ ^ ’ ^ ^ The presence of so-called “superfiiUerides” (for C e o " * , 6 < n < 1 2 )^ indicates that electrons can also transfer into the LUMO + 1 orbitals (tig). Encapsulating metal atom(s) inside fullerene cage forms endohedral species, Mn@C6o .^ ^ Both theoretical and experimental results show that some of the electrons are transferred from the trapped metal atom(s) to the cage. Mn@C6o may display some similarities to their “isomers”, the exohedral metal fuilerides (MnCeo). Mostly, the preparation o f intercalated fuilerides is achieved by the direct reaction o f solid C eo with metals'** or their alloys by vapor-transport technique at high temperature. In addition, there are some other methods: (a) reaction with metals in solution then removing the solvents^*’ ^ (toluene, N H s,'* ® THF); (b) reaction with NaNs, M-M' (M = Na,’® M' = Hg; M = C s," M' = Hg, Tl, Bi; M = Rb, Rb-K,’^ M' = Hg, Tl), MH, M B H *^® ’® ( M = Na, K) in soUd state; (c) reaction with stoichiometric amount of MeCso’^ or M g C e o ;* * * (d) thermal de-intercalation ofMaCeo/^ (e) electrochemical intercalation.” (2 ) by electrolysis or electrocrystalization Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Since the reduction potentials of mono- to hexa-anion of C eo are well separated, bulk C eo anions could be selectively generated by controlled potential electrolysis. Up to have been prepared in this way.’® Using the electrocrystalization technique, single crystals of C eo monoanion have been obtained.’^ (3) with other chemical reductants Fulleride salts can be generated with various reducing chemical reagents. (a) alkali or alkaline-earth metals or their alloys in THF,’* benzene,’® or liquid NH3® ° or other metals (Li, Ba, Mn, Fe, Cu or Zn) in AT-methylimidazole;®* (b) alkali metal (Na) with addition of crown ether;® ^® ^ (c) organic reductant like TDAE®^ or thiolates;®’ (d) organometallic reductant, including: metallocenes CoCpz,®^ CoCp*2, ® ® Fe(Cp)(C6M eg),® ^ MnCp*2,® ‘ [MnCp*2] ',® * NiCp*2; ® ® low-valent metal porphyrin Cr(TPP) in THF,''® Sn(TpTP) in A^-methylimidazole;®^ NaBPlu.^' All the methods in category 2 and 3 are via “wet” chemical routes, through which C eo could usually be reduced up to tri-anion, and in one or two cases, even to Ceo’' ® ° ® and Along with the six reduction waves in its cyclic voltammogram, this demonstrates that in solution phase, electrons can only fill into tia orbitals (LUMO), i.e., C eo can only accept up to six electrons in solution. 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. L2.3. Potential Applications of the Fuilerides The alkali metal fuilerides (M % C g o ) have been studied systematically/^ It is found that for potassium, the conductivity varies with the cation number x. When x = 3, the conductivity reaches a maximum and at low temperature (T== 18 K) the fulleride becomes a superconductor. Fuilerides with some other alkali metal or alkaline-earth, rare-earth metal cations also show super-conductivity at certain temperature. (TDAE)Cgo, another type of fulleride, exhibits unusual molecular ferrom agnetism .^^ For instance, its Curie temperature (Tc= 16 K) is higher than any other nonpolymeric organic ferromagnet and it does not show any remanent magnetization or hysteresis. These two characteristics render C eo fuilerides o f potential applications in the fabrication of new materials. For this purpose, a good understanding of the physical origins of these properties is needed. L2.4. Isolation and Studies of Discrete Fuilerides Both the properties described above are related to the partial occupation o f the degenerate LUMO (tw), or of the conduction bands derived from them. It is necessary to explore their electronic structures in detail. The solid state intercalated fuilerides frequently contain more than one phases and their characterizations are often difGcult. Discrete C eo anions are usually generated under milder conditions and they can be isolated and fully characterized. From their solutions, good quality single 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. crystals^^’ ^ can occasionally be obtained and X-ray data could provide a plenty of structural information. The fuilerides produced by electrolysis unavoidably contain impurities like the supporting electrolytes. With these samples, quantitative measurement o f some properties like magnetic susceptibility is infeasible. To date, the only way to generate fulleride salts of analytical purity is by chemical reduction. The interests of our group have been focused on the chemical synthesis and the isolation of discrete üülerides. These salts were then characterized by means of NER,’* * EPR, NMR, and SQUID. In fulleride studies, the interpretation of some results still remains controversial. For instance, the EPR signals o f C eo’ and C e o ^ * show a “spike”, i.e., a minor, very sharp signal at relatively high temperature, and C eo^' samples show a doublet-like central line. The “spikes” seem to be ubiquitous in all the C eo’ and Ceo^' samples and to explore their origin we need to further explore the electronic structure of fuilerides. The investigation o f their functionalized analogs would hopefully furnish us more information in this respect. References 1. (a) Osawa, E. Kagaku (kyoto) 1970, 25, 854; (b) Yoshida, Z.; Osawa, E Kagakudqjin: Kyoto, 1971, pp. 174-178. 2. (a) Bochvar, D A.; Gal'pem, E. G. Dokl. Akad Nauk SSSR 1973, 209, 610; Proc. Acad. Sci. USSR 1973, 209, 239 (English translation); (b) Stankevich, I. V.; Nikerov, M. V.; Bochvar, D A. Russ. Chem. Rev. 1984, 53, 640. 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3. Davidson, R. A. Theor. Chim. Acta 1981, 58, 193. 4. (a) Haymet, A. D J Chem. Phys. Lett. 1985, 122, 421; (b) Haymet, A. D. J. J. Am. Chem. Soc. 1986, 108, 319. 5. Rohlfing, E. A.; Cox, D. M.; Kaldor, A. J. Chem. Phys. 1984, 81, 3322. 6 . Kroto, H. W.; Heath, J. R.; O'Brien, S. C ; Curl, R. F.; Smally, R. E. Nature 1985, 318, 162. 7. Kratschmer, W ; Fostiropoulos, K.; Huffinan, D. R. Chem. Phys. Lett. 1990, 170, 167. 8 . Kratschmer, W ; Lamb, L. D ; Fostiropoulos, K.; Huffinan, D. R. Nature 1990, 347, 354. 9. (a) Weeks, D. E.; Harter, W. G. J. Chem. Phys. 1989, 90, 4744; (b) Stanton, R E.; Newton, M. D. J. Phys. Chem. 1988, 92, 2141; (c) Cyvin, S. J.; Brendsdal, E.; Cyvin, B. N.; BrunvoU, J. Chem. Phys. Lett. 1988, 143, 377; d) Wu, Z. C ; Jelski, George, T. F. Chem. Phys. Lett. 1987,137, 291. 10. Taylor, R.; Hare, J. P.; Abdul-Sada, A. K.; Kroto, H. W. J. Chem. Soc., Chem. Commun. 1990, 1423. 11. See, for example, (a) Koch, A. S.; Khemani, K. C ; Wudl, F. J. Org. Chem. 1991, 56, 4543; (b) Parker, D. H ; Wurz, P.; Chatteijee, K ; Lykke, K. R.; Hunt, J. E.; Pellin, M- J ; Hemminger, J. C ; Gruen, D. M.; Stock, L. M. J. Am. Chem. Soc. 1991, 113, 7499. 12. See, for example, (a) Ajie, H.; Alvarez, M. M.; Anz, S. J ; Beck, R. D ; Ciederich, F.; Fostiropoulos, K.; Huffinan, D R.; Kratschmer, W ; Rubin, Y.; Schriver, K. E. Sensharma, D ; Whetten, R. L. J. Phys. Chem. 1990, 94, 8630; (b) Diederich, F. Ettl, R.; Rubin, Y.; Whetten, R. L.; Beck, R.; Alvarez, M. M.; Anz, S. J. Sensharma, D ; Wudl, F.; Khemani, K. C ; Kuch, A. Science 1991, 252, 548; (c) Khemani, K. C ; Prato, M.; Wudl, F. J. Org. Chem. 1992, 57, 3254; (d) Atwood, J. L ; Koutsantonis, G. A.; Raston, C. L. Nature 1994, 368, 229. 13. Kroto, H. W. Nature 1987, 329, 529. 14. Diederich, F ; Whetten, R. L. Acc. Chem. Res. 1992, 25, 119. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15. Johnson, R. D.; Yannoni, C. S.; Dom, H. C ; Salem, J. R.; Bethune, D. S. Science 1992, 255, 1235 and the references therein. 16. Liu, S.; Lu, Y.-J.; Kappes, M. M.; Ibers, J. A. Science 1991, 254, 408. 17. Hawkins, J. W.; M ^er, A.; Lewis, T. A.; Loren, S.; Hollander, F. J. Science 1991, 252,312. 18. Haddon, R. C. Acc. Chem. Res. 1992, 25, 127 and the references therein. 19. Beckhaus, H.-D.; Rûchardt, C.; Kao, M.; Diederich, P.; Foote, C. S. Angew. Chem., Int. Ed. Engl. 1992, 31, 63. 20. Hirsch, A. The Chemistry o f the Fullerenes Thieme: NewYork, 1994, pp 15-16. 21. (a) Klein, D. J.; Schmalz, T. G.; Hite, G. E.; Seitz, W. A. J. Am. Chem. Soc. 1986, 108, 1301; (b) Klein, D. J.; Seitz, W. A.; Schmalz, T. G. Nature 1986, 323, 703. 22. (a) Haufler, R. E.; Conceicao, J.; Chibante, L. P. P.; Chai, Y ; Byrne, N. E. Flanagan, S.; Haley, M. M ; O'Brien, S. C ; Pan, C ; Xiao, Z ; Billups, W. E. Ciufolini, M. A.; Hauge, R H.; Margrave, J. L.; l^ so n , L. J.; Curl, R F Smalley, R. E. J. Phys. Chem. 1990, 94, 8634; (b) Banks, M. R.; Dale, M. J. Gosney, I.; Hodgson, P. K. G ; Jennings, R. C. K.; Jones, A. C ; Lecoultre, J. Langridge-Smith, P. R R ; Maier, J P.; Scrivens, J. H.; Smith, M. J. C ; Smyth, C J ; Taylor, A. T.; Thorbum, P.; Webster, A. S. J. Chem. Soc., Chem. Commun. 1993, 1149. 23. Fowler, P. W ; Ceulemans, A. J. Phys. Chem. 1995, 99, 508. 24. Banks, M. R ; Cadogan, J. I. G ; G osn^, I.; Hodgson, P. K. G ; Langridge-Smith, P. R R ; Millar, J. R A.; Parkinson, J. A.; Sadler, I. H ; Talor, A. T. J. Chem. Soc., Chem. Commun. 1995, 1171. 25. Beer, E.; Feuerer, M.; Knorr, A.; \firlach. A.; Daub, J. Angew. Chem., Int. E d Engl. 1994, 33, 1087. 26. Keshavarz-K, M.; Knight, B ; Srdanov, G ; Wudl, F. J. Am. Chem. Soc. 1995, 117, 11371. 27. (a) Ermer, O. Helv. Chim. Acta 1991, 74, 1339; (b) Izuoka, A.; Tachikawa, T.; Sugawara, T.; Saito, Y ; Shinohara, H. Chem. Lett. 1992, 1049. 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28. See, for «cample, (a) Hale, P. D J. Am, Chem. Soc. 1986, J08, 6087; (b) Satpathy, S. Chem. Phys. Lett. 1986, 130, 545; (c) Disch, R. L.; Schulman, J. M Chem. P f^s. Lett. 1986,125, 465. 29. Haddon, R _ C ; Brus, L. E ; Raghavachari, K. Chem. Pfys. Lett. 1986,125, 459. 30. (a) Yang, S. H.; Pettiette, C. L ; Conceicao, J.; Cheshnovsky, O ; Smalley, R. E. Chem. Phys. Lett. 1987,139, 233; (b) Wang, L.-S ; Conceicao, J.; Jin, C ; Smally, R- E. Chem. Pfys. Lett. 1991, 182, 5. 31. Larsson, S.; Volosov, A.; Rosen, A Chem. Phys. Lett. 1987, 137, 501; (b) Rosen, A ; Wastberg, B. J. Chem. Phys. 1989, 90, 2525. 32. (a) Hettich, R. L.; Compton, R. N.; Ritchie, R. H. Phys. Rev. Lett. 1991, 67, 1242; 0)) Limbach, P. A.; Schweikhard, L.; Cowen, K. A ; McDermott, M. T.; Marshall, A G.; Coe, J. V. J. Am. Chem. Soc. 1991,113, 6795. 33. Haufler, R. E.; Conceicao, J ; Chibante, L. P. F.; Chai, Y.; Byrne, N. E.; Flanagan, S.; Haley, M. M ; O’Brien, S. C ; Pan, C ; JGao, Z ; Billups, W. E ; Ciufolini, M. A ; Hauge, R. H ; Margrave, J. L.; Wilson, L. J.; Curl, R. F.; Smalley, R. E. J. Phys. Chem. 1990, 94, 8634. 34. (a) Allemand, P.-M.; Koch, A ; Wudl, F.; Rubin, Y ; Diederich, F ; Alvarez, M. M.; Anz, S. J ; Whetten, R. L. J. Am. Chem. Soc. 1991, 113, 1050; (b) Dubois, D ; Kadish, K. M.; Flanagan, S.; Haufler, R. E.; Chibante, L. P. F.; Wilson, L. J. J. Am. Chem. Soc. 1991, 113, 4364; (c) Meerholz, K.; Tschuncky, P.; Heinze, J. J. Electroanal. Chem. 1993, 347, 425; (d) Dubois, D ; Kadish, K. M ; Flanagan, S.; Wilson, L. J. J. Am. Chem. Soc. 1991, 113, 7773; (e) Dubois, D.; Moninot, G ; Kutner, W.; Jones, M. T.; Kadish, K. M. J. Phys. Chem. 1992, 96, 7137. 35. (a) Xie, Q.; Perez-Cordero, E.; Echegoyen, L. J. Am. Chem. Soc. 1992, 114, 3978; (b) Ohsawa, Y.; Saji, T .J. Chem. Soc., Chem. Commun. 1992, 781. 36. Murphy, D. W ; Rosseinsky, M. J.; Fleming, R. M.; Tycko, R ; Ramirez, A P.; Haddon, R. C ; Siegrist, T.; Dabbagh, G ; Tully, J. C ; Walstedt, R. E. J. Phys. Chem. Solids 1992, 53, 1321. 37. Chen, Y.; Weaver, J. H.; Chibante, L. P. F.; Smalley, R. E. Pfys. Rev. B 1992, 45, 8845. 38. Ozdas, E ; Kortan, A R.; Kopylov, N.; Ramirez, A P.; Siegrist, T.; Rabe, K. M ; Bair, H. E ; Schuppler, S.; Citrin, P. H. Nature 1995, 375, 126. 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39. Chen, X. H.; Roth, G. Pf^s. Rev. B 1995, 52, 21. 40. Kortan, A. R_; Kopylov, N.; Œarum, S.; Gyorgy, E. M.; Ramirez, A P.; Fleming, R. M.; Thiel, F. A ; Haddon, R. C. Nature 1992, 355, 529. 41. Kortan, A. R _; Kopylov, N.; Giarum, S.; Gyorgy, E. M.; Ramirez, A P.; Fleming, R- M.; Zhou, 0.;Thiel, F. A ; Trevor, P. L ; Haddon, R. C Nature 1992, 360, 566. 42. Kortan, A R _; Kopylov, N.; Ozdas, E ; Ramirez, A P.; Fleming, R. M.; Haddon, R. C. Chem. Phys. Lett. 1994, 223, 501. 43. (a) Fleming, R. M.; Ramirez, A P.; Rosseinsky,, M. J.; Murphy, D. W.; Haddon, R- C.; Zahurak, S. M.; Makhija, A V. Nature 1991, 352, 787; (b) Tanigaki, K.; Ffirosawa, I.; Ebbesen, T. W.; Mizuki, J.; Shimakawa, Y.; Kubo, Y.; Tsai, J. S.; Kuroshima, S. Nature 1992, 356, 419. 44. Duggan, A C ; Fox, J. M.; Henry, P. F ; Heyes, S. J.; Laurie, D. E ; Rosseinsky, M. J. Chem. Soc., Chem. Commun. 1996, 1191. 45. Ohno, T. R.; KroU, G. H.; Weaver, J. H.; Chibante, L. P. F.; Smalley, R. E. Phys. Rev. B 1992, 46, 10437. 46. (a) Yildirim T.; Zhou, O ; Fischer, J. E ; Bykovetz, N.; Strongin, R _ A ; Cichy, M. A ; Smith m , A B.; Lin, C. L ; Jelinek, E L Nature 1992, 360, 568; (b) Maxwell, A J ; Bruhwiler, P. A ; Anderson, S.; Martensson, N.; Rudolf P. Chem. Phys. Lett. 1995, 247, 257. 47. Bethune, D. S.; Johnson, E L D.; Salem, J. R.; Vries, M. S.; Yannoni, C. S. Nature 1993, 366, 123 and references therein. 48. Hebard, A F.; Rosseinsky, M J.; Haddon, R _ C ; Murphy, D. W ; Giarum, S. H.; Palstra, T. T. M.; Ramirez, A P.; Kortan, A R. Nature 1991, 350, 600. 49. Bufhnger, D. R.; Ziebarth, E L P.; Stenger, V. A ; Recchia, C ; Pennington, C. H. J. Am. Chem. Soc. 1993, IJ5, 9267. 50. Rosseinslqr, M. J.; Murphy, D. W ; Heming, R _ M.; Tycko, E L ; Ramirez, A P.; Siegrist, T.; Dabbagh, G ; Barrett, S. E. Nature 1992, 356, 416. 51. Kelty, S. P.; Chen, C -C ; Lieber, C M Nature 1991, 352, 223. 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52. Chen, C -C ; Kelty, S. P.; Lieber, C M. Science 1991, 253, 8 8 6 . 53. Fleming, R. M.; Rosseinsl^, M. J.; Ramirez, A. P.; Murphy, D. W.; Tully, J. C.; Haddon, R _ C ; Siegrist, T.; Tycko, R.; Glarum, S. H.; Marsh, P.; Dabbagh, G.; Zahurak, S. M.; Makhija, A. V.; Hampton, C Nature 1991, 352, 701. 54. Yildirim, T.; Fischer, J. E.; Harris, A. B ; Stephens, P. W.; Liu, D.; Brard, L.; Strongin, R. M.; Smith m , A. B. Pf^s. Rev. Lett. 1993, 71, 1383. 55. Chabre, Y.; Djurado, D ; Armand, M.; Romanow, W. R _; Coustel, N.; McCauley Jr., J. P.; Fischer, J. E.; Smith HI, A B. J. Am. Chem. Soc. 1992, 114, 764. 56. (a) Allemand, P.-M.; Srdanov, G ; Koch, A ; Khemani, K.; Wudl, F ; Rubin, Y.; Diederich, F.; Alvarez, M. M.; Anz, S. J.; Whetten, R. L. J. Am. Chem. Soc. 1991, 113, 2780; (b) Greaney, M. A ; Gorun, S. M. J. Pl^s. Chem. 1991, 95, 7142; (c) Dubois, D ; Jones, M. T.; Kadish, K. M. J. Am. Chem. Soc. 1992, 114, 6446; (d) Khaled, M. M ; Carlin, R. T.; Trulove, P. C ; Eaton, G. R_; Eaton, S. S. J. A. Chem. Soc. 1994,116, 3465. 57. (a) Moriyama, H.; Kobayashi, H.; Kobayashi, A ; Watanabe, T. J. Am. Chem. Soc. 1993, 115, 1185; (b) Penicaud, A ; Perez-Benitez, A ; Gleason, R.; Munoz P., E ; Escudero, R. J. Am. Chem. Soc. 1993, 115, 10392; (c) Bilow, U ; Jansen, M. J. Chem. Soc., Chem. Commun. 1994, 403. 58. (a) Bausch, J. W ; Prakash, G. K. S.; Olah, G. A ; Tse, D. S.; Lorents, D. C ; Bae, Y. K ; Malhotra, R. J. Am. Chem. Soc. 1991, 113, 3205; (b) Baumgarten, M.; Gugel, A _; Gherghel, L. Adv. Mater. 1993, 5, 458; (c) Chen, J ; Huang, Z.-E.; Cai, R.-F.; Shao, Q -F ; Chen S.-M.; Ye, H.-J J. Chem. Soc., Chem. Commun. 1994, 2177; (d) Schell-Sorokin, A. J ; Mehran, F.; Eaton, G. R.; Eaton, S. S.; Viehbeck, A.; O’Toole, T. R.; Brown, C. A. Chem. Phys. Lett. 1992, 195, 225. 59. Kukolich, S. G ; HufiOnan, D. R. Chem. Phys. Lett. 1991, 182, 263. 60. (a) Zhou, F ; Jehoulet, C ; Bard, A J. J. Am. Chem. Soc. 1992, 114, 11004; (b) FuUagar, W. K.; Gentle, I. R.; Heath, G. A ; White, J. W. J. Chem. Soc., Chem. Commun. 1993, 525. 61. Selegue, J. P.; Dev, S.; Guarr, T F ; Brill, J W ; Figueroa, E personal communication. 62. Stinchcombe, J ; Penicaud, A ; Bhyrappa, P.; Boyd, P. D. W ; Reed, C. A J. Am. Chem. Soc. 1993, 115, 5212. 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63. Bhyrappa, P.; Paul, P.; Stinchcombe, J.; Boyd, P D W.; Reed, C. A. J. Am. Chem. Soc. 1993, 115, 11004. 64. (a) Allemand, P.-M.; Khemani, K. C ; Koch, A.; Wudl, F.; Holczer, K.; Donovan, S.; Gnmer, G.; Thompson, J. D. Scierie 1991, 253, 301; (b) Stephens, P. W ; Cox, D.; Lauher, J. W ; Nfihaly, L.; Wiley, J. B.; Allemand, P.-M.; Hirsch, A.; Holczer, K.; Li, Q ; Thompson, J. D ; Wudl, F. Nature 1992, 355, 331. 65. Subramanian, R_; Boulas, P.; Vijayashree, M. N.; D'Souza, F.; Jones, M T.; Kadish, K. M. J. Chem. Sac., Chem. Commun. 1994, 1847. 6 6 . Boyd, P. D. W ; Bhyrappa, P.; Paul, P.; Stinchcombe, J.; Bolskar, R. D ; Sun, Y.; Reed, C. A. J. Am. Chem. Soc. 1995, 117, 2907. 67. Bossard, C ; Rigaut, S.; Astruc, D.; Delville, M -H.; Felix, G ; Fevrier-Bouvier, A.; Amiell, J.; Flandrois, S.; Delhaes, P. J. Chem. Soc., Chem. Commun. 1993, 333. 6 8 . (a) Douthwaite, R. E.; Brough, A. R.; Green, M. L. H. J. Chem. Soc., Chem. Commun. 1994, 267; (b) Liu, X.; Wan, W. C ; Owens, S. M.; Broderick, W. E. J. Am. Chem. Soc. 1994, 116, 5489. 69. Wan, W. C ; Liu, X.; Sweeney, G. M ; Broderick, W. E. J. Am. Chem. Soc. 1995. 117, 9580. 70. Penicaud, A.; Hsu, J ; Reed, C. A.; Koch, A.; Khemani, C ; Allemand, P.-M.; Wudl, F. J. Am. Chem. Soc. 1991,113, 6698. 71. Moriyama, H ; Kobayashi, H.; Kobayashi, A.; Watanabe, T. Chem. Phys. Lett. 1995, 238, 116. 72. Wudl, F.; Thompson, J. D. J. Phys. Chem. Solids 1992, 53, 1449. 73. Paul, P.; Jûe, Z ; Bau, R _; Boyd, P. D. W ; Reed, C. A. J. Am. Chem. Soc. 1994, 116, 4145. 74. Bolskar, R D ; Gallagher, S. H.; Armstrong, R S.; Lay, P. A.; Reed, C. A. Chem. Phys. Lett. 1995, 247, 57. 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter n. n i. NMR Characterization of C go Fuilerides II. 1.1. Introduction For the carbon-only cluster Ceo, its "C NMR was reported immediately after its production in macroscopic quantities/ The single sharp line at 143 ppm was the first experimental evidence of its icosahedral structure. Solid state NMR spectroscopy has also been applied to fullerene characterization.^ The observation of a narrow single peak at room temperature with a chemical shift (143 ppm) close to the value in solution indicates rapid and isotropic reorientation o f C eo molecules. The variable temperature C NMR features of the fullerene, along with other results, reveal its dynamic behavior as well as a phase transition in solid state.^ A variety of A xC eo fuilerides (A = alkali metal) have been studied with "C NMR spectroscopy in the solid state.^ All the fulleride chemical shifts are in the range of 150 to 200 ppm, downfield shifted by 1 0 to 55 ppm relatively to neutral Ceo. For all the ‘ ^C chemical shifts reported, there is no simple correlation with the formal charges of the fuilerides. The characteristic "C chemical shifts o f different fuilerides have been employed to determine the phase composition,^ or to detect the presence of impurities.'*'’ The dimeric and polymeric phases of ACeo in solid state have been extensively studied by several methods,® one o f which is the * ^ C NMR spectroscopy.^ Peaks at 50 or 78 ppm 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. are assigned to the sp^-type carbon at the ball-to-ball bonding sites because these values are in the range o f those o f quaternary carbons (30-80 ppm). Table IL l Some typical "C chemical shifts o f in solid state at room temperature fulleride system 5 (ppm) reference C eo N a C 6 o ( T H F > 5 188 È KCeoCTHF) 187 9 N a C 6 o ( T H F > 3 187 10 [ N ( C H 3 > 4 ] C 6 0 • ( T H E ) 1 . 5 185 11 Ceo' KzCeoCTHF^ 183 1 2 NazCgo 173 4 a Ceo'* NazCsCeo 182 4 a K3C60 186 5 ,1 3 R b 3 C 6 0 195 1 3 Ceo^ R b 4 C e o 181 1 4 K 4C 6 0 181 1 3 . 1 5 C e o ® * KeCeo 156 1 3 RbeCeo 154 1 4 Intercalated fuilerides have been studied with "C NMR at various temperatures. At relatively high temperatures their lines are usually narrow, indicating that the isotropic rotations o f anionic buckybails are also rapid, while at low temperature the lines become very broad and obviously the rotation is slowed down.^^'^^ The dramatic change in line-shape and linewidth o f "C peaks of ACeo at different temperatures were believed due to the phase transition. For discrete 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fulleride KCeoCTHF)* in the solid state, the relative intensities of two signals vary substantially at different temperature/^ An ion-pairing model was proposed to rationalize this result as well as the EPR spike at relatively high temperature. Before the work presented here, nearly all the NMR investigations o f C eo fuilerides were in the solid state and most o f them were on the metal-intercalated fuilerides. Because o f the strong environmental effects in these systems, the data may not be the intrinsic ones. Based on a single sharp line at 157 ppm, it was claimed that with excess lithium metal in THF-t/g under sonication, C eo was reduced to a diamagnetic species, probably Cgo^.^^ In our group, discrete fulleride salts (Ceo', Ceo^' and C eo ^* ) have been successfully synthesized and isolated.^" These paramagnetic species have been studied extensively with EPR spectroscopy and SQUID magnetic susceptibility. To further explore their electronic properties, we carried out the '^C NMR characterizations in solution. To investigate the temperature effect on the electronic structures relationship between the isotropic nuclear resonance shift (see Section U. 1.2.3) and temperature, variable temperature NMR experiments were conducted on all these systems. II. 1.2. Results and Discussion n .l.2 .1 . "C NMR of [Na(crown)(THF)2 lm C 6 a (n = 1, 2, 3) at Ambient Temperature 2 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The solvent used for NMR measurements was DMSO-<4, in which all the three fulleride salts are readily soluble. Vis-NIR spectra were used to establish the integrity of the samples before and after the NMR measurements and consequently to ensure the reliability o f the NMR data. The H NMR was checked before every acquisition and there was no detection o f peaks other than crown either and THF. This indicates good chemical purity o f the samples. For all the fulleride salts, including [Na(crown)(THF)2 ]nC6o"' (n = 1, 2, 3), as well as [Na(crypt)*]4C6o '‘ * (see Chapter HI), th^r show similar NMR characteristics: all the peaks are broad and the chemical shifts are around 185 to 200 ppm, i.e., shifted downfield by 40 to 55 ppm relative to neutral Cm. These characteristics are listed in Table Et.2. The result on Ceo^ with Na(crypt)^ as the cation is also included for comparison. As a typical representative, a spectrum of C e o ^ ‘ is shown in Figure I I 1. Compared to those data in Table H I, the solution "C NMR chemical shifts, in general, are consistent with the results in solid state. The result for Cm' has been confirmed in solution investigations by others.^^ Of all these fulleride salts, Cso^' is a little special in respect to its solution ‘^C NMR. While the chemical shifts o f Ceo', Ceo^' and Cm'*' are all around 185 t 2 ppm, the Cm^' resonance is particularly downfield shifted to 197 ppm. Meanwhile, the Cm^' peak is distinctly broader than those of other fuilerides, as indicated by the half height width values in Table n 2 A similar observation has also been made in solid state CM ^' " It is interesting to note that the EPR signal of Cm^' is also the broadest among all the four fulleride salts (see also 2 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table IL2 NMR results of discrete fiillerides in DMSO-<4 at room temperature fuUeride salt chemical shift 5 (ppm) isotropic shift 5ûo(ppm) half height width (ppm) [Na(crown)^C6o ‘ 186 43 3 [Na(crown)^2C6o^* 183 40 1 - 2 [Na(crown)^3C6o^' 197 54 6 [Na(crypt)*]4C«)‘ ^ 185 42 1 - 2 m m A i w «wwr -i 1 -----1 -----1 -----1 -----1 ---- 1 -----1 -----1 ---- 1 ---- 1 -----1 ---- 1 -----1 ---- 1 -----1 ---- 1 -----1 ---- 1 -----1 ---- 1 -----1 ---- 1 ---- r 200 150 100 SO p p m Figure IL l NMR spectrum of [Na(crownXTHF)2*] 3[C«o^ in DMSO-<4 (the peak at 39.5 ppm is from the solvent, the rest are from crown ether and THF). 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter m ), since usually the line widths o f EPR and NMR signals have a reverse relationship. The C chemical shift of solid (PPN^Cgo is at 186 ppm, identical to the solution Geo’ value. Unfortunately, we were not able to obtain reliable "C data for solid C e o ^ * , C e o ^ * or Ceo^, probably due to the sample decomposition during the torch sealing o f the NMR tubes (see Chapter m ). The interpretation of ‘ ^C chemical shifts o f the fUUerides is complex. Ring currents have been proposed to rationalize the relative chemical shifts o f Ceo^ and Cfio-^ This model may also be applied to interpret those o f the other fuUerides. For benzene and monocyclic aromatic ions, their "C chemical shifts are found to correlate linearly to their n electron density.^** The increased k electron density at the carbons will cause an increased paramagnetic shielding and a consequent upfield shifted ‘^C resonance. This correlation does not appear to be applicable to C eo anions, as the chemical shifts o f Ceo', Ceo^' and Cao^ salts are almost identical. The similar chemical shifts may be related to their spin states. For Ceo^' or C e o * * , experimental results suggest that the singlet ground state and the triplet state are probably close-lying in energy and hence the populations in these two states should be at the same level. Therefore, the average spin states o f Ceo^' or C e o '* ’ are close to a doublet, i.e., similar to that o f Ceo' (and of C e o ^ * )- Using neutral C co as the diamagnetic reference (5du =143 ppm), for all the fuUerides, the isotropic shifts (see Section n. 1.2.3) are included in Table n.2. 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For C e o " * salts in solution, the unpaired electrons are delocalized over the whole buckyball and distributed evenly on all carbon atoms. Due to the Jahn-Teller distortion, the symmetry o f the anions will be lowered from icosahedral of neutral C eo and at any given instant the fuUerides are expected to be magneticaUy anisotropic. However, the fuUeride baUs in solution undergo rapid pseudorotations as weU as real rotations. Over the NMR time scale, the anisotropy wiU be averaged out and thus the fuUerides are dynamicaUy isotropic. Although the anisotropy caused by ion pairing has been observed for Ceo" by EPR spectroscopy in 2-Me-THF,“ * this should not be the case under solution NMR conditions since the spectra were recorded at higher temperatures in polar solvents (DMSG, acetonitrile, etc.). Overall, the dipolar shifts (see Section n . 1.2.3) wiU vanish (i.e., Sjip » 0) fr)r the fuUerides in solution at room temperature. B ased on the above discussion, for the fuUeride system s, the contact shifts wiU b e dom inant in the isotropic shifts, i.e., 5uo » 5con- The electrons w ith s character can have probabUity o f being at the nucleus and only in this case the contact interaction b etw een unpaired electron and the nucleus can occur. T heoretical results dem onstrate that in Ceo, the carbon 2 s atom ic orbitals are m ixed into the x-orbital due to the rehybridization.^ The existen ce o f the contact shifts in th e fuUeride system s is an experim ental indicator o f th e s orbital com ponent in th e 7C-orbitals. 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. n.1.2.2. ' c NMR of [Na(crownXTHF)2lnC6o (n = 1, 2, 3) in Solution a t Variable Temperatures To characterize fuUeride salts by NMR spectroscopy at variable temperatures, the solvent should (a) be sufiQciently polar to dissolve the anionic fuUeride; (b) have a wide temperature range between its freezing and boiling point; (c) be stable to the reducing fuUeride salts. THF-f/g has a good temperature range, but only Ceo' salt is soluble in it. After a systematic search (see Table n.3), propionitrile (CH3CH2CN) was found to be the best solvent for our purpose. In propionitrile, aU three o f the fuUerides are stable and have considerable solubilities. Sealed acetone-<4 capUlaries were used as a source o f the lock signal for the spectrometer as weU as the external reference. Table IL3 The solvent search for NMR experiments at variable temperatures solvent freezing point (°C) boUing point (°C) comments THFh/s -106 66 only C eo’ is soluble acetone-<4 -94 56 reaction with C eo^’ (CH3CH2>3N -115 89 Ceo^' is insoluble (CH3)2CHNH2 -101 33 C eo^' is insoluble CH3CH2CN -93 97 aU three are soluble 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. An NMR tube containing frozen propionitrile will crack immediately afrer removed from liquid Nz, probably due to the dramatic thermal expansion of melting CH3CH2CN. For this reason, unlike the usual way o f torch sealing which requires the solutions to be frozen, the NMR samples were sealed in tubes with Young valves and then with paraffin wax. The upper limit of the temperature is about 40°C since the wax starts to melt around 45 to 50® C. Unlike in the solid state, the NMR experiments are carried out in a temperature range limited by the solvent. All the results are presented in Figure n.2 to n.5 (the data are listed in Table n.5 in the experimental section). For all three fuUeride salts, the isotropic shifts vary linearly with HT within the range of the experimental temperatures. The data of Ceo' and Ceo' show a Curie law behavior, indicating that each of these friUerides has only one accessible spin state in the temperature range attempted. However, their Ô û o values will not be zero at infinite temperature (l/T = 0), i.e., the intercepts are not zero. Neutral C eo is probably not an ideal diamagnetic reference for aU these fuUerides. The "C chemical shift of C e o ^ * show a quite different dependence on temperature. The Curie plot shows a negative slope, i.e., its isotropic shift slightly decreases with decreasing temperature. This is typical for a system o f more than one thermaUy accessible electronic state with differing spin multiplicities. This result supports our earlier conclusion that the discrete Cgo^' has very close-lying singlet and triplet states and that the singlet is the lowest in energy.^ 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 6 iso = 6.73x1 OVr + 23 O O 40 2.5 3.5 4.5 5.5 6.5 1 0 0 0 /r ( iC ) Figure IL2 The tem perature dependence o f the isotropic shift o f [N a(crow n)- (TH F)2l [ C 6o1 in TH F-d^. 55 5.49x10^/7 + 27 ^ i s o 50 I j 45 2.5 3.5 4.5 1 0 0 0 /r ( K - ’ ) Figure 0 .3 The tem perature dependence o f the "C isotropic shift o f [N a(crow n)- (TH F)2^ [C 6o * ] in propionitrile. 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5iso = 3.56x10^/7 + 43 58 I a. J 55 52 2.5 3.5 5.5 1000/7 (K ’ ) Figure IL4 The temperature dependence of the * ^ C isotropic shift o f [Na(crown)- (THF)2"l3[C6o^'] in propionitrile. 40.5 5 iso -507/7 + 41.! I J 39.5 38.5 2.5 4.5 3.5 5.5 1000/7 O C ’ ) Figure n.5 The temperature dependence of the isotropic shift o f [Na(crown)- (THF)2l 2[C6o ^ T in propionitrile. 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IL 1.2.3. Supplement: the Related Background Knowledge of NMR of Paramagnetic Species^ Compared to the corresponding diamagnetic resonances, paramagnetic NMR spectra show three notable unique features: the considerable line broadening, large isotropic shifts and frequently, marked temperature dependence In solution studies, the so-called “isotropic shift” is the difference between the observed shifts for the paramagnetic species and for a suitable diamagnetic molecule of the same chemical structure: S iso 5ob« - b d û The isotropic shift includes two contributions, the contact (Fermi) interaction and the dipolar (pseudocontact) interaction, i.e.. S i* , = S c o n + S d jp . The contact shift results from the interaction (through contact) of the nuclear moment with the electric currents arising from the unpaired electron density in s atomic orbitals. This interaction is scalar and isotropic. For a doublet system (electron spin S = 1/ 2 ), the nuclear spin will couple to the spin of the electron and the resonance spectrum is anticipated to show a doublet. The splitting between the two lines of this doublet, a, is the hyperfine coupling constant. However, due to the rapid relaxation of the electron (short 7\e), these two lines will collapse to a single signal. Since the two electron spin levels (Sz = ±1/2) in an external magnetic field are not equally populated, the collapsed NMR line will be at a weighted averaged position and will be shifted from the midpoint o f the doublet. For a simple system o f only one accessible electronic spin 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. State with an isotropic g tensor, the relationship o f contact shift with other factors is expressed as S = _ “ “ A hy^kT where g = 2.002322 for a free electron, p is the Bohr magneton, is the magnetogyric ratio o f the nucleus, and T is the absolute temperature, and h = h/2n. The hyperfine coupling constant between the nucleus and the electron is a = ^ y a tig p |v|/(0)(^ where lv(0)p is the probability of finding the electron within the nucleus. Only for the electrons in s orbitals, |m/(0)|^ ^ 0. Thus only from the spin in s orbitals can isotropic contact interaction arise. From the equation, we know that the hyperfine coupling constant a is proportional to the unpaired electron density on the nucleus of concern. Generally, for a system o f the total electron spin S >1/2, ^ _ A g P S ( ^ S ^ \ ± û c o n 3 h y ^ k T Here, A is the hyperfine coupling constant for a many-electron system. Thus, in simple, well-behaved systems, contact shifts are expected to vary linearly with MT and extrapolate to zero at infinite temperature (l/T = 0), i.e., they will show Curie behavior. Dipolar shift results from through-space dipolar coupling of the nuclear and electron magnetic moments. It depends on the reciprocal o f the third power of the 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dipole-dipole distance and on the angle between the vector connecting the two dipoles and the direction of the external magnetic field. The dipolar interaction only exists in a system with significant magnetic anisotropy and will vanish in an isotropic one. The ideal dipolar interaction also obeys Curie law, but deviations are seen in the systems with unusual magnetic behaviors. The second-order Zeeman efiect or excited- state contributions will result in non-zero intercepts and the hindered rotation of substituent groups or zero-field splitting contributes in curvature. II. 1.3. Conclusion The discrete fuUeride salts o f Ceo', C < s o ^ * and Ceo^' have been characterized by NMR spectroscopy in DMSO-<4 A U the ^ ^ C peaks are broad and the chemical shifts were determined to be 186, 183 and 197 ppm, respectively. Variable temperature NMR experiments have been conducted in propionitrile. The "C chemical shifts o f aU these three fuUerides vary Unearly with l/T. For Ceo ' and Ceo^', Curie law behaviors are observed although the intercepts for isotropic shifts versus l / T are not zero, suggesting that both Ceo' and C go^' have isolated spin states within the range of the experimental temperatures. WhUe for C e o ^ * , the chemical shift decreases sUghtly with increasing temperature, Ulustrating that the singlet and triplet spin states are in close proximity, but the singlet state is sUghtly lower in energy. 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. n.2. Reactivities of with A ll^ l Halides n.2.1. Introduction In 1991, Wudl et al. reported the synthesis o f the diphenyl fulleroid PhzCsi (reassigned to diphenylmethanofUllerene later^) by the reaction o f diphenyldiazo- methane (PhzCNz) with Ceo-^ That was the first organic C so derivative to be isolated. The dihydrofidleroid H2Q 1 has also been prepared in a similar way by reacting diazomethane (CH2CN2) with C eo via a dipolar addition to 6:6 double bond to give (CH2N 2)C6o , which then undergo thermolysis (N2 loss) and a rearrangement to form H2C6 i-^ Its H NMR spectrum shows two doublets (at 2.87 and 6.35 ppm) and based on this its structure is proposed to be 5,6-open, i.e., the addition is over the juncture between a five- and a six-membered ring and the C-C bond is cleaved by this addition. Upon photolyzing (CH2N2)C6o , a mixture of dihydrofulleroid and its isomer the methanofullerene is generated.^” The isolated methanofiillerene (H2C6 1) shows a singlet H NMR signal (at 3.93 ppm) and 16 “ C peaks of fiillerene sp^ carbons. On the basis o f these observations, it is proposed to have a 6,6-closed structure, i.e., the methylene group is added across a 6:6 fullerene double bond. It has been found that the substituted fiilleroids can be converted to the corresponding methanofuHerenes,^ ^ indicating that the latter are probably the thermodynamic products. However, no interconversion between these two isomers can be achieved thermally^®’ ^ ^ or photochemically.^® 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Due to its electron-deficient nature, neutral Cso usually behaves as an electrophile. In contrast, fuUerides (Cso'^ are considered as bases^^ and can be protonated with acids/^ They promote some base-catalyzed organic reactions by deprotonating the organic substances.^^*^^ One of the most typical reactions of fuUerides is with organic halides. The C g o polyanions, generated in situ with excess Li metal, react with CH3I to form (CH3)bC 6o (n < 24).^* This is one of the earliest reactions leading to fuUerene fUnctionlization. The reactions of allqrl halides with Cso^' are the most interesting and th^r fi-equently afford diallqrlated fidlerenes. The reaction of electrochemicaUy generated Ceo^' with excess CH3I leads to the production of a mixture o f isomeric 1,2- and l,4-(CH3>2C6o.^* With the electrochemical method, the formation o f fuUerene adducts by reacting Cso^' with aU cyl monoiodides and aryl diiodides, or by C s o ^ ~ with aryl monoiodides can be detected.^^ The reaction o f Cso^' with alkyl haUdes (RX) is beUeved to be comprised of two steps.^* Since the negative charges o f Ceo^' are delocalized over the whole fuUeride cage, the density on each carbon atom wiU be very low. As a result, the nucleophilic reactivity wiU be weak. The first reaction step is probably via an electron-transfer route^^ and it should be the rate determining step. The subsequent rapid coupling between C eo*' and R * in the radical pair wiU generate monoadduct intermediate R C go Based on theoretical calculations, the negative charge on RCeo is rather localized and its density is highest at the position on C(2), foUowed by C(4). For instance, in AMI molecular orbital calculations, when R = *Bu,^ or MesSiOC-,*^ the charge densities 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. are C(2), -0.29; C(4), -0.15 or -0.14 and are substantially lower on other carbons. Therefore, the second step reaction, i.e., RCeo* with RX, is probably a nucleophilic (5 n2 ) one, and occurs mostly at the C(2) position, or, in case of large steric repulsion, at C(4). RCeo’ has been prepared via the nucleophilic attack o f R (R = organoUthium,"“ ^ Grignard,^*^^ RR'(H3B)PL0 on neutral The reactivities o f RCgo have been extensively studied and the typical reactions include reversible protonation,^^ with organic halides CHsCHal,'*’ PhCHzBr,^ dichloroacetylene^), formaldehyde gas (CHzO),^ benzoyl chloride (PhCOCl),'*^ and tropylium cation These strongly support the nucleophilic reactivities of RCeo*, the intermediate of the reaction o f Ceo^' and RX. In our group, C eo” " (n = 1, 2, 3, 4) have been synthesized as analytically pure discrete salts.^ While most of the functionalization of C eo is done via covalent organic methods,^ we wanted to explore the possibility of using the reactivities of C eo fuUerides as a way to functionalize Ceo We were particularly interested in the reaction o f Ceo^ with alkyl <&halides (RR'CXa) as an alternate route to prepare methanofuUerenes or fUUeroids. After the first step electron transfer and the subsequent coupling of the radicals, the intermediate RR'C(X)C6o* is expected to undergo an intramolecular nucleophilic reaction to form methanofiUlerenes. We note that similar intermediates generated by the nucleophilic attack of a-halocarboanions^^ or phosphonium^*’ or sulfonium ylides^’ on C eo undergo such an intramolecular reaction to produce substituted methanofuUerenes. 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 1,2.2. Results and Discussion The alkyl dihalides RR'CXz used for the reaction with Ceo^' were CH2I2 and Ph2CCl2 . CH2I2 is light sensitive and will decompose under the ambient light. The manipulation o f small amounts of CH2I2 is difficult and in practice it was usually used in excess. Upon adding CH2I2 into C e o ^ * , the color o f the solution changed from red- brown to brown. During the reaction, a precipitation resulted, suggesting low solubility of the product. In the NMR spectra, no fullerene peaks could be observed. For the ‘H NMR spectra, there were frequently many peaks in the region around the resonance o f the expected product (methanofullerene). These peaks may arise from some organic impurities, which are always present in neutral fullerene species. A sample of purified product was redissolved and its NMR spectrum (Figure n.6) shows two peaks (at 3.03 and 1.34 ppm) similar to that of the product of C e o ^ * and CH3I. The products are tentatively assigned to the mixture of 1,2- and 1,4- (ICH2)2C6o - Unfortunately, no further evidence could be obtained to confirm the structures, due to the extremely low solubility of the product. Based on the mechanism proposed for the reaction between C go^' and alkyl halides,^* the second step of the reaction between Cgo^' and CH2I2, i.e., the reaction of intermediate (ICH 2)C6o', can have two paths in competition, intermolecular and intramolecular. The intermolecular one is the 5^2 reaction o f (ICH2)C6o * with another 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.5 3.0 2.5 2.0 1.5 1.0 p p m Figure BL6 H NMR spectrum of the product of the reaction of C e o ^ * with CH2I2 (in toluene-^/g, the peak at 2.09 ppm is from the solvent). CH2I2 molecule, while the intramolecular one is the formation of methanofullerene by the cleavage o f the C-I bond. In our hands, the intermolecular one seems to be favored in the two competition reactions, probably due to the large excess of CH2I2 In order to conduct the reaction in a more manageable way, we decide switch our attention to the dihalides Ph2CCl2. Compared to CH2I2, it is more stable and with a larger molecular weight, is more suitable for a quantitative reaction. The product is expected to be more soluble than that from CH2I2 In practice, one equivalent of PhzCClz was added slowly into a Ceo^' solution, to enhance the intramolecular product ratio in the second step reaction. The NMR spectrum shows two multiplet peaks at 3 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.7 and 7.3 ppm, which are assigned to the expected product diphenylmethanofUllerene (PhzCCeo). In contrast to its first report,^* PhjCCgo in CS2 shows 16 NMR peaks of fullerene sp^ carbon, which is consistent with its expected Czv symmetry. Both the H and C NMR spectra of PhzCCeo show some extra peaks with relatively low intensity around the expected phenyl signal region. These are ascribed to the intermolecular product, most likely, 1,4-(Phz(Cl)C)2C6o The TLC results of the reaction mixture are listed in Table I I 5 For comparison, the results of the mixed product in C6o(CH2)2CPh3 preparation are also included. Clearly, the possible presence of small amount of intermolecular reaction product indicated by the TLC results is consistent with our NMR data. After our work was begun, similar experiments conducted on the reactivities of Ceo^' with organic dihalides by other showed similar results.^^ Table IL4 TLC results for the reaction mixture o f C e o ^ * and Ph2CCl2 (solvent; toluene/hexane 1:4) C spot mixture for PhzCCeo mixture for Ceo(CH3)2CPb 3 Rr assignment R t assignment 1 0.89 C eo 0.89 Ceo 2 0.58 Ph2CCeo 0.51 Ceo(CH2>2CPh3 3 0.26 (Ph2(Cl)C)2Ceo 0.17 bis-adducts 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. II.2.3. Conclusion The reactivity o f with alkyl <£halides ( R 2 C X 2 ) has been studied. According to the previous investigation, the reaction in the first step via electron transfer will lead to the formation of the intermediate (R2C(X)C6o')- In the second step, the competition between intra- and intermolecular nucleophilic reactions is noticed. n.3. Experimental All manipulations of air- and moisture-sensitive compounds were carried out in an inert atmosphere in a Vacuum Atmospheres Drybox under helium (H2O, O2 < 1 ppm). C e o was purchased firom MER Corp. and used without any further purification. Other chemicals were purchased fi’ om Aldrich. [Na(crown)(THF>2]nC 6o “ ’ and (PP>r)„C6o"*(n = 1, 2, 3) were prepared according to the published methods.^® In the drybox, D M S O -< 4 and C D 3 C N were passed through activated alumina right after opening and were stored over 3A molecular sieves. Propionitrile ( C H 3 C H 2 C N ) was gently refiuxed over C a H 2 overnight and distilled both outside and inside the drybox and then passed through activated alumina. THF-c/g was stirred in drybox with N a / K overnight at room temperature until a faint blue color appeared and then distilled. Acetone-tik was distilled after it was kept stirring with K 2 C O 3 while gently warmed for several hrs. Et3N was refiuxed overnight and distilled, in turn, fi^om C a H 2 outside and fi-om N a / K alloy inside the drybox. Isopropylamine ( C H 3 ) 2 C H N H 2 , 3 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I after being refiuxed with LiAlH» overnight, was repeatedly distilled and degassed (freeze/pump/thaw) and then brought into the glovebox. THF was gently refiuxed overnight with sodium benzophenone ketyl and distilled outside and then inside the glovebox. Vis-NIR spectra for integrity checks were recorded on a Shimadzu UV-260 spectrophotometer. Solid state NMR spectra were obtained at room temperature on a Chemagnetics CMX-300 spectrometer in the Department o f Chemistry, Pennsylvania State University. Solution NMR spectra were recorded on a Bruker AC 250, AM-360 or AMX-SOO-MHz spectrometer. A variable temperature unit was used when the NMR experiments were conducted at a temperature other than room temperature. The solid state NMR samples were sealed in glass rotor inserts designed by V^lmad Glass Co. to fit the zirconia Chemagnetics rotors. The solution samples were sealed with a torch in NMR tubes under vacuum on a Schlenk line or in NMR tubes sealed with Young valves and then with paraffin wax. The chemical shifts were calibrated against internal solvents and in the case o f propionitrile as the solvent, to sealed acetone-(4 capUlaries. The temperatures were calibrated with 4% MeOH in MeOH-< /4 when lower than room temperature or 80% glycol in DMSO-dk when higher than room temperature. All the ^ ^ C chemical shifts data of [Na(crown)(THF)2]m C 6o " (n = 1, 2, 3) salts at variable temperatures are listed in Table II S. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table II.5 NMR chemical shifts o f [Na(crown)(THF)2" ]m C 6o"' at variable temperatures Ceo in THF-dg C 60 in CH 3CH 2CN r(K ) 5 (ppm) r(K ) 5 (ppm) 318 187 318 187 301 188 298 188 258 192 269 190 224 195 250 192 196 200 228 194 177 204 207 196 Ceo' in C H 3C H 2CN in CH 3C H 2CN rCK) 5 (ppm) r(K ) 8 (ppm) 318 183.3 318 197 297 183.2 298 198 269 182.9 269 199 250 182.8 249 201 229 182.6 226 202 208 182.4 205 203 187 182.2 References and Notes 1 (a) Taylor, R.; Hare, J.; Abdul-Sada, A. K ; Kroto, H. W. J. Chem. Soc., Chem. Commun. 1990, 1423; (b) Ajie, H ; Alvarez, M. M.; Anz, S. J.; Beck, R. D ; Diederich, F.; Fostiropoulos, K ; Huffinan, D. R,; Kratschmer, W.; Rubin, Y.; Schriver, K. E ; Sensharma, D ; Whetten, R. L J. Phys. Chem. 1990, 94, 8630; (c) Johnson, R. D ; Meyer, G.; Bethune, D S. J. Am. Chem. Soc. 1990, 112, 8983. 2. (a) Yannoni, C. S.; Johnson, R. D.; Meijer, G ; Bethune, D. S.; Salem, J R. J. Phys. Chem. 1991, 95, 9; (b) Tycko, R.; Haddon, R. C ; Dabbagh, G ; Glarum, S. H.; Douglass, D C ; Mujsce, A. M J. P lys. Chem. 1991, 95, 518. 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3. Tycko, R.; Dabbagh, G.; Fleming, R. M.; Haddon, R. C ; Makhija, A V.; Zahurak, S. M Phys. Rev. Lett. 1991, 67, 1886. 4. (a) Murphy, D W ; Rosseinsky, M. J.; Fleming, R _ M.; Tycko, R.; Ramirez, A. P.; Haddon, R. C ; Siegrist, T.; Dabbagh, G.; TuUy, J. C.; Walstedt, R E. J. Phys. Chem. Solids 1992, 53, 1321; (b) Tycko, R. J. Phys. Chem. Solids 1993, 54, 1713; (c) Pennington, C. H ; Stenger, V. A. Rev. M od Phys. 1996, 68, 855. 5. Tycko, R.; Dabbagh, G ; Rosseinsky, M. J.; Murphy, D. W ; Fleming, R. M ; Ramirez, A. P.; Tully, J. C. Sceince 1991, 253, 884. 6. For example; (a) Pekker, S.; Janossy, A.; \fihlay, L ; Chauvet, O ; Carrard, M.; Forro, L. Science 1994, 265, 1077; (b) Martin, M C ; KoUer, D ; Rosenberg, A. Kendziora, C ; Nfihaly, L. Phys. Rev. B 1995, 51, 3210; (c) Zhu, Q ; Cox, D. E. Fischer, J. E P l^s. Rev. B 1995, 51, 3966. 7. (a) Kalber, T.; Zimmer, G ; Mehring, M. Z. Phys. B 1995, 97, 1; (b) Thier, K.-F.; Zimmer, G ; Mehring, M; Rachdi, F. Phys. Rev. B 1996, 53, R496; (c) Thier, K - F.; Mehring, M.; Rachdi, F. Phys. Rev. B 1997, 55, 124. 8. Douthwaite, R E.; Brough, A. R.; Green, M. L H J. Chem. Soc.. Chem. Commun. 1994, 267. 9. Chen, J.; Huang, Z.-E.; Cai, R. F.; Shao, Q -F ; Chen, S.-M ; Ye, H.-J. J. Chem. Soc., Chem. Commun. 1994, 2177. 10. Chen, J ; Cai, R.-F.; Huang, Z.-E.; Shao, Q -F ; Chen, S.-M. Solid State Commun. 1995, 95, 239. 11. Douthwaite, R. E.; Green, M. A.; Green, M L H.; Rosseinsky, M. J. J. M ater. Chem. 1996, 6, 1913. 12. Chen, J ; Huang, Z.-E.; Cai, R.-F.; Shao, Q -F ; Ye, H -J Solid State Commun. 1995, 95, 233. 13. Reichenbach, J.; Rachdi, F.; Luk’yanchuk, I ; Ribet, M.; Zimmer, G ; Mehring, M. J. Chem. Phys. 1994, 101, 4585. 14. Tycko, R.; Dabbagh, G ; Rosseinslqr, M J ; Murphy, D. W.; Ramirez, A. P.; Fleming, R. M. Phys. Rev. Lett. 1992, 68, 1912. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15. Zimmer, G.; Helmie, M.; Mehring, M.; Rachdi, F. Europhys. Lett. 1994, 27, 543. 16. Holczer, K.; Klein, O.; Alloul, H.; Yoshinari, Y.; Hippeit, P.; Huang, S.-M.; Kaner, R. B ; Whetten, R. L. Europhys. Lett. 1993, 23, 63. 17. Tycko, R.; Dabbagh, G.; Murphy, D. W ; Zhu, Q.; Fischer, J. E. Phys. Rev. B 1993, 48, 9097. 18. Chen, J ; Shao, Q.-F.; Huang, Z.-E.; Cai, R.-F ; Chen, S.-M. Chem. Phys. Lett. 1995, 235, 570. 19. Bausch, J. W.; Prakash, G. K. S.; Olah, G. A. Tse, D. S.; Lorents, D. C ; Bae, Y. K.; Malhotra, R. J. Am. Chem. Soc. 1991,113, 3205. 20. (a) Stinchcombe, J ; Penicaud, A.; Bhyrappa, P.; Boyd, P. D. W ; Reed, C. A. J. Am. Chem. Soc. 1993, 1 /5 , 5212; (b) Bhyrappa, P.; Paul, P.; Stinchcombe, J.; Boyd, P. D. W ; Reed, C. A. J. Am. Chem. Soc. 1993, 115, 11004. 21. (a) Chen, J.; Cai, F.-F ; Shao, Q.-F.; Huang, Z.-E.; Chen, S.-M. Chem. Commun. 1996, 1111; (b) Wu, M ; Wei, X.; Xu, Z Tetrahedron Lett. 1996, 37, 7409. 22. Boyd, P. D. W ; Bhyrappa, P.; Paul, P.; Stinchcombe, J.; Bolskar, R. D ; Sun, Y ; Reed, C. A J. Am. Chem. Soc. 1995, 117, 2907. 23. Pasquarello, A.; Schlüter, M.; Haddon, R. C. Phys. Rev. A 1993, 47, 1783. 24. Breitmaier, E.; Voelter, W. Carbon-13 NMR Spectroscopy: High Resolution M ethods and Application in Organic cmd Biochemistry VCH: New York, 3rd ed., 1987, pp. 110-111. 25. Haddon, R. C. Acc. Chem. Res. 1992, 25, 127. 26. lesson, J. P. in NMR o f Paramagnetic Molecules: Principles and Applications La Mar, G. N.; Horrocks Jr., W. DeW ; Holm, R. H. eds. Academic: New York, 1973. 27. Isaacs, L ; Diederich, F. Helv. Chim. Acta 1993, 76, 2454. 28. Suzuki, T.; Li, Q ; Khemani, K. C ; Wudl, F.; Almarsson, O. Science 1991, 254, 1186. 29. Suzuki, T.; Li, Q ; Khemani, K. C ; Wudl, F. J. Am. Chem. Soc. 1992, 114, 7301. 4 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30. Smith HI, A. B ; Strongin, R. M.; Brard, L ; Furst, G. T.; Romanow, W. J.; Owens, K. G.; King, R. C J. Am. Chem. Soc. 1993, 115, 5829. 31. The conversion can be achieved by several ways: thermally: (a) Prato, M.; Lucchini, V.; Maggini, M.; Stimpfl, E ; Scorrano, G ; Eiermann, M.; Suzuki, T.; Wudl, F. J. Am. Chem. Soc. 1993, 115, 8479; (b) Smith m , A. B ; Strongin, R. M.; Brard, L.; Furst, G. T.; Romanow, W. J.; Owens, K. G ; Goldschmidt, R. J.; King, R- C. y. Am. Chem. Soc. 1995, 117, 5492; photochemicaily: (c) Janssen, R. A. J.; Hummelen, J. C ; Wudl, F. J. Am. Chem. Soc. 1995, 117, 544 electrochemicaUy: (d) Arias, F.; Xie, Q ; Wu, Y.; Lu, Q ; Wilson, S. R. Echegoyen, L. J. Am. Chem. Soc. 1994, 116, 6388; (e) Eiermann, M.; Wudl, F. Prato, M.; Maggini, M. J. Am. Chem. Soc. 1994, 116, 8364; acid catalyzed: (f) Gonzalez, R.; Hummelen, J. C ; Wudl, F J. Org. Chem. 1995, 60, 2618. 32. Diederrich, F.; Isaacs, L ; Philp, D. J. Chem. Soc., Perkin Trans. 2 1994, 391. 33. Niyazymbetov, M. E.; Evans, D. H. J. Electrochem. Soc. 1995, 142, 2655. 34. (a) CliSel, D ; Bard, A. J. J. Phys. Chem. 1994, 98, 8140; (b) Meier, M. S.; Corbin, P. S.; Vance, V. K.; Clayton, M.; Mollman, M. Tetrahedron Lett. 1994, 35, 5789. 35. Eastman, M. P.; Wyse, C. L.; Abe, J. P.; Zoellner, R. W ; Kooser, R. G. J. Org. Chem. 1994, 59, 7128. 36. Caron, C ; Subramanian, R.; D'Souza, F.; Kim, J ; Kutner, W ; Jones, M. T.; Kadish, K. M. J. Am. Chem. Soc. 1993, 115, 8505. 37. Mangold, K -M ; Kutner, W.; Dunsch, L.; Frohner, J. Synth. M et. 1996, 77, 73. 38. Subramanian, R_; Kadish, K. M.; Vijayashree, M N.; Gao, X.; Jones, M. T.; MUler, M. D.; Krause, K. L.; Suenobu, T.; Fukuzumi, S. J. Phys. Chem. 1996, 100, 16327. 39. Huang, Y.; Wayner, D. D. M. J. Am. Chem. Soc. 1993,115, 367. 40. Hirsch, A.; Soi, A ; Karfunkel, H. R. Angew. Chem., Int. E d Engl. 1992, 31, 766. 41. Komatsu, K ; Murata, Y ; Takimoto, N.; Mori, S., Sugita, N.; Wan, T. S. M J. Org. Chem. 1994, 59, 6101. 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 42. (a) Fagan, P. J.; Knisic, P. J.; Evans, D. H.; Lerke, S. A.; Johnston, E. J. Am. Chem. Soc. 1992, 114, 9697; (b) IBrsch, A.; Grosser, T.; Soi, A. Chem. Ber. 1993, 126, 1061. 43. Keshavarz-K, M.; Knight, B ; Srdanov, G ; Wudl, P. J. Am. Chem. Soc. 1995, 117, 11371. 44. Yamago, S.; Yanagawa, M ; Nakamura, E. J. Chem. Soc., Chem. Commun. 1994, 2093. 45. Murata, Y ; Motoyama, K.; Komatsu, K.; Wan, T. S. M. Tetrahedron 1996, 52, 5077. 46. Timmerman, P.; Anderson, H. L.; Faust, R_; Merengarten, J.-F.; Habicher, T.; Seiler, P.; Diederich, F. Tetrahedron 1996, 52, 4925. 47. Kitagawa, T.; Tanaka, T.; Takata, Y ; Takeuchi, K. J. Org. Chem. 1995, 60, 1490. 48. Diederich, F.; Thilgen, C. Science 1996, 271, 317. 49. (a) Bingel, C. Chem. Ber. 1993, 126, 1957; (b) Benito, A. M.; Darwish, A. D ; Kroto, H. W.; Meidine, M. F.; Taylor, R.; Walton, D. R. M. Tetrahedron Lett. 1996, 37, 1085. 50. Bestmann, H. J.; Hadawi, D ; Rôder, T.; Moll, C. Tetrahedron Lett. 1994, 35, 9017. 51. Wang, Y ; Cao, J.; Schuster, D. I.; Wilson, S. R. Tetrahedron Lett. 1995, 36, 6843. 52. Boulas, P. L.; Zuo, Y ; Echegoyen, L. Chem. Commun. 1996, 1547. 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter m. Synthesis and Isolation of Discrete €50^ Salts III.1. Introduction in. 1.1. Cgo^amd Its Electronic Structure Highly negatively charged species are usually not very stable in solution under ambient conditions and very few tetra-anionic ones have been reported/ Ceo, as indicated by theoretical results (Figure 1.3), could accept up to six or even twelve electrons. Experimentally, C e o " * * was hrst detected in 1991 by cyclic voltammetry in solution^ and was re-confirmed later in the same way.^ The preparation of C eo '* ' in the solid state as the intercalated alkali-metal fuUeride (MtCeo, M = K, Rb, Cs) foUowed."* In solution, C e o '* * was generated by electrolysis and its vis-NIR spectrum was directly recorded spectroelectrochemicaUy.^ ChemicaUy it was prepared with rubidium in liquid ammonia at low temperature.^ However, no isolation and no characterization other than vis-NIR have been reported. A third report o f Ceo'*' in solution was by consecutive reduction o f Ceo in THF upon contact with potassium mirror.^ It was stated that Ceo'*' is diamagnetic based on a very brief EPR result, but the vis-NIR is inconsistent with the other two reports. TheoreticaUy, two singlet states and one triplet state are “expected” to have similar energies, although no careful calculations have been undertaken.’ The nature o f the fuUeride ion is of interest for a number of reasons. For MtCfio, the four electrons occupy the three-fold degenerate LUMO band and according 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to the band theory, M ^C go compounds should be metallic. The Korringa relaxation mechanism predicts that 7\T (7\ is spin-lattice relaxation time, and T is temperature) should be constant for a metal. In this respect, MtCeo phases are not metallic,^ because for them 7\T is strongly dependent on temperature. Other experimental results also reveal their semiconducting or insulating behavior. Overall, the nature of the electronic properties o f C e o " * * remains not well understood. For neutral C eo and all the fuUerides, C eo” * and Ceo^^^ (n = 0 to 6) can be correlated with each other as “electron-hole pairs”. In Ceo”* , n electrons fiU the riu orbital whereas in C e o ^ * " ”^ there are n “holes”. Since C eo ^’ and C e o * * , the electron-hole equivalents o f Ceo* and Ceo respectively, are foreseen to be much more unstable, Ceo^* and Ceo^ would be the first accessible o f such fuUeride pairs. The electronic structure o f C e o ^ * stiU remains controversial. For better exploring C e o * * * and its relationship with C e o ^ * , we decided to prepare discrete C eo” * * in an analyticaUy pure fi’ om. H L l.2. A ttem pt of Preparing Discrete C«o^ In our group, C g o " * (n = 1, 2, 3) have been synthesized using sodium metal with an addition o f stoichiometric amount o f a crown ether (dibenzo-18-crown-6) and they have been isolated as analyticaUy pure salts. The efifort to extend this series one step higher usuaUy led to the mixture o f C e o ^ * and Cm^, even though more vigorous reaction conditions had been tried: (1) larger amount of compledng reagent, up to 6 equivalents o f crown ether; 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (2) longer reaction time, up to 65 hours; (3) more powerful reducing agent, potassium; (4) 1,4-dioxanes as solvent (boiling point 100°C) to ensure higher reaction temperature. Cw s 450 611 774 923 1060 Wavelength (nm) Figure n L l A mixture of Ceo^ and Ceo^ in DMSO. The reactions were followed with vis-NIR spectroscopy since C w ' and C e o * ' show very characteristic absorbances at 780 nm and 730 nm, respectively (Figure m . 1). Although the purity o f the solvent later proved crucial for obtaining good C so *' spectra, the presence of Cso^ in all the attempted Ceo^ products then made us think that under the above conditions C e o * " was thermodynamically disfavored. 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Na + L + Qso^ ■ " ■ (NaL)^ + Qo^ Scheme IIL l The equilibrium between C go^' and C eo Na + Q o ^ " * Na^(soIv) + Cgo^ Na^(soIv) + L ' (N a L f Scheme HL2 Two sub-equilibria in C^*~ synthesis. The equilibrium between C^a' and C eo” * * in the presence o f complexing agent (Scheme m .1) can be considered to consist of two sub-equilibria (Scheme ni.2), K = Kt K,. While Kt is unalterable under our reaction conditions, we could increase K by choosing a complexing ligand L of larger stability constant A T , . By this means, we hoped to obtain pure Ceo^ Obviously, the complexation of alkali metal with the ligand here acts as a driving force for formation. nLl.3. Cryptand and Crown Ether Compared to crown ethers, the bicyclic analogue cryptands (Figure ni.2) have much larger A T , (Table m .l) for binding alkali metal. This is the so-called 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. macrobicyclic or cryptate eflfect.*^ Using cryptand[2.2.2] in place o f the crown ether in the C eo'*' synthesis, the equilibrium shown in Scheme m .l will be shifted to the right and C e o ^ * " will be more favored thermodynamically. O C ° o o C L JO Figure HL2 Dibenzo-18-crown-6 and cryptand[2.2.2]. Table IIL l Stability constant (logÆ,) in MeOH at 25°C dibenzo-18-crown-6 4.4 5.1 cryptand[2.2.2] 9.0 9.8* Data source: Fenton, D. E. in Comprehensive Coordination Chemistry Wilkinson, G. ed.; Pergamon: Oxford, 1987, vol. 3, pp. 1-80. * in 95.5% methanol 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The complexation o f cryptand[2.2.2] with the alkali metal cation is so strong that it can force Na or K atom to loose one electron to form “sodide” (NaX222-Na‘ )^‘ * or electride (K C222 e") ions.*’ Therefore, in the presence o f cryptand, Na or K is thermodynamically a very powerful reducing agent. Kinetically, the free electron, or that which residues on an electropositive Na atom, should readily transfer to a less electropositive species. Thus, Na or K with addition o f cryptand[2.2.2] will be a good reductant for C e o * * * preparation.** III.2. Results and Discussion IIL2.1. Synthesis and Isolation In the glovebox under an inert atmosphere, the synthesis of C e o * ' is done in the similar method o f preparing (n = 1 , 2, 3). Stirring a suspension of C e o in THF gently warmed with excess sodium or potassium and about 4.2 equivalents of cryptand[2.2.2] leads to precipitation of the fuUeride salts [Na(crypt)]4 [C6o ‘ *l or [K(crypt)]4[C6o‘ ^ in good yield (>90%). A trace o f blue color appears immediately after addition o f sodium or potassium into cryptand solution in THF, indicating the transient formation of solvated electrons.*^ This blue color disappears right away and the solution turns reddish-purple while the black crystalline C g o dissolves gradually, showing that C eo is transformed to Ceo'. A brown precipitate forms and the color of the suspension gets darker increasingly within one hour. C eo is then reduced stepwise to Ceo^', C eo^', and finally to 5 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C eo '* ' salt. These three fUUerides are insoluble in THF and no visible changes could be observed from the conversion of C eo^' to Ceo^. The time needed for the completion of Ceo^ formation varies with several factors, one of which is the solvent purity. With freshly distilled solvent, overnight is normally sufBcient while with THF stored in the box for three to four weeks, the reaction usually takes 20 to 40 h. The final product is a brown, microcrystalline solid and the reaction is essentially quantitative (> 90% yield). The solvent (THF) for the reaction is distilled from sodium benzophenone ketyl. In order to remove any traces of benzophenone left, it is distilled once more from Na/K alloy. Under our reaction conditions, potassium becomes dispersed into fine particles after the reaction since the temperature (40-60°C) is close to its melting point (64°C). In addition, the precipitate from potassium reduction is usually too fine and difficult to filter. These make the isolation of clean fuUeride salt almost impossible. With its melting point much higher (98°C), the chunks o f sodium combine rather than disperse during the warming and stirring period. Also, the product obtained from Na reduction is easier to filter with a medium fiit. The problem is that irregular shaped chunks wUl generate smaU metal particles and it is not easy to remove them. A freshly cut sodium chunk loses its metal luster quickly after it is put into aged THF, presumably due to the formation of a layer of NaOH on the metal surface as a result of the reaction o f sodium with trace water in the solvent. This layer wiU faU off graduaUy during the warming and stirring. To avoid the smaU Na particles and NaOH in the final 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. product, freshly cut sodium metal is pre-treated in freshly distilled THF (from Na/K alloy). This chunk will usually be polished to a shiny metal ball by stirring while warming overnight. After filtration, this ball and this clean THF is used for the synthesis. By all the above efforts, analytically pure fiilleride salt was finally obtained. Most o f the characterizations were carried out with the Na(crypt)"^ salt. UL2.2. Instability and Reactivi^ The instability o f the highly charged Ceo^ is its most characteristic property. This fuUeride salt is exceedingly air and moisture sensitive, oxidizing primarily to Ceo^'. A U manipulations have to be inside the glovebox. Working up under a heUum atmosphere with 3-5 ppm O2 level wiU cause C e o * ' to partiaUy oxidize. Therefore, immediately before each synthesis the drybox catalyst column was switched to a freshly regenerated one to ensure a low O2 level in the box. Isolated Ceo^ wax-sealed in a vial wiU degrade to C go^' within weeks. For this reason, all the samples for characterization were freshly prepared and were handled with extreme caution. This highly charged anionic species could be considered as a “super-reductant”. To dissolve it, the solvent has to be polar enough and be stable to the reduction. Before testing for Cgo^ aU solvents are purified outside and then inside the glovebox rigorously according to the literature methods," distiUed or vacuum distiUed and then passed through activated alumina. is essentiaUy insoluble in "BuNHa and ODCB In C H 3 C N , pyridine and HMPA, it appears to be slightly soluble but will graduaUy give 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ceo^' (in hours). All these three solvents seem not to be polar enough to dissolve C«o^ appreciably, but apparently, the solvents themselves or traces o f impurities in them oxidize the tetra-anionic species to the more soluble leading to the slow dissolution. In the more polar solvent DMF, Ceo^ dissolves readily, but gives only C eo ^ ’- Since the first vis-NIR spectrum of C 6o'^ was recorded in PhCN by the spectroelectrochemical method,’ much effort was made to explore it as a possible as a solvent for isolated Ceo^. PhCN was rigorously purified and with moderate solubility in it, nearly all was converted to C eo’* in 10 to 20 min. Obviously, Ceo^ can not exist in PhCN long enough for spectrophotometeric study. The cathodic potential limit of PhCN is -1.9 V vs. SC E *® while Em for C eo’'/Ceo^ is reported to be -2.35 V vs. Fc/Fc* (-2.01 V vs. SCE)’ and we conclude that thermodynamically C e o '* " is not stable in PhCN. DMSG has high polarity (dielectric constant 46, dipole moment 3.9) and in it Ceo^ dissolves readily to give a brown, relatively stable solution. In our hands, DMSG is the only solvent consistently usable for Ceo^ characterization. Its purity is crucial for obtaining good solution samples and it has to be rigorously dried and deo^genated immediately before use. It is noteworthy that this solvent will not be good after being kept in glovebox for a few weeks, even if well capped, probably due to the slow condensation and accumulation o f traces of water or other solvent vapors fi'om the box atmosphere. 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. All our experimental results show that Ceo'*' is only stable in “good” solvent for a limited time and even in DMSO it decays gradually at room temperature, oxidizing initially to Ceo^' and then to a dark greenish solution. In relatively dilute solution (absorbance at 730 nm between 0.1 to 2 in 1 cm cell, good for vis-NIR measurement), Ceo^will be partially oxidized to Ceo^ in no more than 30 min. In one extreme case, a very dilute Cgo^ solution was observed decaying quickly and after 5 min. only C e o ^ * was left. In concentrated solution (10‘ ^ to 1 0 * ^ M), it shows no sign of degradation in one to two days within the limits of vis-NIR detection (ca. 5%). integration also reveals that 99% of fuUeride is present in the NMR sample. The reason for Ceo'*' degradation in D M S O solution might be the foUowing; (1) Traces of impurities, e.g., water or other solvents. (2) Electron transfer from Ceo^ to DMSO to form C eo^' and DMSŒ radical anion or similar species. With a cathodic potential limit of -1.85 V vs. SCE,^° DMSO potentiaUy could oxidize Ceo^ (Ev2 for Ceo^'/Ceo^ is -2.01 V vs. SCE^). (3) Deprotonation of DMSO by C eo% to form CeoH^' and then Ceo^'. The pA. o f the conjugate acid CeoH^' is estimated by electrochemical methods to be 37,^^ while the p A T , o f DMSO is estimated to be 35.^ According to these estimations, as a base Ceo'*' should be powerful enough to remove a proton from DMSO. Based on aU our observations, the first degradation mechanism is regarded to be fast while the other two are relatively slow. The amount o f impurities in freshly distiUed DMSO is smaU and in concentrated solutions the percentage o f C eo * ' oxidized 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to C e o ^ * by these impurities will be low and may not be detectable by vis-NIR spectra. This is confirmed by the fact that no matter how “pure” a Ceo^ solution (i.e., a concentrated one) is by the vis-NIR criterion, its EPR spectra always show the presence of C « )^ * . While in dilute solution, the amount o f C e o ^ * derived fi-om impurities will be comparable to that of original C eo '* ' and will be easy to identify by vis-NIR spectra. While the instability o f Cgo^ observed in dilute solution is believed to be due to traces o f impurities in the solvent, the “long term” instability is ascribed to the second and/or third mechanism(s). Thermodynamically DMSO will be able to oxidize C e o '* ', but kinetically these processes may be slow. As a result, the concentrated Ceo^ solution in DMSO will be stable, but only for a limited time (one week). The investigation of C e o * ' stability in the presence of salts was initiated while attempting cation metathesis. Of all the salts tested, including °Bu4NCl, PPNCl, NaBPlu and NaCl, none is found to be “inert”. To different extents, they all speed up the decay process. While PPNCl may behave as an electron acceptor^ and therefore as an oxidant here, for the rest, it is difGcult to determine whether the oxidation is due to the salts themselves or some impurities in them. Using pyridine instead of THF as the solvent to synthesize C e o '* " will produce a dark blue mixture, indicating the formation of solvated electron. After the solvent is removed, solid C e o '* ’ remains stable for at least several months if wax-sealed and kept inside the glovebox. Compared to the degradation o f the isolated sample in weeks, the 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. solvated electron obviously plays a role o f stabilizer here. Stirring sodium metal in DMSO leads to its dissolution to form a yellow solution. This solution, presumably to be solvated electron or DMSO radical anion, is stable for several hours before turning green. Using this yellow liquid as a solvent for C e o '* ' gives longer stability to the solutions and no C c o ^ * can be detected. IIL2.3. Characterization As discussed above, Ceo^ is exceedingly vulnerable to oxidation. In order to ensure reliable results, all the containers for characterization are pre-rinsed with C e o '* ' solution. N m Fullerides are formed by filling electron(s) into Ceo fiu orbitals. The electronic transitions to the tig level are around leV in energy and the corresponding absorbances fall in vis-NIR region. The NIK spectra o f Ceo"' (n =1, 2, 3, 4) have been systematically investigated^ and it has found that their absorption bands are very characteristic for each species and can use as fingerprint for the fuUeride identification.Thus NIR spectroscopy also becomes a powerful tool for Ceo^ exploration and is useful for routine integrity checks, reaction progress inspection and stability investigations. For Ceo^ in DMSO in the range of 450 to 1100 nm (limited by the spectrometer), the main absorbance is at 730 nm, accompanied by SOO(sh), 902, and 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I I 450 610 773 923 1059 Wavelength (nm) Figure HL3 Vis-NIR spectrum of Ceo^ in DMSO. 1005 nm peaks (Figure m .3) In DMSO/Na solution from 450 to 1600 nm, C e o * * " shows two characteristic absorbances at 728 and 1195 nm (Figure in.4), as well as other peaks at 501(sh), 904, 1000, 1091(sh), and 1134(sh) nm. DMSO/Na solution is used as a solvent here, because, as discussed before (Section in.2.2), it gives stable solutions of Ceo^ for several hours. In addition, its absorption is below 450 nm and the vis-NIR region is essentially identical to that o f DMSO itself. Subtracting the solvent absorbance, the spectrum is fundamentally the same as in DMSO (Figure in.4, c). Overall, these spectra show no peaks at 780 or 1370 nm, indicating no detectable C eo^', and in DMSO/Na no Ceo*' absorbance at 952 n m .® S8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 450 700 950 1200 1450 Wavelength (nm) F igu re H L 4 V is-N IR spectrum o f [N a(crypt)^4[C6o ^ in DM SO /Na: (a) before subtraction; (b ) th e absorbance o f D M SO /N a; (c) after subtraction o f the ^ sorb an ce ofD M S O /N a. 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. î We note that two earlier reports of Ceo^ are in good agreement with our vis- NIR results, one generated by electrochemical method in benzonitrile (728, 1209 nm), and the other reduced by rubidium in liquid ammonia (714, 1176 nm). NMR As presented in Chapter H, for n = 1, 2 and 3, the C eo” * fullerides have broad, downfield-shifled resonances in their solution ‘ ^C NMR spectra. Interestingly, C e o * " salts also have similar results. As shown in Figure m 5, the fiilleride ‘^C NMR peak of [Na(crypt)]4 [C6o '* * ] is at 185 ppm in DMSO-<4 at room temperature. For [K(crypt)]4- [Ceo^, the fuUeride peak is at the same position, although the cryptand peaks are shifted a little. The chemical shift o f 185 ppm is essentiaUy the same as that o f C e o * (186 ppm) or C e o ^ * (183 ppm). As discussed in Chapter II, the average spin state of Ceo^ is similar to that of C e o * or C go^', i e., the singlet (S = 0) and triplet (S = 1 ) states must be close in energy. Limited by the fi'eezing point o f its only solvent DMSO (18°C) and its instabUity at higher temperature, we were not able to acquire variable temperature NMR data on Ceo^. The independence of the fuUeride chemical shift on the cations in DMSO-ri^ solution is consistent with the absence o f significant interaction between Cgo^ anion and those cations. This is expected because the complexation o f alkaU metal cation with the cryptand is like three-dimensional solvation, but to a much stronger extent. The metal cation is included into a macrobicyclic ligand, leading to a large, stable and 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. i 1 1 1 1 1 1 1 i ! 200 180 1 1 1 1 160 V r ' f ■ r r r - : 140 120 t 1 1 T 100 ■ ' r ' 80 1 1 1 1 1 60 ppm Figure IIL5 NMR spectrum of [Na(crypt)^4[C6o ^ in DMSO-cfe (the peaks at 67.9, 66.9 and 52.1 ppm are from cryptand). 59.63 72.00 200 180 140 160 1 2 0 1 0 0 80 60 ppm Figure IIL6 Confrrmation o f the formula of [Na(crypt)^4[C6o ^ by N M R integration (in DMSO-<üs, NOE-suppressed, pulse delay 15 s, L B = 1.5). 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. organic-like cation with very low surface charge density. As a result, its interaction with fuUeride is expected to be so weak^ that the anion should be considered “naked” from cation surroundings.^ As we discussed above (Section in.2.2.), the concentrated NMR sample is stable at room temperature for several days and an integrity check by vis-NIR afterwards showed no sign of degradation. In order to verify the formulation of the C eo '* ' fuUeride that exists in DMSO-<4 solution, we carried out a quantitative "C NMR experiment by comparing the integration of the friUeride peak and those of cryptand. The direct measurement o f "C spin-lattice relaxation (Ti) is impractical with our instrumental conditions. However, for the large and rigid bicycUc cryptand, it is estimated to be about 1 sec.^ or even shorter, since it is complexing a metal cation and is in a paramagnetic environment. Running the "C NMR by using an inverse-gated sequence to suppress NOE and setting the delay between pulses 15 sec. (to let "C nuclei o f cryptand füUy relax), by integration we were able to determine that the cation/anion ratio was 4:1 (Figure m .6 ) When the weU-sealed NMR sample is kept at room temperature, the fuUeride resonance wiU graduaUy shift downfield from 185 ppm toward C eo ^ ’ position at 197 ppm. EventuaUy the solution wiU turn green and the fWleride peak wiU broaden out. This is another indication o f decay. Also from this result, we know that the electron transfer between and C e o * " is fast on the NMR time scale and the mixture of Ceo^' and C eo * ' wiU give a weighted average resonance. 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Regardless of the cations (Rb \ K \ or Cs^, all reported chemical shifts o f solid state Ceo^ fall in the range of 180-183 which are in close agreement with the chemical shift of discrete C 6 o * ~ in DMSO (185 ppm). The small chemical shift difference may originate from the environmental differences between these fuUeride baUs. Also this difference is possibly caused by a smaU amount (ca. 5%) o f C eo ^ ’ (5 197 ppm) undetectable by NIR but present in C e o * * DMSO solution. While for aU the intercalated fUUerides the interactions between C e o * * anions and the surrounding cations are strong, soUd state [Na(crypt)]4 [C6o * * ] and [K(crypt)]4[C6o*'] may provide us with the intrinsic C e o * * chemical shift. Unfortunately, limited by the design o f soUd state NMR tubes, C eo * " samples wiU decompose under the heat of the flame during the torch sealing. EPR EPR spectroscopy has been widely used to investigate the electronic structures o f fUUerides. In our group, Ceo“* (n = 1, 2, 3) have been intensively investigated with EPR For Ceo**, the EPR signals are always accompanied with Ceo^* peaks, unlike vis- N IR which with extreme caution it is possible to obtain “pure” Ceo** spectra. Obviously, EPR spectroscopy is a more sensitive technique to detect the trivial amount ofCeo^' caused by traces of impurities in DMSO left by the limitation o f our purification method. 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For comparison, the [Na(crypt)]3 [C6o ^ T spectra under the same conditions are included here along with those o f [Na(crypt)]4[C6o^ The EPR spectra of C e o * * * at 4 K using different microwave powers are shown in Figure m .7, from which we could see essentially three signals, from left to right (g value from high to low), a very broad signal, a peak in the middle (g = 2.0024), and a very sharp line {g = 2.0012, = 2 G). The middle and the sharp signals are considered to be o f Cgo^ origin, while the broad line is assigned to C^a'. This is supported by the fact that the intensity of the broad signal is higher in a purposeful mixture o f Cgo^ and Ceo^. It is not clear now whether the other two doublet-like signals arise from two electronic states of C e o " * * or from two different species, like C e o '* ' and CeoH^'. Since these signals appear to have integrated intensities that are less than trace signal, the corresponding EPR active states (or species) must be very low populated. As in Figure HI. 8 , the Ceo^ doublet shows little power dependence In the Ceo^ spectra, the other two signals ascribed to C e o '* ' overlap with each other and the sharp signal seems easier to saturate. At low power level, the sharp line dominates, but when the microwave power is increased, it is increasingly saturated. With 20 mW power, the middle signal becomes easier to discern. The Ceo^-related signals broaden as the temperature increases and th ^ overlap with the “spike” that simultaneously grows into the C eo ^ ’ signal (Figure DI.9). This is most apparent when the temperature reaches ISO K or higher. As a result, the relative intensity of the central line and its peak-to-peak width value (Mfpp) cannot be 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 200 nW 2 nW 20 nW 200 p . W 2m W 20 mW DPPH 2mT Figure in.7 Microwave power dependence of the EPR spectra of [Na(crypt)^4 [C6o '* * l in DMSO (at 4 K, modulation 0.25 G). 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 200 nW 2 nW 20 nW 200 fiW 2mW 20 mW 2mT Figure in.8 Microwave power dependence of the EPR spectra of [Na(crypt)‘ ^ 3 [C6o ^ ] in DMSO (at 4 K, modulation 0.25 G). 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4K 6 5 K 1 1 7 K 149 K 168 K 2 0 9 K D P P H 2mT Figure in.9 Temperature dependence of the EPR spectra of [Na(crypt)^4 [Cgo'^] in DMSO (modulation 0.25 G, microwave power 2(X ) |iW). 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4K 5 6 K 8 3 K 9 4 K 130 K yi w » * t » > i V ^ i % i | 160 K 194 K 2mT Figure H I.10 Temperature dependence of the EPR spectra of [Na(crypt)‘ ^ 3 [C6o ^ ‘ ] in DMSO (modulation 0.25 G, microwave power 200 p.W). 6 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2mT Figure I IL ll A typical EPR spectrum o f powder [Na(crypt)'l4 [C6o ‘ ^ salt (4 K, modulation 0.5 G, microwave power 0.2 mW, gain 8 x 10^, frequency 9.4388 GHz). accurately determined. For C e o ^ * , its value is 3.4 G for the spike, and gives the appearance o f increasing to 4.8 G for the mixture o f Ceo^' and Ceo^ For “pure” C eo” * * it gives the appearance o f being 6.4 G, but this is an artifact of the overlap o f two signals. The temperature dependence of the EPR spectrum o f [Na(crypt)]3[C6o^l is shown in Figure HI. 10 and they show no difference from those o f [Na(crown)]3[C6o^l. Since frozen solution EPR spectra are always complicated by the presence of C e o ^ * , the solid state EPR spectra o f C e o '* ' were measured. The single doublet-like signal of powder [Na(crypt)]4C6o^ (Figure HI. 11) with little power or temperature dependence manifests little information. However, its g value (2.0024) and shape are 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. similar to those of the middle line in the frozen solution EPR spectra, and probably these two signals have the same origin, i.e., the triplet state o f C 6 o * ~ . To date, the only other EPR report of is by Baumgarten et al.^ They concluded that Ceo^ is diamagnetic because no EPR signal were found. This is probably in error, since their vis-NIR result is inconsistent with literature reports®’ ® and with results presented here. It is possible that their Ceo'*' product, if formed, was completely insoluble under the conditions used. SQUID Preliminary SQUID magnetic susceptibility data o f [Na(crypt)^4 [C6o ‘ *l show results similar to those o f The significant magnetic moment at all temperatures (4-300K) reveals that Ceo^ salt is paramagnetic. The magnetic susceptibility decreases while the temperature drops, implying that the diamagnetic singlet state is probably lower in energy than the triplet state. However, we were not able to carried out quantitative assessment due to the extreme sensitivity to oxidation of this salt and the possible presence of ferromagnetic impurities.^® III.3. Conclusion Using sodium or potassium as the reductant with addition of cryptand[2.2.2], we were able to synthesize discrete Cco^ for the first time and isolate it as an analytically pure salt, [Na(crypt)]4 [C6o^. It is exceedingly sensitive to oxidation and 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. can only exist for a limited time in solution or as a isolated solid. According to our NMR results, the tetra-anionic fuUeride is paramagnetic and approximately its spin density a v e rte d from different electronic states may be equivalent to one unpaired electron per buckybaU. The high sensitivity of EPR and SQUID susceptometry to magnetic impurities and the extreme difficulty of performing quantitative experiments with these techniques leave the question of the ground state (singlet or triplet) open. This is a similar problem to that o f the Ceo^' ion, its electron-hole equivalent. Nevertheless, the data are most consistent with a singlet ground state and a smaU singlet/triplet splitting. III.4. Experimental in.4.1. Physical Measurements Vis-NIR spectra were recorded on a Shimadzu UV-260, an Ocean Optics or a Varian Cary 2415 spectrometer. C NMR spectra were recorded at room temperature on a Broker AM-360 or AMX-500 MHz spectrometer and the chemical shifts were calibrated against internal solvent values. EPR spectra were recorded in frozen solution on a Broker ER 200D-SRC spectrometer equipped with an Oxford Instruments ESR 900 cryostat, and g values were determined by calibration to DPPH powder. SQUID data were collected on a Quantum Design MPMS magnetometer at Caltech. Elemental analyses were performed by the U.C. B erkel^ Nficroanalyticai Laboratory. 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in .4 .2 . Solvent Purification The purification o f the three most important solvents used for the experiments presented in this chapter is described in detail here For most of the other solvents used, they were purified and dried similarly by heating under reflux or near reflux overnight with the agent indicated in Table in.2, both inside and outside the glovebox, followed by distillation or vacuum distillation. DMSO was refluxed with CaHz overnight under Ar atmosphere, followed by vacuum distillation and three cycles o f degas (fi'eeze/pump/thaw). It was then brought into glovebox. After vigorously stirred with CaHz while warmed (70-90®C) overnight, it was vacuum distilled, passed through activated alumina, and stored over 3 A molecular sieves.^ The purification o f PhCN was done in a similar way, but it was washed with concentrated HCl first and processed with sodium outside and PzOs inside the drybox, respectively. THF was refluxed overnight with sodium benzophenone ketyl, then distilled both outside and inside the glovebox. To remove the trace of benzophenone left, this solvent was distilled once more fi'om Na/K alloy in.4.3. Synthesis and Isolation A piece o f fireshly cut sodium metal was stirred while gently warmed overnight with THF distilled from Na/K alloy. After it was polished to a shiny metal ball, this suspension was filtered to remove the dispersed small sodium particles. This sodium 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ball was put back into clean THF (25-40 ml) along with addition of C eo (20.0 mg, 0.0278 mmol) and 4.2 equivalents o f cryptand[2.2.2] (44.0 mg, 0.117 mmol). This mixture was gently stirred near reflux overnight. After the reaction was finished, the brown precipitate was filtered, the ball of excess sodium removed with a spatula, and the product was washed with fi'eshly distilled THF and dried under vacuum. Elemental analysis: calcd. for Ci32Hi44Ng024Na4: C, 68.38; H, 6.26; N, 4.83. Found: C, 67.69; H, 6.16; N, 4.88%. Table IIL2 The purification o f the solvents used or tested for C é o '* * solubility solvent pre treatment outside box purification inside box purification afterward treatment DMSO CaHz CaHz alumina, 3A THF Na/PhzCO Na/PhzCO Na/K alloy PhCN HCl Na P2O5 alumina, 3A Pyridine LiAttL Na CHsCN P2O5 CaHz alumina HMPA CaHz Na DMF MgS04 alumina alumina alumina ODCB H2SO4 CaHz Na "BuNHa CaHz CaHz in .4.4. Sample Sealing Elemental analysis samples were triple sealed in vials with wax and shipped to Berkeley \ficroanalytical Laboratory. NIR samples were sealed in a special cuvette 7 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I with a teflon valve, carried over to NIR spectrometer in a container under reduced pressure. NMR samples were sealed under vacuum with torch and kept frozen in liq. Nz until experiment time. EPR samples were flame sealed and kept similarly before being transferred directly from liq. N % to the cryostat cavity. References 1. (a) Mullen, K.; Oth, J. F. M ; Engels, H.-W.; Vogel, E Angew. Chem., Int. Ed. Engl. 1979, 18, 229; (b) Huber, W ; Unterberg, H ; MüUen, K. Angew. Chem., Int. E d Engl. 1983, 22, 242; (c) Ayalon, A.; Rabinovitz, M ; Cheng, P C ; Scott, L. T. Angew. Chem., Int. Eld Engl. 1992, 31, 1636; (d) Müller, U ; Baumgarten, M J. Am. Chem. Soc. 1995, 117, 5840; (e) Bock, H ; Gharagozloo-Hubmann, K ; Nâther, C ; Nagel, N.; Havlas, Z. Angew. Chem., Int. E d Engl. 1996, 35, 631. 2. (a) Dubois, D ; Kadish, K. M.; Flanagan, S.; Haufler, R. E.; Chibante, L. P. F.; Wilson, L J. J. Am. Chem. Soc. 1991, 113, 4364; (b) Dubois, D ; Moninot, G ; Kutner, W ; Jones, M. T.; Kadish, K. M. J. Phys. Chem. 1992, 96, 7137. 3. (a) Dubois, D ; Kadish, K. M.; Flanagan, S.; Wilson, L. J. J. Am. Chem. Soc. 1991, 113, 7773; (b) Jûe, Q ; Perez-Cordero, E.; Echegoyen, L. J. Am. Chem. Soc. 1992, 114, 3978; (c) Ohsawa, V.; Saji, T. J. Chem. Soc., Chem. Commun. 1992, 781. 4. Fleming, R M.; Rosseinsky, M. J.; Ramirez, A. P.; Murphy, D. W ; Tully, J. C ; Haddon, R. C ; Siegrist, T.; Tycko, R.; Glarum, S. H ; Marsh, P.; Dabbagh, G ; Zahurak, S. M ; Makhija, A. V.; Hampton, C. Nature 1991, 352, 701. 5. Lawson, D R.; Feldheim, D. L ; Foss, C. A.; Dorhout, P. K ; Elliott, C. M ; Martin, C. R.; Parkinson, B. J. Electrochem. Soc. 1992, 139, L6 8 . 6 . Fullagar, W. K ; Gentle, I. R.; Heath G. A.; White, J. W. J. Chem. Soc., Chem. Commun. 1993, 525. 7. (a) Baumgarten, M.; Gûgel, A.; Gherghel, L. Adv. M ater. 1993, 5, 458; (b) Baumgarten, M ; Gherghel, L. Appl. Magn. Reson. 1996,11, 171. 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 . Green, W. H., Jr.; Gorun, S. M.; Fitzgerald, G ; Fowier, P. W ; Ceulemans, A.; Titeca, B. C. y. Phys. Chem. 1996,100, 14892. 9. (a) Murphy, d. W.; Rosseinsky, M. J.; Fleming, R. M.; Tycko, R_; Ramirez, A. P.; Haddon, R C.; Siegrist, T.; Dabbagh, G ; Tully, J. C ; Walstedt, R. E. J. Phys. Chem. Solids 1992, 53, 1321; (b) Reichenbach, J.; Rachdi, F.; Luk’yanchuk, I.; Ribet, M ; Zimmer, G ; Mehring, M. J. Chem. Phys. 1994, 101, 4585. 10. Boyd, P. D. W.; Bhyrappa, P.; Paul, P.; Stinchcombe, J.; Bolskar, R. D ; Sun, Y.; Reed, C. A J. Am. Chem. Soc. 1995,117, 2907. 11. Trulove, P. C ; Carlin, R. T.; Eaton, G. R.; Eaton, S. S. J. Am. Chem. Soc. 1995, 117, 6265. 12. (a) Stinchcombe, J.; Penicaud, A ; Bhyrappa, P.; Boyd, P. D. W ; Reed, C A. J. Am. Chem. Soc. 1993, 115, 5212; (b) Bhyrappa, P.; Paul, P.; Stinchcombe, J.; Boyd, P. D. W ; Reed, C. A. J. Am. Chem. Soc. 1993, 115, 11004. 13. Lehn, J. M ; Sauvage, J. P. J. Am. Chem. Soc. 1975, 97, 6700. 14. Tehan, F. J.; Bamett, B L.; Dye, J. L. J. Am. Chem. Soc. 1974, 96, 7203. 15. Huang, R. H.; Faber, M. K.; Moeggenborg, K. J.; Ward, D. L.; Dye, J. L. Nature 1988, 331, 599. 16. Connelly, N. G ; Geiger, W. E. Chem. Rev. 1996, 96, 877. 17. Dye, J. L. in Prog. Inorg. Chem. Lippard, S. J. éd., l^ e y & Sons; New York, 1984, 32, 327. 18. Perrin, D. D ; Armarego, W. L. F. Purification o f Laboratory Chemicals Pergamon: Oxford, 3rd ed. 1988. 19. Bard, A J.; Faulkner, L. R. Electrochemical M ethods: Fundamentals and Applications Wiley & Sons: New York, 1980, inside back cover. 20. Sawyer, D. T.; Roberts Jr., J. L. Experimental Electrochem istry fo r Chem ists W il^ & Sons: New York, 1974, pp. 65, 170-171. 21. Myazymbetov, M. E.; Evans, D. H.; Lerke, S. A.; Cahill, P. A ; Henderson, C C J. P l^s. Chem. 1994, 98, 13093. 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22. Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456. 23. Tüset, M.; Zlota, A. A.; Folting, K.; Caulton, K. G. J. Am. Chem. Soc. 1993, 115, 4113. 24. Fraenkel, G.; Ellis, S. H.; Dix, D. T. J. Am. Chem. Soc. 1965, 87, 1406. 25. Lehn J. M. Pure Appl. Chem. 1980, 52, 2303. 26. Bock, H.; Ansari, M.; Nagel, N.; Claridge, R. F. C. J. Organomet. Chem. 1995, 501, 53. 27. Breitmaier, E.; Voelter, W. Ccarbon-13 NMR Spectroscopy: High Resolution M ethods cmd Application in Organic and Biochemistry VCH: New York, 3rd ed., 1987, pp. 163-183. 28. (a) Zimmer, G ; Heimle, M.; Mehring, M.; Rachdi, F. Europhys. Lett. 1994, 27, 543; (b) Zimmer, G ; Mehring, M ; Goze, C ; Rachdi, F. Phys. Rev. B 1995, 52, 13300; (c) Goze, C ; Rachdi, F ; Mehring, M. Phys. Rev. B 1996, 54, 5164. 29. (a) Burfield, D. R.; Smithers, R. H. Org. Chem. 1978, 43, 3966; (b) Burfield, D. R.; Gan, G.-H.; Smithers, R. H. J. Appl. Chem. Biotechnol. 1978, 28, 23. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter IV. Fullerides of Pyrrolidine-Functionalized IV. 1. Introduction The fullerides of unfimctionalized C eo have been extensively investigated. However, some of their properties are still not well understood and the interpretation o f some results remains controversial. For instance, the EPR spectra o f C e o * and Ceo^ samples show a minor sharp signal (“spike”) ubiquitous at relatively high temperature, and Cgo^' shows a doublet-like central line. The origin o f these features is not well explained to date even though various causes have been proposed: impurities, disproportionation, a thermally accessible excited state, ion pairing or aggregation. ‘ Parallel studies on fullerides o f fUnctionalized C so may provide more insight into these problems. Also, investigation o f derivatized fullerides may help us to understand the perturbation that functionalization has on the Cm electronic structure. Compared to their parent fullerides, the fUnctionalized counterparts have received little attention. Methoxylated fullerides have been detected by negative ion mass spectrometry.^ Anions o f several substituted 1,2-dihydridofullerenes have been generated chemically^’ '* or electrochemically^ and briefly studied. More systematic investigations have been carried out but only on fullerides generated transiently by a photochemical method.^ No isolation o f any fUnctionalized fUUerides has been reported so &r. 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It is well known that fuHerenes behave chemically like an alkene and can be fUnctionalized by a variety o f methods/ Azomethine ylides are a class of so-called 1,3- dipoles which react readily with olefins/ They also add to the C e o C=C double bond via 1,3-dipolar cycloadditions,^ forming pyrrolidine fUnctionalized fUUerenes, the so- called fUlleropyrrolidines/” ’^ ^ This method has become a versatile route to derivatize C eo. Via a pyrrolidine ring, the electron deficient fUUerene moiety has been linked to other firagments, like ferrocene,tris(bipyridine)ruthenium,‘ ^ porphyrins, tetrathia- fUlvalene,** benzoquinone,*® iV^-dimethylaniline,‘^ and to a nitroxide radical/* The interactions o f the Qo ball with these electroactive or photoactive moieties have aroused wide interests. Also by this method, unnatural a-amino acid and peptide derivatives could be obtained.** To date, the fUlleropyrrolidine derivatives have shown a variety o f potential applications in medicinal,^ material^* and surface sciences.^ Figure IV .l iV'-methylfUlleropyrrolidine I. 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The parent “Prato-type” compound, ^-methylfiilleropyrrolidine 1 (Figure IV.l), is readily synthesized under mild reaction conditions with a reasonable yield. The functionalization occurs at a 6:6 ring juncture and only one o f the double bonds is saturated, consequently the fullerene characteristics are expected to be largely retained. It has good stability and for most of the fullerene reactivities, the pyrrolidine ring will not interfere. The redox properties of this Ceo derivative have been studied electrochemically. Like most of the organofullerenes,^ its reduction pattern is similar to that o f free Ceo, but with the potentials shifted cathodically.*^^ Owing to all the above reasons, we decided to explore the anions o f 1, as representative of fUnctionalized fullerides. Interactions between C eo spheres have received considerable attention. FuHerenes can aggregate to form clusters in solution^ and under photochemical conditions they can be cross-linked to form polyfullerenes via presumably [2 + 2] cycloaddition. In the solid state, fuHerenes can undergo photo- or pressure-induced polymerization.^^ The interactions between anionic and neutral buckybaHs are also of interest. The formation o f a dimeric anion (€ 60)2' had been proposed to rationalize some experimental phenomena.^ (Cso)^' and (Ceojz^" were considered as a possible explanation for the EPR behavior of fuHeride s a l t s . I n the solid state, the dimerization” ^ and polymerization^’ ^ of singly charged anion C e o * have been extensively studied at different temperatures and it is usuaHy beUeved that these processes are via thermaHy aHowed cycloadditions. 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For any of the interactions to occur, fullerene balls have to be in a close distance, probably in the range o f van der Waals contact. For some cross-linkings between Ceo’s, in addition, the approaching C=C double bonds on adjacent balls must align in parallel so that the [2 + 2] cycloaddition can occur. To probe the interactions in a more manageable manner, covalently linked Ceo’s may be a good model to bring two fullerene balls close to each other. Quite a few fullerene “dimers”, i.e., covalently linked C eo balls have been synthesized,^ ‘ and in some cases, the investigation of these interactions have been initiated with electrochemical studies. To bring two Adlerene balls within interaction distance, they were linked meta to benzene via pyrrolidine amides (see 4, in Scheme IV.2) in our lab. Molecular modeling suggests that the C eo moieties in 4 are held in van der Waals contact and, in addition, they are in good alignment. It should be a superior model to explore the potential ball-to-ball interactions. We are interested in studying its properties when the two C eo balls are reduced, to the same or different states. For instance, we like to examine any possible perturbation on the properties (vis-NIR, EPR, etc.) of one monoanionic ball by the neighboring neutral or anionic buclqrball. Two fullerene balls were also linked para to benzene via same organic groups to form dimer 5, in which two Cso's are held relatively remote from each other. Comparatively studying these two isomers may hopefully help us to differentiate the ball-to-ball contact interaction from the long range communication. 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV.2. Results and Discussion IV.2.1. Synthesis of Neutral Starting Materials IV.2.1.1. Compound 1 and C6 o(CH2 )2 NCPh3 and Applications A/'-methylfulleropyrroIidine C6o(CH2)2NCH3 1 was synthesized according to the reported method:*** G e o was refluxed with 2 equivalents o f sarcosine (^-methylglycine) and 5 equivalents of paraformaldehyde in toluene. The product 1 is readily separated with a silica gel column from the unreacted C e o and a third fraction (presumably the mixture of the bis-adducts^^ and even multi-adducts). CF3 SO3H H H Scheme IV .l The cleavage o f the trityl group o f 2 C6o(CHz)2NCPh3 2 was prepared in a similar way according to Prato et a l .* * * In our hands, separation of the product from unreacted Cm was troublesome by the 8 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reported method (flash chromatography, eluent; toluene/hexanes 2 :8 ), and in addition, the product would frequently crystallize inside the silica gel column. After systematically searching for a better eluent, these problems were overcomed by using p-xylene/hexanes (2:5). The trityl group in 2 is quite labile and it can be cleaved by an addition of triflic acid to C6o(CH2)zNCPh3 to form C6o(CHz)2NH2^ 3 (Scheme IV .l). Neutralizing this salt with a base will afford C6o(CHz)2NH, a versatile starting material for preparing other Prato-type fiillerene derivatives. Since secondary (and primary) amines will react with fUUerenes, this neutralization is usuaUy carried out in situ immediately befr>re the next step reaction. IV.2.1.2. I^ D im e r 4 and 1,4-Dimer 5 The preparations of these two dimers are shown in Scheme IV.2. The neutralization of the ammonium salt 3 is accomplished by the addition of pyridine. The key steps in these syntheses are the quantitative reactions of the resulting V-H fuUeropyrrolidine as a nucleophile with benzenedicarbonyl dichlorides. The clean formation o f these dimers could be achieved by precise control o f the stoichiometry (2:1) o f the two reactants. Due to the moisture sensitivity of benzenedicarbonyl dichlorides, these reactions are carried out in a dried solvent inside the glovebox. The structures of these two dimers are confirmed by their and NMR spectra. For the 1,3-dimer 4, the integrations of H resonances of methylene (5.56 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ppm) and the phenyl group show an ideal ratio of 8 :1:2:1 (Figure IV.2). In its C NMR spectrum (Figure IV.3), the assignment of the four phenyl ring peaks was made by comparison with Sadtler standard spectra. There are 16 fullerene peaks as well as the C =0 peak at 168 ppm, but no peaks can be convincingly assigned to CHz groups and the fullerene sp^ carbons. For the 1,4-dimer 5 (Figure IV.4), the H integrations of methylene and phenyl signals show the expected ratio of 8 : 4. In its NMR (Figure FV.5), two peaks were assigned to the phenyl ring by comparison with Sadtler standard "C spectra. There are 16 fullerene sp^ peaks and their pattern (i.e., the chemical shifts, relative positions and intensities) is almost identical to that in 1,3-dimer 4. The methylene groups in pyrrolidine rings as well as the sp^ fullerene carbons and the carbonyl is not detectable. The solubility o f the dimer 4 is very low in all the solvents tried and hence its chemistry was limited. For instance, all attempts to characterize its sandwich complexes with metal Pd^^ and Rh^^ etc. failed. For fullerene compounds, their solubilities usually seem lower when redissolution is attempted. The dimer 4 shows a more dramatic behavior in this respect. The isolated sample is soluble in CSz or ODCB, but once the solvent is removed, it is impossible to dissolve in any solvent any more. The origin o f this phenomenon is possibly the cross linkage o f buckyballs to form polymeric fuHerenes via [2 + 2] cycloadditions under ambient light. The intramolecular linkage may promptly occur, since the two adjacent C g o balls are conformationally ready. 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. « » A 2) 1) Py 1) Py 1,4-dimer Scheme IV.2 The synthesis of 1,3-dimer 4 and 1,4-dimer 5. 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. w —I —1 —1 —I —■ — I —, — 1 —I— I— 8.0 5.5 5.0 O B # 8.0 7.5 ? !o 8 ^ 5 Figure IV.2 NMR spectrum of 4 in CSj/CDCU (the peak at 7.24 ppm is from the solvent, and that at 7.29 ppm is from impurity). 135 130 ppm 150 140 145 Figure IV.3 "C NMR spectrum of 4 in C S % (acetone-d^ as the insert, not shown is the C=0 peak at 168.3 ppm). 8 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8.01 “ I I I I I I I I I I I I I I I I I I I I I— I— r ~ i — I— I— t — r — I— I— I— I— I— I— I— I — I— I— I— I— I— I — I— I— I— | — 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 ppm Figure IV.4 H NMR spectrum of 5 in CS2 (acetone-dk as the insert). I » I I I I I I I I I I 1 I r I I ■ I I— I I I ! I I \ 150 145 140 135 130 ppm Figure IV.5 NMR spectrum of 5 in CS2 (acetone<4 as the insert). 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV.2.1.3. Structural Fluxionality of the Fullerene Dimers In the parent "Prato-type" compound I, the pyrrolidine ring is not planar (see Figure IV. 19), but the nitrogen inversion is fast compared to the NMR time scale. As a result, at room temperature only one signal is observed for the methylene protons and the number o f fullerene sp^ peaks in "C NMR is 16 rather than 32. C6o(CH2)zN(CO)R (R = alkyl groups), which could be regarded as a model for the two fullerene dimers, has two singlet 'H peaks for the methylene groups on the pyrrolidine ring, and for most R groups, 32 "C peaks for the fullerene sp^ carbons.^ But for each of the dimers, there is only one methylene resonance in NMR spectrum and in C spectrum there are 16 fullerene sp^ peaks. Obviously, the amide groups in these dimers are more fluxional on the NMR time scale compared to those in their model compounds. The amide group in organic compounds has been extensively studied by NMR spectroscopy.^® It is well known that the substituents on the nitrogen atom (e.g., the two methyl groups in Figure IV 6) are frequently magnetically nonequivalent, even when they are chemically identical. As shown in Figure IV 6, there are two resonance structures for the amide R(CO)N(CH3)z and due to the contribution of structure b, the C-N bond in amide functionality has partial double-bond character. This bond is usually rigid to some extent and the barrier to the rotation around it sometimes is quite large. Due to the anisotropy o f the diamagnetic susceptibility o f the amide group, when the rotation around C-N bond is relatively slow compared to the NMR time scale, the two 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. methyl groups will experience different shielding and will show two distinct resonances in H NMR spectrum. This is also true for the two methylene groups in the pyrrolidine ring o f C6o(CH2)2N(CO)R and in some cases, in addition, the fullerene moieties would experience magnetic anisotropy as well and show 32 sp^ peaks in the "C NMR spectra. ^M ei Mea + c Figure IV.6 The resonance structures of amides. With increasing temperature, more and more amide molecules can cross the rotation barrier, i.e., more of the methyl groups can exchange their positions with the others. At a certain point (the coalescence temperature T^, the rotation will be sufBciently rapid and the two NMR resonances will be averaged out to a single signal. For R(C0)N(CH3)2, when R = Me, Et, the coalescence temperature is about 75 and 60®C, respectively.^’ The 7^ value for the amide groups in C6o(CH2)2N(CO)R should 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. be at the same level and therefore, at room temperature the rotation of fuUeropyrrolidine moiety around the C-N amide bond wUl be slow in the NMR time scale. The two different methylene signals in the reported H NMR spectra are consistent with the above interpretation. In R(C0)N(CH3)2, when R is a phenyl group, the cross conjugation o f the amide group with the phenyl ring results in a competitive delocalization (see structure c in Figure IV.6). The double-bond character of C-N bond, hence its strength, w iU decrease. As a result, the rotation barrier and the coalescence temperature 7^ w iU be lowered. For C6H5(CO)N(CH3)z, 7^ is reported to be 3 to 17"C, depending on the solvent and analysis method. In dimer 4 and S, two buckybaUs are linked to a benzene ring via pyrrolidine amides. The coalescence temperatures for these amide groups are expected to be o f similar value, i.e., around 10 ± 10°C. In contrast to the situation in C6o(CHz)2N(CO)R for R = alkyl, the fuUeropyrrolidine moieties in dimer 4 and 5 wiU be able to rotate with certain freedom around the amide C-N bonds at room temperature. The H NMR spectrum of either 4 or S shows a single broad methylene signal. Due to the rapid environment exchange o f the two methylene groups, the two distinct H NMR resonances wiU be averaged to a single broad U ne. Under our experimental conditions, the "C resonance of the methylene groups is expected to have an intensity higher than those o f the fuUerene peaks (the Nuclear Overhauser effect). Nevertheless, these peaks are not observable at room temperature. Possibly, they are too broad to 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. detect due to the environmental exchange. JV^diethyl-w-toluamide (m- CH3C6H4(CO)N(CHzCH3)2) may be used as a model to explain this phenomenon. In its NMR spectrum (Sadtler standard), the only one CH2 signal is a sharp, well- splitted quartet while the corresponding resonance is very broad. That is consistent with the slightly shorter time scale o f C NMR^® In our dimer systems, the peaks of the methylene groups are expected to be broader (even not observable) since the corresponding H signals are already broad. IV.2.1.4. Other Prato-Type Cm Derivatives H H 1) 2eq. (Me3Si)2 NLi 2) Br Br 1,3-dimer 6 Scheme IV.3 The synthesis of 1,3-dimer 6. 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The synthesis of another 1,3-dimer 6 was attempted (Scheme IV.3). With its structure similar to that of 4 but more fluxional, it would be interesting to compare its properties with those of 1,3-dimer 4. The preparation is via the nucleophilic reaction of C6o(CH2)2N ' with 1,3-CgH4(CH2Br)2. This strong nucleophile could be obtained by treating C(so(CH 2 )2NH2" ^ 3 with 2 equivalents of base. The bulky bis(trimethylsilyl)- amide (Me3 Si)2NLi was chosen as the reagent in the preparation to avoid the possible nucleophilic attack on the electron-deficient fullerene ball. The exact equivalency of l,3-C6H4(CH2Br>2 (2 equivalents) and a long reaction time were applied to ensure the clean fi)rmation o f 6. It is even less soluble than the dimer 4 and the reduction with cobaltocene leads to inconclusive EPR results. The low solubility o f the dimer 6 is probably due to its low polarity, since dimer 6 has two methylene groups in place of the two C=0 groups o f dimer 4. This is consistent with our knowledge o f the solubilities of fullerene derivatives.^^ PPh, 1) Py 2) PhjPCHjOH Scheme IV.4 The synthesis of Prato-type derivative 7. 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The phosphonium sait [PhzPCCHzOHbJCl treated with EtsN will generate the phosphine PhzPCHzOH, which reacts with primary or secondary amines to form aminomethylphosphines/^ Nfixing this phosphine with iV'-H fuUeropyrrolidine C6o(CH2)2NH is expected to give C6o(CH2)zNCH2PPh2 . Unfortunately, we were not able to fuUy characterize the product due to its extremely low solubility, as in the case o f 1,3-dimer 6 The only data obtained was a weak ^ P NMR signal at 24.4 ppm after a very long acquisition time. IV.2.2. M onoanions of Fuileropyrrolidines, T and 2 Synthesis Among aU of the reagents capable o f reducing Cso to the monoanion, cobaltocene and Na (with the addition o f 1 eq. crown either) are two o f those used to generate analyticaUy pure C e o * salts.C o b alto cen e, a 19e' compound, is a convenient one-electron reductant which can be oxidized cleanly to the robust cation. Its reduction potential is reported to be -1.3 V (vs. Fc^/Fc),^* lying between the first and second reduction potential o f 1 (-1.05, -1.44 V vs. Fc"'/Fc).*“ Therefore, unlike Cm, which can be partly overreduced by excess cobaltocene,^" 1 can be selectively reduced to form T. The redox behavior o f 2 is believed to be similar to that of 1 and with cobaltocene we should also be able to prepare 2 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Experimentally, [Cp2C o'^[l'] is readily isolated as a salt o f analytical purity by simply mixing one equivalent o f CpaCo with 1 in appropriate solvent. According to the elemental analysis result, in an isolated sample the content o f N is slightly higher than expected, probably due to a trace o f PhCN (bp 191 ® C ) remaining even after vigorous washing with hexane/toluene. The presence of PhCN was also confirmed by H and ‘^CNMR. The preparation o f 1 can be achieved with Na/crown ether as well. Unlike with the reduction o f C eo, both 1 and 1 are fairly soluble in THF. The only appearance diftërence between them is the darker color of 1 Naturally, judging the completion of the reaction is a little intricate and vis-NIR. spectroscopy has to be frequently employed. Vis-NIR As indicated in Figure IV. 7, the vis-NIR spectrum of 1 shows three peaks (1002, 882, 780 nm) as well as a broad bump between 500 and 620 nm. Notably, the distinctive fingerprint pattern (relative position and intensity of vis-NIR bands) of Cgo' is essentially retained. Compared to the spectrum o f Ceo', the three absorbances o f 1 fiilleride are broadened and probably due to this broadening the peaks at 1072 and 1044 nm in Cgo' spectrum coalesce to the one at 1002 nm o f 1 All the three absorbances at 1002, 882, and 780 nm are blue shifted relatively to the corresponding Ceo' peaks (1072, 1003, and 935 nm). A similar blue shift has been noticed in the only other NIR report of fimctionalized fullerene anions.^ These 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. s 563 726 879 1020 Wavelength (nm) Figure IV.7 Vis-NIR spectra of [CP2C0 * ] [1*] and [CpzCo^] [C eo"] in PhCIN. 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. observations suggest that after ftmctionallzation the electronic structure o f fullerene monoanion is essentially retained, but the HOMO/LUMO gaps have increased. For example, a shift by 70 nm ft^om 1072 nm band o f Cgo' to 1002 nm band o f 1 indicates that the energy gap corresponding to this transition in 1 is ca. 2 kcal/mol larger than that in C e o * . The above arguments based on the vis-NIR results are in qualitative agreement with the electrochemical data of 1. As already known, the reduction pattern o f 1 is similar to free C eo and its potentials are shifted cathodically compared to those of These also imply that after the functionalization, the electronic structures of fullerene are essentially retained and the energy o f LUMO increases. NMR With a sample of regular concentration, none of the peaks o f 1 in ‘H and "C NMR were discernible. In DMSO-<i^, a saturated sample o f [Cp2Co^[l*] is needed to detect all the and resonances. The fiilleride C peak can be observed after 85800 scans. The signal of the methylene group in the pyrrolidine ring of [CpzCo'^P'] can be observed in pyridine-ef; or PhNOa-tffs, but not in DMSO-ri^, presumably due to the limited solubility in the latter. The ‘H NMR spectrum of [CpzCo^Cl*] in DMSO-rig is shown in Figure IV.8. All the peaks are very broad due to the paramagnetism. The CpzCo"^ resonance appears as a singlet at 5.8 ppm and the methylene and W-methyl peaks in the pyrrolidine ring are 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table IV .l H chemical shifts o f I and 2 and their monoanions solvent 6 (CHz) (ppm) 5 (CHs) (ppm) neutral compounds 1 2 CS2/CDCI3 CS2/CDCI3 4.39 4.17 3.00 NA monoamons [Cp2Co1[n [Cp2Co1[2T DMSO<4 PhN02-<4 1.4 1.2 4.1 NA at 1.4 and 4.1 ppm, respectively. The assignments were first made based on their integrations relative to CpzCo"^ peak and were confirmed by comparing with [Cp2Co*][2 " ] which shows methylene resonance at 1.2 ppm. Choosing the neutral parent 1 as the diamagnetic reference, the isotropic shifts (ô iM o = ô o b i - Sdu, see section n . 1.2.3.) for methylene and iV-methyl in 1 are -3.0 and 1.1 ppm, respectively, i.e., the CHz (upfield) and iV-CH3 (downfield) are shifted in opposite directions. The small isotropic shift values imply that the contact shifts must be small. As we know fi’ om Section n . 1.2.3, the contact shift is in connection with the unpaired electron density on the nucleus of concern. Therefore, we can conclude that in T, the unpaired electron density is principally localized on the fullerene cage and little is leaked into the pyrrolidine ring. The "C NMR spectrum o f [CpzCoHCIl (Figure IV.9) shows only two detectable peaks, that o f CpzCo^ at 85.1 ppm and a broad fuUeride peak centered at 187 ppm. The methyl and methylene in the pyrrolidine ring are not observable due to 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ppm Figure IV.8 H NMR spectrum o f [Cp2 Co^ [1*] in DMSO-<4 (the peak at 2.49 ppm is from the soivem). 100 80 ppm 120 140 160 200 180 Figure IV.9 NMR spectrum o f [CpzCo*] [1*] in DMSO-<4 (the peak at 85.1 ppm is from CpzCo , and those at 110-133 ppm are from a trace o f PhCN). 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the paramagnetic broadening. In the diamagnetic parent 1, the sp^ buckyball signalsspan the range 136-154 ppm, with the majority cluster around 143 ± 3 ppm. Surprisingly, the peak of functionalized fulleride resides in a position almost identical to that of Ceo' (186 ppm), revealing that the unpaired electron spin is delocalized over the buckyball in a similar pattern as in Cgo'. For [CpzCo'][2'], the isotropic shift o f the methylene H signal (-3.0 ppm) is the same as that for 1 (see Figure IV. 10). In addition, the broad C peak o f fulleride is at 188 ppm, almost identical to that of 1 fulleride resonance. These results strongly suggest that the electronic structure of 2 is similar to that of 1 p p m Figure IV.IO H NMR spectrum of [Cp2C o^[2 " ] in PhNOj-fife (those peaks at 7.50, 7.67 and 8.11 ppm are from the solvent). 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EPR Compared to Cm , the EPR spectrum o f 1 * is much more simple and shows only a narrow doublet signal (A//pp = 1-3 G) with little anisotropy. No hyperfine coupling to N and H atoms on the pyrrolidine ring could be observed. The g value is determined to be 2.0006 ± 0.0002, higher than that of Cm' (1.9963) but lower than free electron value (2.0023), typical o f fuUeride-originated EPR signals. The EPR spectra of [CpzCo^Cl*] under different microwave powers are shown in Figure IV. 11 and Figure IV. 12. Unlike G e o * spectra which show evident cation- dependent anisotropy at low temperature, the EPR signal o f 1 is nearly isotropic and its line-shape is almost independent of the microwave power over a wide range of temperature. No strong temperature dependence in the 1 EPR was observed. With increasing temperature, the line-width o f 1' signal only increases a little (by 1-2 G) (see Figure T V . 13). Notably, [Cp2C o')[l‘ ] does not show a “spike” in its EPR spectrum at relatively higher temperature. The spikes have been observed in all the reported EPR spectra of Ceo' and they have been proposed to arise from the thermal population o f a low-lying excited state. Schematically, in Figure IV. 14, A represents the ground electronic state of Ceo' and hence corresponds to the main doublet line in the EPR spectrum, while B denotes the excited state related to the EPR “spike”. A ' and B ' are the respective counterpart o f 1 * . Based on the blue shift o f all vis-NIR bands o f 1 relatively to those o f Cm', S ' will be 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 200 nW 2^W 20 nW 200 2mW ImT Figure IV .ll EPR spectra of [CpzCo^[l'] in pyridine under different microwave powers (4 K, modulation 0.25 G). 200 nW 2^W 2mW ImT Figure IV .ll EPR spectra of [Cp2Co^[l*] in PhCN under different microwave powers (157 K, modulation 0.25 G). 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. B ' energy B Æ ' AE A' '-'60 * ■ AE' — AE = 2 kcal/moi Figure IV.14 Schematic representation o f the electronic states of Ceo' and 1 higher in energy (relative to A') compared to B (relative to A). Since the minimum shift 70 nm (from 1072 to 1002 nm) is equivalent to an energy difference of 2 kcal/mol, it is natural to assume that is at least 2 kcal/mol higher than B in relative energy. Under the Boltzmann distribution conditions, the population of state will be 0.1% or less o f that o f ^ at 150 K. In this situation, it would be appropriate to consider the excited state B* thermally inaccessible at this temperature and therefore in 1 EPR spectra no “spike” could be detected. The line shapes o f [Cp2C o^[l‘] EPR spectra appear to be solvent independent, but the peak-to-peak separations vary slightly with solvent. As indicated in Figure IV. 15, in PhCN the A^pp value of the doublet is 1 G, while in pyridine and DMSO its value increases to 2.7 and 2.9 G, respectively. 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4K 59 K 102 K 157 K ImT Figure IV.13 EPR spectra of [Cp2Co^[l*] in PhCN at diffèrent temperatures (modulation 0.25 G, microwave power 20 pW). PhCN DMSO ImT Figure IV. 15 EPR spectra o f [CpzCo'][l'] in three diffèrent solvents (4 K, modulation 0.25 G, microwave power 20 pW). 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 mT Figure IV. 16 EPR spectra o f 1 in DMSO (4 K; microwave power 20 pW; modulation for Na(crown)* 0.0125 G, for CpzCo* 0.25 G). 900 600 300 nm Figure 1V.17 Vis-NIR spectrum o f [Na(crownXTHF)2 " l2 [ l^ in DMSO. 1 0 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The 1 salt with Na(crown)^ as the cation has also been systematically studied with EPR and it shows similar behavior to the Cp^Co salt under all the conditions (microwave power, temperature and solvent). The cation dependence is seen in the respect o f the EPR line widths and is most evident in DMSO. Figure IV. 16 shows the line width o f EPR signal o f [CpzCoHCll is about 10 times as large as that of [Na(crown)"][!"] In DMSO, the EPR signals o f [Na(crown)"^[l*] are so sharp that the modulation had to be set as low as 0.0125 G in order to record the whole spectrum. The solvent and cation dependences o f the EPR line widths of 1 indicate the possible existence o f ion pairing effect, but to a much less extent compared to the situation of Ceo - IV.2.3. Dianion of iV-Methylfulleropyrrolidine, Synthesis and NMR Sodium metal with addition o f two equivalents of crown ether is used to prepare 1^ , since cobaltocene is not sufSciently powerful to generate the dianion of 1, as discussed in Section IV.2.2. The reaction is done in a manner almost identical to the synthesis o f the Ceo^' salt.'"^ Since is slightly soluble in THF, hexane was added after the reaction to effect the complete precipitation o f the product. In the H and NMR spectra of [Na(crown)(THF)2n 2[l^'], the crown ether and THF peaks are readily seen. The observation of no other peaks suggests good purity o f the product. The signals o f the pyrrolidine ring and fulleride dianion are not 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. detectable in both spectra, no matter how concentrated the samples are. These resonances might be too broad to detect due to the strong paramagnetism. Vis-Nm The vis-NIR spectrum (Figure IV. 17) of [Na(crown)(THF)2'l 2[l^T shows two broad absorbances at 730 and 860 nm. Compared to the spectrum of the fingerprint pattern, i.e., the relative positions and intensities of these peaks, is essentially retained. However, relative to the peaks o f C^o^' at 830 and 950 nm, the absorbances of 1^' are blue shifted by 90-100 nm. Similar to the circumstances of 1’ , the electronic structures o f and C e o ^ * resemble each other, but the excited states in are higher in energy relatively to the ground state than those in C e o ^ * . EPR The EPR spectrum o f at 4 K essentially consists o f two signals; a central doublet-like feature and triplet-like “wing”. The intensities of these signals are so low that to observe them very high signal gains have to be used. In Figure IV. 18, the central doublet-like signal is more readily saturated than the triplet, as indicated by the diminished intensity o f the central line with increasing microwave power. Limited by the low intensity of the signals, we were not able to study 1^' at higher temperature with EPR spectroscopy. 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 200 nW 2nW 200 nW 2mW ImT Figure IV.18 EPR spectra of [Na(crown)(THF)2^ 2 [l^T in pyridine (4 K, modulation 2 G). 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A U these EPR characteristics resemble those of C eo ^ This fact apparently does not support the assumption that aU EPR signals of C g o ^ " salts are due to some impurities/'' Combined with the phenomenon that its vis-NIR pattern is similar to that o f C eo^', the electronic structures are probably retained in In any o f the EPR spectra of C e o ^ * , the intensity o f aU signals by integration is only a few percent o f C eo ^ ’ concentration and it is believed that its electronic ground state is a singlet and the singlet-triplet splitting is estimated to be 600 cm''/® The intensities of the EPR signals of 1 ^ * appear to be even lower. The vis-NIR bands o f 1^' are blue shifted relative to those of C e o ^ * , indicating that the excited states o f 1^' are higher in energy, i.e., the energy gaps between the excited and ground state are larger, similar to the case of 1 As a result, in 1 \ the populations of these excited states and consequently the intensities of the corresponding EPR signals are lowered. X-ray Structure Like most o f fuUerene species, the single crystal structures of fWleride salts (Cgo ,^ and C e o ^ * are frequently disordered. However, ordered structures of discrete Ceo' and Ceo^' * * are also known and the X-ray results exhibit the direct evidences for the expected Jahn-TeUer distortion in fuUeride anions. To date, no structure of functionalized fuUeride has been reported. It is usuaUy believed that ions o f simUar size tend to pack better in solid state and subsequently give better crystals. Slow difrusion of a PPNCl solution (in CH3CN) 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. into [Na(crown)(THF)2 l 2 [C6o ^ T affords a good-shaped single crystal [PPN*]2[C6o ^ 1 - For 1^ , the functionalized analog of C e o ^ * , the large-sized PPN* may also be a good cation to produce crystals. Compared to C go^', salt has higher solubility. No solid was formed upon adding toluene into the mixture o f PPNCl and [Na(crown)- (THF)2^2[1^T in benzonitrile. Vapor diffusion of hexane (1 to 2 weeks) into the above solution in PhCN/toluene yielded black shiny single crystals suitable for X-ray analysis. The structure was initially solved only in P\ space group; then, a center of symmetry was found and the space group was corrected to P 1 .* ^ The solution with the best figure of merit was then refined (full-matrix least-squares on F) and the entire molecule was located, but in certain portions within 1^' ball the adequate bond lengths and angles were not obtainable. The C g o 5:6 and 6:6 distances respectively were then restrained to be equal within the fuUerene baU with the use o f DFIX routine. A few cycles o f the refinement o f this model revealed in difference electron-density map a second Ceo-V-Me-pyrrolidine orientation. In order to take into account the “derivatization” (in a fuUerene derivative, the 5:6 and 6:6 bonds respectively are not equivalent any more), the DFIX restraints were specified for the respective 5:6 and 6:6 junctions radiating fi-om the pyrrolidine ring, with starting distance values being fi-ee to refine. The site occupancies were aUowed to refine and converged to 0.56/0.44 for the two orientations o f the fuUeride anion. Figure IV. 19 shows the two orientations of 1 ^ , one o f which is depicted by solid Unes, while the other is by open Unes.^ 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure IV. 19 Dual disorder: the two orientations of 1 [PPNnzCl^'] toluene. 2- in the single crystal of 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I ' / O i Figure IV.20 The unit ceil of (PPN*]2[l^l'tcluene (for clarity all H atoms are omitted). Besides the above described anion disorder, there are two other kind of disorder in this structure; (a) Solvent disorder: at each site of the toluene molecule, no Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. there are two possible orientations, each in 50% occupancy. For clarity, only one of the orientations is shown in Figure IV.20. (b) Cation disorder; in each [PP>T]2[1^T molecule, one o f the 12 phenyl rings on the two countercations is disordered. As shown in Figure IV.20, each unit cell contains two inversion-related [PPhT]2 [l^T molecules. This structure confirms the expected formula o f [PPhT]2[l^T and also shows that no disruption^^ of the pyrrolidine group occurs during the reduction with sodium metal. This is the first known X-ray structure of a functionalized fulleride. Due to the disorder problem, no accurate bond lengths and angles could be obtained. However, fi-om Figure IV. 19, the fuUerene baU is apparently elongated at the fimctionalization site, as seen in most o f the neutral functionalized fuUerene structures.^^ This is also the first solved structure of a Prato-type compound, confirming that the addition o f the azamethine yUde is across a 6:6 C=C bond, as previously deduced firom NMR spectroscopy.^® Also revealed by this structure is that the conformation o f V-methyl pyrroUdine ring. The equatorial orientation o f the V-methyl substituent is consistent with the other reported pyrroUdine X-ray structures.^* IV.2.4. Anions of 13-Dimer and 1,4-Dimer Sample Preparation The reduction behaviors of dimer 4 have been studied by Prof. P. D. W. Boyd with cycUc voltammetry.^® Its first two reduction potentials were determined to be 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -1.00 and -1.41 V (vs. Fc/Fc), respectively (solvent 1:1 PhCN/ODCB, supporting electrolyte 0.1 M B U 4NPF6, scan rate 50 mV/s). Electrochemical results also indicate that the reduction of the two buckyballs in 4 occurs simultaneously, implying their independence. With its reduction potential (-1.3 V vs. Fc/Fc) falling between the first and second wave o f the dimer 4, cobaltocene is employed to prepare 4^ ', i.e., to reduce both C go balls to monoanions. Unlike the electrochemical reduction, by stoichiometric control o f the reductant, 4 could be selectively reduced to 4 , i.e., haveing only one of the two bucl^balls reduced to Cm'- Cobaltocene is not a reductant sufficiently powerful to generate 4^ or 4% but its decamethyl analog Cp*2Co'" will be a good choice for this purpose, since its reduction potential is ca. 0.5 V more negative. The possible reactivity o f the amide groups in 4 and 5 with sodium metal prohibits us fi*om reducing them with Na/crown. Relative to 4, the reduction potentials (-1.05 and -1.45 V vs. Fc^/Fc) o f 5 are cathodically shifted by approximately 50 mV. Nevertheless, the overall reduction properties o f these fuUerene dimers are similar and the above discussion for 4 also applies to the 1,4-dimer 5. In an identical way, mono- and dianion o f 5 are prepared and investigated. IdeaUy, 4 could be generated by simply mixing one equivalent CpzCo with 1,3- dimer 4. But for a reaction in such a smaU scale, it is difficult to avoid overreduction simply by accurate measurements. EPR spectroscopy is the main tool here for fuUeride characterization. With this sensitive technique, the presence o f a trace o f 4^' wiU be 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. detected along with 4 To prevent the possible overreduction in sample preparation and the subsequent interference in EPR characterization, a deficit of reducing agent (< O S eq.) is used in practical preparation. Benzonitrile is chosen as the solvent here, since it is one o f the few that can dissolve both neutral and anionic fullerenes to some extent. In PhCN, the solubility of starting material 4 is limited. To ensure the completion o f the reaction, the mixture was usually stirred overnight. Most of unreacted 4 could be removed by filtration and the small amount of 4 remained in the filtrate would not interfere 4 EPR spectra. \^ th two equivalents o f cobaltocene, 4 is reduced to its dianion 4^', in which both C g Q balls are in the monoanionic form. Limited by the reduction potential of CpzCo, no overreduction to 4^ or 4^ could occur in this process. To avoid underreduction (i.e., the contamination o f 4^' by 4*), slightly excess cobaltocene was used. The minute amount of Cp^Co left will not interfere the EPR characterization of 4^', since this 19e* paramagnetic species has no EPR absorption under our experimental conditions.^* N m Limited by the low solubility o f 4, the quantitative titration of 4 with CpzCo is impractical. The vis-NIR spectra (Figure IV.21) of neutral 4 and the successively 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I I 410 573 735 888 1028 Wavelength (nm) Figure IV.21 Vis-NIR spectra o f neutral 4 and the progressive reduction of 4 with cobaltocene in PhC T N . 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. s I 410 573 735 888 1028 Wavelength (nm) Figure IV.22 Vis-NIR spectra of neutral 5 and the progressive reduction of 5 with cobaltocene in PhCN 1 1 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reduced 4 are recorded after each preparation. Neutral 4 has no strong absorbance in the vis-NIR re^on and the only noticeable feature is a small and sharp peak at a position (431 nm) identical to the corresponding peak of 1. With successive addition of CpzCo, all the absorbances grow up progressively. After 4 has been completely reduced to 4^', further addition of cobaltocene has no effect on the spectrum, supporting the idea that CpzCo can only reduce 4 to its dianion. The fingerprint pattern o f the 4 ^ * spectrum is similar to that o f 1 (Figure IV.7). The three bands of at 1015, 897, 782 nm resemble those o f 1 at 1002, 882, 780 nm, respectively. Relative to the absorbances of 1, the peak at 1015 nm is blue shifted by 13 nm, and the 897 nm peak is red shifted by 15 nm while the one at 782 nm remains nearly unchanged. These suggest that the electronic features o f Cso' balls in 4^' are essentially the same as those o f 1, but the energy gaps between the electronic states vary. It is not clear whether the small shifts of the vis-NIR bands is due to the ball-to- ball interaction in 4^' or to the fimctionalization difference. However, the similar vis- NIR pattern hints that the interaction between the two Ceo’ balls in 4^', if any, is not significant. The vis-NIR spectra of 5 and reduced 5 (Figure IV.22) are identical to the corresponding ones of 4 and reduced 4, respectively. For instance, the three bands of 5 ^ * (1015, 896, 784 nm) are at the same positions as those o f 4^ peaks. With the fact in mind that the distances between two Ceo' balls in 4^' and 5^' vary considerably, their identical vis-NIR spectra suggest that the through-space ball-to-ball interactions, if 1 1 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. present, are undetectable by vis-NIR and the “other*' Cm ball essentially acts as a spectator and has little perturbations on the electronic features of the anionic buckyball. The spectrum o f 4 (reduced with I eq. CpzCo) appears to be the combination of those o f 4 (0 eq. CpzCo) and 4 ^ * (2 eq. CpzCo). It seems that, in 4 , the neutral and the anionic C go balls essentially behave independently and the interaction between them, if any, should be weak. It is noteworthy that all these spectroscopic results are consistent with the cyclic voltammetry behaviors of the dimers 4 and 5.^^ The simultaneous reduction of the two buckyballs illuminates that reducing the first ball has no effect on the other C e o moiety and the ball-to-ball interaction is not detectable by the electrochemical method. EPR of M onoanions For both monoanions o f dimer 4 and S', single doublet EPR spectra are seen under all the conditions. These doublets have g values (2.0006) identical to that of 1 doublet. Their line shapes also show characteristics resembling those o f 1 little power dependence (Figure IV.23, 24), the slight increase o f line width (by 1 to 2 G) with increasing temperature (Figure IV.2S, 26), and the solvent dependence o f the line width (Figure IV.27, 28). All these features indicate that, in the respect o f EPR properties, 4 and S' are almost indistinguishable firom [Cp2C o^[l']. Under these conditions the intramolecular interactions between Cm and C m * among 4 and S’ should be considered negligible, i.e., 4 and S' behave like firee 1'. 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. J\ 2 0 0 n W J\ Y 2p iW V 20 n W J\ \f 200 n W 1 mT V Figure IV.23 EPR spectra of [CpzCo^[4'] in PhCN under different microwave powers (4 K, modulation 0.05 G). 2 hW 20 nW 200 nW 2m W 39 m W ImT Figure IV.24 EPR spectra of [Cp2Co'^[5*] in PhCN under different microwave powers (4 K, modulation 0.025 G). 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4K 57 K 131 K 183 K 1 mT Figure IV.25 EPR spectra of [Cp2Co^[4*] in PhCN at different temperatures (modulation 0.05 G, microwave power 20 (iW). 4K 58 K 154 K 205 K ImT Figure IV.26 EPR spectra of [Cp2Co"^[5^ in PhCN at different temperatures (modulation 0.025 G, microwave power 20 p-W). 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PhCN DMSO ImT Figure IV.27 EPR spectra of [CpzCoHM in three different solvents (4 K; modulation in PhCN 0.05 G, in pyridine and DMSO 0.25 G; microwave power 20 |iW). PhCN DMSO Im T Figure IV.28 EPR spectra of [Cp2Co"][5^ in three different solvents (4 K; modulation in PhCN and pyridine 0.025 G, in DMSO 0.25 G; microwave power 20 |iW). 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EPR of Dianions The EPR spectra o f 4^' are a little more complicated. They show two signals: a central doublet line and a triplet line. Compared to 4', not surprisingly, the doublet in 4^' signal has the same g value (2.0006) and similar peak-to-peak separation (AHpp = 1- 3 G), since they are believed to have the same origin, i.e., monoanion of the functionalized Ceo. For the triplet line of 4 ^ * , the g value (2.0005) is slightly lower than that o f the doublet and the peak-to-peak separation (D value) is 29 G. As shown in Figure H.29, in the EPR spectra o f 4^', the triplet signal is easier to saturate compared to the doublet. As the microwave powers increase, the triplet signal diminishes gradually. The intensity o f this triplet is so low that its parallel portion is not discernible and at 4 K, a very high gain has to be used in order to observe the perpendicular part. With increasing temperature, its relative intensity will increase (Figure IV.30). The triplet EPR signal suggests that in 4^' the electronic triplet state is thermally accessible and its energy should be close to that o f the doublet. The increasing population at higher temperature implies that, in addition, the triplet is slightly higher than the doublet in energy. Based on the EPR results, we conclude that 4^' behaves like a biradical. In most o f the cases, the interaction between the two electron spins is weak and the biradical is equivalent to two independent doublet radicals. But to some extent, intramolecular spin-spin coupling does exist and it shows a relatively weak triplet EPR 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2m W 20 m W ImT Figure IV.29 EPR spectra of [Cp2Co^2[4^'] in PhCN under different microwave powers (42 K, modulation 0.25 G). 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4K 42 K 83 IC 127 K 199 K ImT Figure IV.30 EPR spectra o f [Cp2Co^2[4^T in PhCN at different temperatures (modulation 0.25 G, microwave power 200 pW). 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. signal. In other words, the doublet and triplet are the two accessible electronic states of the biradical 4^'. They are close to each other in energy, but the triplet state is probably a little higher. 1 mT Figure IV.31 EPR spectra o f a mixture o f 4^ and 4^' prepared by 4 and 2.7 equivalents o f Cp*zCo in PhCN (modulation 0.25 G; microwave power 200 p.W; temperature: top: 4 K, bottom: 106 K). Reduction o f 4 with ca. 2.7 eq. of Cp*2Co will diminish the intensity o f the triplet signal in 4 ^ * (Figure IV.31). This verifies that the triplet EPR signal in 4^' originates fi-om the spin-spin coupling between the two Cw balls. Since Cp*zCo (2 eq.) is able to reduce fullerene derivative to its dianion, with 2.7 eq. of it, a mixture of 4^‘ (30%) and 4^ (70%) will be produced. The fact that the intensity o f the triplet EPR signal is diminished is consistent with the reduced amount o f 4^' in the sample. In 4^', the ball-to-ball coupling between Cgo' and C^o^' may exist and will be a little complicated, since the fimctionalized C e o ^ * has more than one electronic state. But 1 2 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4fC 48 K 116 K 186 K ImT Figure IV.32 EPR spectra o f [CpzCo^zCS^l in PhCN at different temperatures (modulation 0.025 G, microwave power 200 ixW). 1 mT Figure IV.33 Vertically expanded EPR spectrum of [Cp2Co^2[5^*] in PhCN (188 K, modulation 0.025 G, microwave power 2 mW). 1 2 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. these states are either EPR silent (the singiet(s)) or very low populated (the triplet(s)) (see section IV.2.3.), so consequently, the coupling between the Ceo' and C e o ^ * units overall should be negligible, i.e., 4 ^ * essentially behaves like a doublet 4 * . The EPR doublet signals o f 4^' and 4^ should overlap with each other, since they are of similar origins (i.e., the monoanion of pyrrolidine-functionalized fullerene). 5^', the doubly reduced salt o f 5, shows EPR spectra (Figure IV.32) fundamentally identical to those of 5" (and 1 * ) (see Figure IV.26 and Figure IV. 13). These doublet signals demonstrate that in 5^' the two Ceo' balls are essentially independent and act as two non-interacting doublet radicals. However, with a suitable gain, a very weak triplet signal is also observable at high temperature (Figure IV.33), indicating that the spin-spin coupling also exists. This coupling is notably weaker than that in 4 ^ * , probably due to the farther distance between the two Cm' balls in 5^'. rv.3. Conclusion The mono- and dianions o f Prato-type fimctionalized fullerene, fUllero- pyrrolidine, have been investigated systematically. T h ^ retain most of the respective electronic structures o f their parent Cm' and Cm^', but their LUMOs are typically shifted to higher energy. In the monoanion, the electronic density is mainly localized in the fuUeride ball and little is distributed on the pyrrolidine ring. A pair o f fullerene dimers have been prepared by linking two Cm balls meta and para to benzene via pyrrolidine amide groups. Their structures show considerable 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fluxionality by rotating around the amide C-N bond. The neutral dimers have very limited solubilities in most of solvents. They have been reduced to different stages and the interactions between two C eo " balls or Ceo' and Cm, if any, are weak, and essentially the moieties behave independently. The spin coupling among the two C m * balls in 1,3- dimer (and occasionally in 1,4-dimer as well) is the only evidence o f ball-to-ball interaction. IV.4. Experimental All manipulations o f air- or moisture-sensitive compounds were carried out in an inert atmosphere in a Vacuum Atmospheres Drybox under helium (H2O, O2 < 1 ppm). C(o was purchased from MER Corp. and used without any further purification. All other chemicals were purchased from Aldrich. l,3-CgH4(CH2Br)2 was recrystallized from acetone/H2 0 and dried under vacuum overnight. Ph2P(CH2 0 H)2Cl was prepared by Prof. P. D. W. Boyd according to a literature method.^* The solvents were refluxed overnight and distilled or vacuum distilled both outside and inside the glovebox before passed through an activated neutral alumina column and stored with 3 A molecular sieve. Pyridine was distilled from sodium metal. Toluene, THF and hexane were all distilled from sodium benzophenone ketyl. Benzonitrile and o-dichlorobenzene were vacuum distilled from sodium metal and DMSO was distilled from CaHz 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EPR spectra were recorded on a Broker ER 200D-SRC spectrometer equipped with an Oxford Instruments ESR 900 cryostat, and ^-values were determined by calibration to DPPH powder. Vis-NIR spectra were recorded on a Shimadzu UV-260 or an Ocean Optics Spectrometer. NMR spectra were recorded at room temperature on a Broker AC-250, AM-360 or AMX-SOO-MHz spectrometer and the chemical shifts were calibrated against internal solvent or a sealed acetone-d^ capillary. C g o (200 mg, 0.0278 mmol) was refluxed with sarcosine (A^-methylglycine) (50.0 mg, 0.561 mmol) and paraformaldehyde (42.0 mg, 1.40 mmol) in 150 ml toluene for about 7 h. After the solvent was removed under reduced pressure, the product mixture was dissolved in ca. 5 ml ODCB and loaded into a silica gel (70-230 mesh) column. Using toluene as the eluent, the product was isolated as a brown solid (72 mg, 33%). ‘H NMR (CS2/CDCI3) Ô 3.00 (s, 3H, A T-CH s), 4.40 (s, 4H, CHj). NMR (CS2/CDCI3) Ô 41.29 69.81 (-CH2-), 70.95 (Cgo-sp"), other fullerene peaks: 136.02, 139.95, 141.65, 141.84, 141.98, 142.40, 142.87, 144.30, 145.03, 145.21, 145.42, 145.75, 145.81, 145.99, 147.03, 154.55. 2 , C fio(C H 2)2NC Ph3 The starting material 3-triphenylmethyl-5-oxazolidine was synthesized according to the literature method." In ethanol (45 ml), AT-tritylglycine 1 2 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PhsCNHCHzCOzH (0.615 g, 1.94 mmol) and 37% aqueous formaidhyde solution (0.90 ml, 12.0 mmol) were mixed and stirred at room temperature for 3 h. After all the solvent was evaporated, the white residue was dissolved in CH2CI2 (ca. 90 ml) and dried over MgS0 4 . Recrystallization o f the solid from benzene/hexanes afforded crystalline product (0.474 g, 74%). NMR (CDCI3) Ô 3.58 (s, 2H, 2 -CH2), 5.14 (s, 2H, 4 -CH2), 7.1-7.5 (m, 15H, trityl). NMR (CDCI3) 5 48.70 (2 -CH2), 76.08 (trityl quaternary), 83.44 (4 -CH2), phenyl ring; 127.22, 128.32, 128.70, 142.38. C eo (200 mg, 0.0278 mmol) was refluxed with 3-triphenylmethyl-5-oxazolidine (55.0 mg, 0.167 mmol) in PhCl (ca. 200 ml) overnight. After the solvent was removed under reduced pressure, the product mixture was dissolved in ca. 5 ml ODCB and loaded into a silica gel (70-230 mesh) column. With mixed solvents of /7-xylene and hexanes (2:5 v/v) as the eluent, the product was isolated as a brown solid (90 mg, 32%). 'H NMR (CS2/CDCI3), Ô 4.17 (br, 4H, CH2), phenyl ring. 7.29 (m, 3H), 7.42 (m, 6H), 7.89 (d, 6H, / = 6.96 Hz). "C NMR (CS2/CDCI3), 5 60.76 (CH2), 69.38 (trityl quaternary), 73.97 (Cgo-sp^), three o f the phenyl peaks: 126.77, 128.04, 129.15; one o f the phenyl peaks and other fullerene peaks: 136.42, 140.10, 141.77, 141.94, 142.22, 142.53, 142.95, 144.44, 145.18, 145.35, 145.52, 145.96, 146.04, 146.15, 147.16, 154.63. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (C p2C oi[n 1 (25 mg, 0.032 mmol) was stirred with cobaltocene (6.2 mg, 0.033 mmol) in benzonitrile (2 ml) for several hrs. Half of the solvent was removed under vacuum and toluene was added to effect precipitation o f the product. This solid was filtered off, washed with toluene several times and pumped to dryness (30 mg, 97%). NMR (DMSO-ajs) 5 1.4 (br, 4H, CHz), 4.1 (br, 3H, CH3), 5.8 (lOH, CpzCo"). NMR (DMSO-dk) 5 85.1 (CpzCo^, 187 (Ceo-pyrrolidine anion). Vis-NIR (in benzonitrile) X b u x (e) = 1002 (7.8 X 10^), 882 (3.6 x 10^), 780 (3.0 x 10^) nm. Elemental analysis; calcd. for C73H17NC0: C, 90.68; H, 1.77; N, 1.45. Found: C, 85.04; H, 1.48; N, 1.83%. [Na(crown)(THF)2^ [I*] The dissolution o f 1 (10.0 mg, 0.013 mmol) in THF (20-30 ml) afforded a reddish brown solution. After dibenzo-18-crown-6 (4.6 mg, 0.013 mmol) and a piece of fi*eshly cut Na metal was added, the solution was stirred while gently warmed for one hour before the color turned darker. Half of the solvent was evaporated under reduced pressure and after hexane was added, a microcrystalline precipitate was formed. After filtration, the piece of unreacted Na metal was removed and the solid product was washed several times with hexane and pumped to dryness. 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [CpîCo*][21 This is prepared in the similar manner to [Cp2C o^[l'] using 10.0 mg o f 2 (9.9 x 10'^ mmol) and 2.0 mg o f cobaltocene (0.011 mmol). NMR (PbNOi-f^) 5 1.2 (CHz), 5.9 (CpzCol, 7.2 and 7.6 (trityl). "C NMR (PhNOz-f^) 8 86.3 (CpzCo"), 188 (Ceo-pyrrolidine anion). [Na(crown)(THF>2^z[l* l To a solution o f 1 (40.0 mg, 0.0514 mmol) in THF (30-40 ml), dibenzo-18- crown-6 (37.1 mg, 0.103 mmol) and a piece o f freshly cut Na metal was added. The mixture was stirred while gently warmed for 12 h. The solution first turned dark reddish brown, finally to a dark greenish brown suspension. Part of the solvent was removed and hexane was added to effect complete precipitation. After the mixture was filtered and sodium was removed, the solid product was washed several times with hexane and pumped to dryness. X-ray Structure of 1^' Black single crystals of [PPN^zCl^l were obtained by vapor diftusion o f hexane into a mixture o f [Na(crown)(THF)2l 2[l^*] and PPNCl in benzonitrile/toluene. The X- ray data collection was accomplished by R S. Lu and Prof R Bau and the structure was solved by Dr. T. Drovetskaya and Prof R. Bau. 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The crystal data of [PPhT]2[l^l*toluene are listed in Table IV.2. Intensity measurement were performed at -110®C on a SIEMENS P4/RA difoatometer using graphite-monochromated Cu K a radiation (X = 1.54178 A) to a maximum 20= 102.5® . Data were collected in omega scan mode. A total of 10366 reflections were recorded, of which 8894 were unique. The structure was solved by direct methods (SHELXTL IRIS) and the partially restrained model was refined against F (SHELX 76) yielding R = 10.41%. Table IV.2 Crystal data o f [PPN^zCl^"] toluene Formula C 142H7 5. 5N3P 4 c(A ) 19.801 (2) Formula weight 1947.6 a(deg) 80.21 (2 ) Crystal size (mm) 0.3 X 0.2 X 0.25 >9(deg) 71.03 (2) Crystal system Triclinic r(deg) 84.25 (2) Space group P T Z 2 a (A) 14.869 (2) V (k^) 4637.9 (10) *(A ) 16.923 (2) Density (calc.) 1.395 g cm' l^D m m er, 4 To a suspension of C6o(CH2)2NCPh3 2 (150 mg, 0.149 mmol) in CH2CI2 (30 ml), triflic acid (0.4 ml) was added and the mixture was stirred for 2 h. After the solution volume was reduced, Et2 0 (10 ml) was added and the precipitate was 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. centrifuged, washed several times with EtiO, benzene and EtzO again. The brownish solid 3 was pumped to dryness, taken into glovebox. Its suspension in ODCB (IS ml) was converted to a reddish brown solution upon adding pyridine (2.4 ml). To this, 1,3- C6H4(C0C1)2 (15.1 mg, 0.0744 mmol) was added, and after 8 h stirring, the solvent was removed under reduced pressure and product was washed with methanol and hexane several times then pumped to dryness (119 mg, 96%). NMR (CSz, CDCU) 5 5.56 (br, 8H, GHz), 7.80 ( t,J = 7.7 Hz, IH), 8.10 (dd,J=7.6H z,y = 1.6 Hz, 2H), 8.36 (t, y = 1.6 Hz, IH). NMR (CSz, acetone-dk as insert) Ô phenyl ring; 129.06, 129.45, 131.35, 135.65; fullerene peaks: 136.14, 140.55, 142.18, 142.33, 142.44, 142.96, 143.36, 144.68, 145.57, 145.63, 145.71, 145.91, 146.37, 146.58, 147.53, 153.39; C=0 : 168.26. 1,4-Dimer, 5 This was prepared in a similar manner to 4 using l,4-C6H4(COCl)2. ‘H NMR (CSz, acetone-<4 as insert) Ô 4.84 (br, 8H, GHz), 7.32 (s, 4 H, GgH»). * ^ G NMR (GSz, acetone-dk as insert) Ô phenyl ring: 129.15, 137.61; ftiUerene peaks: 136.18, 140.57, 142.20, 142.35, 142.47, 142.96, 143.38, 144.70, 145.58, 145.64, 145.71, 145.92, 146.37, 146.58, 147.55, 153.35. 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 (20.0 mg, 0.0199 mmol) was treated with triflic acid in the same way as preparing 4. The brown residue 3 was taken into the glovebox and suspended in ODCB (ca. 5 ml). After the addition o f (MegS^NLi (7.0 mg, 0.042 mmol) followed by stirring for 20 min., the suspension turned to a reddish-brown solution. To this, 0.98 ml o f l,3-C6H4(CH2Br)2 solution in ODCB (1.00 x 1 0 * ^ M, 9.8 x 10'^ mmol) was added, and after overnight stirring, the reaction mixture removed from the drybox and the solvent was evaporated under vacuum. The solid was washed several times with methanol, hexane and methanol again and pumped to dryness. 'H NMR (CSz. acetone- ( U as insert) Ô 2.73 (s, 8H, C H % in pyrrolidine ring), 7.32 (s, 4H, benzylic CHi), the signals o f the phenyl ring are overlaped with that of the ODCB residue. "C NMR (CS2, acetone-d5s as insert) Ô 65.0 (CH2 in pyrrolidine ring), 79.8 (Ceo-sp^?), phenyl ring (?): 125.9, 132.5, 136.0; fullerene peaks: 137.4, 140.5, 142.2, 142.3, 142.4, 142.8, 142.9, 143.4, 144.7, 145.5, 145.6, 146.1, 146.3, 146.5, 147.3, 155.1. [Ph2P(CH20H)2]Cl (5.6 mg, 0.020 mmol) was dissolved in a mixed solvent (6 ml) o f water-methanol (2:1). To the resulting clear colorless solution, EtsN (0.1 ml, 0.7 mmol) was added and a white suspension was formed. After stirring for 0.5 h, all the volatile species in the mixture were removed under vacuum. The nearly colorless solid (Ph2PCH20H) was then dissolved in ODCB (4 ml). 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 (20.0 mg, 0.0199 mmol) was treated with triflic acid in the same way as preparing 4. The brown residue 3 was suspended in ODCB (ca. 5 ml), then pyridine (0.3 ml) and the PhzPCHzOH solution was added consecutively. This mixture was refluxed under Ar atmosphere for 3 h. After removing all the solvent under vacuum, the solid residue was washed with methanol and hetane for several times. After work up, the product was barely soluble in any solvents. EPR Sample of 4 4 (5.0 mg, 3.0 X 10'^ mmol) and 2.50 x 10'^ M cobaltocene PhCN solution (0.50 ml, 1.3 X 10'^ mmol) was mixed in PhCN (2 ml) and stirred for several hours. The suspension was filtered and the supernatant was sealed in an EPR tube. EPR Sample of 4^ 4 (5.0 mg, 3.0 X 10'^ mmol) and 2.50 x 10'^ M cobaltocene PhCN solution (0.275 ml, 6.88 X 1 0 * ^ mmol) was mixed in PhCN (2 ml) and stirred for several hours. The suspension was filtered and the supernatant was sealed in an EPR tube. EPR samples o f S’ and 5^' were prepared in an identical manner. References and Notes 1 . (a) Stinchcombe, J.; Penicaud, A.; Bhyrappa, P.; Boyd, P. D. W.; Reed, C. A. J. Am. Chem. Soc. 1993, 115, 5212; (b) Khaled, M. M.; Carlin, R T.; Trulove, P. C ; Eaton, G. R ; Eaton, S. S. J. Am. Chem. Soc. 1994,116, 3465; (c) Boyd, P. D 1 3 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. w ; Bhyrappa, P.; Paul, P.; Stinchcombe, J ; Bolskar, R. D.; Sun, Y.; Reed, C. A. J. Am. Chem. Soc. 1995,117, 2907 and re&rences therein. 2. Wilson, S. R-; Wu, Y. J. Am. Chem. Soc. 1993,115, 10334. 3. (a) Baumgarten, M ; Gherghel, L. in Progress in Fullerene Research Kuzmany, H.; Fink, J.; Mehring, M.; Roth, S. eds. World Scientific; Singapore, 1994, pp. 384-388; (b) CHierghel, L.; Baumgarten, M Synth. Met. 1995, 70, 1389; (c) Baumgarten, M.; Gherghel, L. Appl. M agn. Reson. 1996, 11, 171. 4. lyoda, M.; Sasaki, S.; Sultana, F ; Yoshida, M.; Kuwatani, Y.; Nagase, S. Tetrahedron Lett. 1996, 37, 7987. 5. Jones, M. T.; ECadish, K. M ; Subramanian, R.; Boulas, P.; Vijayashree, M. N. Synth. M et. 1995, 70, 1341. 6. Brezova, V.; Gugel, A.; Rapta, P.; Stasko, A. J. Phys. Chem. 1996,100, 16232. 7. (a) Diederich, F.; Thilgen, C. Science 1996, 271, 317; (b) Hirsch, A. Synthesis 1995, 895; (c) Hirsch, A. The Chemistry o f the Fullerenes Thieme: Stuttgart, 1994. 8. Tsuge, O ; Kanemasa, S.; Ohe, M.; Takenaka, S. Bull. Chem. Soc. Jpn. 1987, 60, 4079. 9. 1,3-dipolar cycloaddition is the addition to an alkene double bond o f a three-atom, four-electron k system that has dipolar character. In most o f fullerene literature, this addition is classified to a [3 + 2] addition (See Ifirsch, A. The Chemistry o f the Fullerenes Thieme: Stuttgart, 1994, pp.87-107 and the references therein). Strictly speaking, it belongs to the category of [4 + 2], since the numbers should refer to those of it electrons rather than the atoms in the reacting components. Similarly, the so-called [1 + 2] cycloadditions should be [2 + 2]. (See Lowry, T. H ; Richardson, K. S. Mechanism and Theory o f Organic Chemistry Harper & Row: New York, 3rd. ed., 1987, pp. 848-971.) 10. Maggini, M.; Scorrano, G ; Prato, M. J. Am. Chem. Soc. 1993, 115, 9798. 11. (a) Zhang, X.; Willems, M.; Foote, C. S. Tetrahedron Lett. 1993, 34, 8187; (b) lyoda, M.; Sultana, F ; Komatsu, M. Chem. Lett. 1995, 1133; (c) Wilson, S. R.; Wang, Y.; Cao, J ; Tan, X. Tetrahedron Lett. 1996, 37, 775. 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12. (a) Maggini, M.; Karlsson, A.; Scorrano, G.; Sandona, G ; Famia, G ; Prato, M. J. Chem. Soc., Chem. Commun. 1994, 589; (b) Guldi, D M.; Maggini, M.; Scorrano, G ; Prato, M J. Am. Chem. Soc. 1997, 7/9, 974. 13. (a) Maggini, M.; Dono, A.; Scorrano, G ; Prato, M. J. Chem. Soc.. Chem. Commun. 1995, 845; (b) SariciAd, N. S.; Wudl, P.; Heeger, A J ; Maggini, M.; Scorrano, G ; Prato, M. Bourassa, J.; Ford, P. C. Chem. Phys. Lett. 1995, 247, 510. 14. (a) Drovetskaya, T.; Reed, C. A ; Boyd, P. D. W. Tetrahedron Lett. 1995, 36, 7971; (b) Imahori, H.; Sakata, Y. Chem. Lett. 1996, 199; (c) Akiyama, T.; Imahori, H ; Ajawakom, A.; Sakata, Y. Chem. Lett. 1996, 907. 15. Martin, N.; Sanchez, L ; Seoane, C ; Andreu, R.; Garin, J ; Orduna, J. Tetrahedron U tt. 1996, 37, 5979. 16. lyoda, M.; Sultana, F ; Kato, A ; Yoshida, M ; Kuwatani, Y.; Komatsu, M ; Nagase, S. Chem. Lett. 1997, 63. 17. Williams, R. M ; Koeberg, M ; Lawson, J. M.; An, Y.-Z.; Rubin, Y.; Paddon-Row, M N.; Verhoeven, J. W. J. Org. Chem. 1996, 61, 5055. 18. Corvaja, C ; Maggini, M ; Prato, M ; Scorrano, G.;Venzin, M. J. Am. Chem. Soc. 1995,117, 8857. 19. (a) Maggini, M ; Scorrano, G ; Bianco, A ; Toniolo, C ; Sijbesma, R.; Wudl, F ; Prato, M. J. Chem. Soc., Chem. Commun. 1994, 305; (b) Bianco, A ; Maggini, M.; Scorrano, G ; Toniolo, C ; Marconi, G ; Villani, C ; Prato, M. J. Am. Chem. Soc. 1996, 118, 4072. 20. (a) Friedman, S. H.; DeCamp, D. L.; Sijbesma, R. P.; Srdanov, G ; Wudl, F ; Kenyon, G. L. J. Am. Chem. Soc. 1993, 115, 6506; (b) Sijbesma, R _; Srdanov, G ; Wudl, F.; Castoro, J. A ; VHlkins, C ; Friedman, S. H ; DeCamp, D. L.; Kenyon, G. L. J. Am. Chem. Soc. 1993, 115, 6510; (c) Toniolo, C ; Bianco, A.; Maggini, M ; Scorrano, G ; Prato, M.; Marastoni, M ; Tomatis, R.; Spisani, S.; Palu, G ; Blair, E. D. J. M ed. Chem. 1994, 37, 4558. 21. (a) Signorini, R.; Zerbetto, M.; Meneghetti, M ; Bozio, R.; Maggini, M.; Faveri, C. D ; Prato, M.; Scorrano, G. Chem. Commun. 1996, 1891; (b) Agostini, G ; Corvaja, C ; Maggini, M ; Pasimeni, L ; Prato, M. J. Phys. Chem. 1996, 100, 13416. 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22. (a) Maggini, M.; Karlsson, A.; Pasimeni, L.; Scorrano, G.; Prato, M.; Valli, L. Tetrahecb’ on Lett. 1994, 35, 2985; (b) Maggini, M.; Pasimeni, L.; Prato, M.; Scorrano, G ; Valli, L. Langnmir 1994, 10, 4164. 23. (a) Suzuki, T.; Maruyama, Y.; Akasaka, T.; Ando, W.; Kobayashi, K.; Nagase, S. J. Am. Chem. Soc. 1994, 116, 1359; (b) Ohno, T.; Martin, N.; Knight, B ; Wudl, F.; Suzuki, T.; Yu, H. J. Org. Chem. 1996, 61, 1306. 24. Prato, M.; Maggini, M.; Giacometti, C ; Scorrano, G ; Sandona, G ; Famia, G Tetrahedron 1996, 52, 5221. 25. (a) Sun, Y -P ; Ma, B ; Bunker, C. E ; Liu, B. J. Am. Chem. Soc. 1995, 117, 12705; (b) CoUigiani, A.; Taliani, C. Chem. M ater. 1994, 6, 1633. 26. Fischer, J. E. Science 1994, 264, 1548 and references therein. 27. (a) Compton, R. G ; Spackman, R. A.; Wellington, R. G ; Green, M. L. H.; Turner, J. J. Electroanal. Chem. 1992, 327, 337; (b) Moriyama, H.; Kobayashi, H.; Kobayashi, A.; Watanabe, T. Chem. Phys. Lett. 1995, 238, 116. 28. Zhu Q ; Cox, D. E ; Fischer, J. E. Phys. Rev. B. 1995, 51, 3966. 29. (a) Martin, M. C ; KoUer, D ; Rosenberg, A.; Kendziora, C ; Mihaly, L. Phys. Rev. B. 1995, 51, 3210; (b) Peit, P.; Robert, J.; Fischer, J. E. P lys. Rev. B. 1995, 51, 11924. 30. (a) Stephens, P. W ; Bortel, G ; Faigel, G ; Tegze, M ; Janossy, A.; Pekker, S.; Oszlanyi, G ; Forro, L. Nature 1994, 370, 636; (b) Pekker, S.; Janossy, A.; Nfihaly, L ; Chauvet, O ; Carrard, M ; Forro, L. Science 1994, 265, 1077. 31. (a) Suzuki, T.; Li, Q ; Khemani, K. C ; Wudl, F.; Almarsson, O J. Am. Chem. Soc. 1992, 114, 7300; (b) Belik, P.; Gugel, A.; Spickermann, J.; Mullen, K Angew. Chem., Int. Ed. Engl. 1993, 32, 78; (c) Gugel, A.; Kraus, A.; Spickermann, J ; Belik, P.; Mullen, K. Angew. Chem., Int. E d Engl. 1994, 33, 559; (d) Anderson, H. L.; Faust, R.; Rubin, Y ; Diederich, F. Angew. Chem., Int. E d Engl. 1994, 33, 1366; (e) Diederich, F.; Dietrich-Buchecher, C ; Nierengarten, J.-F.; Sauvage, J.-P. J. Chem. Soc., Chem. Commun. 1995, 781; (f) Paquette, L. A.; Graham, R. J. J. Org. Chem. 1995, 60, 2958; (g) Smith E Q , A. B ; Tokuyama, H.; Strongin, R. M.; Furst, G. T.; Romanow, W. J ; Chait, B. T.; \firza, U. A.; Haller, I. J. Am. Chem. Soc. 1995, 117, 9359; (h) Lawson, J. M ; Oliver, A. M ; Rothenfluh, D. F ; An, Y.-Z.; Ellis, G. A _; Ranasinghe, M. G ; Khan, S. I ; Franz, A G ; Ganapathi, P. S.; Shephard, M. J.; Paddon-Row, M. N.; Rubin, Y. J. Org. 1 3 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chem. 1996, 61, 5032; (i) Timmerman, P.; Witschel, L. E ; Diederich, F ; Boudon, C ; Gisselbrecht, J.-P.; Gross, M. Helv. Chim. Acta 1996, 79, 6; 0) Komatsu, K.; Takimoto, N.; Murata, Y.; Wan, T. S. M ; Wong, T. Tetrahedron Lett. 1996, 37, 6153; (k) Osterodt, J ; Vogtle, F. Chem. Commun. 1996, 547; (1) de Lucas, A. I.; Martin, N.; Sanchez, L.; Seoane, C. Tetrahedron Lett. 1996,52, 9391. 32. Lu, Q ; Schuster, D. I.; Wilson, S. R. J. Org. Chem. 1996, 61, 4764. 33. Nagashima, H ; Nakaoka, A.; Saito, Y.; Kato, M.; Kawanishi, T.; Itoh, K. J. Chem. Soc., Chem. Commun. 1992, 377. 34. Ishi, Y ; Hoshi, H ; Hamada, Y ; K da, M. Chem. Lett. 1994, 801. 35. Stewart, W. E.; Siddall m , T. H. Chem. Rev. 1970, 70, 517. 36. Bryant, R. G. J. Chem. Soc. 1983, 60, 933. 37. Fawcett, J.; Hoye, P. A. T.; Kemmitt, R. D. W ; Law, D. J ; Russell, D. R. J. Chem. Soc., Dalton Trans. 1993, 2563. 38. ConneUy, N. G ; Geiger, W. E. Chem. Rev. 1996, 96, 877. 39. Trulove, P. C ; Carlin, R. T.; Eaton, G. R.; Eaton, S. S. J. Am. Chem. Soc. 1995, 117, 6265. 40. (a) [CpzColCCgoH CSz: Balch, A. L.; Lee, J. W ; Noll, B C ; Olmstead, M. M. Personal communication; (b) [PluP^CCeol'PluPCl: Bilow, U.; Jansen, M. J. Chem. Soc., Chem. Commun. 1994, 403; (c) [PPN*][C6oTPhCl. Kobayashi, H ; Moriyama, H ; Kobayashi, A.; Watanabe, T. Synth. M et. 1995, 70, 1451. 41. (a) [Na(crypt)"]2[C6o ^ * ] : Sun, Y ; Drovetskaya, T.; Bau, R.; Reed, C. A Unpublished result; (b) [K(crypt)'l2[C6o^']*4toluene: Fassler, T. F.; Spiekermann, A.; Spahr, M. E ; Nesper, R. Angew. Chem., Int. E d Engl. 1997, 36, 486. 42. K3C6o(THF)m: Janiak, C ; Mûhle, S.; Hemling, H. Polyhedron 1996, 15, 1559. 43. Wan, W. C ; Liu, X.; Sweeney, G. M ; Broderick, W. E. J. Am. Chem. Soc. 1995, 117, 9580. 44. Paul, P.; Jfle, Z ; Bau, R.; Boyd, P. D. W ; Reed, C. A. J. Am. Chem. Soc. 1994, 116, 4145. 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45. A similar problem was encountered in a fullerene derivative structure: Paulus, E. F ; Bingel, C. Acta Cryst. 1995, C57, 143. 46. Similar disorder was observed in C eo 4CgH6: Balch, A. L.; Lee, J. W.; Noll, B. C.; Olmstead, M. M. J. Chem. Soc., Chem. Commun. 1993, 56. 47. For example: (a) Rubin, Y.; Khan, S.; Freedberg, D I.; Yeretzian, C. J. Am. Chem. Soc. 1993, 115, 344; (b) Balch, A. L.; Cullison, B ; Fawcett, W. R.; Ginwalla, A S.; Olmstead, M. M. Winkler, K. J. Chem. Soc., Chem. Commun. 1995, 2287. 48. (a) Murray-Rust, P.; Murray-Rust, J. Acta Cryst. 1980, B36, 1678; (b) Tetzlafi^ C ; Butz, V.; Mlsmaier, E ; Wagemann, R ; Maas, G.; von Onciul, A R.; Clark, T. J. Chem. Soc., Perkin Trans. 2 1993, 1901; (c) Seibel, J.; VUsmaier, E ; Frohlich, K ; Maas, G ; Wagemann, R. Tetrahedron 1994, 50, 715. 49. Boyd, P. D. W. Unpublished results. 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter V. The Search for Hydridocobaloximes V .l. Introduction Transition m etal hydrides, a class o f com pounds w ith hydrogen atom (s) as the ligand, are o f extrem e im portance in preparative and catalytic rea ctio n s/ For instance, all the processes for storing energy by the photochemical^ o r electrochem ical^ hydrogen production from w ater splitting require catalysts, due to the extrem e high activation e n e rg y / Cobalt com plexes w ith N4 m acrocyclic ligands frequently act as efhcient hom ogenous catalysts in th ese processes and it is generally believed that the corresponding cobalt hydrides play a critical role. Am ong all the cobalt-N 4 com pounds, cobaloxim es are a class w ith a core unit o f bis(dim ethyIglyoxim ato)cobalt [Co(DH )2 or Co(dm gH )z] (se e Figure V.l). This equatorial m oiety is form ed by com plexing cobalt w ith tw o dim ethylglyoxim e (DH2 or dmgHz) ligands. Each ligand is m ono-deprotonated and is connected to the other through tw o intram olecular O -H O hydrogen bonds. T he resulting N4 unit is nearly co-planar and the cobalt atom resides in the center. The tw o axial coordination sites can be occupied by one neutral (L ) and one m ono-anionic ligand (X ^, or less com m only, by tw o L, or tw o X" ligands. Since they show sim ilar chem ical properties, cobaloxim es are w idely considered as m odel com pounds o f Vitam in B12 and have been extensively stu d ied / H ydridocobaloxim es, the hydride com plexes o f cobaloxim es (X = H ), are believed to 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. L = axial ligand, e g. phosphine, py, Im X = CI, Br, alkyl, etc., Co(III) n = I, Co(I) X = H, hydrides n = 0,CoÔl) F igure V .l The structures o f cobaloxim es. O - H - ® XCo(PBF2)2PBu3 XCo (DH)(DBF2)PBu3 F igu re V.2 The m ono- and bis-BFz bridged cobaloxim es. 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. be involved in photochem ical hydrogen generation,*^ or many chem ical reactions^ as interm ediates. They m ay only exist as transient species and a particular hydride H C o(D H )zPy w as concluded to be thermodynamically unstable.’* ’ H ow ever, w ith tributylphosphine (P B us) as the axial ligand, Schrauzer et al. reported the synthesis o f a stable hydridocobaloxim e HCo(DH)2PBu3 under ambient conditions.' A ccording to E spenson et al., hydrogen w ould evolve from HCo(DH)zPBu3 in acidic solutions and the kinetics o f this process had been studied.^ The unusual stability and the interesting reactivities served as an im petus for us to reinvestigate this important chem ical com pound. The Co(DH )2 m oiety is not very stable under certain conditions. For instance, in acidic solution, one (and probably both) o f the tw o O -H O bridges can be cleaved b y protonation, * * * and th e com plexes are subject to com plete decom position by acid to produce free Co^ .* * W hile in basic solution, the bridging H (pA* - 11) can be deprotonated.*^ Therefore, it is desired to find a m ore stable system w ith a sim ilar structure. For transition m etal bis(dim ethylglyoxim ato) com plexes, the bridging H could b e replaced by B F j,”’ BRz (R = alkyl," phenyl*'* *^), BHz,*^ BBN ,*^ C H z*** or even other metal com plex m oieties.*’ W e are particularly interested in the BFz-bridged cobaloxim es (X C o(D B Fz)2L ). The BFz-bridged cobaloxim es can be readily prepared and are more resistant to th e protonation and acid decom position com pared to their parent com plexes.*'*^ The electron-withdrawing characteristics o f the BFz groups w ill low er the electron density o f the cobalt center and m ay probably alter the reactivity o f 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the corresponding hydride com plex. It is known that th e Co(DBp2)2 species could act as a catalyst in the process o f hom ogeneous proton reduction and Hz evolution. During the study, a convenient method w as foim d for the preparation o f C1C o(D H )(D BF2)PB u3, in w hich only one o f the bridging H is replaced by BFz group. It w ill be interesting to conduct parallel studies on Co(DH>2, Co(D H )(D BF2) and C o(D B F 2)2 species and to investigate the electronic and steric effects o f the BFz group on th e stability o f the corresponding hydride com plexes. V.2. Results and Discussion V.2.1. Starting Materials: Synthesis and Reactivities The synthesis m ethod o f XCo(DH)2L reported by Schrauzer et al. is simply m ixing Co^^ salt w ith dim ethylglyoxim e and the base L in the air.^ H owever, according to M arzilli et al., the products obtained in this w ay are usually [L2C o(DH )2][C o(D H )2X2].^^ Our results confirm this statem ent, and therefore, the preparation o f the starting m aterials presented here generally follow ed M arzilli’s m ethod. The synthesis o f ClCoCDBFzhPBus follow s th e reported m ethod by reacting C1Co(DH )2PB u3 w ith BF3 OEtz in EtzO .^ The product generated after a long reaction tim e contains tw o portions, the CH2CI2 soluble part and insoluble part. They are roughly in equal am ount. The orange, CHzClz-soluble part is our goal product, as indicated by the N M R spectra. W hereas, the main com ponent o f the yellow , CH2CI2- 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. insoluble part is probably protonated species, since fT released upon BFz substitution w ill be a strong acid in E tzO A conductivity measurement (in CH3CN) reveals its ionic nature. C onducting the synthesis w ith addition o f a base, such as E tsN or pyridine, no m ore CH zClz-insoluble portion is form ed and the product is quite clean after washing w ith E tzO and HzO N o axial ligand exchange is observed during this process, which is consisten t w ith the report that PBus coordination is therm odynam ically favored.^ T he substitution o f the bridging H by the BFz group cou ld b e follow ed by H N M R , since th e three cobaloxim e sp ecies involved in this reaction, including the starting material ClCo(DH)zPBu3, the interm ediate ClCo(DH )(DBFz)PBu3 and the final product C1Co(DBFz)zPBu3, show distinct dm g methyl resonances. In E tzO , overnight reaction tim e is necessary for com plete conversion o f ClCo(DH)zPBu3 to the corresponding bis-BFz species. Frequently m ono-BFz bridged species C1Co(DH)(DBFz)PBu3 could be obtained after several hours reaction tim e. In other solvents, e .g ., CHCI3 or THF, in w hich all three com plexes are readily soluble, a long reaction tim e is also needed. An attem pt to prepare C1Co(DBFz)zPBu3 in CHCI3 by strictly follow in g the published m ethod (3-4 h)^* only afforded a m ixture o f m ono- and bis-BFz com pounds. T he reported synthesis o f ClCo(DH )(DBFz)PBu3 w as on ly in 13% yield^^ and frequently produces mixture o f tw o or three species. A fter a system atic search, suitable reaction conditions w ere found to prepare this m ono-BFz bridged species in a 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reasonable yield. This opens the w ay to parallel study o f the three cobaloxim e com pounds. Reactivities T he reaction o f ClCo(DBF2)2L (L = PBus, Py) w ith organic bases, e.g ., E tsN , leads to its reduction. The solution color changes to purple and then im m ediately to blue after th e bases are added, suggesting the form ation o f C o(II) and C o(I) species. The addition o f the inorganic base OIT to C lC o^ B F 2)zL (L = PBus, Py) in E tO H also causes a colo r change to blue. The reduction probably originates in the stabilization o f the cobaloxim e at low er oxidation states by the strongly electron-withdrawing BF2 groups. T he lack o f similar reactivities for ClCo(DH )(DBF2)PBus or ClCo(DH)2PBu3 is consisten t w ith this statem ent. It has been reported that the reaction o f a variety o f bases, including organic and inorganic ones, w ith (CHs)Co(DBF2)2Py w ill cleave th e CH3-C0 bond and generate dark blue C o (I).^ It is also noticed that the organic bases w ill induce auto-reduction o f C o(II), N i(III) and Fe(III)-N 4 m acrocyclic com plexes.^ C1Co(DBF2)2PBu3 is unstable in protic media. Its BF2 groups w ill dissociate, i f kept in m ethanol overnight, resulting in ClCo(DH )(DBF2)PBu3. Longer standing w ill even generate ClCo(DH)2PBu3 B y follow in g its U V -vis spectra and by th e com parison to that w ith addition o f Ag^ salt, the axial chloride w ill slow ly dissociate after ClCo(DBF2)2PBu3 is dissolved in M e0 H /H 2 0 and within 30 min. a considerable 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. amount of solvolyzed species will form. Similar behavior has also been observed for C1Co(DH)2PBu3, which is consistent with an earlier report.^ V.2.2. Reinvestigation of ^^Hydridocobaloximes”: Synthesis and Characterization ofCo(H) V.2.2.1. Reduction of CICo(DH)2 PBu3 with NaBH* Hydridocobaloximes have been reported to be synthesized by the reduction of Co(iU) cobaloximes with NaBH» in MeOH/HzO at pH 7.^ Our studies on these hydrides began by reexamining them under the identical synthetic conditions. In our hands, the preparations essentially followed the report but were carried out in a more cautious manner. Standard Schlenk-line techniques were employed, in order to prevent the possible oxidation of the desired product by the air. Freshly collected deionized water was used as the solvent. Before the reaction was started, the mixture (the starting material and the buffer solution) was degassed by bubbling Ar through the solution. A Schlenk flask containing this mixture was then assembled with a filter and a side arm loaded with solid NaBH». After this Schlenk assembly was thoroughly purged by a continuous Ar flow, solid NaBH4 was added fi-om the side addition arm, thereby eliminating the possible introduction of 0 % that could occur had the aqueous NaBH* solution been added via syringe. To reduce the possibility of hydride decomposition, the reaction was immediately filtered and after 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. filtration no washing o f the product with water was attempted. The solvents were removed under vacuum and the whole sealed assembly was transferred into a glovebox. The product of reduction o f ClCo(DH)2PBu3 by NaBH» essentially contains two portions: the soluble part and the insoluble part in MeOH/HiO mixed solvent. They could be separated by a simple filtration. After removing the solvents, the soluble part solid sample was purified by washing with toluene or hexane. No iQrdride peak (5 6.0 ppm) could be detected in its H NMR. The sharp resonances (Figure V.3) indicate that the complex is diamagnetic. This species can only be dissolved in polar solvents, giving dark blue solutions. The blue color and the intense UV-vis absorbance at 610 nm in MeOH (Figure V.4) strongly suggest that this is a Co(I) species. It has been extensively reported that the mono-valent cobaloxime species are dark blue in color and the most intense UV-vis band is around 610 n m .^ ^ * ^ * A recent Co(I) x-ray structure confirms this general belief that the deep blue color is associated with the Co(I) cobaloximes.^ V.2.2.2. Characterization of **Hydridocobaloxmie’ ’ The other portion of the reduction product is insoluble in the mixed solvent of MeOH/HzO and could be separated fi’ om the remaining phosphate, NaBH» and possibly the unreacted starting material by detraction with hexane. A dark violet solid was obtained after removing the solvent. This was done very quickly and no indication 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.5 2.0 1.5 0.5 0.0 1.0 -0.5 ppm Figure V.3 ‘ H NMR of Co(DH)zPBu3' (generated by NaBHt) in MeOH-rf» (the missing bridging H signal is presumably due to the deprotonation or the proton exchange with the solvent). 618 330 403 474 543 692 Wavelength (nm) Figure V.4 UV-vis spectrum of Co(DH>2PBu3’ (generated by NaBH*) in MeOH. 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of decomposition was observed. Compared to the properties described in the original report,' the violet product obtained in this way is probably the so-called hydridocobaloxime “HCo(DH)2PBu3”. The isolated product appears to be indefinitely stable under anaerobic conditions. It is readily soluble in many solvents, including hydrocarbons (alkanes, benzene, toluene), ethers, alcohols, nhriles, ketones, halocarbons (CH2CI2), etc. It has very limited solubility in DM SO and is sparingly soluble in H2O. The UV-vis spectra o f this product show several interesting properties. In methanol, the solution color appears to be dependent on the concentration. The relatively concentrated solutions (e.g., 4 mM) are of a purple color and the spectra show an intense absorption band around 570 nm, while the dilute solutions (e.g., 0.4 mM) have a pink color and the spectra show a shoulder peak at 510 nm (Figure V.5). The appearance o f the 570 nm band is reversible since this band will develop upon adding solid ‘ TICo(DH)2PBu3” to the dilute solution while adding more methanol could cause this peak disappear. This interconversion is not due to the small amount of oxidizing impurity in the solvent, since cooling the dilute solution will also reversibly convert the pink species to the one showing the 570 nm band. In other donor solvents, such as CH3CN, THF, acetone and benzene, “HCo(DH)2PBu3” exhibits similar spectral concentration dependence. In contrast, in hexane or Et2 0 , no matter how low the concentration of “HCo(DH)2PBu3” is, an absorbance band around 570 nm always exists and a dilution only causes its relative intensity to decrease. 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I I 1.5 4 mM, in 1 mm cell 1 0.5 0.4 mM, In 1 cm cell 0 350 422 493 563 638 711 Wavelength (nm) Figure V.5 UV-vis spectra o f “HCo(DH)2PBu3” in MeOH at different concentrations. dry Ice/EtOH RT 330 403 474 543 618 692 Wavelength (nm) Figure V. 6 UV-vis spectra of “HCo(DH)2PBu3” in MeOH at different temperatures. 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Another spectroscopic characteristic o f this species is its thermochromism, i.e., its UV-is spectra are temperature dependent.^° The pink color of a dilute solution in a donor solvent will turn to purple when the temperature decreases. As shown in Figure V.6 , a absorption band at 570 nm developed dramatically in intensity when the solution in methanol was cooled down. For a relatively concentrated solution or a solution in a non-donor solvent, the intensity of the absorption band around 570 nm will also increase dramatically and the solution color becomes much darker (from purple to blue- purple) (Figure V.7). Although the concentration or temperature variation alters the relative intensity o f band at 570 nm substantially, little effect could be observed on the rest o f the UV-is spectrum. dry ice/EtOH i I < RT 250 324 397 468 538 Wavelength (nm) 612 686 Figure V.7 UV-is spectra o f "HCo(DH)zPBu3" in hexane at different temperatures. 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Under certain conditions, for instance, at high concentration or at a low temperature, the color o f "HCo(DH)zPBu3" is purple or blue purple. In this situation, the color and the absorption band around 570 nm are close to those o f Co(DH)zPBu3 , but it is definitely discernible fi'om this species. Besides other differences, with sufficiently caution th ^ are distinguishable by their colors. The color of “HCo(DH)2PBu3” is either pink, purple or blue-purple while Co(DH)2PBu3 is (greenish-) blue. The intense absorption band of “HCo(DH)2PBu3” is around 570 nm while that of C o(D I^PBu3' is at 610 nm. The ‘ H NMR spectrum of "HCo(DH)zPBu3" shows only broad phosphine peaks at room temperature. The spectra remain unchanged as the temperature increases (up to 50°C). When the sample was cooled down, however, the dmg methyl and the bridging H resonances sharpened up (Figure V.8 ). In contrast to the original report,' no hydride peak at 6 ppm or any other peak between 40 to -100 ppm could be observed under all conditions attempted. The variable temperature NMR spectra are shown in Figure V.9. At room temperature, two peaks could be detected at 43 and 82 ppm after an overnight data accumulation (53840 scans). A very broad peak around - 8 ppm could be seen at -45°C and it sharpened up with decreasing temperature. It should be noted that those room temperature peaks are much lower in intensity relative to the peak ( - 8 ppm) at low temperature. The former peaks become noticeable only after a very long acquisition time while at low temperature (e.g., -75®C) the peak at - 8 ppm is readily discernible 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -60*C -40"C -30"C -3”C A RT 16.0 P P S JL _ _ y x _ A p m ! jTT r i ’ ! I j !"! I rj : i I I j I ! n j : : : I j : : I 1 j 1 1 : I j : : ! I j Ï : : 1 j ! : : I <.£ 4.0 3.5 3.0 2.5 2.0 1.5 1.0 .5 0.0 -.5 ppy. Figure V.8 Variable temperature ‘H NMR spectra of "HCo(DH)zPBu3" in toluene-d, (the peak at 2.09 ppm is from the solvent). 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 3 CD ■ o O Q . C s Q . ■ o CD K ' O 3 O CD 8 ■ O 3. C O 3 CD CD CD ■ D O Q . C & . o D ( D Q _ O C T 3 ( D U ) o' D I 1 I I I I I I I I I I I I I 1 ------- 1 ------- 1 ------- 1 ------- 1 ------- 1 ------- 1 ------- 1 ------- 1 ------- 1 ------- 1 ------- r 150 100 50 0 -50 ppm Figure V.9 Variable temperature ^ ‘P NMR spectra of “HCo(DH)2?Bu3” in toluene-<4 (from bottom to top: temperature, number of scans: RT, 53840; -45°C, 6008; -60°C, 8200; -65°C, 2056; -75°C, 800; -85® C , 600). L A after several hundred scans. Apparently, the latter peak is from the main component in the NMR sample while those at 42 and 82 ppm can be ascribed to a trace o f impurities. Both H and ^ P NMR indicate that “HCo(DH)2PBu3” is likely a paramagnetic species. Its magnetic moment in solution was measured by the Evans method (see below) and at room temperature fMs was found to be 2.4 Bohr magneton (BM). This value decreases slightly with decreasing temperature (Figure V. 10). 2.4 i 2 . 1 1.8 210 240 270 Temperature (K) 300 Figure V.iO Magnetic moments o f “HCo(DH)2PBu3” in acetone-dk measured by the Evans method at different temperatures. Molecular weight determinations o f "HCo(DH)zPBu3" were attempted. No reliable data could be obtained in cyclohexanol by the freezing point depression method 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. due to the supercooling o f the solvent. The measurement by solution vapor pressure equilibrium with a known compound appears not to be very precise either (Table V. 1). Table V .l Molecular weight of “HCo(DH)2PBu3” determined by the method of vapor pressure equilibrium of solutions (solvent: acetone) Volume o f 10.0 mg ClCo(DH)zPBu3 (ml) 2.77 Volume of 10.0 mg “HCo(DH)2PBu3” (ml) 3.50 MW of “HCo(DH)2P 417 2.95 3.31 470 2.83 3.43 435 2.91 3.33 460 2.80 3.45 428 2.78 3.47 422 2 . 6 8 3.56 396 average MW 433 Shiny, well-shaped crystals of "HCo(OH)zPBu3" were obtained by slow solvent evaporation from its hexane solutions. Unfortunately, none o f these dif&acted X-ray very well. Supplement: The Evans Method for Determining Magnetic Moment of Paramagnetic Species in Solution^* The H NMR chemical shift of a molecule depends on the bulk susceptibility of the medium. This property can be employed to determine the magnetic moment of a paramagnetic species in solution. An inert compound is used as a reference. A sealed 1 5 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. capillary with this compound in a pure solvent is placed in a solution in an NMR tube containing the reference at the same concentration and the paramagnetic species of concern. The reference compound in the two different media will exhibit distinct H NMR resonances. For high field NMR spectrometers with superconducting magnets when the operating fi-equency is vb, the fi-equency separation Av of the reference signals and the volume susceptibility % of the NMR solution will be related by the equation (in cgs units) Av _ 4;z% V o 3 The mass susceptibility Xe the dissolved paramagnetic species can be calculated by the following ^ 3000AV Zb =^o*g+ ------— A tw q cM where % b * g is the mass susceptibility of the solvent, c is the concentration o f the NMR sample in moles/liter, and M is the molecular weight o f the paramagnetic molecule. The molar susceptibility could be obtained after the diamagnetic correction Finally, the magnetic moment (in Bohr magneton) o f the paramagnetic species o f the concern will be Hes — 2.828 where Tis the temperature (in Kelvin). The diamagnetic susceptibility of a given molecule can be obtained^^ by 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. j j T o = + nxX i where n\ is the number o f atoms o f each type and is the contribution to the susceptibility o f each of the constituent atoms, A . is a constitutive correction that depends on the type of the bonds in the molecule. Tables o f Xi and A . are available in the literature.^^ V.2.2.3. b the Reported ‘‘H ydride” a Co(H) Species? Almost all the known octahedral C o(in) complexes are diamagnetic.^^ If the structural assignment of the species obtained by reduction of ClCo(DH)zPBu3 with NaBHt to “HCo(DH)2PBu3” was correct, formally it is a Co(III) complex and it will most likely be diamagnetic. In contrast, the broadness of the and ^^P NMR signals and the magnetic moment demonstrate that this compound is paramagnetic. All the properties suggest that it is more likely to be Co(DH)2PBu3, a Co(II) complex. The magnetic moment of 2.4 BM at room temperature fits well in the range of a low spin = 1/2) C o(n) system with considerable orbital contribution.^^ This value is the same as the reported fies of Co(DH)zPBu3 determined by the Gouy method.^* To rationalize all the results presented in Section V.2.2.2, an equilibrium in solution between Co(DH)2PBu3 and its dimer is proposed here (Scheme V .l.). The dimerization through Co-Co bond appears to be feasible since Co(DH)zPBu3 is a five- coordinate, 17 e' species.^^ A crystal structure o f a Co(II) dimer with planar N-donor ligands is also known.^^ Nevertheless, as indicated by all the experimental results, the 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. equilibrium constant (K) for the dimer formation is by no means large at room temperature, i.e., the equilibrium lies to the left and the concentration of the dimer is low. 2 C o (D H )(2PB u3 —----- — B u3P(D H >2C o------C oC D H ^zPB ug Scheme V .l The equilibrium between monomeric and dimeric Co(DH)2PBu3 . The Co-Co bond here is probably very weak, but its formation should be an exothermic process, i.e., < 0 . According to the van’t Hofif equation à ln K cT RT^ the equilibrium constant K will increase with decreasing temperature, i.e., the dimer formation is more favored at lower temperature. The monomeric d^ species Co(DH)2PBu3 is paramagnetic {S = 1/2). As two monomers approach each other, a Co-Co bond is formed via the pairing of the two electrons from the two Co(II) centers, leading to a diamagnetic dimer. The overall magnetic moment o f Co(DH)aPBu3 solution should be the weighted average of those values of the monomer and the dimer. At room temperature, fics is essentially the value of the monomer due to the low concentration o f the dimer. As the temperature decreases, the percentage o f the dimer will increase and therefore /tg will decrease. That is exactly what was observed in the measurement o f the magnetic moment by the 160 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Evans method (Figure V.IO). It is noticeable that a similar / 4 g-temperature curve shape has been obtained for the dimerization of a Co(II) hydride system in soiution.^^ The variable temperature 'H and ^ P NMR results are consistent with the decrease of the magnetic moment at lower temperature. The fact that the high temperature NMR spectra are indistinguishable from those at room temperature imply that all Co(DH)2PBu3 is essentially in monomeric form at these temperatures, i.e., the value o f K is very small. As the temperature decreases, more diamagnetic dimer will form and the NMR signals become sharper. A very weak EPR signal of the frozen solution of Co(DH)2PBu3 near 4 K suggests that at this temperature almost all the free Co(DH)2PBu3 has been converted to the EPR-silent dimer. For the UV-vis spectra, the shoulder peak around 510 nm may be due to the monomeric Co(DH)zPBu3 whereas, the absorption band around 570 nm is considered to be associated with the Co(II) dimer, probably related to the Co-Co bond. Relative to the 510 nm band (s = 2.3 x 10^ cm'^M"*), the one around 570 nm is much more intense. In a dilute solution in donor solvents (concentration < 0.4 mM), Co(DH)2PBu3 displays a light pink color, that o f the monomer. The presence o f even a trace of the dimer will be readily recognized, because it exhibits a much more intense absorption and therefore a much darker color (purple). In non-coordinating or weakly coordinating solvents (e.g., hexane or EtzO), the dimer will exist in a certain amount even when the concentration is very low. That is why a dilute solution of Co(DH)zPBu3 in hexane (or EtzO) shows a purple rather than 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pink color. In contrast, in a donating solvent, the free Co(DH)2PBu3 will be better solvated and stabilized so that in very dilute solutions, nearly all Co(DH)2PBu3 is in the monomeric form and hence the color is pink. As the concentration o f Co(DH)zPBu3 increases, more dimer will form and the solution will display a purple color. The dependence o f the UV-vis spectra of Co(DH)zPBu3 on the solvent (solvatochromism) and the concentration is reasonably rationalized. As discussed above, as the temperature decreases, the equilibrium in Scheme V .l will shift to the right and the concentration o f the dimer will increase and the solution color will turn from pink or purple to blue purple. This color change is reversible and the intensity increases dramatically with decreasing temperature. These spectroscopic phenomena are consistent with the NMR and magnetic moment results at variable temperature. A similar thermochromism has been observed in the Co(DBF2)zPBu3 sy ste m .* * ® All our data on Co(DH)zPBu3 support the mechanism o f the monomer-dimer equilibrium. Interestingly, a recent paper also presented a dimerization equilibrium of a cobalt system and similar characterizations and observations have been made.^^ V.2.2.4. Synthesis of Co(DH>2PBu3 : the **Hydride** is Co(II)! All the properties of ^HCo(DH)2PBu3" are well explained by the monomer- dimer equilibrium o f Co(DH)2pBu3 . In order to further confirm the structural 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. assignment o f “hydridocobaloxime”, several experiments were carried out. The reduction of ClCo(DH)zPBu3 was investigated by the titration with cobaltocene (CpzCo). As shown in Figure V.ll, the UV-vis spectrum obtained by the addition o f 1 equivalent of Cp^Co resembles that o f a dilute “HCo(DH)2PBu3” solution in MeOH. Further addition o f Cp^Co produces blue Co(I) which has a strong absorption band at 610 nm. Independent synthesis o f Co(DH)zPBu3 may furnish direct evidence for the assignment. This had been attempted by several methods, including; a) direct mixing of a Co(II) salt, dmgHz, and PBus with addition o f NaOH;^° b) adding one equivalent o f PBu3 into Co(DH)z(H2 0 ) 2 solution;"*^ c) reduction of ClCo(DH)2PBu3 by Zn or Zn(Hg). The purification of these products was difBcult. However, the major species in them, indicated by the *H NMR, is the same as “HCo(DH)2PBu3” and their UV-vis spectra show characteristics identical to those o f “HCo(DH)zPBu3” in every respect. Synthesis of a clean product o f authentic Co(DH)2PBu3 was successfully achieved by reducing ClCo(DH)zPBu3 with CpzCo. In typical preparation, cobaltocene was used in an amount of slightly less than 1 equivalent, to avoid possible overreduction and to ensure all the CpzCo was reacted. Prepared in this way, after the reaction the only species that could be dissolved in hexane was the desired product, Co(DH)zPBu3 Extraction with hexane will readily separate Co(DH)2PBu3 fi'om the 163 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.5 2eq S 1.5 eq 0.5 1 eq 350 422 493 563 638 Wavelength (nm) Figure V .ll Spectroscopic titration of ClCo(DH)2PBu3 with CpzCo in methanol (CICo(DH)2PBu3 2.0 X 1 0 " * M, Cp^Co 2 . 0 x 1 0 '^ M). 2.5 2.0 1.5 1.0 0.5 0.0 ppm Figure V.12 NMR spectrum o f Co(D l^PBu3 in toluene-d, at room temperature (the peak at 2.09 ppm is from the solvent). 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. rest, i.e., the starting material CICo(DH)2PBu3 and Cp^Co salt. Not surprisingly, all the properties o f Co(DH)2PBu3 are indistinguishable from those o f "HCo(DH)zPBu3" For instance, its ‘H NMR in toluene-<4 (Figure V.12) is identical to that at room temperature shown in Figure V.8 . This is a direct evidence that the so-called “hydridocobaloxime” is actually Co(DH)2PBu3 . The p A ", o f “HCo(DH)2PBu3” is reported to be 10.5,® indicating that the hydridic proton is somewhat acidic and should be able to be deprotonated by NaOH solution with a reasonable concentration (Scheme V.2.a). On the other hand, in strongly basic media, Co(DH)2L undergoes a disproportionation reaction to form equi- molar Co(lll) and Co(I) (Scheme V.2.b).^^ Upon reaction with hydroxide, the hydride “HCo(DH>2PBu3” will lead to an equal amount o f Co(I) while Co(DH)2PBu3 will produce Co(I) in half o f its original concentration. Relative to Co(I), the absorbance of Co(m ) is very low. Particularly, in the region o f the Co(I) strong band (610 nm), Co(in) does not show significant absorption. Therefore, quantitative measurement of Co(I) by the 610 nm band is possible. After determining the extinction coefBcient s ( 1 . 8 X 1 0 ^ cm 'M "^ in methanol) o f Co(DH)zPBu3 independently, we were able to measure the concentration o f Co(DH)zPBu3 solution. In the dark blue reaction mixture o f “HCo(DH)zPBu3” and NaOH in methanol, the concentration of Co(DH)zPBu3 was found to be half o f the original “HCo(DH)2PBu3” solution. This is another piece of evidence showing that “HCo(DH)2PBu3” is actually Co(DH)zPBu3 165 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. O H * HCo(DH)2PBi^ ------------------------► Co(DH)2PBi; ^ ‘ O H * Co“(D H )2PB i^ ----------- ► 1/2 C o*(DH)2 P B V + 1/2 Co“ (DH)2PBi^" Scheme V.2 The reactivities o f hydride and Co(II) with hydroxide. All our results leave little doubt that “HCo(DH)2PBu3” prepared from C1Co(DH)2PBu3 should be reformulated as Co(DH)zPBu3 However, as pointed out by Marzilli et al.,^‘ the starting material used in the original preparation might be [Co(DH)2(PBu3)2] [Co(DH)zCl2] instead o f ClCo(DH)2PBu3 as claimed. “ The idea that HCo(DH)2PBu3 might be obtained from a different starting material lead to conduct the reduction o f [Co(DH)z(PBu3)2] [Co(DH)zCl2] with NaBH* in pH 7 phosphate-bufrered aqueous methanol. The isolated product is the seme as from ClCo(DH>2PBu3. V.2.2.5. Comments on Schrauzer s" and Espenson s^ Works "HCo(DH)zPBu3", i.e., Co(DH)2PBu3 generated by reducing ClCo(DH)2PBu3 with NaBHt at pH 7, has been fully characterized with great caution. Since the widespread “hydridocobaloximes” have been included in several important text books,^ it seems necessary to make a few comments, based on our results, on such a class o f important compounds. 1 6 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The product in the original report, purified simply by washing with water, appears to be a mixture o f Co(DH)zPBu3 and Co(DH)2PBu3. In the solid state, a dark blue and a violet species will give the appearance o f the reported black color. In hexane, Co(I) is insoluble and hence the color o f the mixture will be that o f the Co(II), i e., a purple or violet color. The absorption band at 565 nm is consistent with our observation o f Co(II). In a polar solvent (e.g., methanol), in contrast, both Co(I) and Co(II) are readily soluble. Since the color and the UV-vis band (around 610 nm) of Co(I) are much more intense at room temperature, in the mixture Co(DH)2PBu3 will not be discernible and the solution will be like that o f Co(I). Purposely mixed Co(I) and Co(n) shows a spectrum in methanol (Figure V.13) almost indistinguishable fi'om that of Co(DH)2PBu3 (Figure V.4). The reported absorption o f “HCo(DH)2PBu3” at 614 nm in methanol^ indicates the presence of Co(DH)2PBu3 in their product. I I < 300 374 445 515 587 662 Wavttlenth (nm) Figure V.13 UV-vis spectrum of a mbcture o f Co(DH)2PBu3 and Co(DH)2PBu3 in MeOH. 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The reported reactivities toward electrophiles like alkyl halides (RX) in different solvents are consistent with a mixed product. The reaction rates o f "HCo(DH)zPBu3" in protic media with RX comparable to those observed with free Co(I) will not be unexpected considering what they claimed "HCo(DH)zPBu3" contains significant amount of Co(DH)2PBu3 In carefully dried and methanol-firee hexane or benzene, “HCo(DH)2PBu3” was unreactive with RX. This is not surprising since Co(I) is not soluble in these solvents and under those conditions, Co(DH)2PBu3 is not reactive to RX, etc. It had also been claimed that HCo(DBF2)2PBu3 could be synthesized under the same conditions of “HCo(DH)2PBu3” preparation, i.e., reducing ClCo(DBF2)2PBu3 with NaBHt at pH 7. The attempt to repeat this experiment generally produced two species; a blue one (Xm« = 620 nm in MeOH) and a purple one = 520 nm in benzene). Both of these are insoluble in hexane. By comparing these to the original report,® we conclude that “HCo(DBF2)2PBu3” was also a mixture o f Co(II) and Co(I). The H C o^B F 2)2PBu3, if it exists, may have a lower p A T . value compared to HCo(DH)2PBu3. The attempted preparation o f HCo(DBF2)2PBu3 at lower pH, e.g., pH 4-6, usually generated more Co(II) in the mixed product. Espenson et al. reported the kinetic investigation o f “HCo(DH)2PBu3” in the presence of acid. In this study, the “hydride” was prepared according to the original procedure,' and no other characterization was described. The concentration o f “HCo(DH)2PBu3” was derived fi'om the UV-vis absorbance at 610 nm. However, this 1 6 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. band is o f C o(I) origin, as discussed in Section V.2 .2 . 1. A papei^^ published tw o years later from the same lab specifically pointed out that 610 nm band w as due to Co(DH)2PBu3 Apparently, th ose kinetic equations and the mechanism o f "HCo(DH)zPBu3" reactivities w ere actually deduced from the intensity variation o f a C o(I) absorption and subsequently the reliability is in doubt. V.2.2.6. Synthesis and Characterization of Co(DH)(DBF2 )PBu3 and Co(DBF2 )2 - FBU3 T he synthesis o f C o(D H )(D BFz)PBu3 was readily achieved w ith CpzCo in the similar manner to that o f Co(DH)2PBu3. Its solubility in hexane is lim ited, therefore, EtzO w as used to extract the product from the reaction m ixture. This com pound usually exhibits a blue color and th e intense absorption band is around 620 nm, which are alm ost indistinguishable from Co(DH )(DBFz)PBu3 But th e product obtained here is definitely C o(II), since it w as prepared from 1 equivalent o f CP2C0 and it is soluble in EtzO, toluene and slightly soluble in hexane. The therm ochrom ism (presented b elow ) only exists in C o(II) system s and can also be used to dififerentiate Co(DH )(DBF2)PBu3 from its C o(I) counterpart. N o t surprisingly, the properties o f Co(DH)(DBF2)PBu3 are very similar to those o f Co(DH )2PB u3. In a donating solven t, for instance, THF, CH3CN, M eOH , or even toluene, th e color o f a dilute solu tion is pink w hile that o f a relatively concentrated one is usually purple. In EtzO or hexane, in all the concentration attem pted, the color is 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. i J B cooled with cry ice/EtOH RT 290 400 506 615 725 Figure V.14 temperatures. Wavelength (nm) UV-vis spectra of Co(DH)(DBF2)2PBu3 in toluene at different I cooled with dry ice/EtOH RT 230 342 450 554 666 Figure V.15 temperatures. Wavelength (nm) UV-vis spectra of Co(DH)(DBp2)2PBu3 in Et2 0 at different 170 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. blue. All the solutions, when cooled down, will turn blue As shown in Figure V. 14, in a dilute solution in toluene, the UV-vis band is at 520 nm while at lower temperature, a 620 nm band (related to the blue color) will evolve. While in EtzO, the absorbance of the 620 nm band is relatively low at room temperature but its intensity will increase dramatically when the sample is cooled down (Figure V. 15). The VT NMR behavior of Co(DH)(DBF2)PBu3 are similar to those of Co(DH)zPBu3 . For the NMR in toluene-c/g, only the phosphine peaks are visible at room temperature. The dmg methyl peak becomes sharper while shifts slightly upfield with decreasing temperature. At -SO°C, this broad peak is barely discernible (at 4 ppm). At -60°C, it turns to be a bump at 3.5 ppm and at -80°C, this peak is relatively sharp and shifts to a position at 3 ppm. The resonance of the bridging hydrogen is invisible even at -50®C, and it becomes very broad peak at -60®C. Finally at -80°C, it is readily discernible (at 17.5 ppm). The ^ P NMR signal is quite weak in all the temperature range. At -50°C, a broad and weak signal at 27 ppm is detectable. It becomes quite sharp as the temperature is lowered to -80®C. Compared to the situation for Co(DH)zPBu3, the temperature required to see those and ^^P signals which is non-detectable at room temperature are considerably lower, implying that the dimerization of Co(DH)(DBF2)PBu3 is less favored. The lower electron density on the cobalt center and the steric effect of the BF2 group may be the two fectors responsible for this 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. i .(O ( O Î 8 s m % S §• I I I I Ô 172 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. phenomenon. The shift o f resonance from - 8 of Co(DH>2FBu3 to 27 ppm of Co(DH)(DBFz)PBu3 reflects the electron-withdrawing nature o f the BFz bridge. The NMR at room temperature Co(DH)(DBFz)PBu3 shows only one broad signal at -148 ppm (Figure V.16). As the temperature decreases, this peak shifts downfield gradually, to -136 ppm at -90°C. Meanwhile, a new peak at higher field evolves as the sample is cooled down. At -35°C, this peak (around -155 ppm) is barely discernible. However, at a temperature 8 degrees lower, it becomes apparent. As the temperature further decreases, both peaks become sharper and finally of the same intensity, suggesting the increasing dimerization of Co(DH)(DBF2>PBu3. These results also indicate that the two fluorine atoms of the bridging BFz group are in different environments^ and experience the effect of the paramagnetic cobalt center to different extents. Attempts to prepare Co(DBF2)zPBu3 in an identical manner, i.e., reducing C1Co(DBF2)2PBu3 with CpzCo with an equivalency slightly less than one, however, always generate mixtures which contain the desired product. The ^^P and * ® F NMR at lower temperature show several peaks and we were not able to make the assignment. Nfixing equa-molar ClCo(DBF2)2PBu3 and Co(DBF2)2PBu3* (see below) fare no better. The reduction had been attempted also by using NaBHi, Na2 S2 0 3 , Fe powder. They all produce Co(DBF2)2PBu3 , containing some other minor species. The reduction of C1Co(DBF2)2PBu3 in toluene with Zn(Hg) affords essentially two major species. One is probably Co(DBF2)zPBu3, which at room temperature shows a broad signal 173 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. around -162 ppm and broad phosphine peaks in NMR. The other is probably a diamagnetic species and its amount appears to increase with the reaction time. The mixture after 10 h reaction time shows a clean ‘ H NMR spectrum, including sharp signals o f all the protons on the phosphine and a triplet of the dmg methyl. The NMR shows two sharp peaks with both FF and BF coupling, along with a broad signal at -162 ppm. A signal in NMR at 7 ppm also suggests its diamagnetism at room temperature Judging from its color, this species is not Co(DBF2)2PBu3* (see Section.V.2.3.1). Further experiments are needed to determine its structure. Although no pure Co(DBFz)2PBu3 has been obtained, its properties are quite similar to those of Co(DH)(DBF2)PBu3 . It is soluble in Et2 0 but insoluble in hexane. In toluene its color is pink or purple while in Et2 0 , it is usually blue purple. It is thermochromie and at low temperature the color turns blue, as observed by Mita et al.^ V.2.3. Attempts of the Synthesis of Hydrides hy Protonating Co(I) species Hydridocobaloximes generated by NaBHt reduction may be very unstable and hence short-lived. Therefore, attempts to isolate hydrides in this way only generate mixtures o f Co(I) and Co(II). Cyclic voltammetry (CV) results on the complexes C1Co(DH)2PBu3, C1Co(DH)(DBF2)PBu3 and ClCo(DBF2)2PBu3 at different pH indicate that the hydridocobaloximes may be stable on the CV time-scale.^’ Based on these results, we attempted to prepare the hydridocobaloximes by protonating the corresponding Co(I) species with acids, frequently at low temperature. 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. v.2.3.1 . Synthesis and Isolation of the Co(I) Species In order to achieve the goal described above, a pure C o(I) sp ecies is needed. The d eep blu e or greenish blue m onovalent cobaloxim es sp ecies have been w idely noticed as reaction interm ediates.^ H ow ever, the only isolated C o(I) w as generated by electrolysis m ethod several years a g o .^ In our lab, the isolation o f m onovalent cobaloxim es w as first achieved for the bis-BFa sp ecies, because CoCDBFzhPBug could be readily prepared by reducing C1 C o(D B F2 )2 PBu3 with tw o equivalents o f cobaltocene. The synthesis o f Co(DH)2PBu3* and Co(DH)(DBF2)PBu3* w ere found much m ore difGcult. A s show n in Figure V.ll, in a very dilute methanol solution (ca. 0.2 mM), C1 C o(D H )2 PB u3 could be reduced to C o(I) by CP2C0, but m ore than tw o equivalents o f cob altocen e is needed. For sim ilar titration in CH3CN, even w ith as m uch as five equivalents o f CP2C0, only very sm all am ount o f C o(I) could b e generated. In a practical preparation (concentration is considerably higher), w ith m ore than 2 eq. CP2C0 in m ethanol, m ost product is obtained as Co(DH)2PBu3. Sim ilar reduction o f C1 C o(D H )(D B F2 )PBu3 w ith m ore than tw o equivalents o f cobaltocene (in M eOH or CH3CN) also afford a m ixture o f C o (lI) and C o(I), although the percentage o f C o(I) appears to b e higher com pared to the reduction o f ClCo(DH)2pBu3. O bviously, cobaltocene here is not a sufGciently strong reductant to reduce C1 C o(D H )2 P B u3 or ClCo(DH )(DBF2)PBu3 to pure C o(I) sp ecies. The formal reduction potential (£®') for CP2C0VCP2C0 is -0.9 V vs. SC E /^ A lthough the 175 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. £® '(C o“/Co^) values for ClCo(DH)2PBu3, ClCo(DH )(DBF2)PBu3 and CICo(DBF2)2PBu3 w ere determ ined to be -0 .765, -0 .563, and -0.330 V (v s. SCE), respectively, how ever, th ese data w ere obtained in solutions containing supporting electrolytes and pH buffers in a m ixed solvent M e0H/H20.^^ N aBH « cou ld reduce ClCo(DH)2PBu3 to Co(DH)2PBu3*, but the product alw ays contains considerable am ount o f Co(DH)2PBu3, even in a strongly basic solu tion .^ In addition, w hile the C o(II) com plex is easy to separate, the unreacted BH* is difG cult to rem ove. Judging from the ^ ^ P N M R signal o f the free PBu3 at -31 ppm, part o f th e cobaloxim e com plex w ould decom pose during the reduction process. T he H and N M R also indicate the presence o f PBu3-containing species other than Co(DH)2PBu3*. T he reduction o f C lC o(D I^ P B u 3 w ith sodium benzophenone ketyl^^ also produced a m ixture show ing m essy NM R, presumably due to radical reaction w ith the dm g C = N bonds. Sodium am algam N a(H g) has been em ployed in reducing Co(II)^®^® o r N i(II)- com plexes to the corresponding m onovalent com pounds in non-aqueous solvents. Its reduction potential (approxim ately -2.05 V vs. SCE in CH3CN and THF***) indicates that N a(H g) should be a g o o d reductant to prepare m onovalent cobaloxim e com plexes. R eduction o f ClCo(DH)2PBu3 w ith excess 1-2% N a(H g) generates a deep blue solution. T he *H N M R (Figure V .17) o f the blue product is qu ite clean. H ow ever, th e dm g m ethyl and the a-m ethylene on phosphine are are to o broad to be observed. T his is presum ably due to the presence o f small amount o f Co(DH)2PBu3. 176 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. O w ing to the rapid electron exchange w ith the paramagnetic C o(II), the protons closest to the cobalt center in C o(I), i.e., th ose on the a-C H z o f th e axial phosphine and the dmg m ethyl, w ould directly experience the effect o f the unpaired electron spin. The absence o f the ^ P signal is also consistent with this argum ent. Extraction o f the product w ith hexane affords a purple solution, indicating the presence o f Co(DH)2PBu3. H ow ever, the C o(II) species could not be com pletely removed even after vigorous w ashing w ith hexane. L onger reaction tim e (e .g ., overnight) or w ith large am ount o f N a(H g) causes com plete decom position o f cobaloxim e, judging from a gray precipitate and the absence o f blue color. R eduction o f ClCoCDHjaPBus w ith sodium m etal in THF produces m ixture o f C o(I) and C o(II) as w ell. Neither N a(H g) nor N a could reduce ClCo(DH )(DBp2)PBu3 or ClCo(DBF2)2PBu3 com pletely to C o(I). C ooling the m ixture (to -40 °C ) obtained from reduction o f ClCo(DH)(DBF2)PBu3 w ith N a(H g) w ould sharpen the ^H and N M R signals. T his is in accord w ith the presence o f a sm all am ount o f C o(II) in Co(DH)(DBF2)PBu3 It is unusual that the pow erful reducing agent N a or N a(H g) cannot com pletely reduce cobaloxim es to C o(I). The reason may probably be that the Na^ is not very w ell solvated in the aprotic solvent under the synthetic conditions. A fter a system atic survey, Co(DH)2PBu3* could be obtained via reducing ClCo(DH)2PBu3 by N a(H g) w ith addition o f a quaternary ammonium salt,’^ e.g., M e^NCl (Figure V .18). The cobaloxim e com plex undergoes partial decom position during the reaction, as indicated by the ^ 'P N M R signal (-31 ppm) o f free PBu3 in the product. It is desired to keep the 177 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. À A _ a . / 1 I I I 1 1 I I 1 1 1 1 1 1 1 i 2.0 1.5 1.0 I 1 I 1 1 1 0.5 ppm Figure V.17 H N M R spectrum o f th e reduction product o f ClCo(DH)2PBu3 by 1% N a(H g) in CD3CN (n ot show n is the peak o f the bridging H at 19.6 ppm). 2.0 1.5 1.0 0.5 ppm Figure V.18 NMR spectrum of [(CH3)4N^[Co(DH)2PBu3T in CD3CN (not shown is the (CH3> 4N^ peak at 3.09 ppm). 178 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reaction tim e short to reduce the possibility o f decom position. H ow ever, i f the reaction tim e is to o short, som e o f the reduction interm ediate C o(II) w ould still b e present and a mixture o f C o(I) and C o(II) w ill b e obtained. Therefore, tim e control seem s to be critical in preparing Co(DH)2PBu3 In addition, the relative am ount o f M e4NCl also appears to be important in controlling the reduction process. R eaction w ith °Bu4NCl instead o f M e4NCl seem s to be less tim e-consum ing, presumably due to the higher solubility o f "BU 4NCI. The ‘H N M R o f the product clearly show s sharp peaks o f dm g m ethyl and CH3 on PBU 3. H ow ever, th o se peaks o f CHz on phosphine overlap w ith butyl group o f “ B U 4NCI, the characterization is therefore restricted. W ith addition o f M e4NCl, C1C o(D H )(D BF2)PB u3 cou ld be reduced to C o(I) as w ell. T he cobaloxim e decom position is also seen, as indicated by the ^ ^ P NM R. T he C o(I) cobaloxim e com plexes are am ong the m ost reactive nucleophiles,^** how ever, th e difficulties in obtaining Co(DH)2PBu3 and Co(D H )(D B F 2)PB u3 * are unexpected, especially when the £® '(C o“/Co^) values (see above) are n ot very high. The addition o f chloride salt m ay precipitate Na" and drive the reaction to com pletion, since tetra-alkyl ammonium salts could b e better solvated and w ith them as the cation, the C o(I) m ay b e relatively stabilized in CH3CN. The C o(I) is probably to o basic to coexist w ith the bridging hydrogen. In the m ixtures o f C o(I) and C o(II), both m ono- and non-Bp2 system s show the bridging hydrogen in 'H NMR. In contrast, no such signal could b e detected in pure Co(DH)2PBu3* or Co(DH)(DBF2)PBu3* sam ple. The deprotonation o f these bridging hydrogen is probably the main ob stacle for C o(I) 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. form ation. It m ay also be the starting point o f the decom position during N a(H g) reduction. V.2.3.2. The NMR o f BFi-Bridged Cobaloximes: the Unusual B-F Coupling The N M R o f the BFj-bridged cobaloxim es exhibit several interesting features. The results are sum m arized in Table V .2 . T a b le V.2 The N M R results o f som e BFz-bridged cobaloxim e com plexes FI F2 Complex Solvent 5 f (ppm) * /b f (Hz) 5 f (ppm) (Hz) V ff 0 ClCofDBFzhPBus CDCI3 -145.2 0 -147.8 0 62 C1C o(DH)(DBF2)2PBu3 CDCI3 -139.2 2 0 -154.9 0 65 ClCoCDBFjzPy acetone- < /6 -140.1 2 2 -156.7 1 0 59 CofDBFzhPBu; CD3CN -149.7 23 -153.4 0 70 Co(DH)(DBF2)2PBu3 ' CD3CN -149.0 2 2 -152.1 0 73 O K D B F zh P y' CD3CN -148.4 26 -151.5 0 75 In all th ese system s, tw o signals are observed in the N M R spectra, revealing the inequivalency o f th e tw o fluorine atom s o f each B F % group. The inequivalency has been illuminated by th e X -ray structure o f C o(D BFz)2Py \^ t h no exception, the tw o fluorine nuclei are coupled to each other and in general, the coupling constants (V ff) in 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C o(I) are approxim ately 8-15 H z larger than those o f the corresponding C o(III) com plexes. The coupling betw een the boron and fluorine nucleus is unique here. For ClCo(DBp2)2PBu3, the N M R sh ow s tw o doublets (F igure V .19). In this system , th e tw o fluorine nuclei in each BFz bridge are coupled to each other and no B -F coupling is observable. This is in agreem ent with the singlet signal in "B NM R. W hen ClCo(DBF2)2PBu3 is reduced to the C o(I) state, the ^ ® F N M R (Figure V .20) show s tw o signals; a doublet o f quartet and a doublet. The doublets originate from the F-F coupling, w hile 1:11:1 quartet show s that one o f the tw o F nuclei is coupled to "B ( / = 3/2 ). A doublet is seen in the “ B N M R o f ClCo(DBF2>2PBu3' and the coupling constant (V b f) is the same as that obtained from the N M R . T hese facts strongly support that the boron is coupled to only o n e o f the tw o F nuclei. For all the system s presented in Table V .2 , the low er field F ’s are usually coupled to boron, w hile those at higher field are not. Only in the case o f ClCo(DBF2)2Py, B is coupled to both F’s, but still the coupling constant to the higher field F is smaller. That the nuclei o f the tw o F atom s attached to the same B exhibit distinct behaviors o f cou p lin g to B is uncom mon. T his m ight be due to the dififerent fractional s character in th e tw o B -F bonds. The other unusual feature o f th e NM R is the iso to p e effect. For those F’s w hich do n ot show coupling to B , th e signal also sh ow s a doublet at a field slightly low er than that o f main doublet. This low er-field doublet is believed to originate from th ose fluorine attached on the ^°B atom s. The relative intensities o f th ese tw o doublets 181 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F igu re V .19 ‘® F N M R s p e c tru m o f C lC o (D B F 2)2P B u 3 in C D C I3. -149 -150 -151 -152 -153 -154 ppm F ig u re V .20 ‘® F N M R spectrum o f [Cp2Co^[Co(DBF2)2PBu3'] in CD3CN. 182 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ( 1:4 ) are in g o o d agreem ent w ith the natural abundances o f the boron isotop e 19.78% and B 8 0 .22% ). The previous reports o f boron isotop e effect on NMR are either not w ell resolved®^ or can only be seen at low tem perature.’’ V.2.3.3. Reactions of Co(l) with Acids The attem pts to prepare hydridocobaloxim es w ere initiated by the observation o f the product o f the reduction o f ClCo(DBFz)2PBu3 w ith NaBH* at pH 7 is m ostly dark blue C o(I). That made us to consider that due the strongly electron-w ithdraw ing characteristic o f the BFz groups, the hydridic proton o f HCo(DBFz)2PBu3, if form ed, is probably to o acidic and may dissociate com pletely at pH 7 . H ow ever, at room temperature the acidification o f Co(DBFz)2PBu3 at low er pH only produce a pink or purple color, that o f C o(II). This suggests to us that hydridocobaloxim es m ight be unstable at such tem perature. E xtensive efforts had been made in the attem pt to synthesize th ese hydrides by protonating the C o(I) species at low temperature. The acids used include acetic CH3CO2H, trifluoroacetic CF3CO2H, and benzoic acid C6H3CO2H and the C o(I) com plexes include Co(DH)2PBu3 , Co(DH )(DBF2)PBu3 and Co(DBF2)2PBu3*. A typical experim ent is described here. Co(DH )(DBF2)PBu3 (6.0 m g) and benzoic acid (3 m g) w ere m ixed in an NMR tube with a teflon valve. V ia a vacuum line, toluene-dg (ca. 0.4 ml) (or CD3OH, CD3CN) w as vacuum transferred into the mixture and the sealed tube w as kept in a liq. N2 Dewar until an NMR spectrom eter 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. w as available. T he N M R probe w as p re-cod ed to -90 ®C, then the NM R tube w as rem oved from liq. N z to let the frozen sample partially m elt or w as put in a heptane/liq. Nz bath (-91 °C ) to pre-equilibrate. It w as then transferred into th e N M R spectrom eter and the ‘H , and spectra w ere recorded. The tem perature w as increased gradually and m ore N M R spectra w ere obtained. U nder all the conditions attem pted, the H spectra sh ow ed th e existence o f more than o n e sp ecies and no signal could be convincingly assign ed to th e hydride. The broadness o f all th e N M R spectra suggested the presence o f param agnetic species. The H N M R o f on e o f the main species resem bles to th o se o f Co(DH )(DBFz)PBu3. The ^ ^ P N M R sh ow ed a peak at 27 ppm w hich is identical to that o f Co(DH )(DBFz)PBu3 at lo w tem perature. The acidification o f C o(I) always generate sp ecies o f pink or purple color. M ore im portantly, th ese products exhibit therm ochrom ism , a unique property o f C o(II) species. A ll the N M R and U V -vis results suggest that reaction o f C o(I) w ith acid produce C o(II) instead o f hydrides. V.2.3.4. Are Hydridocobaloximes Accessible? The failure to prepare hydridocobaloxim es in our lab as w ell as in others^^ naturally drive us to contem plate, whether these hydrides are accessib le at all or if they exist, whether th ey are sufGciently stable to be identified. The instability o f hydride H C o (D H )zL m ay b e depicted by equation (a) in Schem e V .3 . T he enthalpy o f reaction ( A f P ) for this s e lf decom position can be 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. deduced by using the thermodynamic cycle o f eq (b) and (c), i.e., = àH h + HCo(DH>2L — Co(DH>2L + 1/2 Hz M P (a) HC o(DH>2L — ^ C o(DH>2L + H A^b“ (b) H 1/2 Hz A/fc“= - 5 2 kcal/m oI (c) S ch em e V .3 Thermodynamic cycle o f s e lf decom position o f hydridocobaloxim es. T he value o f A^c° can be obtained from the bond dissociation energy o f Hz, i.e., = - 1/2 B D E (H -H ) = -52 kcal/m ol. N o direct inform ation on the H-Co(DH)2L bond strength is available, i.e., AA/i,° is unknown. H ow ever, 50-55 kcal/m ol as the upper lim it o f might be a reasonable approxim ation. This value is slightly higher than th o se obtained from an estim ation*’ based on electrochem istry results.^* Furtherm ore, the alkyl analogs o f hydrido-cobaloxim es, R C o (D I^ L , have been w ell studied*^ and th e Co-C BD E values are determined to be ca. 20-25 kcal/m ol.” It is generally know n that metal hydrogen bond M -H is normally 30 kcal/m ol stronger than the corresponding M -R bond.*^ I f this applies to cobaloxim es as w ell, then the BD E value for H -C o(D H )zL (A^b®) w ould b e clo se to 50-55 kcal/m ol. A s a result o f above discussion, for reaction (a) w ould be up to -2 to +3 kcal/m ol. On tlie other hand, the entropy effect for reaction (a) is large. I f w e assume the entropy values o f Co(DH>2L and HCo(DH)2L to be approxim ately equal, for 1 8 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reaction (a), A S ° w ill be clo se to h a lf o f the S ° value (31 e.u .) o f The value o f AG° w ill not exceed that o f à H °, and at room temperature (ca. 300 K ), AG° is -7 to -2 kcal/m ol. Even at the low est tem perature o f our NM R experim ents (ca. 200 K ), AG° w ill be -5 to 0 kcal/m ol. Judging from th ese data only, hydridocobaloxim es may still be detectable under certain conditions, e .g ., under a substantial pressure o f Hz H owever, the value o f AG® obtained in this w ay is believed to be the upper lim it, i.e., the real values could be low er. M eanwhile, in very few cases, an atm osphere o f Hz has been maintained in the efforts to identify H Co(DH )zL. I f the system is open, the liberated Hz w ill escape and hence the s e lf decom position o f the hydrides w ill be irreversible. That was exactly the situation w hen reducing ClCo(DH)zPBu3 w ith N aB H ,, since the reaction w as carried out under a continuous Ar flow . In this case, all the cobalt(III) hydride generated w ill hom olytically decom pose to form C o(II). In addition, the dim erization o f the decom posed product C o(II) at low tem perature also acts as a driving force for reaction (a). In summary, H Co(DH )zL is probably not a thermodynamically favorable species and once formed, it has the tendency o f se lf decom position to generate C o(II) and Hz Regarding the protonation o f C o(I), from another point o f view , the C o(I) may act as a pow erful reductant. A s proton approaches it, an electron may transfer directly from cobalt center to I f , even befr>re the C o-H bond is form ed. The products generated in this manner are identical to th ose from the hydride decom position. 186 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T he therm odynam ic instability o f HCo(DH>2L is supported by the fact that all the attem pts to prepare hydrides only resulted in C o(II) species. A previous paper also concluded that H C o(D H )2Py is thermodynamically unstable.^ The M -H bond strength seem s to play a crucial role in the stability o f hydrides. T hose w ith H -C o bond several kcal/m ol stronger (58 kcal/m ol), such as HCo(CO>4“ and HCo(CN)5^',**‘ are readily identified. W ith sim ilar strength o f Rh-H bond, rhodium porphyrin hydrides have been prepared and extensively investigated.®^ In contrast, hydridocobalt porphyrin com plexes have only been speculated to exist as transient species and no isolation has been achieved.®^ Com bining with our observations on cobaloxim e system s, H -C o(DH )2 and H -C o(P or) bonds may have similar strength. This is also in accordance w ith the conclusion that M -H bond energies are about 5-11 kcal/m ol greater for the second-row m etals than for first-row metals. A lthough the B D E s o f R-Co(DH)2L are only 20-25 kcal/m ol, they are usually therm odynam ically stable. For exam ple, for the system s o f C6Hs(CH3)CHCo(DH)zL, the decom position reaction described in Schem e V .4 is energetically unfavorable because the A C value (calculated from the known A C and AS°) is positive. C6H5(CH3)CHCo(DH)2L Co(DH)zL + C 6H 5C H C H 2 + 1/2 H2 Scheme V.4 The equilibrium for C o-R dissociation in alkycobaloxim es. 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T he decom position o f H C o(D H )zL is also kineticaily feasible. I f eq (b ) is the initial step, it w ould be rate determ ining and the activation energy or the enthalpy o f activation is then expected be clo se to the C o-H BD E value (AHh°) and as a result, the hydrogen evolution should be quite slow . H ow ever, the reaction could occur in a manner that the bond betw een the tw o hydrogen atom s is being formed directly,® as show n below . C o H H C o T his linear four-centered transition state has been proposed to account for the m echanism o f the reaction o f R h(II)(P or) w ith Hz to form tw o o f HRh(Por).®^ A ccording to the principle o f m icroscopic reversibility, that tw o HCo(DH )zL m olecules react w ith each other to generate C o(II) and Hz via the similar transition state is plausible. T he form ation o f a strong H -H bond essentially com pensates for breaking the tw o C o-H bonds and hence the enthalpy o f activation (AH^) for this reaction is expected to be very small. The instability o f hydridocobaloxim es relative to R C o(D H )zL is probably also due to the low er activation barrier for hydride decom position, since no similar transition states for RCo(DH )zL are possible. In summary, hydridocobaloxim es appear to b e unstable therm odynam ically and once form ed, their decom position is kineticaily feasible. Therefore, they are probably synthetically inaccessible. Our experim ental results strongly support this speculation. 188 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. v. 3. Conclusion T he so-called hydridocobalcxinie “HCo(DH)2PBu3” is not accessible under all the conditions attem pted. The previously reported isolation o f "HCo(DH)zPBu3" is probably in error and all our experim ental results indicate that what they obtained w as in fact a mixture o f Co(DH)zPBu3 and Co(DH)2PBu3*. For confirm ation, Co(D H )2PB u3 w as prepared independently by reducing ClCo(DH)2PBu3 w ith cobaltocene. This C o(II) species is essentially m onom eric at room temperature but it w ill increasingly dim erize via the form ation o f a C o-C o bond as the tem perature decreases. The tem perature dependences o f all properties o f Co(DH)zPBu3 can rationalized by the m onom er-dim er equilibrium. The m ono-BFz and bis-BFz bridged cobaloxim es have been studied in a parallel manner. T hey show behaviors sim ilar to th o se o f the non-BFz bridged counterparts. For exam ple, the C o(II) species are also therm ochrom ie. H ow ever, the electron w ithdraw ing effect o f the BFz groups is also reflected in their properties. Co(DBF2)zPBu3 could be readily obtained up on reducing Co(IH) w ith cobaltocene w hile in order to prepare Co(DH)zPBu3' or C o(D H )(D BF2)PBu3, a much stronger reductant has to be used. The tw o fluorine atom s on each B are in different environm ents and for all th e diam agnetic BFz-bridged species the coupling betw een tw o fluorine nuclei are observed. In contrast, coupled to boron could be none, on e or both o f th e tw o fluorine nuclei. 189 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. v.4. Experimental All the air sensitive compounds are prepared and manipulated on a Schlenk line (N2 atmosphere) or in a Vacuum Atmospheres glovebox under Ar atmosphere. NMR spectra were recorded on a Bruker AC-250, WP-270 SY, or a AM-360 spectrometer. H NMR chemical shifts were calibrated against internal solvent, referencing all to TMS. The NMR chemical shifts were referenced to a sealed CFCI3 capillary, NMR chemical shifts to 85% H3PO4, and "B NMR to BF3 OEt UV-vis measurements were performed on a Milton Roy Spectronic 3000 diode spectrophotometer. EPR spectra were recorded on a Bruker ER 200D-SRC spectrometer equipped with an Oxford Instruments ESR 900 cryostat. Elemental analysis was performed by Oneida Research Services Inc., Whitesboro, NY. Air-sensentive NMR or EPR samples were fi'ozen with liquid N2, evacuated, and then sealed with a flame torch. Air-sensitive UV-vis samples were prepared in a glovebox and added to cuvettes which were then sealed with rubber septa. The septa turnover flanges were flapped over the cuvette necks and then wrapped with copper wires. Tributylphosphine (PBus) was distilled under reduced pressure and kept under a Ar atmosphere. Other chemicals were used as received. Water was purified by using a Bamstead Nanopure system (18 M O cm) Other solvents were of analytical purity and no purification were attempted for usage outside the drybox. For those used inside glovebox, they were purified and dried by heating under reflux overnight with the 190 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reagent indicated below, then were distilled or vacuum transferred into a Schlenk flask, after degassed (freeze/pump/thaw) for three cycles and taken into the glovebox. Toluene, benzene, hexane, THF, or EtzO was distilled from sodium benzophenone ketyl, CHsCN from CaHz, methanol from magnesium turnings. Acetone was purified by distillation from its Nal addition compound. CD3CN or toluene-^/g was purified by stirring with Na(Hg) for a very long time period followed by vacuum transfer. CD3OH or CD3OD was purified by distillation from a blue Co(DH)zPBu3 solution. C o(D H )(D H 2)02 C0 CI2- 6 H 2O (10.0 g, 0.042 mol) was dissolved in acetone (300 ml) and dimethylglyoxime (9.8 g, 0.084 mol) was added. The mixture was stirred for 30 min. before the precipitate was filtered off. The remaining dark green solution was kept in the air overnight, during which green crystals were formed. The product was filtered and washed with cold acetone and pumped to dryness (12.0 g, 79%). aCo(DH>2PBu3 Under N2 at 55-60°C, the suspension o f Co(DH)(DH2)Cl2 (10.0 g, 0.0277 mol) in MeOH (100 ml) was converted to a dark brown solution upon adding EtsN (4.0 ml, 0.029 mol). PBus (7.0 ml, 0.028 mol) was added with syringe and the solution was kept stirring fisr 30 min. before water (30 ml) was added. The solution was then cooled down while stirring. After most o f the solvents were removed, the crystalline 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. product was filtered and washed with water and acetone. Recrystallization (twice) fi-om a mixed solvent (CH2CI2, MeOH and hexane, 1:1:1) afforded a clean product (6.8 g, 47%). 'H NMR (CDCI3) 5 0.87 (t, 9H, CH3 of P B U 3), 1.3 (m, 18H, CH2 o f P B U 3 ), 2.32 (d, V pH = 1.3 Hz, 12H, dmg methyl), 17.9 (s, hr, 2H, bridging H). "C NMR (CDCI3) Ô 12.8 (dmg methyl), 13.5 (CH3 o f P B u j ) , 20.3 (d, Vpc = 24 Hz, a-CH2 of PBu3), 24.6 (d, Vpc 13 Hz, ^C H 2 o f P B U 3) , 24.7 (d, Vpc = 6 Hz, y-CHz of P B U 3 ), 151.6 (oxime C=N). "P NMR (CDCI3) 631 (s, hr). ClCo(DBF%)2PBw3 To a suspension of ClCo(DH)2PBu3 (0.50 g, 0.95 mmol) in Et20 (20 ml), BF3 OEt2 (2.0 ml) and Et3N (1.0 ml) was added. After stirring for 24 h, the orange precipitate was filtered and washed with Et20 and water. A clean product (0.42 g, 71%) was obtained by recrystallization fi-om mixed solvent of CHCI3, acetone and heptane (1:1:1). ^H NMR (CDCI3) 6 0.88 (t, 9H, CH3 of PBU 3); CH2 of PBus. 1.16 (m, 6H), 1.33 (m, 6H), 1.64 (m, 6H); 2.50 (d, Vph 1.4 Hz, 12H, dmg methyl). NMR (CDCI3) 6 13.5 (dmg methyl), 14.1 (CH3 o f PBU3), 21.0 (d, Vpc = 22 Hz, a -C lh of PBu3), 24.4 (d, Vpc = 12 Hz, /3-CHz of PBU3). 25.8 (d, Vpc = 6 Hz, ;^CH2 of PBU 3), 160.4 (oxime C=N). NMR (CDCI3) 5 -147.8 (d), -145.2 (d), Vff = 62 Hz. ^ ‘B NMR (CDCI3) 6 2.6 (s). ^^P NMR (CDCI3) 6 34 (s, br). Elemental analysis: calcd. for C20H39B2CIC0 F4N 4O4P: C, 38.59; H, 6.31; N, 9.00; found: C, 38.50; H, 6.14; N, 8.89. 192 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. O C o(DH)(DBF2)PBu3 To a solution of ClCo(DH)2PBu3 (0.25 g, 0.47 mmol) in THF (10 mi), BFg OEtz (1.0 mi) and pyridine (0.4 mi) were added consecutively. After stirring for 4 to 10 h, the volume o f solution was reduced (to about 2 ml). A white precipitate (presumably to be PyH* and Py-BFs) was removed by filtration, then 2 ml acetone and 1 ml water was added. The crystalline product obtained upon solvent evaporation was washed with water and EtzO and recrystallized fi’ om a mixed solvent (CHCI3, acetone and heptane, 2:1:3) (0.15 g, 56%). NMR (CDCI3) ô 0.87 (t, 9H, CH3 o f PBus); CH2 of PBu3 : 1.16 (m, 6H), 1.30 (m, 6H), 1.53 (m, 6H); dmg methyl: 2.38 (d, Vph = 1.2 Hz, 6H, H-bridge side dmg methyl), 2.45 (d, Vph =1.4 Hz, 6H, BFz-bridge side dmg methyl); 17.5 (s, br, bridging H). "C NMR (CDCI3) Ô 13.2 (H-bridge side dmg methyl), 13.5 (CH3 o f PBU3), 13.7 (BFz-bridge side dmg methyl), 20.6 (d, ‘ Jpc = 23 Hz, a-CHz ofPBua), 24.5 (d, Vpc = 12 Hz, ^CH z ofPBu;), 25.5 (d, Vpc = 6 Hz, ^ CH2 of PBus), 152.6 (H-bridge side oxime C=N), 159.4 (Bp2-bridge side oxime C=N). * ® F NMR (CDCI3) Ô -154.9 (d, Vff = 65 Hz), -139.2 (dq, Vff = 65 Hz, ‘ Jbf = 20 Hz). ^'P NMR (CDCI3) Ô 32 (s, br). aco(D H >2Fy The mixture of Co(DH)(PH2)Cl2 (1.4 g, 3.9 mmol) and pyridine (0.90 g, 0.011 mol) in CHCI3 (45 ml) was vigorously stirred for 10 min. and then H2O (15 ml) was added. After stirred for two more hrs, the aqueous layer was discarded. The dark 193 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. brown CHCU layer was washed with H2O for several times before the solvent was evaporated. Recrystallization from CHzCk/EtOH (1:1) will afford a clean product (1.1 g, 70%). NMR (CDCI3) S 2.37 (s, 12H, dmg methyl); pyridine ring: 7.2 (m, 2H), 7.7 (tt, IH), 8.2 (dd, 2H); 18.1 (s, br, 2H). aCo(DBF2)2Py After BF3 OEtz (2 . 0 ml) was added, the suspension o f ClCo(DH)2Py (0.50 g, 1.2 mmol) in EtzO (40 ml) was kept stirring for 24 h. The reaction mixture was filtered and washed with H2O and then with EtOH. The product was reciystallized from hot CH3CN (0.27 g, 45%). ‘H NMR (CD3CN) S dmg methyl: 2.64 (s, 12H), pyridine ring: 7.15 (m, 2H), 7.63 (dd, 2H), 7.73 (tt, IH). NMR (acetone<4) S 14.7 (dmg methyl); pyridine ring: 126.4, 140.4, 150.5; 162.9 (oxime C=N). NMR (acetone- de) Ô -156.7 (dq, V ff = 59 H z, V = 10 Hz), -140.1 (dq, V ff = 59 Hz, = 22 Hz). [Co(DH)2(PBU3)2] [Co(D H )2C l2l This was prepared by Prof. G. M. \fiskelly according to a literature method.®^ ^H NMR (CDCI3) S 0.88 (t, Vhh = 7 Hz, 18 H, CH3 ofPBua), 1.12 (m, 12H, CH2 of PBu3), 1.29 (m, 24H, CH2 o f PBua), 2.33 (t, Vph = 2 Hz, 12H, dmg CH3 of Co(DH)2(PBu3)2"), 2.45 (s, 12H, dmg CH3 of Co(DH)2Cl2'), 17.6 (s, br, 2H, bridging H of Co(DH)2(PBu3)2l . 18.8 (s, br, 2H, bridging H o f Co(DH)2Cl2'). "*P NMR (CDCI3) S 24 (s, br). 194 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. “HCo (DH)2PBu3 ’ ’, Co(DH)zPBu3 by NaBBL, An aqueous bufifer solution (pH 7) was prepared by dissolving NaH2P0 4 Hz0 (0.68 g, 4.9 mmol) and Na2HP0 4 THzO (0.82 g, 3.1 mmol) in H2O (15 ml). With this buffer, a methanol solution (15 ml) o f ClCo(DH)2PBu3 (0.50 g, 0.95 mmol) was mixed. After bubbling Ar through for ca. 30 min., this mixture was loaded into a double-neck Schlenk flask, to which a Schlenk filter and a solid addition side arm with NaBH4 were attached. This assembly was then kept strictly anaerobic by maintaining an Ar flow. After 30 min. (the air trapped in the assembly was expected to be completely expelled), NaBHi (0.15 g, 4.0 mmol) was added in small portions by gentle tapping. With vigorous stirring, immediately the mixture turn dark blue, with bubbles and brown foams formed. After stirring for another 20 min., the mixture was filtered and all the solvents were removed and the product was dried under vacuum. The assembly with the product was transferred into the glovebox and the product was extracted with hexane. Upon removing the solvent, a dark violet solid was obtained. Magnetic Moment Measurement by the Evans Method “HCo(DH)2PBu3” (assumed to be Co(DH)2PBu3) ( 1 0 mg, 0 . 0 2 0 mmol) was dissolved in acetone-dk (0.40 ml) and loaded into an NMR tube. To this solution, a drop o f benzene and a sealed capillary with a benzene solution in acetone-dé was added. The NMR tube was then flame-sealed with a torch. 195 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Diamagnetic corrections for acetone-<4 and Co(DH)2PBu3 were calculated to be -69 X 1 0 " ® and -283 x 10^ emu-mol'*, respectively. The related data are listed in Table V.3. Table V.3 Magnetic moments of "HCo(DH)zPBu3" at diffèrent temperatures in acetone-<4 T(K) Av/vo (x 1 0 ^) /Z c ff (BM) 293 5.60 2.36 283 5.44 2.29 273 5.23 2 . 2 0 268 5.19 2.17 263 5.11 2.13 258 5.15 2 . 1 2 253 5.14 2.09 248 5.19 2.08 243 5.23 2.07 238 5.31 2.07 233 5.35 2.05 M olecular W eight Determination by Solution Vapor Pressure Equilibrium M ethod The measurement was conducted in a specially designed apparatus. Two parallel long tubes, of which the volume have been calibrated and marked, were jointed to two round bottom Schlenk flasks (with teflon valves). These flasks were connected to each other so that the solvent vapor could diffuse through. C1Co(DH)2PBu3 (10.0 mg) was dissolved in approximately 3.5 ml acetone and added to one tube o f the above-described flask while the solution of "HCo(DH)zPBu3" 196 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (10.0 mg) in acetone (ca. 3.0 ml) was added into the other tube. After three cycles of freeze/pump/thaw degas, the flask was left in a glovebox undisturbed for a week. After this period it was considered that the vapor pressure o f the two solutions had reached an equilibrium. The solution volumes in the two tubes were recorded once every day for another week. Spectroscopic Titration o f OCo(DH>2 PBu3 with CpiCo Inside a glovebox, the methanol solution o f ClCo(DH)2PBu3 (2.0 ml, 2 x 1 0 "* M) in a 1 cm UV-vis cuvette was sealed with a septa and copper wire. A cobaltocene (CpzCo) methanol solution (100 pi, 2.0 x 10^ M) was loaded into a gastight syringe and the needle was pre-pierced into the UV cuvette through the septa. The whole setup was cautiously removed out o f the drybox. UV-vis spectrum o f the sample was recorded. After every addition of 10 pi o f CpzCo solution, the cuvette was shaken for the solution to mix and then UV-vis spectrum was recorded. Co(DH)2 PBu3: Generated by CpzCo To a solution o f ClCo(DH)2PBu3 (53 mg, 0.10 mmol) in CH3CN (ca. 10 ml), CpzCo (17 mg, 0.090 mmol) was added and the solution turned purple immediately. The solvent was removed under reduced pressure after the mixture had been kept stirring for 1-2 h. The residue was extracted with hexane (ca. 10 ml). Upon removal o f the solvent, a dark purple product could be obtained. 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Co(DH)(DBF2)PBu3 To a solution o f CICo(DH)(DBF2)PBu3 (50 mg, 0.087 mmol) in CH3CN (ca. 1 0 ml), a CP2C0 (15 mg, 0.079 mmol) solution in CH3CN (ca. 8 ml) was added slowly while stirring. The solution turned to red purple first and then to blue. After 2 h, the solvent was removed under reduced pressure. The product was then extracted with EtzO A blue solid was obtained after the evaporation ofEtzO Co(DBF2)2PBU3 Cobaltocene (48.0 mg, 0.254 mmol) was added to the solution of C1Co(DBF2)2PBu3 (76.0 mg, 0.122 mmol) in CH3CN (15 ml). The color turned red purple immediately and then to blue within 1-2 min. After stirring for ca. 30 min., the solvent was removed under reduced pressure and the residue was washed with toluene and then hexane and pumped to dryness. NMR (CD3CN) Ô 0.82 (t, 9H, CH3 of PBua); CH2 ofPBu3 : 1.04 (m, 6 H), 1 . 2 0 (m, 6 H), 1.29 (m, 6 H); 1.87 (d, Vm = 6.5 Hz, 12H, dmg methyl). NMR (CD3CN) Ô 12.0 (d, Vpc = 3 Hz, dmg methyl), 13.9 (CH3 o f PBus), 21.5 (d, ‘ ypc = 16 Hz, a-CH2 of PBU 3), 25.2 (d, Vpc = 3 Hz, ^C H 2 of PBU3), 25.5 (d, Vpc = 11 Hz, yS ^C H z of PBU 3), 142.5 (oxime C=N). ‘® F NMR (CD3CN) 5 -153.4 (d, Vff = 70 Hz), -149.7 (dq, Vff = 70 Hz, = 23 Hz). “ B NMR (CD3CN) Ô 4.0 (d, Vbf = 23 Hz). ^* P NMR (CD3CN) Ô 19 (s, br). 198 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C o(DH)2PBU3' To a solution o f ClCoCDH)2PBu3 (12 mg, 0.023 mmol) and Mc4NCl (13 mg, 0.12 mmol) in CHsCN (10 ml), Na amalgam (1%, 0.6 g) was added and kept stirring for 36 h. After removing the solvent under reduced pressure, the solid was stirred with hexane overnight (this was necessary to remove the free PBus generated by decomposition) and pumped to dryness after the filtration. NMR (CD3CN) 5 0.81 (t, 9H, CH3 of PBu3 ); CHz of PBu3 . 1.02 (m, 6 H), 1.17 (m, 6 H), 1.24 (m, 6 H); 1.77 (d, V pH = 6.1 Hz, 12H, dmg methyl); 3.09 (Me4N ^. ^^P NMR (CD3CN) Ô 17 (s, br). Co(DH)(DBF2>FBu3' This was prepared in a maimer similar to Co(DH)2PBu3* . Na(Hg) (1%, 0.7 g) was added to a mixture o f ClCo(DH)(DBF2)PBu3 (12 mg, 0.021 mmol) and Me^NCl (15 mg, 0.14 mmol) in CH3CN (ca. 10 ml). After stirring for 24 h, the solvent was removed under reduced pressure. The blue residue was extracted with hexane and then pumped to dryness. NMR (CD3CN) Ô 0.81 (t, 9H, CH3 o f PBU3); CH2 of PBU3 : 1.00 (m, 6 H), 1.18 (m, 6 H), 1.27 (m, 6 H); 1.80 (d, Vph = 6.9 Hz, 6 H, H-bridge side dmg methyl), 1.84 (d, Vph = 5.2 Hz, 6 H, Bp2-bridge side dmg methyl); 3.07 (Me4N*) NMR (CD3CN) Ô -152.1 (d, Vff = 73 Hz), - 149.0 (dq, Vff = 73 Hz, ‘ JeF = 22 Hz) ^‘ P NMR (CD3CN) Ô 18 (s, br). 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References and Notes I- Dedueu, A., ed.. Transition M etal H ydrides VCH: New York, 1992. 2. Sutin, N.; Creutz, C ; Fujita, E. Comments on Inorganic Chemistry 1996, 19, 67 and the references therein. 3. Kellett, R_; Spiro, T. G. Inorg. Chem. 1985, 24, 2373. 4. Koelle, U. New. J Chem. 1992, 16, 157. 5. (a) Schrauzer, G. N. Ace. Chem. Res. 1968, 1, 97; (b) Schrauzer, G. N. Angew. Chem., Int. E d Engl. 1976,15, 417; (c) Marzilli, L. G ; Toscano, P. J.; Ramsden, J. H.; Randaccio, L ; Bresciani-Pahor, N. in Catalytic Aspects o f M etal Phosphine Complexes Aiyea, E. C ; Meek, D. W. eds.. Advances in Chemistry Series Vol. 196, 1982, p. 85; (d) Bresciani-Pahor, N.; Forcolin, M.; Marzilli, L. G ; Randaccio, L.; Summers, M. F.; Toscano, P. J. Coord Chem. Rev. 1985, 63, 1 ; (e) Randaccio, L ; Bresciani-Pahor, N.; Zangrando, E ; Marzilli, L. G. Chem. Soc. Rev. 1989, 18, 225. 6. Hawecker, J ; Lehn, J.-M.; Ziessel, R. Nouv. J. Chim. 1983, 7, 271. 7. See, for example, (a) Naumberg, M ; N-V-Duong, K.; Gaudemer, A. J. Organomet. Chem. 1970, 25, 231; (b) Szeverenyi, Z ; Budo-Zahonyi, E.; Simandi, L. I. J. Coord Chem. 1980, 10, 41; (c) Pasto, D. J.; Timmers, D. A_; Huang, N.- Z. Inorg. Chem. 1984, 23, 4117; (d) Derenne, S.; Gaudemer, A.; Johnson, M. D. J. Organomet. Chem. 1987, 322, 229. 8. Schrauzer, G. N.; Hilland, R. J. J. Am. Chem. Soc. 1970, 93, 1505. 9. Chao, T.-H.; Espenson, J. H. J. Am. Chem. Soc. 1978, 100, 129. 10. (a) Crumbliss, A. L.; Gaus, P. L. Inorg. Chem. 1975, 14, 486; (b) Crumbliss, A L ; Gaus, P. L. Inorg. Chem. 1975, 14, 2745; (c) Espenson, J. H ; Chen, J.-T. J. Am. Chem. Soc. 1981, 103, 2036; (d) Szczepura, L. F ; Muller, J. C ; Bessel, C. A ; See, R. P.; Janik, T. S.; Churchill, M. R.; Takeuchi, K. J. Inorg. Chem. 1992, 31, 859. 11. (a) Adin, A _; Espenson, J. H. Inorg. Chem. 1972, 11, 686; (b) Gjerde, H. B ; Espenson, J. H. Organometallics 1982, /, 435. 200 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12. Wilmarth, W. K.; A shl^, K. R.; Harmon, J. C.; Fredericks, J.; Crumbliss, A. L. Coord. Chem. Rev. 1983, 51, 225. 13. Schrauzer, G. N. Chem. Ber. 1962, 95, 1438. 14. (a) Stynes, D. V.; Leznofi^ D. B ; Harshani de Silva, D. G. A. Inorg. Chem. 1993, 32, 3989; (b) Vemik, I.; Stynes, D. V. Inorg. Chem. 1996, 35, 6210. 15. (a) Dreos, R.; Tauzher, G.; Vuano, S.; Asaro, P.; Pellizer, G ; Nardin, G. Randaccio, L ; Geremia, S. J. Organomet. Chem. 1995, 505, 135; (b) Asaro, F. Dreos, R.; Geremia, S ; Nardin, G ; Pellizer, G ; Randaccio, L.; Tauzher, G. Vuano, S. J. Organomet. Chem. 1996, 525, 71. 16. van ArkeL B.; van der Baan, J. L ; Balt, S.; de Bolster, M. W. G ; van Delft, R. J ; Klumpp, G. W ; de Koning, H ; van den Winkel, Y. R ed. Trav. Chim. Pays-Bas 1988,107, 23. 17. (a) Espenson, J. H.; McHatton, R. C. Inorg. Chem. 1981, 20, 3090; (b) Zhong, Z. J.; Okawa, H.; Matsumoto, N.; Saldyama, H ; Kida, S. J. Chem. Soc., Dalton Trans. 1991, 497; (c) Ruiz, R.; Sanz, J.; Cervera, B.; Lloret, F ; Juive, M ; Bois, C ; Faus, J ; Munoz, M C J. Chem. Soc., D alton Trans. 1993, 1623. 18. Bakac, A.; Brynildson, M. E.; Espenson, J. H. Inorg. Chem. 1986, 25, 4108. 19. Connolly, P.; Espenson, J. H. Inorg. Chem. 1986, 25, 2684. 20. Schrauzer, G N. in Inorganic Synthsis Jolly, W. L. éd., McGraw-Hill: New York, 1968, II, p. 6 1 . 21. Trogler, W. C ; Stewart, R. C ; Epps, L. A.; Marzilli, L. G. Inorg. Chem. 1974, 13, 1564. 22. Das, J. K ; Dash, K. C. Polyhedron 1986, 5, 1857. 23. Jensen, F. R.; Kiskis, R. C. J. Am. Chem. Soc. 1975, 97, 5820. 24. Ramasami, T.; Espenson, J. H. Inorg. Chem. 1980,19, 1523. 25. Stadlbauer, E. A.; Holland, R. J ; Lamm, F. P.; Schrauzer, G. N. Bioinorg. Chem. 1974, 4, 67. 2 0 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26. (a) Barefîeld, E. K.; Mocella, M. T. J. Am. Chem. Soc. 1975, 97, 4238; (b) Gaudio, J. D ; La Mar, G. N. J. Am. Chem. Soc. 1978, J O O , 1112; (c) Küdahi, N. K.; Vîriyanon, P. Inorg. Chem. 1987, 26, 4188; (d) L ev^, G ; Sweigart, D A.; Jones, J. G ; Prignano, A. L. J. Chem. Soc., D alton Trans. 1992, 605 and the references therein. 27. Trogler, W. C ; Marzilli, L. G. Inorg. Chem. 1975, 14, 2942. 28. (a) Schrauzer, G. N. Ann. N. Y. Acad. Sci. 1969, 158, 526; (b) Schrauzer, G. N.; Weber, J. H.; Beckham, T. M. J. Am. Chem. Soc. 1970, 92, 7078. 29. Shi, S.; Daniels, L. M ; Espenson, J. H. Inorg. Chem. 1991, 30, 3407. 30. Day, J. H. Chem. Rev. 1968, 68, 649. 31. (a) Evans, D. F. J. Chem. Soc. 1959, 2003; (b) Crawford, T. H ; Swanson, J. J. Chem. Edu. 1971, 48, 382; (c) Sur, S. K. J. Magn. Reson. 1989, 82, 169. 32. O’Connor, C. J. in Progress in Inorganic Chemistry Lippard, S. J. ed., Wiley & Sons: New York, 1982, 29, p. 203. 33. Boudreaux, E. A. ; Mulay, L. N. eds. Theory and Applications o f M olecular Paramagnetism Wiley & Sons: New York, 1976, p. 477. 34. Cotton, F. A _; Wilkinson, G. Advanced Inorganic Chemistry \^ley & Sons: New York, 1988, 5th ed., p. 733. 35. See 33, pp. 211-225. 36. Nfihichuk, L. M.; Mombourquette, M. J.; Einstein, F. W. B.; Willis, A. C. Inorg. Chim. Acta 1982, 63, 189. 37. The dimerization of RCo(DH)z via Co-dmg oxygen linkage had been proposed based on variable temperature NMR results. See (a) Ludwick, L. M ; Brown, T. L. J. Am. Chem. Soc. 1969, 91, 5188; (b) Herlinger, A. W ; Brown, T. L. J. Am. Chem. Soc. 1972, 94, 388. 38. Peng, S.-M.; Liaw, D.-S.; Wang, Y ; Simon, A. Angew. Chem., Int. E d Engl. 1985, 24, 210. 39. Fryzuk, M. D ; Ng, J. B ; Retdg, S. J ; Huffinan, J. C ; Jonas, K. Inorg. Chem. 1991, 30, 2437. 202 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40. (a) Mita, K. Nippon Kagaku Kaishi 1986, 771; CA: 106: 175643; (b) Mita, K. Nippon Kagaku Kaishi 1987, 823; CA: 107: 49307; (c) Mita, K.; Hotta, K.; Takui, T. Bull. Chem. Soc. Jpn. 1988, 61, 3740. 41. Reinaud, O. M ; Yap, G. P. A.; Rheigold, A L.; Theopold, K. H. Angew. Chem., Int. Ed. Engl. 1995, 34, 2051. 42. (a) Halpem, J.; Phelan, P. F. J. Am. Chem. Soc. 1972, 94, 1881; (b) Labauze, G.; Raynor, J. B. J. Chem. Soc., Dalton Trans. 1981, 590. 43. (a) Schrauzer, G. N.; Windgassen, R. J. Chem. Ber. 1966, 99, 602; (b) Simandi, L I.; Budo-Zahonyi, E.; Nemeth, S. Inorg. Chim. Acta 1983, 77, L203. 44. See, for example, CoUman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R G. Principles and Applications o f Organotransition M étal Chemistry University Science: NCll Valley, 1987, p. 91. 45. Miskelly, G. M. Unpublished results. 46. For example, Adams, T. A ; Welker, M. E.; Liable-Sands, L M ; Rheingold, A L. Organometallics 1997, 16, 1300. 47. Connelly, N. G ; Geiger, W. E. Chem. Rev. 1996, 96, 877. 48. Finke, R. G ; Smith, B. L.; McKenna, W. A ; Christian, P. A. Inorg. Chem. 1981, 20, 687. 49. Kobayashi, H.; Hara, T.; Kaizu, Y. Bull. Chem. Soc. Jpn. 1972, 45, 2148. 50. Whitlock Jr., H. W ; Bower, B. K. Tetrahedron Lett. 1965, 52, 4827. 51. (a) Stolzenberg, A M.; Stershic, M. T. J. Am. Chem. Soc. 1988, 110, 6391; (b) Lahiri, G. K ; Stolzenberg, A M. Angew. Chem., Int. E d Engl. 1993, 32, 429. 52. It was recently reported that Co(DH)2Py' could be allqrlated by stirring with quaternary ammonium salts at room temperature. However, the ammonium salts they used usually contain benzyl substituents. The lack o f similar reactivity for normal quaternary ammonium salts is consistent with our observation here. See (a) Hilhorst, E.; Iskander, A S.; Chen, T. B. R. A ; Pandit, U. K. Tetrahedron Lett. 1993, 34, 4257; (b) Mlhorst, E ; Iskander, A S.; Chen, T. B. R. A ; Pandit, U. K. Tetrahedron 1994, 50, 8863. 2 0 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53. With Na(Hg) in addition of M6 4NC1, the reduction of ClCo(DBF2)2Py to Co(DBF2^Py' is not very clean indicated by the NMR. However, the ^ NMR of Co(DBF2)2Py is quite clean and the coupling pattern is similar to that o f Co(DBF2)2PBu3 54. (a) Kuhhnann, K.; Grant, D. M. J. P f^s. Chem. 1964, 68, 3208; (b) Belcher, W. J.; Boyd, P. D. W.; Brothers, P. J.; Liddell, M. J.; Rickard, C E. F. J. Am. Chem. Soc. 1994, 116, 8416. 55. Gillespie, R. J.; Hartman, J. S. Can. J. Chem. 1967, 45, 859. 56. Simandi, L. I ; Budo-Zahonyi, £.; Szeverenyi, Z ; Nemeth, S. J. Chem. Soc., Dalton Trans. 1980, 276. 57. Tilset, M.; Parker, V. D. J. Am. Chem. Soc. 1989, III, 6711. 58. For example, (a) Randaccio, L.; Bresciani-Pahor, N.; Toscano, P. J.; Marzilli, L. G. J. Am. Chem. Soc. 1980, 102, 7372; (b) Randaccio, L.; Bresciani-Pahor, N.; Toscano, P. J.; Marzilli, L. G. J. Am. Chem. Soc. 1981, 103, 6347. 59. (a) Halpem, J. Acc. Chem. Res. 1982, 15, 238; (b) Halpem, J. Inorg. Chim. Acta 1985, 100, 41. 60. Ungvary, F. J. Organomet. Chem. 1972, 36, 363. 61. De Vries, B. J. Catal. 1962,1 ,489. 62. (a) Setsune, J.; Yoshida, Z ; Ogoshi, H. J. Chem. Soc., Perkin 1 . 1982, 983; (b) Faraos, M. D ; Woods, B. A.; Wayland, B. B. J. Am. Chem. Soc. 1986, 108, 3659; (c) Wayland, B. B.; Balkus Jr., K. J.; Faraos, M. D. Organometallics 1989, 8, 950; (d) Sherry, A. E ; Wayland, B. B. J. Am. Chem. Soc. 1990,112, 1259. 63. (a) Setsune, J ; Ishimaru, Y.; Moriyama, T.; Khao, T. J. Chem. Soc., Chem. Commun. 1991, 555; (b) Gridnev, A A ; Ittel, S. D ; Fryd, M ; Wayland, B. B. J. Chem. Soc., Chem. Commun. 1993, 1010. 64. Wayland, B. B.; Ba, S.; Sherry, A E. Inorg. Chem. 1992, 31, 148. 65. Costa, G ; Tauzher, G ; Puxeddu, A Inorg. Chim. A cta 1969, 3, 45. 204 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A study of the double sulphates of some rare-earth elements with sodium, potassium, ammonium and thallium

PDF

A synthetic model approach to H-bonded oxyhemoglobin

PDF

Crystallographic studies of metal hydrides and other transition metal complexes

PDF

A photophysical study of pyrene adsorbed onto silicas of variable surface area and porosity

PDF

CO2 Laser-induced chemistry of C23CN

PDF

Anionic synthesis of highly isotactic polystyrene in the presence of lithium hydroxides or lithium alkoxides, and, Controlled "living" emulsion polymerization of methyl methacrylate by atom trans...

PDF

A study of the interactions of polar gases with solid proteins and some simple organic compounds

PDF

alpha-Fluoromethanediphosphonic acids and derived ATP analogs

PDF

Binuclear porphyrin and imidazolate bridged complexes as models for cytochrome c oxidase.

PDF

Carbocations, carboxonium dications and their role in superacid catalyzed reactions

PDF

A critical study of the hydrogenation of talloel

PDF

A study of the structures and properties of triaryl compounds of boron

PDF

A highly efficient and general method for the chemical synthesis of beta,gamma-fluoro- and difluoromethylene analogs of purine and pyrimidine 5'-ribo- and deoxyribonucleotides

PDF

A determination of ionic diffusion coefficients

PDF

AMERICAN DIPLOMATIC RELATIONS WITH AUSTRALIA DURING THE SECOND WORLD WAR.

PDF

A critical study of the methods for the quantitative determination of beryllium

PDF

A classical mechanical method of analysis and periodic orbit search: Applications to HCN/HNC, C1HC1, FH2, and their deuterates

PDF

A study of the kinetics of the reaction of sulfenyl halides with olefins

PDF

A study of the Mohr and Volhard methods of determination of chloride content in brines

PDF

Crystallographic stydies on Zn(2), Mn(2) and Pt(2) complexes of phosphonates

Asset Metadata

Creator Sun, Yongping (author)

Core Title A.The anions of C(60) and its pyrrolidine derivatives B.The search for hydridocobaloximes

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Chemistry

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag chemistry, inorganic,chemistry, organic,OAI-PMH Harvest

Language English

Advisor Reed, Christopher (committee chair), [illegible] (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-300446

Unique identifier UC11353101

Identifier 9816073.pdf (filename),usctheses-c17-300446 (legacy record id)

Legacy Identifier 9816073.pdf

Dmrecord 300446

Document Type Dissertation

Rights Sun, Yongping

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA