UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Charge transfer in conjugated organometallic materials Zhu, Yongbao 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-566560.pdf [ 6.74MB ]
Metadata
JSON: 831-1.0061484.json
JSON-LD: 831-1.0061484-ld.json
RDF/XML (Pretty): 831-1.0061484-rdf.xml
RDF/JSON: 831-1.0061484-rdf.json
Turtle: 831-1.0061484-turtle.txt
N-Triples: 831-1.0061484-rdf-ntriples.txt
Original Record: 831-1.0061484-source.json
Full Text
831-1.0061484-fulltext.txt
Citation
831-1.0061484.ris

Full Text

CHARGE TRANSFER IN CONJUGATED ORGANOMETALLIC MATERIALS by YONGBAO Z H U B.Sc, Zhejiang University, 1987 M.Sc., Chinese Academy of Sciences at Fuzhou, 1990 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF GRADUATE STUDIES (Department of Chemistry) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA May 2000 ©Yongbao Zhu, 2000 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of frHZMlST&X The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract This thesis examines factors which influence charge delocalization in conjugated organometallic materials. This delocalization is examined by electrochemical and spectroscopic characterization of model complexes, oligomers and polymers. The complex [c/5-Ru(dppm)2(CsCFc)2]CuI (dppm = Ph2PCH2PPh2, Fc = ferrocenyl) (52) is prepared by the coupling of FcOsCSn(«-Bu)3 (60) and RuCl2(dppm)2 in the presence of excess Cul, while frans-Ru(dmpe)2(C=CFc)2 (dmpe = Me2PCH2CH2PMe2) (54) is obtained from RuCl2(dmpe)2 and 60 using catalytic Cul. Removal of the coordinated Cul from 52 with excess P(OMe)3 yields rraHs-Ru(dppm)2(G=CFc)2 (53). The mono- and dications of both 52 -54 and ^fl«5(/rfl«5,^a/75-Ru(PBu3)2(CO)(L)(C=CFc)2 complexes (L = CO (55); py (56); P(OMe)3 (57)) are prepared by oxidation with FcPF6. A l l the neutral and oxidized species are characterized using UV-Vis-near-IR spectroscopy and cyclic voltammetry (CV). The data are interpreted according to the Hush model of electron transfer, and the results indicate that ruthenium bisacetylide bridges facilitate electronic interactions between the two terminal ferrocenyl groups. Charge delocalization between the F e m and R u n centers in the oxidized species is enhanced when the ancillary ligands on the Ru center are electron donors and is lessened when the ligands are acceptors. Complexes ?rans-Ru(dppm)2(Cl)(C=CR) (62a - c), /ra^-Ru(dppm)2(GCR)2 (63a -c), and Fc-=-R (68a - e), Fc-=-R-=-Fc (69a - e) and R - F c - R (40, 50 and 73) (R = 1 - 3 linked thiophene rings with various substituents) are prepared to elucidate electronic interactions between the metals and oligothienyl groups. The complexes are all redox-active due to the Ru117™ and Fe11™ couples and oligothiophene-based oxidations. The CVs of 62a - c i i and 63a - c show that the Ru oxidation process becomes more reversible with an increase in the conjugation length of the oligothienyl group. Complexes 68a - e, 69a - e, 40,50 and 73 all contain a reversible F e ^ oxidation wave and an irreversible oligothiophene-based wave. When oxidized past the oligothiophene-based oxidation potential, and by careful exclusion of water, the complexes with terminal bi- and terthienyl groups (50, 62b, 62c, 68c, 68e and 73) electropolymerize or dimerize, resulting in the deposition of an electrochemically active film on the electrode surface. The monocations 62c+ and 63c+ prepared in solution at -20 °C exhibit intense L M C T absorption bands at 500 - 700 nm and 900 - 1700 nm, indicative of significant charge derealization from the R u m to the conjugated oligothienyl group. Electrochemical oxidation of the Fe11 centers in the ferrocene-oligothiophene complexes yields the corresponding monocations and dications, which all have oligothiophene-to-Fera charge-transfer transitions in the near-IR region. For each series, the energy and intensity of these low-energy transitions correlate to the difference in the oxidation potentials of the ferrocenyl and oligothienyl groups, showing that charge delocalization in these compounds is enhanced when the conjugated organic group and the metal are close in oxidation potential. The complex fra/w-RuCi2(dppm)2 (19) is converted to its cis isomer (59) at 20 °C in the presence of catalytic CuCl or Cul, and [{cw-RuCi2(dppm)2}2Cu][CuCi2] (74) is isolated when 19 or 59 reacts with excess CuCl. Addition of a small amount of 74 to a solution of 19 results in isomerization of 19 to 59. Complex 74 can be converted quantitatively to 59 by reaction with excess [(n-Bu)4N]Cl or HC1 (aq). A mechanism for the catalytic isomerization of 19 to 59 and the formation of 74 is proposed on the base of in situ 3 1 P N M R results. iii s V Fe P h 2 P . J >C T ^ PhaP / P h 2 U P P h > ^ & P V v P P h 2 ^ Fe 52 53 Me^P PMe2 L PBu 3 ^ > — R u - = - ^ > ^ = Ru & M e ^ P M e , ^ ^ B U 3 P ' ^CO ^ 54 55 L = CO 56 L = py 57 L = P(OMe) 3 Me 2 P v ™ e 2 nJV?a ^ ^ ^ - S n ( « - B u ) 3 C l - R i i - C l "'2^<Z\ Fe / \ P h 2 P ' | " C l Me 2 P PMe 2 U-PPh2 \_7 58 59 60 iv / ^ P P h Ph 2P/,.. J ..*C1,.._ . P h 2 P ^ | %C1 XC1 U P P h 2 P h 2 P ^ "I C U ^ P ^ P P h 2 P > J + C u C l . 74 v Table of Contents Abstract ii Table of Contents vi List of Tables viii List of Figures ix List of Schemes xii List of Equations xiii List of Symbols and Abbreviations xiv Acknowledgements xviii Chapter 1 General Introduction 1 1.1 Organic Conjugated Polymers 2 1.2 Poly- and Oligothiophenes and Their Derivatives 7 1.3 Charge Transfer and Hush Theory 12 1.4 Transition-Metal a-Acetylide Polymers 21 1.5 Conjugated Polymers with Ferrocene in the Backbone 26 1.6 Electropolymerization to Prepare Metal-Thiophene Hybrid Polymers 32 1.7 Goals and Strategies 35 1.8 Scope 37 Chapter 2 Charge Derealization in Ruthenium(II) Bis(ferrocenylacetylide) Complexes 39 2.1 Introduction 39 2.2 Experimental 40 2.3 Results and Interpretation 45 2.3.1 Syntheses and Structure 45 2.3.2 Electrochemistry.. 50 2.3.3 Spectroscopic Characterization 53 2.3.3.1 Visible and IR Spectroscopies 53 2.3.3.2 Near-BR. Spectroscopy and IVCT 56 2.4 Discussion 67 2.5 Conclusions 72 Chapter 3 Models for Conjugated Metal Acetylide Polymers: Ruthenium Oligothienylacetylide Complexes 74 vi 3.1 Introduction 74 3.2 Experimental 75 3.3 Results and Interpretation 82 3.3.1 Syntheses and Structure 82 3.3.2 Electrochemistry 85 3.3.3 Spectroscopic Characterization 91 3.4 Discussion 96 3.5 Conclusions 97 Chapter 4 Charge Delocalization in (Ferrocenylethynyl)oligothiophene Complexes 98 4.1 Introduction 98 4.2 Experimental 99 4.3 Results and Interpretation 105 4.3.1 Syntheses 105 4.3.2 Electrochemistry 106 4.3.3 Spectroscopic Characterization 112 4.4 Discussion 116 4.5 Conclusions 121 Chapter 5 Charge Delocalization in Oligothienylferrocene Monomers and Polymers 122 5.1 Introduction 122 5.2 Experimental 123 5.3 Results and Discussion 125 5.3.1 Syntheses 125 5.3.2 Electrochemistry 126 5.3.3 Spectroscopic Characterization 133 5.3.4 Spectroelectrochemistry of Electropolymerized Films 135 5.4 Conclusions 139 Chapter 6 Copper(I) Halide Catalyzed Trans - Cis Isomerization of RuCkCdppmh 141 6.1 Introduction 141 6.2 Experimental 142 6.3 Results and Interpretation 144 6.3.1 Catalytic Isomerization 144 6.3.2 Synthesis and Structure of 74 145 6.3.3 N M R Studies 148 6.4 Discussion 152 6.5 Conclusions 155 Chapter 7 Suggestions for Future Work 156 References 158 vii List of Tables table 1. Oxidation Potentials of Thiophene Derivatives and Corresponding Polymers 10 Table 2. Band Gaps of Pt-Containing a-Acetylide Polymers 25 Table 3. Summary of Crystallographic Data for 52-2(CHCl3)fl 45 Table 4. Selected Bond Lengths in 52-2(CHCl3) (A) 48 Table 5. Selected Bond Angles in 52-2(CHCl3)-(deg) 48 Table 6. Electrochemical Data for 52 - 57 and 61 50 Table 7. Visible and IR Spectroscopic Data for 52 - 57 55 Table 8. Near-IR Spectroscopic Data for Monocations 52+ - 57* 57 Table 9. Near-IR Spectroscopic Data for Dications 522+ - 572+ 58 Table 10. Selected Bond Lengths (A) and Angles (deg) for 63c 84 Table 11. Spectroscopic and Electrochemical Data for 62a - c 91 Table 12. Electrochemical and UV-vis Spectroscopic Data for 68a - e and 69a - e 106 Table 13. UV-vis-Near-IR Spectroscopic Data for 68a+ - e+ and 69a2+ - e2+ 114 Table 14. UV-vis - near-IR Spectroscopic and Electrochemical Data for 40,50 and 73... 128 Table 15. Crystallographic Data for 74-solvent 143 Table 16. Selected Bond Lengths in 74-solvent (A), 147 Table 17. Selected Bond Angles in 74-solvent (deg) 147 viii L is t of Figures Figure 1. Examples of some common organic conjugated polymers 3 Figure 2. Evolution of the L U M O and HOMO gap energy of oligo- and polythiophenes ".. 4 Figure 3. Solitons, polarons and bipolarons in polyacetylene and polythiophene 5 Figure 4. Evolution of the band structure of polythiophene upon oxidation: (a) the neutral form, (b) a polaron, (c) a bipolaron and (d) subbands. 6 Figure 5. a - a and a - P couplings in polythiophene 7 Figure 6. Proposed mechanism of thiophene electropolymerization 9 Figure 7. Potential-energy diagrams of initial and final states for (a) a symmetric mixed-valence complex and (b) an asymmetric mixed-valence complex 13 Figure 8. General structures of monomers targeted in this thesis 36 Figure 9. ORTEP diagram of the solid-state molecular structure of 52 (30% probability ellipsoids shown). Hydrogen atoms are omitted for clarity 47 Figure 10. Cyclic voltammograms of 52 (a) -0.2 - 1.2 V and (b) -0.2 - 1.4 V vs SCE, and 54 (c) -0.3 - 1.1 V and (d) -0.3 - 1.5 V vs SCE in CH 2 C1 2 containing 0.1 M [(«-Bu)4N]PF6. Scan rate = 100 mV/s. 52 Figure 11. Near-ER spectra of 56* and 56 2 + in CH 2C1 2 . Sharp absorptions are due to vibrational overtones from the solvent 60 Figure 12. Potential-energy diagrams for initial and final states for (a) states A and B (Scheme 13), and (b) states C, D and E (Scheme 14) 62 Figure 13. Plots of (a) vs /^,a(3) - Em(2) (R = 0.989) and (b) a 2 vs 7^,a(3) -Em(2) (R = 0.983) for the dications 5 3 2 + - 5 7 2 + '. 63 Figure 14. Relative energy diagrams for states A - D 64 Figure 15. Plot of (near-IR) vs l/n2 - l/Ds for 5 3 2 + with the best-fit line (R = 0.925). (a) trichloroethylene; (b) chlorobenzene; (c) o-dichlorobenzene; (d) C1CH 2CH 2C1; (e) CH 2C1 2; (f) nitrobenzene; (g) C H 3 C O C H 3 ; (h) C H 3 N 0 2 ; ( i )CH 3 CN 66 Figure 16. Plot of tsE\a = E\n(2) - E\r>(l) vs Avoc the difference in the vc=c between the neutral and dicationic complexes with the best-fit line (R = 0.999) 70 ix Figure 17. Potential energy diagrams for electron derealization in (a) 61 + and in a hypothetical molecule in which a R11L4 is inserted into the central C-C bond in61 + 71 Figure 18. Solid-state molecular structure of 63c 85 Figure 19. Cyclic voltammograms of (a) 62a (2.2 x if/" 3 M), (b) 62b (1.6 x IO - 3 M) and (c) 62c (1.8 x IO - 3 M) in CH 2C1 2 containing 0.1 M [(«-Bu)4N]PF6. Scan rate = 100 mV/s 88 Figure 20. Cyclic voltammograms of (a) 63a, (b) 63b (4.0 x IO - 4 M) and (c) 63c (3.5 x 10"4 M) in CH 2C1 2 containing 0.1 M [(n-Bu)4N]PF6. The scan rate = 100 mV/s. The dotted lines show the cyclic voltammograms in the range of -0.2 - 0.6VvsSCE 89 Figure 21. Multiple scan cyclic voltammograms of (a) 63b (4.0 x io-4 M) and (b) 63c (1.8 x io-3 M) in CH2CI2 containing 0.1 M [(/J-BU)4N]PF6 Scan rate = 100 mV/s 90 Figure 22. UV-vis spectra of 62a (—), 62b (—) and 62c (•••) in CH 2C1 2 92 Figure 23. UV- vis spectra of 63a (—), 63b (—) and 63c (•••) in CH 2C1 2 93 Figure 24. Vis-near-IR spectra of 62c+ (•••) and 63c+ (—) in CH 2C1 2 at -17 °C 94 Figure 25. Cyclic voltammograms of (a) 68a, (b) 68d, (c) 68b, (d) 68c and (e) 68e in CH2CI2 containing 0.5 M [(/i-Bu4)N]PF6. Scan rate = 100 mV/s 107 Figure 26. Cyclic voltammograms of (a) 69a, (b) 69d, (c) 69b, (d) 69c and (e) 69e in CH 2C1 2 containing 0.5 M [(n-Bu)4N]PF6. Scan rate = 100 mV/s 108 Figure 27. Cyclic voltammograms of (a) 70, (b) 71 and (c) 72 in CH2CI2 containing 0.5 M [(n-Bu) 4N]PF 6. Scan rate = 100 mV/s 111 Figure 28. Vis-near-IR spectra of 68a +- e +in CH2C12containing 0.1 M [(n-Bu) 4N]PF 6... 115 Figure 29. Vis-near-IR spectra of 69a 2 + - e2 + in CH 2C1 2 containing 0.1 M [(«-Bu ) 4 N ] P F 6 116 Figure 30. Absorption maxima Vmax (near-IR) vs the oxidation-potential difference AE = £p> a(2) -£1/2 (1) for 68a +- e+ 118 Figure 31. Absorption maxima v ^ (near-IR) vs the oxidation-potential difference AE = £ p , a (2) - £ 1 / 2 ( 1 ) for 69a 2 +- e2 + 119 x Figure 32. Oscillator strength / (near-IR) vs the oxidation-potential difference AE = £p,a(2) - £ I / 2 (1) for 68a+- e+ and 69a2 +- e2+ 120 Figure 33. Cyclic voltammogram of 50 at 20 °C in CH2CI2 containing 0.6 M [(«-Bu)4N]PF6 at a Pt working electrode (a) between 0 - 0.8 V and (b) multiple scans between 0 -1.6 V. Scan rate = 50 mV/s 129 Figure 34. Cyclic voltammogram of poly-50 on a Pt working electrode at 20 °C in CH2CI2 containing 0.6 M [(n-Bu)4N]PF6. Scan rate = 50 mV/s 130 Figure 35. Cyclic voltammogram of 73 at 70 °C in C1CH2CH2C1 containing 0.6 M [(/*-Bu)4N]PF6 at a Pt working electrode (a) between 0 - 0.7 V and (b) multiple scans between 0 - 1.3 V. Scan rate = 50 mV/s 132 Figure 36. Cyclic voltammogram of poly-73 on a Pt working electrode at 20 °C in CH2CI2 containing 0.6 M [(rc-Bu)4N]PF6. Scan rate = 50 mV/s 133 Figure 37. Vis - near-IR spectra of (a) 40+, (b) 50+ and (c)73+ in CH 2 C1 2 containing 0.13M[(«-Bu) 4N]PF 6 134 Figure 38. Spectroelectrochemistry of poly-50 on an ITO electrode at the oxidation potentials (a) -0.1 V , (b) 0.8 V and (c) 1.7 V vs SCE in CH 2 C1 2 containing 1.3M[(«-Bu) 4N]PF 6 137 Figure 39. Spectroelectrochemistry of poly-73 on an ITO electrode at the oxidation potentials (a) 0 V, (b) 0.7 V and (c) 1.5 V vs SCE in CH 2 C1 2 containing 1.3 M [(«-Bu)4N]PF6 138 Figure 40. ORTEP diagram of the solid-state molecular structure of 74-solvent. The solvent molecules and the phenyl groups, except the ipso carbon atoms, are omitted for clarity. The thermal ellipsoids are depicted at 30% probability 146 Figure 41. 3 1 P N M R spectrum of 74 in CD 2C1 2 149 Figure 42. Absorption spectra of 59 (...) and 74 (—) in CH 2C1 2 152 xi List of Schemes Scheme 1 ^ Scheme 2 ^ Scheme 3 ^ 19 Scheme 4 22 Scheme 5 22 Scheme 6 22 Scheme 7 • Scheme 8 ^ Scheme 9 : 3 ^ 31 Scheme 10 46 Scheme 11 49 Scheme 12 Scheme 13 ^ 64 Scheme 14 82 Scheme 15 • 82 Scheme 16 c . 1 7 105 Scheme 17 Scheme 18 ^ Scheme 19 ^ Scheme 20 Scheme 21 ^ Scheme 22 xii List of Equations Equation 1. v o p = 4 E& 14 Equation 2. Avi/2 = (2310 v m a x ) 1 / 2 14 Equation 3. Avm = [2300 ( v m a x - AE °)]m 14 Equation 4. /= 1.085 x 1015 G v ^ M V 14 Equation 5. M2 = a2e2d2 14 Equation 6. /= 4.6 x 10"9 e Av\a 15 Equation 7. a 2 = (H^ Aw) 2 = (4.2 x 10"24 6 Avx^lv^d2 15 Equation 8. v m a x = Xi + Xo + + AE° 15 Equation 9. Xo = (m2e2/h c) (1/r - \ld) (l/n2 - \/Ds) 15 Equation 10. M w = [Em(2) - Em(\)] 16 Equation 11. ln(/Q = nF (AEm)IKY 16 Equation 12. AE °= (e/h c) [Em(2) - Em(\)] +D = (e/h c) AEm +D 19 Equation 13. V m a x = Xi + Xo + A£" + D + (e/h c) [EPA(3) - £ 1 / 2 (2)] 63 Equation 14. = Xi + Xo + bE' + D + (e/h c) AE 117 Equation 15. M2 = 9.22 x 10"16 e2fhm^ G 117 xiii List of Symbols and Abbreviations Symbol Description Units c speed of light cm/s d intermetallic distance m transition dipole length m D correction factor in eqs 12-14 cm - 1 Ds static dielectric constant of solvent AE oxidation potential difference between two redox couples V AE ° ground-state energy difference (Figures 7b, 12) cm - 1 AE' additional energy due to spin-orbit or ligand-field splitting cm - 1 Avi/2 bandwidth at the half-peak height in spectrum cm - 1 e electric charge C E\a half-wave potential V Eg band gap eV £ p , a anodic peak potential V EPfi cathodic peak potential V Eh thermal activation energy (Figures 7a, 17) cm - 1 F Faraday constant C/mol / oscillator strength cm - 1 G degeneracy of states h Plank constant J s /4d resonance exchange integral (Figure 7a) cm - 1 z'P;a anodic peak current pA j ' P i C cathodic peak current pA Kc comproportionation constant J coupling constant Hz M transition dipole moment C m m number of electron transferred in an IVCT process n number of electron involved in a redox process xiv refractive index of solvent R gas constant J/(molK) correction coefficient r radius of a metal coordination sphere m T temperature K a delocalization coefficient or interaction parameter e extinction coefficient PvT'cm -1 ^max . wavelength at band maximum nm Vmax absorption energy at band maximum cm - 1 v o p optical-transition energy (Figures 7,12) cm - 1 v frequency (wavenumber) cm - 1 8 chemical shift ppm Xi inner rearrangement parameter cm - 1 Xo outer reorganization parameter v cm - 1 Abbreviation Description A angstrom OAc acetate Anal. analysis aq aqueous bpy bipyridine Bu butyl C coulomb Calcd calculated °C degrees Celsius CB conduction band cm centimeter Cp cyclopentadienyl C V cyclic voltammetry Cy- cyclohexyl A heat at reflux X V dd doublet of doublets deg degree dmpe Me2PCH2CH2PMe2 or l,2-bis(dimethylphosphino)ethane dppm Ph2PCH2PPh2 or bis(diphenylphospliino)methane dppp Ph2PCH2CH2CH2PPh2 or l,3-bis(diphenylphosphino)propane e electron eq equation Et ethyl eV electronvolt Fc ferrocene; ferrocenyl h hour HOCO highest occupied crystal orbital HOMO highest occupied molecular orbital IR infrared ITO indium tin oxide rVCT intervalence charge transfer K Kelvin L liter L M C T ligand-to-metal charge transfer LUCO lowest unoccupied crystal orbital L U M O lowest unoccupied molecular orbital M metal; mol/L (molar) m multiplet meter max maximum Me methyl mg milligram MHz megahertz min minimum mL milliliter M L C T metal-to-ligand charge transfer xvi mmol millimole mol mole mV millivolt near-IR near-infrared N M R nuclear magnetic resonance nm nanometer uA microampere pm micrometer o ortho OTf ~OS0 2 CF 3 or triflate p para ppm parts per million Ph phenyl py pyridyl ref reference S Siemens (conductance) s singlet; second SCE saturated calomel electrode sh shoulder SSCE saturated sodium chloride calomel electrode t triplet TCNE tetracyanoethylene TCNQ (2,2'-(2,5-cyclohexadiene-1,4-diylidene)bispropanedinitrile THF tetrahydrofuran U V ultraviolet V volt V B valence band vis visible vs versus v/v volume-to-volume ratio xvii Acknowledgements First and foremost, I would like to thank my supervisor, Dr. Michael Wolf, for helping me to be a good chemist. Without his continuous support, guidance, patience and encouragement I would never be in a position to present this thesis. He taught me chemistry and cared about my career development. I would also like to thank Dr. Peter Legzdins who took his time to read the thesis and make corrections. I must thank all present and past members in the Wolf group for their great friendship which made my life more cheerful. I specially appreciate Olivier Clot, Dylan Millet and Nathan Jones for their work which is related to my thesis. I would like to thank the departmental support staff, especially Liane Diarge, Marietta Austria (NMR), Peter Borda (elemental analysis), and the late Dr. Steve Rettig (X-ray). I appreciate the Thompson and Orvig groups for allowing me to extensively use their spectrometers. I would also like to thank Dr. Glenn Yap in the Department of Chemistry at the University of Windsor for the determination of two solid-state molecular structures. Finally I thank my wife Xuequn for her unlimited support and my little girl, DuoDuo, who used to wonder why Dad liked spending more time on work than with her. xviii 1 Chapter 1 General Introduction The miniaturization of microelectronic devices has been one of the most important scientific ventures in the latter part of the twentieth century. Enormous success has been achieved in the fabrication of integrated circuits for high-speed computers, encouraging continuing efforts towards further miniaturization. An important aspect of these efforts has been focused on the construction of devices on the nanometer scale. Forty years ago, when the construction of such devices was proposed in a speech entitled "Plenty of Room at the Bottom" by physicist Richard Feynman,1 it seemed to be far from reality. This reality is gradually corning nearer as research both on materials and tools for such devices is progressing. The availability of STM (scanning tunneling microscopy) and A F M (atomic force microscopy) has allowed for the observation and manipulation of single atoms or molecules. Achievements in molecular self-assembly, metal nanoparticles and molecular wires means that the materials needed for such devices are no longer out of reach.2 Fabrication of devices from one molecule or a small assembly of molecules is a possible route to nanoscale devices.3-4 This concept has been considered within an interdisciplinary field called molecular electronics.2 Molecular electronics was originally denned as the application of a single molecule to process signals, and now is generally recognized as the use of molecule-based materials for electronic applications.5 Many molecule-based devices, such as transistors, sensors and switches have been designed, and prototypes have been built.2-3'6 The use of molecular materials provides a huge opportunity for device miniaturization because the functionality of the device arises from the intrinsic properties of individual molecules. It has even been reported that a single, specifically-2 designed molecule is able to perform some simple mechanical functions,7'8 but such molecules are complex to design and synthesize. The preparation of a device using an assembly of such molecules is more practical. Carrying electricity through a single polymer chain or a small assembly is important for the fabrication of nanometer scale electronic devices. Organic conducting polymers are very useful for such applications due to their high electrical conductivity upon oxidation or reduction. Research into conducting polymers, triggered by the discovery of high conductivity in oxidized polyacetylene in 1977,9 involves many fields including chemistry, physics, and materials science. A comprehensive summary of research results in this area can be found in two editions of the Handbook of Conducting Polymers, published in 1986 1 0 and 1998,11 respectively. Chemists have discovered many new conducting polymers and tailored existing ones for specific purposes, such as the enhancement of conductivity, stability, solubility or processibility. The following introduction covers the fundamentals of organic conducting polymers, as well as recent advances in research on conjugated polymers in which transition-metals are integrated into the polymer backbone. Background on mixed-valence complexes and intervalence charge-transfer is also included, due to its relevance to the work reported in this thesis. 1.1 Organic Conjugated Polymers Organic conjugated polymers possess a one-dimensional backbone which consists of alternating single and double bonds (or triple bonds). This structural characteristic allows overlap between orbitals with p (or 7i) symmetry on adjacent atoms, and results in their unique poly-/>-phenylene electronic properties. The structures of several of the more well-studied organic conjugated polymers are shown in Figure 1. polyacetylene polythiophene H polypyrrole poly-/?-phenylenevinylene polyaniline Figure 1. Examples of some common organic conjugated polymers. Polyacetylene, which took central stage in the early phase of the evolution of this field, shows an increase in electrical conductivity of 12 orders of magnitude when it is oxidized (also called doping).10 Neutral polyacetylene is semiconducting or insulating with electrical conductivity between -10 - 10 S/cm depending on the ratio of cis to trans double bonds and the number of defects in the backbone. Doping polyacetylene causes a dramatic increase in conductivity up to ~1 - 105 S/cm. 1 0 The highest conductivity ever measured in the defect-free polyacetylene doped with iodine, is 105 S/cm, 1 2- 1 3 comparable to that of copper (106 S/cm). The conductivity of polyacetylene varies with doping agents used. Other organic conjugated polymers show similar electrical behavior to polyacetylene. For example, polythiophene is semiconductive (10 S/cm) in the neutral state, while its conductivity is as great as 103 S/cm upon doping.1 0 Hiickel theory predicts that extended conjugation along the polymer backbone causes a decrease in the energy gap between the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO). The absorption maxima in electronic spectra may be used as an estimate of the HOMO - L U M O gap. Polythiophene has ^max at ~ 450 nm, while thiophene has at 243 nm, 2,2'-bithiophene at 302 nm and 2,2':5',2"-terthiophene at 355 nm. 1 4 Neutral polythiophene has a band gap of -2.2 eV, and a conductivity of 10 - 8 S/cm. 1 5 L U M O Energy HOMO H- w / H + + + Conducting band (CB) E„ = - 2.2 eV Valence band (VB) n = 1 n = 2 n = 3 n = oo Figure 2. Evolution of the L U M O and HOMO gap energy of oligo- and polythiophenes. Theories have been developed to understand the behavior of doped organic conjugated polymers. 1 0 ' 1 1 One explanation involves the formation of solitons, polarons or bipolarons, as shown in Figure 3, in doped polymers. 1 0' 1 1 Neutral polyacetylene has two degenerate ground states, thus oxidation of polyacetylene results in the formation of solitons which are associated with an electronic state in the mid-gap. Other conjugated polymers, such as polythiophene, have a non-degenerate ground state. Lifting the ground-state degeneracy leads to the formation of polarons which are singly-charged paramagnetic states, and bipolarons, spinless bound states of double charges.10 (a) Polyacetylene Soliton (b) Polythiophene Polaron Bipolaron Figure 3. Solitons, polarons and bipolarons in polyacetylene and polythiophene. Both polarons and bipolarons in polythiophene are associated with two new electronic states between the HOMO and the L U M O (Figure 4). Upon oxidation of polythiophene both 6 polarons and bipolarons form, and they are associated with two new states, one 0.60 - 0.65 eV above the HOMO, and the other 0.65 - 0.70 eV below the L U M O . As the level of doping is increased these intergap states overlap to form subbands. The subbands have been observed by electronic spectroscopy. Oxidation of polythiophene results in the appearance of a new, very broad band extending from 650 to 2000 nm, indicative of the formation of subbands in doped polythiophene.10 Subbands (a) (b) (c) (d) Figure 4. Evolution of the band structure of polythiophene upon oxidation: (a) the neutral form, (b) a polaron, (c) a bipolaron and (d) subbands. Organic conjugated polymers have been proposed for many applications due to their unique optical and electrical properties. Applications have been discovered in many fields, such as in batteries, microelectronics, display devices, optical devices and coatings;1 0'1 1 however, most commercial applications require the polymers to be chemically and thermally stable both in the neutral and doped states, as well as to be easily processed. Although doped 7 polyacetylene is highly conductive, its applications are limited by its instability in air and water, and poor processibility. Recently, more attention has been paid to polythiophene and its derivatives, since they are more stable in different oxidation levels, and can be easily modified chemically. Oligothiophenes often have similar properties to polythiophene, and are readily available. 1.2 Poly- and Oligothiophenes and Their Derivatives Polythiophene has been prepared both by chemical and electrochemical methods.1 0'1 1 Both methods give primarily a - a coupling with some a - p coupling, depending on the routes used (Figure 5). Many polythiophene derivatives have been obtained using readily available p-substituted thiophenes, and their solubility and conductivity are significantly affected by the substituents.10-11 ^ 3 o r p a - a coupling a - P coupling Figure 5. a - a and a - p couplings in polythiophene. There are two common chemical methods for the preparation of polythiophene and its derivatives (Scheme 1): (a) oxidative coupling of thiophene or its dianion using oxidizing reagents such as FeCb, 1 6 AsFs 1 7 and C u C ^ . 1 8 This approach yields doped polythiophene directly, and (b) transition-metal catalyzed Grignard coupling of 2,5-dihalothiophene.11 The latter method avoids a - P coupling, which occurs frequently with oxidative coupling. 8 Scheme 1 Electropolymerization has also been extensively used to prepare polythiophene and its derivatives. In this method catalysts are not required, avoiding the impurities arising from transition-metal catalysts. Electropolymerization can produce high quality films deposited directly on electrode surfaces. Film thickness can be easily controlled through control of current, voltage and deposition time, and the deposited films can be characterized directly by cyclic voltammetry. The major drawback is that a - p coupling may occur due to over-oxidation of thiophene or the resulting intermediates during the polymerization process. Electropolymerization is typically carried out in a single-compartment cell with a three-electrode configuration. Platinum, gold and indium tin oxide (ITO) surfaces are typically used as working electrodes. Acetonitrile and dichloromethane are commonly used as solvents and tetraalkylammomum salts as supporting electrolytes. Insoluble polythiophene deposits as a film on the working electrode when a sufficiently high potential is applied to the working electrode, or the potential is repeatedly scanned beyond the thiophene-oxidation potential. The latter deposition method involves cyclic voltammetry (CV). C V is a versatile electroanalytical technique, consisting of cycling the potential at the working electrode in an analyte solution, and simultaneously measuring the resulting current. The current-potential plot yields thermodynamic and kinetic information about the redox processes of the analyte, 9 mcluding redox potentials and reversibility. When C V is used for electropolymerization, the polythiophene film can be obtained either in the neutral or oxidized form. The mechanism which has been proposed for electropolymerization of thiophene is detailed in Figure 6. Oxidation of the thiophene monomer produces a radical cation, which dimerizes and deprotonates to yield a neutral dimer. Oxidation of the dimer to its radical cation, followed by coupling of the dimer radical cation with another radical cation, allows for propagation of the polymer chain. Since electropolymerization proceeds via radical cation intermediates, it cannot occur in electrochemical media containing nucleophilic species. Figure 6. Proposed mechanism of thiophene electropolymerization.10 Oligothiophenes can also be electropolymerized to yield polythiophene, but this occurs at a lower oxidation potential than for thiophene. Interestingly, polythiophene prepared from an oligomer has a lower oxidation potential than polymer prepared from thiophene, possibly due to a lower number of a - P linkages. Table 1 gives the oxidation potentials of 10 thiophene, some thiophene derivatives, oligothiophenes and the corresponding polymers. Electron-donating substituents result in a decrease in the oxidation potentials of both monomers and the resulting polymers. Table 1. Oxidation Potentials of Thiophene Derivatives and Corresponding Polymers a Monomer Ep>a (monomer) (V vs SCE) £ p , a (polymer) (V vs SCE) Thiophene 1.65 1.1 2,2'-Bithiophene 1.20 0.70 3-Methylthiophene 1.35 0.77 3,4-Dimethylthiophene 1.25 0.98 3-Bromothiophene 1.85 1.35 3,4-Dibromothiophene 2.0 1.45 "Refs. 1 9- 2 0 In addition to their use as monomers for polythiophene, oligothiophenes are of interest since they are suitable model compounds to elucidate structure-property relationships for organic conjugated polymers, and they are electronic materials in their own right. Oligothiophenes up to 10 thiophene units long have been prepared using either chemical or electrochemical synthetic methods (Scheme 2). 1 4> 2 1" 2 6 Longer oligomers can be prepared by transition-metal catalyzed Grignard coupling,1 4 or by electrochemical dimerization.25 Tetrathiophene, sexithiophene, and octathiophene have all been chemically synthesized by lithiation of a shorter oligomer followed by dimerization with copper chloride. 2 4 ' 2 6 These 11 oligomers have been used in device applications such as thin film transistors, light-emitting diodes, photovoltaic cells and light modulators.27"29 Scheme 2 In solution, the absorption maximum (kmax) of the n - n* transition in oligothiophenes increases linearly as chain length increases from 1 to 5 units. For longer oligothiophenes, X^ax varies only very slightly with length, and is close to that of polythiophene.14'21 Similarly, a linear relationship between oxidation potential and chain length has been observed in this series.21 Since long oligothiophenes can be oxidized at lower potentials, and the resulting charge can be delocalized over a more extended conjugation system, longer oxidized oligothiophenes are more stable. The oxidized species of quinquethiophene and sexithiophene can be spectroscopically characterized in highly dilute solution before any coupling reaction occurs. 2 3 ' 3 0 C V has been used to characterize both poly- and oligothiophenes.10 The C V of polythiophene displays a broad redox wave due to the range of conjugation lengths present in the material, while oligothiophenes typically exhibit discrete oxidation processes. For instance, the C V of didodecylsexithiophene demonstrates two reversible redox waves with 12 is 1/2 = 0.34 V and 0.54 V referenced to Fc+/Fc, assigned to an one-electron transfer in each step leading to the radical cation and the dication.23 1.3 Charge Transfer and Hush Theory Originally developed for interpretation of charge-transfer spectra of mixed-valence complexes, Hush theory has also been demonstrated to be useful for the evaluation of other charge-transfer processes.31 These include charge transfer between organic donors and acceptors,32'33 as well as ligand-to-metal (LMCT) and metal-to-ligand charge-transfer (MLCT) processes.34"36 In Hush theory, charge transfer is considered as a process in which an electron is coupled between a donor and an acceptor via a single oscillator which has the same frequency in both initial and final states. 3 1' 3 7' 3 8 The electronic and optical behavior of polymers consisting of metal centers linked via conjugated organic groups depends on the extent of charge derealization between the metal centers, as well as between the metal and organic groups. It may therefore be beneficial to use Hush theory to understand charge derealization in such polymers. /=\ 1 5 + / = X " l 5 + ( N H 3 ) 5 R u - N s ^ N - R u ( N H 3 ) 5 ( N H 3 ) 5 R u - N ^ N - R u ( N H 3 ) 3 ( b p y ) la lb Mixed-valence complexes like the Creutz-Taube ion (la) 3 9 have been the subject of numerous theoretical and experimental investigations.37-38'40 This class of complexes contains metal centers in different oxidation states. Mixed-valence complexes have been classified as symmetric, in which the metal centers are identical (e.g. la), and asymmetric, in which the metal centers are different (e.g. lb). 4 1 Charge transfer is often observed between metal centers in mixed-valence complexes. On the basis of the degree of charge 13 delocalization Robin and Day divided mixed-valence complexes into three classes:42 Class I, completely valence-trapped (no charge delocalization between metal centers); Class II, partically delocalized (weak coupling); and Class III, completely delocalized (strong coupling). Figure 7. Potential-energy diagrams of initial and final states for (a) a symmetric mixed-valence complex and (b) an asymmetric mixed-valence complex. In Hush model, potential-energy diagrams are used to explain thermal and optical charge-transfer processes in binuclear mixed-valence complexes (Class II) as shown in Figure 7 31,37,38 The curve on the left represents the initial state (A) and the curve on the right the final state (B) of the charge-transfer process. A symmetric complex has the same energy for both the initial and final ground states (AE° = 0), while an asymmetric complex has AE° > 0. The vertical transition between the two states with transition energy v o p represents an optical intervalence charge-transfer (F/CT) process. Broad IVCT bands typically appear in the visible or the near-IR region (400 - 3000 nm). Thermal electron transfer can occur by 14 vibronic coupling of the initial and final states with activation energy In this model, the resonance-exchange integral (//ad) gives the degree of coupling between the two ground states. For Class II complexes, the following relationships have been derived to evaluate the degree of charge delocalization based on the features of their IVCT bands. 3 1 ' 3 7- 3 8 For symmetric mixed-valence complexes: v0p = 4 £ t h (1) Avi/ 2 = (2310v I M X) 1 / 2 (2) For asymmetric mixed-valence complexes: Av 1 / 2 = [2300(v m a x-A£ o ) ] 1 / 2 (3) Where Avm is the bandwidth at the half-peak height in cm - 1 , and v m a x is the energy of an IVCT band maximum in cm - 1 . The oscillator strength/of an IVCT band in a Class II mixed-valence complex can be theoretically derived as given in eq 4. Here, G refers to the degeneracy of the states concerned, e is the electric charge in C, and M is the charge-transfer transition dipole moment in C m . / = 1.085 x 1 0 1 5 G v m a x M 2 / e 2 (4) M2 = a2e2d2 (5) The oscillator strength / of a Gaussian band can be determined experimentally using eq 6. The delocalization coefficient or interaction parameter a 2 , which is proportional to the amount of time spent by an electron in a given state, can then be derived from eqs 4, 5 and 6 as given in eq 7. Here e is the extinction coefficient in rVT'cm - 1, and d is the transition dipole length in m. /=4 .6x 10- 9 sAv I / 2 (6) 15 a 2 = (HJv^f = (4.2 x 1(T24 £ A v ^ A w d2 (7) The dependence of the absorption maximum of an IVCT band on solvent has been used as a diagnostic test for a class II species. The solvent effect on the IVCT energy is given by eq 8. vmax = ii + x0 + A£' + AE° (g) Xo = (m2 e2lh c)(\lr - \ld)(\ln2 - 1/A) (9) AE' is the additional energy due to either a spin-orbit or a ligand-field splitting, Xi is the inner rearrangement parameter and Xo is the outer reorganizational parameter. The dielectric continuum treatment defines the outer-sphere reorganization energy Xo according to eq 9, where m is the number of electrons transferred, r is the radius of the metal coordination sphere in m, n is the refractive index and Ds is the static dielectric constant of the solvent. If x0 is assumed to be the only solvent-dependent term in eq 9, then vmso^ may be expected to vary linearly with (l/n2 - 1/A) for a Class II complex. Mixed-valence complexes can often also be characterized by CV. The electrochemical potential difference between two sequential redox processes, AE\a = [E\a(2) - £1/2(1)], in a symmetric mixed-valence complexes such as l a is related, to the degree of charge delocalization in the complexes. The comproportionation constant Kc of a symmetric binuclear complex such as L m M n M m L m (eq 11) can be calculated from AE\a (eq 10). Here, £1/2(1) and £1/2(2) are the first and second oxidation potentials of L j ^ M 1 1 ! ™ , n is the number of electrons in each redox process, and F is the Faraday constant. Charge delocalization makes contribution to the magnitude of A£ 1 / 2 (therefore Kc) along with several other factors, such as statistical distribution, electrostatic repulsion, inductive factor.38 Comparsion of A£ 1 / 2 16 (Kc) between the complexes having similar structures is often instructive with respect to the extent of charge derealization. The magnitude of Kc has been used as a criterion for the classification of mixed-valence complexes.38 Complexes with Kc> 106 are considered to belong to Class III complexes. In Class III complexes, the even distribution of charge between metal centers results in the breakdown of Hush model and solvent-independent absorption features. Most asymmetric mixed-valence complexes belong to either Class I or Class II because of the energy difference AE° between the initial and final ground states (see Figure 7b). Scheme 3 '"^ iSf> L"M"M",L° L™MmM'"1 Kc L m M n M n L m + L m M m M m L m L m M n M i n L m &Em = [Exl2{2)-Em{\)} (10) \n(Kc) = nF(AEm)/RT (11) Charge derealization between the metal centers in mixed-valence complexes is influenced by the nature of the bridge linking the two metal centers. Conjugated bridges enhance the electronic interactions between the two metal centers. For example, the C V of diferrocenylacetylene Fc-CsC-Fc shows two redox waves with AE\n of 0.13 V . The monocation [Fc-C=C-Fc] + has an IVCT band with Xmax at 1560 nm. 4 3 On the other hand, only one redox wave occurs in the C V of 1,2-diferrocenylethane FC-CH2CH2-FC, in which the bridge is saturated.44 The length of a bridge also greatly affects derealization of charge between the two metal centers. Ribou and coworkers synthesized and characterized a series of 17 diferrocenylpolyenes Fc-(CH=CH) n-Fc (n = 1 - 6) . 4 5 They showed that as n increases from 1 to 3, M1/2 decreases from 0.17 to ~ 0.10 V, and that the compound with a longer bridge (n = 4, 5 and 6) has only one redox wave. Other symmetric, bimetallic complexes of the type I ^ M - ( C s C ) n - M L m show similar effects of the bridge length on charge derealization. 4 6- 5 1 Gladysz and coworkers have synthesized and electrochemically characterized a series of linear symmetric birhenium complexes 2 . 4 8 - 5 1 They showed that lengthening the carbon chain results in a smaller difference in the oxidation potentials between the two Re centers: AEm -0.53 V for n = 2, 0.38 V for n = 3, 0.28 V for n = 4, 0.19 V for n = 6, 0.1 V for n = 8 and 0 V forn= 10. NO ON Re- ( = n Re PPh 3 Ph 3P 2 n = 2, 3, 4, 6, 8 and 10 P h 2 P ^ P P h 2 _ P h 2 P ^ P P h 2 P h 2 P ^ P P h 2 ^ P h 2 P ^ P P h 2 C l - R u - = — ( * \ = - R u - C l r\-Ru-==—ff \ = _ R „ . r i ^ = & J C l - R u - = = — ^ J - R u - C l P h 2 P / v ' p P h 2 X = = / P h 2 P ^ > P h 2 P h 2 P / v P P h 2 P h 2 P ^ P P h 2 3 4 P h . P ^ P P h , P h 2 P ^ P P h 2 c l - ^ - & - ( V ^ - ^ - c l P h 2 P v P P h 2 ' P h 2 P v P P h 2 5 Colbert and coworkers have studied the effects of the electron density in the bridge on charge derealization by the synthesis and characterization of a series of bimetallic ruthenium complexes 3 - 5 containing conjugated bridges of differing electron density.52 The CVs of 3 - 5 all show two separated redox waves, assigned to sequential oxidations of the two ruthenium centers with a potential difference tsE\a~ 0.30 V for 3, 0.33 V for 4 and 0.36 V for 18 5. The monocations of these complexes all have an TVCT band in the near-IR region. The calculated a 2 is 4.4 x IO"3 for 3 +, 6.5 x 10"3 for 4 + and 8.9 x 10~3 for 5+. Both results indicate that the more electron-rich thienyl bridge favors charge derealization in this series. There are some interesting bimetallic complexes 6 and 7 in which two ferrocenyl groups are linked via a metal bisacetylide bridge. The observation of two separated ferrocene redox waves with AE\a = 0.22 V in 6 indicates a strong electronic interaction between the two ferrocenyl groups. 5 3 ' 5 4 Complex 7, on the other hand, has only one ferrocene redox wave, suggesting only a weak electronic interaction between the two terminal ferrocenyl groups.55 P h 2 P ^ P P h 2 PPh 3 ^ > = Ru — ^ > ^ > = ^ P t = ^ > Pe P h P " N p P h Fe Fe I Fe ^h? 2 V 2 PPh 3 6 7 As shown in Figure 7b, charge transfer in an asymmetric mixed-valence complex should be affected by the energy difference AE° between the initial and final ground states as well as by the nature of the bridge between two metal centers. It is impossible to measure AE° experimentally, but the relative value of AE° can be estimated from the electrochemical data. As shown in Scheme 4, M i n M 2 m corresponds to the higher energy state (B) in Figure 7b, and the oxidation potential of this species can be given as A£,i/2(1) + D, in which D is a correction factor resulting from the presence of M 2 n i instead of M21 1. As given in eq 12 A£° of an asymmetric complex should be linearly related to the potential difference AEm between the two sequential oxidations of the metal centers. The CVs of complexes 8a and 8b show two reversible redox waves with A£ 1 / 2 = [£1/2(2) - £1/2(1)] = 0.59 V for 8a, and 0.80 V for 8b. The electronic spectra of 8a+ and 8b+ in 19 CH 2C1 2 have IVCT bands with A™* at 1590 nm (e = 3060 NT'crn"1, a 2 = 2.31 x IO - 3) for 8a+, and at 1295 nm(e= 1630 NT'cm - 1 , a 2 = 0.98 x 10 -3) for 8b + . 5 6 The complex with the smaller value of M 1 ^ has the lower energy absorption and a higher value of a 2 . Scheme 4 M ln M 2 n M / ' V 1 M i m M i i ImM M m M m AE° M / W " £ i / 2 0) EV2(l) + D AE° = (£ 1 / 2(2) - Em(l)) +D = AEm+D (12) ^ = F e ^ ^ — F e ^ Fe / PPh 2 Fe I PMe 2 ^gb» P h 2 P ^ ^g±fr M e ^ F ^ 8a 8b The observation of low-energy absorptions due to L M C T transitions, as well as charge transfer from an organic donor to an organic acceptor, has been reported recently. Crutchley and coworkers synthesized a series of Ruffl-cyanamide complexes 9, which have a L M C T band appearing in the near-IR region.3 6 Lambert and Noll prepared a set of six bistriaiylamine derivatives with varying organic conjugated bridges 10 - 12 . 3 3 The CVs of these complexes contain two reversible redox waves due to the sequential oxidations of the amine groups. Oxidation of only one amine group results in the appearance of low-energy absorption bands in the near-IR region, indicative of effective charge delocalization between 20 the two organic redox centers. In complexes 9 and 10+ - 12+ Hush model has been used to evaluate the degree of charge derealization. Low-energy absorptions in the near-IR region have also been observed in conducting polymers. Oxidized organic conjugated polymers show very broad and strong near-IR absorption bands, 1 0 ' 1 1 as do polymers such as 13, 5 7 which contain Fe, Ru, or Os 21 phthalocyanines bridged by tetrazine or 2,5-dimethyltetrazine. These polymers exhibit high conductivities (0.05 - 0.3 S cm - 1) even without oxidative doping. L4 Transition-Metal a-Acetylide Polymers Transition-metal a-acetylide polymers with the general structure [ - M L m - C s C - R - C s C - ] , , (M = Pd, Pt, Ru, Os etc., m = 2 or 4, R = conjugated organic groups), are of significant interest due to their rigid-rod backbone, high stability and possible extended n-conjugation.58'76 It has been speculated that transition-metal a-acetylide polymers may possess extended ^-conjugation along the polymer backbone due to the overlap of metal dn and alkyne pn orbitals; thus, these polymers may have similar optical and electrical properties to organic conjugated polymers.70 Experimental results have shown that some transition-metal a-acetylide polymers have third order non-linearities and electrical conductivities 7 7 - 7 9 More interestingly, transition-metal a-acetylide polymers may combine the properties of the metal and the organic conjugated bridge, and could be superior to organic conjugated polymers in some applications. The nature of the ligands on the metal centers can alter the chemical and physical properties of the polymers, such as the solubility, stability, electrical conductivity and non-linearity. The rigid-rod structure also gives some of these polymers liquid-crystalline properties. Many transition-metal a-acetylide polymers have been synthesized by the Hagihara and Lewis groups. 5 8 - 6 7 ' 7 1> 7 2 ' 7 5 Over 20 years ago, Hagihara and coworkers first synthesized soluble Pd and Pt polymers such as 14 via the copper halide-catalyzed coupling of trans-M(PBu 3 ) 2 Cl 2 (15) and zrans-M(PBu 3 ) 2 (C=C-C=CH) 2 (16) in the presence of Et 2 NH (Scheme 5). 5 8 22 Scheme 5 PBih I PBu, C l - M - C l + H = — M — — i PBu, 15 M = Pd,Pt PBu 3 16 Cul -H • E t 2 N H A PBu 3 - f - M — = 1 I PBu, 14 Lewis and coworkers have developed a new synthetic route using bis(alkynylstannyl) reagents, thus avoiding the need for amine solvents which may lead to the decomposition of the resulting polymers . 6 4 ' 6 5 ' 6 7 ' 7 1 . 7 2 . 7 5 ' 8 0 - 8 5 Reaction of Me3SnC=C-/?-CoH4-C=CSnMe3 (17) with 15 affords polymers 18 as shown in Scheme 6. Scheme 6 PBu, I + 17 C l - M - C l PBu 3 15 M = Pd,Pt 17 = Me 3 Sn-toluene PBu 3 -j-M—= I PBu 3 18 •SnMe, P h 2 P ^ P P h 2 C l - R u - C l + 17 Scheme 7 cu. l 4 I ! i P i > L y / w C 1 C H 2 C H 2 C 1 I ^ n P h 2 P v P P h 2 — z ^ P h 2 P v P P h 2 19 20 The Lewis approach has also been very successful for the preparation of polymers with the general structure [-M(PR 3 ) 4 -C=C-R-C=C-] n (M = Fe, Ru and Os) 6 4 . 7 1 .80,82 m m e presence of catalytic copper halide, /ra«5-Ru(dppm)2Cl2 (19) reacts with 17 yielding the soluble polymer 20 (Scheme 7).71 23 Many polymers analogous to 14 and 18 have been prepared using different acetylides. Lewis and coworkers have prepared polymers 21 - 26 with various conjugated bridges. 6 7 ' 7 5 ' 8 4 These polymers have been spectroscopically characterized, and their band gaps have been derived from their electronic spectra. They are wide band-gap semiconductors or insulators in their neutral states with band gaps around 3 eV (Table 2). The band gaps decrease with an increase in the conjugation length of the organic bridges. Polymer 24 has a terthienyl bridging group, and 26 has an anthracene bridge, and these polymers show lower band gaps than the others. Doping is expected to change the band gaps of metal acetylide polymers. There is only one example of this in the literature in which doping of 22 by exposure to iodine vapor or nitric acid results in the appearance of two new absorption bands with energy at 2.8 eV and ~2 eV respectively.78 PBu 3 PBu, , k t? = - =]. - H — O — f c 25 m= 1,2 24 26 Efforts in this field have also been directed toward the preparation of a-acetylide polymers containing different metal centers in their backbones. Takahashi and coworkers have prepared polymer 27, which contains alternating Pd and Pt centers, by the Cul-catalyzed coupling of 15 ( M = Pd) and 16 ( M = Pt) in the presence of E t 2 N H . 5 9 Using a similar approach Dixneuf et al. have synthesized the mixed Ru - Pd a-acetylide polymer 28.86 PBu, PBu, . ? ? | d PBu, PBu, 27 PBu, / / V H ^ ^ W A -Ph 2P^J>Ph 2 V ~ / PBu 3 28 Elucidation of the extent of rc-conjugation along the a-acetylide polymer backbone is of central interest in these materials. Since extension of 7i-conjugation in these polymers depends on the overlap of metal dn orbitals with alkyne pn orbitals, both the metal and organic bridge can affect the extent of conjugation. Frapper and Kertesz have carried out a theoretical study on [ - M L m - C s C - R - C s C - j n ( L = PR 3, CO or H; M = Ni , Pd, Pt, Rh, Co, Fe or Ru; m = 2 or 4; R = C=C orP-CGHA) using extended Htickel theory.70 They concluded that derealization of the highest occupied crystal orbital (HOCO) extends over the metal centers 25 for both m = 2, 4, and that polymers containing 4-coordinate metal centers are better candidates for conduction by intrinsic doping than polymers with 6-coordinate metal centers. Table 2. Band Gaps of Pt-Containing a-Acetylide Polymers Polymer Band gap (eV) [-Pt(PH 3) 2-(C=C) 2-] n 2.92* [-Pt(PH 3) 2-(C=C) 4-]n 2.42° [-PtiPHsh-C^C-p-CelU-C^C-]* 2.56a [-Pt(PH 3) 2-C=C-^-(C6H4) 2-C SC-] n 2.38° [-Pt(PBu 3 ) 2 -(C S C) 2 -]„ 14 3.23* 18 3.26* 23 3.11* 24 (m = 1) 2.80* 24(m = 2) 2.55* 24(m = 3) 2.40* 25(m=l) 2.70* 26 2.48* a Calculated.70 * Experimentally determined. 6 7 ' 7 5 ' 8 6 The band gaps of the polymers, which they defined as the energy difference between the lowest unoccupied orbital (LUCO) and the highest occupied orbital (HOCO), vary between 2.16 eV and 3.18 eV, depending on the nature of the metal, ligands and organic 26 linkers. Without doping, these polymers are semiconducting or insulating. The calculated band gaps of some Pt-containing a-acetylide polymers are summarized in Table 2, as well as the experimentally determined band gaps of closely related polymers. 1.5 Conjugated Polymers with Ferrocene in the Backbone Ferrocene-containing polymers are attractive electrical and magnetic materials due to the well-known thermal and redox stability of ferrocene. Ferrocene-containing polymers have been prepared either by attaching ferrocene as a pendant group on the polymer backbone or incorporating it into the backbone. The latter approach has yielded polymers containing directly linked ferrocene centers,8 7'9 1 as well as those containing saturated hydrocarbon,9 2 - 9 5 alkene 9 6 aromatic, 6 9. 9 7" 1 0 2 ER 2 (E = Si, Ge, and Sn), P and S bridges. 1 0 3" 1 0 7 Poly(l , l ' -ferrocenylene) 29, in which ferrocene centers are directly linked, can be prepared using several different synthetic routes. Polycondensation of ferrocene radicals in the presence of peroxides yields a polymer with a molecular weight greater than 5000 g/mol, but the polymer also contains other residues such as - C H 2 - and - O - in the main chain. 8 7 ' 8 9 Upon oxidation, this polymer becomes semiconducting with a conductivity of IO - 6 - 10~8 S/cm. More structurally well-defined poly(l,l'-ferrocenylene) has been prepared either via the condensation of l,r-dilithioferrocene with l,r-diiodoferrocene,90 or the reaction of dihaloferrocene with magnesium.91 The latter yields crystalline poly(l,r-ferrocenylene), which, upon oxidation with TCNQ, exhibits a conductivity of IO - 2 S/cm. 9 1 Soluble poly- and oligo(l,r-ferrocenylenes) have been characterized by CV. The cyclic voltammogram of poly(l,l'-ferrocenylene) with a molecular weight less than 900 g/mol contains a broad redox feature between 0.2 and 0.8 V vs S S C E , 1 0 8 while short 27 oligo(l,r-ferrocenylene) with less than 7 ferrocene units has a discrete redox wave for each ferrocene unit. Nishihara and coworkers have prepared a series of oligo( 1,1'-dihexylferrocenylene) complexes 30 with up to 7 ferrocene units by reaction of disodium dihexylfulvalene with FeCl2(THF)2.108 The C V of these oligomers reveals closely spaced Fe11™ waves, indicative of strong electronic interactions between adjacent iron centers. Ring-opening polymerization has been used with great success for the preparation of polymers with the main chain consisting of ferrocene and various ER2 (E = Si, Ge and Sn), P and S n bridges. 1 0 3 - 1 0 7 Ring-opening polymerization generally proceeds via a chain-growth mechanism, and yields polymers of high molecular weights. Manners and coworkers discovered that thermally induced ring-opening polymerization of the strained silaferrocenophanes (31), shown in Scheme 8, yields poly(ferrocenylsilanes) (32) with a molecular weight of 520,000 g/mol for R = M e . 1 0 4 Using this method they were able to prepare analogous polymers with other linking groups.1 0 7 The C V of 32 exhibits two reversible oxidation waves of equal intensity, corresponding to initial oxidation at alternating iron sites followed by oxidation at the remaining sites.1 0 3 Similar features occur in the cyclic voltammograms of analogous polymers with other linkers. 1 0 5 The appearance of two redox waves indicates that electronic interactions between adjacent iron centers exist in these materials. Fe - r ^ ' J n R R 29 30 n = 1 - 7; R = «-hexyl 28 Scheme 8 <^f R Fe R Jn 31 32 R = Me, Et, n-Bu and «-hexyl Although the CV of such polymers shows that some electronic interactions exist among adjacent iron centers, the possibility of appreciable electron delocalization along the polymer chain is excluded by both electronic and Mossbauer spectroscopies. 1 0 6 ' 1 0 7 ' 1 0 9 The electronic spectra of polymers 32 with different R groups show similar absorption features in the visible region to those of ferrocene and monomers such as 31, and The Mossbauer spectra of the partially oxidized 32 reveal discrete Feu and F e m environments. Conductivity measurements further support essentially localized electronic structures in these polymers. Polymers 32 are insulators with conductivities of 1CT13 - 10~14 S/cm in the neutral state. 1 0 6 ' 1 0 7 When doped with I 2 they become semiconductors with conductivities of 10~7 - 10"8 S/cm, consistent with a mechanism in which conductivity is due to electron hopping between localized redox sites. ^ ^ C H 2 - r -Fe " t o R 4 ^ ^ C H 2 - C H 2 T -Fe R 33 34 35 R = H , Me Analogous polymers with saturated hydrocarbon bridges can be prepared either by step-growth or ring-opening polymerization. Polymer 33 with a methylene bridge and a 29 closely related polymer 34 were prepared via step-growth polymerization, 8 8 ' 9 2 ' 9 3 and 35 (R = H, Me) with an ethylene bridge was prepared from the corresponding ferrocenophanes via the ring-opening route 9 4 The conductivity of the neutral 34 is 5 x 10~12 S/cm, and the partially oxidized polymer has a conductivity of 5 x 1 0 - 2 x 1 0 S/cm depending on the percentage of iron centers oxidized.8 8 The C V of polymer 35 (R = Me) shows the presence of a single reversible oxidation wave, indicative of virtually no interaction between iron centers.103 , All Al A H-Fe 36 37 38 R = w-hexyl, w-dodecyl; m = 0, 1,2 and 3 There has been significant interest in polymers with the main chain consisting of ferrocene groups linked via organic conjugated bridges such as aromatic, alkene or alkyne groups. 6 9 ' 9 6 - 1 0 2 Such hybrid polymers may possess extended 7t-conjugation along the polymer chain, thus combining the properties of 7i-conjugated polymers and ferrocene. Rosenblum and coworkers first prepared oligomeric or low molecular weight polymer 36 by coupling zincated ferrocene with 1,8-diiodonaphthalene in the presence of a Pd(0) catalyst.97 Later they developed a new route leading to the preparation of a closely related polymer 37, which is soluble, and has high molecular weight (14,363 g/mol) 9 8 Both polymers consist of 30 stacked ferrocene moieties held together by a 7t-conjugated naphthalene bridge. Using Pd(0)-catalyzed polycondensation Rehahn and coworkers prepared a series of poly(l , l ' -ferrocenylene-alt-p-oligophenylenes) 38, in which the ferrocene moieties are linked via p-oligophenylenes with the number of the phenyl rings ranging from 1 to 7 . 1 0 1 Scheme 9 Jn 39 R = H, w-hexyl The Gamier group has prepared hybrid polymers 39 (R = H, H-hexyl) by the Pd(0)-catalyzed coupling of zincated l,l'-dithienylferrocene 40 and the corresponding dibromothiophene (Scheme 9 ) . 6 9 ' 1 1 0 The main driving force behind this work has been the application of such hybrid polymers as magnetic materials. In the neutral state polymers 39 are diamagnetic, but oxidation of Fe11 to F e m results in the appearance of antiferromagnetic properties. For instance, when oxidized by NO2BF4, polymer 39 (R = H) can be attracted to a magnetic stirring bar. The bulk conductivities of the oxidized polymers have been measured, but the results are inconsistent. When doped with TCNE the measured conductivity of 39 (R = n-hexyl) is approximately 10~8 S/cm, 1 1 0 while a four-point probe conductivity measurement on 39 (R = H) when doped with FeCb, gives a value of 1 x IO - 3 S/cm. 6 9 31 Scheme 10 (a) Fe Fe ^ N > ^ — j Pd(PPh 3) 4, Cul, diisopropylamine _|__^*4^, 41 42 Jn (b) 41 (c) 41 B r M g — = — — E = = — M g B r Pd(OAc) 2, THF 17 Pd(PPh 3) 4 42 42 Fe Fe =_J/ w s' C*H 6 n 1 3 43 44 C 1 2 H 2 5 Fe Q2H25 45 Jn Yamamoto and coworkers have synthesized a poly(arylenethynylene)-type polymer 42 via the palladium-catalyzed coupling of l,r-diiodoferrocene (41) with HC=C-p-C6H4 -CSCH or B r M g C s C - p - C ^ - C s C M g B r (Scheme 10a, b ) . 1 0 0 Lewis and coworkers have obtained the same polymer using alkynyltrimethylstannane 17 (Scheme 10c). 1 1 1 By varying aromatic groups Yamamoto was able to obtain a series of closely related polymers 43 - 45.100 The cyclic voltammograms of these polymers all contain a single broadened Fe n/Fe r a redox wave, indicating that there is insignificant interaction between iron centers. Mossbauer characterization on the iodine adducts of polymer 42 yields the same conclusion. These 32 polymers all show an increase in conductivity when oxidized. Polymers 42 - 45 in the neutral state have conductivities of approximately 1.0 x 10~12 S/cm, while their iodine adducts have a conductivity of 1.3 x 10™* S/cm for 42,1.3 x 10"6 S/cm for 43 and 6.0 x 10~7 S/cm for 44. 1.6 Electropolymerization to Prepare Metal-Thiophene Hybrid Polymers Electropolymerization has been extensively used in the synthesis of organic conjugated polymers such as polythiophene and polypyrrole.10 Recently this approach has also been used to prepare hybrid polymers containing transition metal and thiophene units. This approach allows the direct preparation of polymer-modified electrodes for applications in electrocatalysis, chemical sensors and electrochromic displays.1 1 2 Electropolymerization typically produces polymers which are insoluble and deposit as a film on the electrode surface. Thus electropolymerization and film growth require that the resulting polymer has moderate to high conductivities. Hybrid metal-thiophene polymers can be obtained by anodic oxidation of transition-metal complexes bearing oligothiophenes with unsubstituted a positions. In order to obtain hybrid polymers with well-defined structures it is very important that the monomers be electropolymerized at potentials at which the metal centers are stable. Thus an important prerequisite for electropolymerization of such monomers is a relatively low oxidation potential. This may be achieved by the use of monomers which have either longer oligothienyl groups or electron-donating substituents. Shimidzu and coworkers have prepared porphyrin-containing polymers by electropolymerization of monomers 46 and 47 , 1 1 3 and Swager has prepared polythiophene-Ru(bpy)3n+ hybrids of monomer 48 , 1 1 4 in all cases, the monomers bear bi- or terthienyl groups. 33 46 M =Zn, Pd 47 n = 2, 3 Swager and coworkers have also focused on the synthesis and subsequent polymerization of Schiff-base complexes such as 49a and 49b . 1 1 5 ' 1 1 6 Electropolymerization of 49a (M = Co) results in the deposition of a yellow film on the electrode surface. The C V of the thin polymer film contains quasireversible Co1 1 7 1 1 1 wave at -0.1 V vs Fc+/Fc and two larger polymer-based waves at 0.3 and 0.6 V , but only a fraction of the Co centers are electroactive in a thick film. In situ conductivity measurements indicates that there is no contribution from Co11™ to the conductivity, and that the polymer has a maximum conductivity of 34 S/cm at the second polymer oxidation potential. By matching the oxidation potentials of the metal and organic groups in these polymers the electroactivity of the metal centers is enhanced. Electropolymerization of 49b (M = Co) which has ethylenedioxy substituents results in a polymer which has broad redox waves due to both the Co1 1 7 1 1 1 (-0.05 V vs Fc+/Fc) and polymer-based oxidation processes (0.1 V). The conductivity profile of this polymer shows a broad trace increasing from 34 approximately -0.4 V through a maximum at -0.05 V and a subsequent decrease to a plateau at higher potentials. The maximum conductivity of this polymer is 44 S/cm. N ' - R U v N N I * N 48 49a S" ^S M = Co, Cu, N i , U 0 2 49b Concurrent with the work reported in this thesis, Higgins and coworkers also reported electropolymerization of 50 and 51. 1 1 7 In both cases they obtained dark films in which the ferrocenyl groups are not electrochemically accessible. 51 R = /z-hexyl 35 1.7 Goals and Strategies As discussed in this Chapter many hybrid polymers with a backbone consisting of transition-metal centers and organic conjugated fragments have been prepared. Such polymers possess extended ^-conjugation along the polymer backbone due to the dn - pn orbital overlap between the metal and organic fragments, and in some cases demonstrate similar electrical properties to organic conjugated polymers. Many show a significant increase in their electrical conductivity when oxidized; however, the conductivity is often in the range of 10"2 - IO"7 S/cm, far lower than that of doped organic conjugated po lymers . 6 9 ' 8 4 ' 8 8 ' 9 1 ' 1 0 0 ' 1 1 0 Conductivity in such hybrid polymers can result from two possible mechanisms: electron hopping between adjacent metal centers, or charge delocalization over both metal and organic fragments. The contribution from electron hopping is expected to be least dependent on the dn - pn overlap between the metal and organic fragments, and to decrease dramatically with increasing distance between adjacent metal centers. On the other hand, charge delocalization is expected to have a greater impact if the energy levels and symmetries of the frontier orbitals of the metal and organic fragments favor the dn - pn overlap. The work presented in this thesis focuses on understanding the factors which influence the extent of charge delocalization and conductivity in such hybrid polymers. The strategies exploited in this thesis include the synthesis of polymerizable monomers containing both transition metal and conjugated organic fragments which show significant charge delocalization between the metal and organic components. Complexes, containing oligothienyl groups and metal centers, such as ferrocene, bisethynylferrocene or ruthenium bisacetylide (Figure 8), are targeted as monomers since they can be 36 electropolymerized via the oxidative coupling at the unsubstituted a positions of thiophenes. Such monomers may be synthesized by coupling a metal halide complex with an oligothiophene bearing suitable functional groups in the presence of Pd(0), Pd(II) or/and Cul catalyst, as shown in Schemes 5,6,7,9 and 10. ( M J Fe or Fe or — = R u L m - = — Figure 8. General structures of monomers targeted in this thesis. Polymers prepared from these monomers may exhibit several relevant charge-transfer processes such as IVCT, L M C T and MLCT. Analysis of these charge-transfer processes can also be carried out on the monomers as well as on model complexes such as Fc + /Fc-s -RuLn -=-Fc /Fc + and Fc + /Fc-^-R-=-Fc/Fc + (R = oligothiophene). The electronic interactions between the two ferrocenyl groups over a ruthenium bisacetylide bridge and an oligothiophene can be probed by electrochemical and spectroscopic methods. Analysis of electrochemical and spectroscopic data using Hush theory is expected to give insight into the extent of charge derealization in these complexes. The targeted monomers are expected to be electroactive due to redox processes involving both the metal and oligothienyl groups. The charge derealization due to L M C T and M L C T processes in these monomers at different oxidation levels can then be assessed using electronic spectroscopy and CV. These results can be used to guide the optimization of 37 charge delocalization in these monomers. Varying the ligands on the metal centers, and altering the conjugation length of the oligothienyl groups allows for the optimization of charge delocalization. 1.8 Scope The syntheses and characterization of a series of ruthenium bis(ferrocenylacetylide) complexes RuL4(CsCFc)2, and their mono- and dications are described in Chapter 2. The results show that the ruthenium bisacetylide bridge allows electronic interaction between two terminal ferrocenyl groups, and that the interaction can be enhanced by an increase in the electron density on the ruthenium center. Chapter 3 covers the syntheses and characterization of ruthenium monoacetylide (trans-Ru(dppm)2(C\)(C=C-R) (R = 2-oligothienyl) and bisacetylide (fra«.s-Ru(dppm)2(C=C-R)2) complexes. Complexes which contain sufficiently long conjugated oligothienyl groups electropolymerize and form electroactive and electrochromic films. An intense low-energy absorption due to an oligothienyl-to-Ru(III) charge-transfer transition in the monooxidized complexes occurs, indicating electronic delocalization between the metal and organic fragments in these monocations. The preparation, electrochemical and spectroscopic characterization of a series of mono- and bis(ferrocenylethynyl)oligothiophene complexes are described in Chapter 4. Changing the conjugation length and substituents of the oligothienyl group affects the degree of charge delocalization between the ferrocenyl and oligothienyl groups. Matching the oxidation potential of the metal center to that of the oligothienyl group enhances charge delocalization. Chapter 5 describes the preparation and electropolymerization of a series of 38 bis(oligothienyl)ferrocene complexes, and the spectroscopic and electrochemical characterization of these compounds and their electropolymerized films. In Chapter 6, the copper(I) halide-catalyzed trans - cis isomerization of RuCl2(dppm)2 is described. A mechanism for the catalytic isomerization is proposed on the basis of the solid-state molecular structure of a reaction intermediate [{c^-RuCl2(dppm)2}2Cu][CuCl2], and the identification of several other intermediates by N M R spectroscopy. Finally, Chapter 7 summarizes some suggestions for future work. 39 Chapter 2 Charge Delocalization in Ruthenium(H) Bis(ferrocenylacetylide) Complexes 2.1 Introduction Conjugated transition-metal a-acetylide polymers will only be highlyconductive if charge carriers can be delocalized over both the metal and organic fragments. Since many organic conjugated oligomers and polymers delocalize charge very well, it is important to find suitable metal bridging groups. Metal bisacetylide bridges are good candidates because they can be stable under the oxidizing conditions which are typically used to dope conducting polymers. They are also synthetically accessible for a large number of different metal-ligand combinations.118 Charge delocalization over a metal bisacetylide bridge can be probed by examining the electrochemical and spectroscopic behavior of molecules in which the bridge spans two redox-groups. The Wolf group and others have recently demonstrated that ruthenium bisacetylide complexes show electronic interaction between terminal redox groups, suggesting that such bridges are capable of delocalizing charge. 5 3 ' 5 4 This Chapter examines the effect of the ancillary ligands on the ruthenium on the electronic interactions between the terminal ferrocenyl groups in a series of ruthenium bis(ferrocenylacetylide) complexes 52 - 57. The effect of isomerization at the ruthenium is also evaluated by comparing the electronic properties of the cis complex 52 with those of the closed related complex 53. 40 Fe r P P h 2 ^ ^ P h P ^ P P h , P h 2 P ^ J P h 2 ° / * 2 Ph 2P | >&<T / % p 0 UPPh 5,g Ph 2 P v PPh 2 f , e Fe 52 53 Me 2 P PMe 2 L PBu 3 E E E - R U = ^ > — R u - = ^ M 0 ^ ^ Bu3r' \ » ^ 54 55 L = CO 56 L = py 57 L = P(OMe) 3 2.2 Experimental General. ?ra«s-RuCl2(dppm)2 (19), 1 1 9 frarts-RuCl2(dmpe)2 (58), 1 1 9 cis-RuCl2(dppm)2 (59), 1 2 0 FcPF 6 , 1 2 1 , ethynylferrocene122 and FcC=CSn(«-Bu)3 (60) 5 3 were all prepared using literature procedures. Complex 53 and its dication were previously prepared by the Wolf group, but were not characterized by near-IR spectroscopy.53 Complexes 55 - 57 were prepared by Olivier Clot in the Wolf group. A l l other reagents were purchased from either Strem Chemicals or Aldrich and used as received. Electronic absorption spectra were recorded on a U N I C A M UV2 UV-vis spectrometer. Near-IR spectroscopic data were obtained on a Varian Cary 5 spectrometer. Extinction coefficients and absorption maxima for overlapping near-IR bands were determined by fitting the data using multiple Gaussian curves. ER. data were collected on a U N I C A M Galaxy Series FTIR 5000 spectrometer. 'H , 1 3 C and 3 1P{'H} N M R experiments were performed either on a Bruker CPX-200, Varian X L -300 or Bruker WH-400 spectrometer. Spectra were referenced to residual solvent ( [ H, 1 3C) or 41 external 85% H3PO4 (3 1P). Elemental analyses were performed by Peter Borda in this department. Electrochemical measurements were conducted on a Pine AFCBP1 bipotentiostat using a Pt disc working electrode, Pt wire-coil counter electrode and saturated calomel reference electrode (SCE). The supporting electrolyte was 0.1 M [(w-BuUNJPFg, which was purified by triple recrystallization from ethanol and dried at 90 °C under vacuum for three days. Decamethylferrocene (-0.12 V vs SCE) was used as an internal reference. CH2CI2 used in C V was dried by refluxing over CaH.2. [c/s-Ru(dppm)2(CsCFc)2]CuI (52). To a solution of 60 (1.10 g, 2.20 mmol) in CH2CI2 (70 mL) under a nitrogen atmosphere was added 59 (0.82 g, 0.87 mmol) and Cul (0.26 g, 1.4 mmol). The red-brown suspension was stirred at 25 °C for 72 h. The cloudy solution was filtered through Celite 545, the volume of the solution was reduced to approximately 4 mL, and hexanes (100 mL) were added to precipitate a yellow-brown solid. The solid was dissolved in CH2Ci2 (50 mL), and a solution of sodium iodide (2.1 g, 14 mmol) in acetone (30 mL) was added. The solution turned cloudy after the solution was stirred at room temperature for 2 h. After the solvent was removed chloroform (20 mL) was added. The undissolved solids were removed by filtration, the filtrate was concentrated, and hexanes were added until the solution was almost saturated. A yellow-orange crystalline solid precipitated after the solution was cooled to -10 °C. The solid was dissolved in a small amount of choroform and reprecipitated by adding hexanes. The resulting powder was dried in vacuo at room temperature for 3 days. Yield: 0.94 g (72%). ' H N M R (400 MHz, CD 2C1 2): 5 8.23 (m, 4H, Ph), 8.03 (m, 4H, Ph), 7.50 - 7.15 (m, 24H, Ph), 6.86 (t, JHH = 7.1 Hz, 4H, Ph), 6.39 (t, JHH = 8.2 Hz, 4H, Ph), 4.67 (m, 2H, C5H4), 4.72 - 4.62 (m, 2H, CH 2 ) , 4.40 - 4.30 (m, 2H, CH 2 ) , 4.05 (s, 10H, Cp), 3.95 (m, 4H, C5H4), 3.88 (m, 2H, C5H4). 3 1P{'H} (81.015 MHz, 42 CDC13): 5 -16.9, -17.5 (AA 'BB' , J P P - 28 Hz). Anal. Calcd for C T ^ ^ C U F ^ R U : C 60.12; H 4.23. Found: C 60.38; H 4.25. [*rans-Ru(dmpe)2(X>CFc)2] (54). To a solution of 60 (1.00 g, 2.00 mmol) in chlorobenzene (70 mL) under a nitrogen atmosphere was added 19 (0.31 g, 0.66 mmol) and Cul (8 mg, 0.04 mmol). The red-brown suspension was heated at reflux overnight. The cloudy solution was filtered through Celite 545, the volume of the solution was reduced to approximately 4 mL, and diethyl ether (100 mL) was added to precipitate a yellow-brown solid, which was washed thoroughly with diethyl ether. The product was recrystallized from CH2Cl2/haxanes, and the crystalline solid contained 0.5 equivalent CH 2C1 2 . Yield: 0.38 g (69%). ' H N M R (400 MHz, CDC13): 8 5.29 (s, 1H, CH 2C1 2), 4.03 (s, 10H, Cp), 3.97 (s, 4H, C5H4), 3.89 (s, 4H, C5H4), 1.68 (s, 8H, CH 2 ) , 1.55 (s, 24H, CH 3 ). Anal. Calcd for C36.5H5iClP4Fe2Ru: C 50.86; H 5.96. Found: C 50.79; H 6.12. {[m-Ru(dppm)2(C=CFc)2]CuI}[PF6]2 (522+[PF6]2) A solution of 52 (67 mg, 0.046 mmol) in CH 2 C1 2 (4 mL) was cooled to -78 °C. To this solution was added a solution of FcPF6 (30 mg, 0.091 mmol) in CH2CI2 (3 mL). The solution turned brick-red immediately. After the solution was stirred for 2 min at -78 °C, hexanes (40 mL) were added to precipitate a brick-red solid. The solid was collected by filtration, washed with hexanes, and dried in vacuo at 80 °C for 3 days. Yield: 71 mg (88%). Anal. Calcd for C74H62CuFi2Fe2IP6Ru: C 50.26; H 3.53. Found: C 50.30; H 3.53. [frafls-Ru(dmpe)2(C=CFc)2][PF6]2 (542+[PF6]2). 54 (42 mg, 0.051 mmol) was dissolved in hot CH2CI2 (15 mL) under N2. To this solution was added FcPF6 (34 mg, 0.103 mmol). The solution turned blue-purple immediately. After the solution was stirred for 20 min at room temperature, hexanes (50 mL) were added to precipitate a dark-blue solid. The 43 solid was collected by filtration, washed with hexanes, and dried in vacuo at 80 °C for 3 days. Yield: 54 mg (95%). Anal. Calcd for CaeHsoF^Fe^eRu: C 38.98; H 4.54. Found: C 38.74; H 4.50. [trans,trans,trans-Ru(PBu3)2(CO)2(C^C¥c)2][V¥6h (552+[PF6]2). To a solution of 55 (56 mg, 0.057 mmol) in CH 2C1 2 (2 mL) was added a solution of FcPF 6 (38 mg, 0.12 mmol) in CH 2C1 2 (2 mL). The solution turned red immediately. After the solution was stirred for 30 min at room temperature, hexanes (40 mL) were added to precipitate a brick-red solid. The solid was collected by filtration, washed with hexanes, and dried in vacuo at 85 °C for 4 days. Yield: 64 mg (88%). Anal. Calcd for C 5oH 7 2F 1 2Fe 20 2P4Ru: C 47.29; H 5.71. Found: C 47.15; H5.55. [fra«5,rra«s,frfl«s-Ru(PBu3)2(CO)(py)(C=CFc)2][PF6]2 (562+[PF6]2). To a solution of 56 (47 mg, 0.046 mmol) in CH 2C1 2 (2 mL) was added a solution FcPF 6 (30 mg, 0.091 mmol) in CH 2 C1 2 (2 mL). The solution turned purple-red immediately. After the solution was stirred for 10 min at room temperature, hexanes (30 mL) were added to precipitate a dark purple solid, which was isolated by filtration and washed with hexanes. The solid was collected by dissolving it in a small amount of CH 2C1 2 , removing the solvent and drying in vacuo at 90 °C for 7 days. Yield: 48 mg (80%). Anal. Calcd for C54H77Fi2Fe2OP4Ru: C 49.10; H 5.88, N 1.06. Found: C 49.19; H 6.03; N 1.04. ^fl«s,frfl«s,fra«s-Ru(PBu3)2(CO)(P(OMe)3)(C=CFc)2] [PF6]2 (572+[PF6]2). This complex was prepared as described for 562+[PF6]2. Yield: 77%. Anal. Calcd for C52H8iF12Fe204P5Ru: C 45.73; H 5.98. Found: C 45.27; H 5.85. Reaction of 53 with Cul. To a solution of 53 (130 mg, 0.10 mmol) in CH 2C1 2 (30 mL) was added Cul (38 mg, 0.20 mmol). After the suspension was stirred at 25 °C under 44 nitrogen for 24 h, the solution was filtered through Celite 545. The volume of the filtrate was reduced to approximately 1 mL, and hexanes (30 mL) were added to the solution to precipitate a yellow solid. The solid was collected by filtration, washed with hexanes, and dried in vacuo at room temperature overnight to obtain pure 52. Yield: 130 mg (88%). Reaction of 52 with P(OMe) 3. To a solution of 52 (150 mg, 0.10 mmol) in CH 2C1 2 (25 mL) was added P(OMe)3 (50 mg, 0.40 mmol). The solution was stirred at 25 °C under nitrogen for 2 days. During this time the solution became deeper red and cloudy. After this period, the volume of the solution was reduced to approximately 2 mL, and diethyl ether (20 mL) was added to the solution to precipitate an orange solid. The solid was collected by filtration, washed with diethyl ether, and dried in vacuo at 85 °C for 2 days. N M R indicated that the product was 53. Yield: 82 mg (64%>). Crystallographic Study. Data collection and structure determination were carried out by Dr. Glenn Yap (Department of Chemistry and Biochemistry, University of Windsor, Ontario). Single crystals of 52-2(CHCi3) were obtained by slow crystallization from layered CHCI3 and hexanes. Crystal data and refinement parameters were summarized in Table 3. Suitable crystals were mounted on thin, glass fibres with epoxy cement. The systematic absences in the diffraction data and the unit-cell parameters were uniquely consistent with the reported space group. The structure was solved by direct methods, completed by Fourier syntheses, and refined by full-matrix least squares procedures based on F2. The data were corrected for absorption by using redundant data at different effective azimuthal angles. Two symmetry-unique molecules of cocrystallized chloroform solvent were located in the asymmetric unit of 52-2(CHCl3). A l l non-hydrogen atoms were refined with anisotropic displacement coefficients. A l l hydrogen atoms were treated as idealised contributions. A l l 45 software and sources of atomic scattering factors are contained in the SHELXTL (5.03) program library. 1 2 3 Table 3. Summary of Crystallographic Data for 52-2(CHCl3)a Empirical formula C76H64Cl6CuFe2lP4Ru ^max/^min 1.695 Formula weight 1717.06 T,K 296 Crystal system Monoclinic Radiation M o K a (0.71073 A) Space group P2yc Diffractometer Siemens CCD Crystal color Yellow-orange Z 4 a, A 17.7522(6) / / (MoKa) , cm -1 16.4 b,k 17.3529(5) e 1.45-22.50 c, A 23.8798(7) R(F), % 5.24 A deg 90.221(1) RfwF2), % 12.5 F, A 3 7350.7(4) NJN0 10.5 D (calcd), g cm 1.552 GOF onF2 1.08 a Quantity minimized = RfwF2) = Z[w(F02 - £ c 2 ) 2 ]/S[w(F 0 2 ) 2 ] 1 / 2 ; R = E|(F 0 - FC)\/Z(F0), w = [a2(F02) + (0.0563P)2 + 23.7444F]"1; GOF = Goodness-of-fit. 2.3 Results and Interpretation 2.3.1 Syntheses and Structure The ruthenium bisacetylide complex 53 has previously been synthesized by the Wolf group by the Cul-catalyzed coupling of 59 and 60.53 The synthesis of 53 is sensitive to both the amount of Cul used and to the reaction temperature. When catalytic Cul (5 mol%) is used 46 and the reaction is carried out in 1,2-dichloroethane at reflux, 53 is obtained in good yield. When a stoichiometric amount of Cul is used a mixture of two new complexes with very similar N M R spectra is obtained. Elemental analysis of the mixture suggests that the two products are [Ru(dppm)2(C--CFc)2]CuCl and [Ru(dppm)2(C--CFc)2]CuI (Scheme 11). By metathesis of the product mixture with Nal pure 52 is isolated. Complex 52 is obtained in 72% yield when the reaction is carried out at room temperature in CH 2C1 2 for 3 days. Complex 52 may also be prepared in 70% yield from FcC=CSn(«-Bu)3 (60) and trans-RuCl2(dppm)2 (19) in the presence of excess Cul at room temperature. Scheme 11 P h 2 P ^ P P h 2 p h - p v p p h - & 53 p - P P h 2 Ph 2 P^ J ^ C l = Sn(»-Bu)3 ^ R u ^ + Fe Ph 2 P" | " C l \IL^ V p P h 2 ^ ^ ^ Q ^ F e 59 60 r p p h 2 v Ph,P^ I ™. ^ R u C / C u - I Ph 2P I ^ U P P h 2 Fe 52 Less Cul and higher temperatures favor the formation of the trans complex, while excess Cul and lower temperatures produce the cis complex. The cis complex is found to decompose in C1CH2CH2C1 at reflux. Complex 54 is prepared in 71% yield from the reaction of 60 and fran.s-RuCl2(dmpe)2 (58) at 132 °C using Cul (6 mol%) as catalyst. 47 Figure 9. ORTEP diagram of the solid-state molecular structure of 52 (30% probability ellipsoids shown). Hydrogen atoms are omitted for clarity. The solid-state molecular structure of 52-2(CHCi3) shows that the ruthenium center is in a distorted octahedral environment, and that the ferrocenylacetylide ligands are in a cis orientation around the ruthenium (Figure 9). The Cul unit is bonded in an n 2 fashion to both acetylide bonds. Many complexes in which Cu is bonded in this manner to an acetylide bond are known; 1 2 4 * 1 2 8 however, there are only a few examples of organometallic bidentate bis(alkyne) ligands which chelate Cu1, such as in [(r\5-C5H4SiMe3)2Ti(C--CSiMe3)2]CuOTf.129-132 48 Table 4. Selected Bond Lengths in 52-2(CHCl3) (A) Ru - P ( l ) 2.340(2) R u - C(4) 2.055(7) C u - I 2.529(1) Ru -P(2) 2.368(2) C u - C(l) 2.229(7) C(l) -C(2) 1.20(1) Ru -P(3) 2.370(2) C u - C(2) 2.143(6) C(l) -C(15) 1.467(8) Ru -P(4) 2.330(2) C u - C(3) 2.264(8) C(3) -C(4) 1.218(9) Ru -C(2) 2.062(8) C u - C(4) 2.140(8) C(3) -C(35) 1.452(7) Table 5. Selected Bond Angles in 52-2(CHCl3)-(deg) P(l) - R u - P(2) 71.62(6) C ( l ) - C u - C ( 4 ) 110.1(3) P(l) - R u - P(3) 102.81(7) C(2 ) -Cu-C(3 ) 110.0(3) P(l) - R u - P(4) 172.04(7) C(2 ) -Cu-C(4 ) 78.8(3) P(2) - R u - P(3) 95.08(8) C(3 ) -Cu-C(4 ) 31.9(2) P(2) - R u - P(4) 102.78(7) C(2) -C( l ) -C(15) 160.9(7) P(3) - R u - P(4) 71.64(7) C ( 2 ) - C ( l ) - C u 70.1(5) P(l) - R u - C(2) 95.6(2) C(15) - C(l) - Cu 128.7(4) P(l) - R u - C(4) 89.0(2) C ( l ) - C ( 2 ) - R u 168.9(7) P(2) - R u - C(2) 93.1(2) C ( l ) - C ( 2 ) - C u 78.0(4) P(2) - R u - C(4) 159.7(2) R u - C ( 2 ) - C u 99.2(3) P(3) - R u - C(2) 161.4(2) C(4) - C(3) - C(35) 161.1(7) P(3) - R u - C(4) 95.1(2) C ( 4 ) - C ( 3 ) - C u 68.4(5) P(4) - R u - C(2) 90.2(2) C(35) - C(3) - Cu 128.9(5) 49 P(4) -Ru-C(4) 97.1(2) C ( 3 ) - C ( 4 ) - R u 168.7(6) C(2 ) -Ru-C(4) 82.6(3) C(3) - C(4) - Cu 79.6(5) C ( l ) - C u - C ( 2 ) 31.8(3) R u - C ( 4 ) - C u 99.5(3) C ( l ) - C u - C ( 3 ) 141.8(3) Isomerization of 52 and 53. In order to examine the behavior of 52 in the absence of the chelated Cul, P(OMe)3 is used to remove the Cul. This method has been used successfully to remove a CuOTf unit from the chelating bis(ri2-alkyne) unit of [(n5-CsH4SiMe3)2Ti(C=CSiMe3)2]CuOTf.130 The major product of the reaction of 52 with excess P(OMe)3 is 53 (Scheme 9) (64% isolated yield). When 53 is allowed to react with stoichiometric Cul for 24 h at room temperature, 52 is obtained in 88% isolated yield (Scheme 12). Scheme 12 Fe Fe 52 ^ ± 5 ? 53 2 The Cul acts to "lock" the complex in the cis form with the bis(r) -alkyne) units chelating the Cul. When the Cul is removed from 52 by complexation with P(OMe)3, isomerization to the trans isomer occurs. It is possible that 53 and its cis isomer are in 50 equilibrium, with 53 being the favored form for steric reasons. The chelated Cul must stabilize the cis isomer and trap the cis form as 52. 2.3.2. Electrochemistry Table 6 summarizes the electrochemical data of the series 52 - 57, as well as the literature data for 61 4 3 The electrochemical data of 55 - 57 are obtained by Olivier Clot. Fe Fe 61 Table 6. Electrochemical Data for 52 - 57 and 61 Complex £ i /2 ( l ) a (±0.01 V) Em(2)a (±0.01 V) £ 1 / 2 (3)or£p , a (3) a (±0.01 V) A£ 1 / 2 b (±0.02 V) Kf 52 0.20 0.34 0.97 0.14 260 53d 0.04 0.26 0.92 0.22 6100 54 -0.04 0.27 0.80 0.31 210000 55* 0.29 0.38 1.52e 0.09 35 56* 0.22 0.35 1.24e 0.13 170 57s 0.20 0.35 1.30e 0.15 370 61' 0.58 0.68 0.10 52 a Volts vs SCE, Pt working electrode, CH 2C1 2 containing 0.1 M [(n-Bu)4N]PF6,20 °C. B L\EV2 = [Em(2) - £i/ 2(l)]. c \n(Kc) = nF(AEM)/KT. ^Ref. 5 3 e £ p , a (irreversible wave). f Ref. 4 3. 8 Measured by Olivier Clot. 51 The cyclic voltammogram of complex 52 in CH2CI2 containing 0.1 M [(«-Bu)4N]PF6 at 20 °C, as shown in Figure 10a and b, contains three reversible waves of equal area between 0 and +1.1 V vs SCE. The waves at +0.20 V and +0.34 V are assigned to oxidation of the two ferrocenyl centers in 52. The potentials for the two waves are close to those observed for the two ferrocenyl centers in 5353 and consistent with results observed for related complexes.133 The potential difference between the first and second ferrocenyl oxidation waves in 52, A£ 1 / 2 , is 0.14 V (AEm = £1/2(2) - £1/2(1)). The redox wave at +0.97 V vs SCE in the C V of 52 is assigned to oxidation of the ruthenium center. This oxidation occurs at a slightly higher potential than the corresponding wave in the C V of complex 53. When the scan range is extended to +1.4 V vs SCE, a smaller quasireversible wave (wave 4) is observed at +1.18 V , along with a reduction wave (wave 5) at +0.81 V. The relative intensity of wave 4 compared to waves 1 - 3 in the C V increases as the scan rate is decreased, suggesting that wave 4 may be due to oxidation of a product resulting from decomposition of oxidized 52. Wave 5 may also be due to the decomposition of 52 upon oxidation. This feature is found to increase in intensity upon repeated scanning, possibly due to the deposition of the decomposition product on the working electrode. The C V of 54 is very similar to that of 52 or 53, containing three reversible redox waves with Em at -0.04, 0.27 and 0.80 V vs SCE in the range -0.4 - 1.10 V vs SCE (Figure 10c and d). A l l three redox potentials in 54 shift to lower potentials compared with those for 53 due to electron-donation from the dmpe ligands. The difference in the first and second redox potentials A£i/2 is 0.31 V for 54. When the scan range is extended to 1.50 V , an irreversible oxidation wave (wave 4) at 1.44 V vs SCE is observed, along with reduction waves (waves 5 and 6) at 0.66 and 0.88 V vs SCE. This oxidation wave is possibly due to a 52 RvF1™ oxidation process, and the waves 5 and 6 may result from reduction of the products formed in this process. Figure 10. Cyclic voltammograms of 52 (a) -0.2 - 1.2 V and (b) -0.2 - 1.4 V vs SCE, and 54 (c) -0.3 - 1.1 V and (d) -0.3 - 1.5 V vs SCE in CH 2C1 2 containing 0.1 M [(«-Bu)4N]PF6. Scan rate = 100 mV/s. 53 The cyclic voltammograms of complexes 55 - 57 contain two closely spaced waves which are assigned to oxidation of the two ferrocenyl centers, and the third irreversible wave which is assigned to the Ru1 1 7 1 1 1 oxidation. The Ru117™ oxidation potential is higher than that observed for 52 -54 due to increased backbonding with the carbonyl ligands in these complexes. 2.3.3 Spectroscopic Characterization Chemical oxidation of one or both ferrocenyl groups in complexes 52 - 57 allows for the spectroscopic properties of the oxidized species to be determined and compared with those of the neutral analogs. Shifts in diagnostic absorptions in the IR region as well as the appearance of intervalence charge-transfer bands in the near-IR region are useful in evaluating the extent of electronic derealization in these complexes. The dications 52 2 +- 572+ are prepared by oxidation with two equivalents of FcPF<6 and are isolated as stable solids which are characterized by elemental analysis. The dications are shown to be paramagnetic at room temperature using the Evans method. 1 3 4 The monocations 52+ - 57+, prepared by dissolving equimolar masses of the neutral complex and the corresponding dication in an appropriate solvent, are in equilibrium with the corresponding neutral and dicationic species (Scheme 3). The equilibrium constants (Kc) for 52+- 57* at 20 °C are calculated from AE\a using the Nemst equation (eq 11) and are shown in Table 6. 2.3.3.1 Visible and I R Spectroscopies The visible and IR spectroscopic data for the neutral and oxidized species are shown in Table 7. The visible spectra of the neutral complexes 52 - 57 all contain ligand-based 54 absorptions (data not shown) in addition to weaker bands (430 - 460 nm) which are assigned to metal d - d transitions. The monocations 52 +- 57* all exhibit medium-intensity transitions in the visible region between 500 - 625 nm which are assigned to a ferrocenium-based ligand-to-metal charge-transfer (LMCT) excitation. This assignment is based on the similarity of the energy and intensity of the band to that observed for ferrocenium (620 nm). 1 3 5 The energy of (the L M C T absorption of ferrocenium blue-shifts with the introduction of electron-withdrawing substituents,135 and the absorption is observed at 510 nm in [Fc-C=C -GsC -Fc] + ( 6 1 V 3 The L M C T absorptions for the monocations 52+- 57+ appear at higher wavelengths than the absorption for 61+, consistent with electron donation from the ruthenium moiety to the ferrocenium. In the visible spectra of the dications 522+ - 572+ the LMCT absorption blue-shifts with respect to the corresponding monocation, consistent with competition between the two ferrocenium groups for the electron density of the ruthenium center. Significantly, the absorption maximum for the L M C T transition is sensitive to the electronic nature of the ancillary ligands around the ruthenium center. Within the series of monocations of the trans substituted complexes, Xmsa for the L M C T absorption decreases in the order: 53+ (616 nm) * 54+ (625 nm) > 56+ (570 nm) * 57+ (562 nm) > 55+ (514 nm). The same trend is observed within the series of dications. The strong ER absorptions due to the C=C and C=0 groups are sensitive to the electron density at the ruthenium center. The energy of the acetylide absorption (voc) increases in the series of trans bisacetylide complexes as the number of electron-withdrawing carbonyl ligands on the metal increases (vc=c: 53 « 54 < 56 « 57 < 55). The energy of the absorption due to the carbonyl group (v<>o) is sensitive to the degree of backbonding with the ligand trans to the carbonyl. Thus vc=o in 56, which has a a-donor (py) trans to the carbonyl, 55 is lower than vc=o in 57 in which the trans P(OMe)3 is a weak Tt-acceptor. The carbonyl absorption in 55 has the highest wavenumber in the series as the two trans carbonyls compete for backbonding with the same Ru d-orbital. The acetylide absorption in the spectra of the • 2"!" 2"t" • dications 52 - 57 shifts to lower energy than the absorption for the analogous neutral complexes as more electron density from the ruthenium is transferred to the acetylide bond via increased backbonding. This is also observed in the carbonyl stretching frequency in 552+ -572+which increases slightly relative to the neutral analogs as less electron density is available for backbonding with the carbonyl groups upon oxidation of the ferrocenyl groups. Table 7. Visible and IR Spectroscopic Data for 52 - 57 Complex Visible A™* (nm) (±5 nm); IR (KBr) 8 (NT1 cm - 1) (±5%)" v ( c m _ 1 ) ( ± 2 c m _ 1 ) "52 460 (850) 1994 ( O C ) 52+ 500 (4700) (±10%) 522+ 378 (7700), 476 (sh) (8500) 1948 ( O C ) 53 434 (2000) 2067 ( O C ) 53+ 448 (sh) (2100), 616 (4100), 820 (sh), (1800) 532+ 420 (sh) (5500), 560 (13 000) 1997 ( O C ) 54 350 (12000), 455 (1150) 2057 ( O C ) 54+ 470 (sh) (3100), 625 (4100), 780 (4800) 542+ 430 (sh) (5800), 585 (13000) 1964 ( O C ) 556 450(760) 2103 ( O C ) , 1986 ( 0 0 ) 56 55 + 390 (sh) (5100), 514 (6900) (±25%) 55 2 + 378 (10 000), 488 (12 000) 2067 ( O C ) , 1994 ( O O ) 56* 450(920) 2079 ( O C ) , 1940 ( O O ) 56" 420 (sh) (3400), 570 (5300) (±15%) 56 2 + 394 (sh) (8500), 524 (12 000) 2030 ( O C ) , 1954 ( O O ) 51b 454 (810) 2084 ( O C ) , 1974 ( O O ) St 414 (sh) (3700), 562 (5300) (±10%) 57 2 + 394 (sh) (7500), 524 (9100) 2032 ( O C ) , 1979 ( O O ) a Solvent: CH2CI2, ±5% except indicated. Measured by Olivier Clot. 2.3.3.2 Near-IR Spectroscopy and I V C T The near-IR spectroscopic data for the monocations and dications are shown in Table 8 and 9 respectively. None of the neutral complexes absorb in the near-ER region. The spectra of the monocations and dications all contain multiple absorption bands in the near-IR region. For these absorptions, the band widths at half maximum (Avi/2) are measured directly when possible, and using Gaussian peak fitting when the band is overlapped. The spectra are obtained in several solvents in order to study the solvent dependence of the observed transitions. The lower-energy band is obscured by solvent overtones in many solvents; therefore, the data for this band are shown only for spectra taken in CH2CI2. The spectra of 52 + and 52 2 + are only obtained in CH2CI2 and 1,2-dichloroethane because the complexes decompose or react with other solvents used. For 54 + and 5 4 2 + the measurements are only carried out in CH2CH2. 57 Table 8. Near-IR Spectroscopic Data for Monocations 52+ - 57+ Complex Vmax (cm l) (±50 cm '); (Xmax, nm)" s O l " 1 cm - 1) A\'i/2 (cm - 1) (±50 cm - 1) a 2 (x 10~3) 52+ 8420(1190) 440 ±40 2400 1.4 ±0.2 5630 (1775) 290 ±30 2100 1.2 ±0.2 4320(2315) 310±30 810 53 + 4770 (2095) 6700 ±300 3300 4 7 ± 5 4380 (2285) 2700 ±100 700 54+ 4914(2035) 11000 ±600 2625 6 0 ± 6 55+ 8030 (1245) 2400 ±600 3500 12±4 4380 (2285) 410 ±100 540 56" 6430 (1555) 3200 ±500 2900 16±3 4410 (2265) 1100 ±200 510 57" 6520 (1535) 2900 ±300 3100 15 ± 2 4440 (2250) 1100 ±100 530 a CH 2C1 2,20 °C. 58 Table 9. Near-IR Spectroscopic Data for Dications 522+ - 572+ Complex Solvent" Vmax e a 2 (x 103) (±50 cm"1) (M^cm - 1 ) (±50 cm - 1) (±10%) (kmax, nm) (±5%) 522+ CH 2 C1 2 9620 (1040) 1600 3200 5.8 4260 (2350) 150 950 C1CH2CH2C1 9260 (1080) 1800 3600 7.5 532+ CH 2 C1 2 6650 (1505) 7400 2800 33 4470 (2240) 2400 480 CH3COCH3 6850(1460) 7500 2800 32 C1CH2CH2C1 6540(1530) 6400 2800 29 o-Dichlorobenzene 6210(1610) 5500 3800 36 Chlorobenzene 6120(1635) 5600 4400 42 CH3CN 6950(1440) 7800 2600 31 Nitrobenzene 6780 (1475) 10000 2100 34 Trichloroethylene 6170(1621) 6300 3500 37 C H 3 N 0 2 6900 (1450) 8400 2600 33 542+ CH 2C1 2 6270(1595) 9600 2560 40 4400 (2270) 4700 430 552+ CH 2 C1 2 9220 (1085) 3700 2800 12 4290(2330) 340 840 o-Dichlorobenzene 9130(1095) 3405 2900 12 59 CH3COCH3 9620 (1040) 3230 2830 11 C H 3 C N 9620 (1040) 4200 2870 14 Nitrobenzene 9440 (1060) 3980 2760 13 CH3NO2 9520(1050) 4075 2810 13 CH2CI2 7940 (1260) 5200 2900 21 4360 (2295) 860 570 o-Dichlorobenzene 7880 (1270) 5340 2880 22 CH3COCH3 8400(1190) 3875 3160 16 C H 3 C N 8370(1195) 4950 2980 20 Nitrobenzene 8130(1230) 4635 3151 20 CH3NO2 8300 (1205) 4340 3060 18 CH2CI2 7880 (1270) 3500 3200 16 4370 (2290) 620 640 o-Dichlorobenzene 7810(1280) 2825 3420 14 C H 3 C O C H 3 8200 (1220) 2900 3130 12 C H 3 C N 8230(1215) 2375 3190 10 Nitrobenzene 8030 (1245) 3505 3150 13 CH3NO2 8130(1230) 3190 3160 14 a 20 °C. 60 6000 1000 1500 2000 2500 3000 Wavelength (nm) Figure 11. Near-IR spectra of 56* and 56 2 + in CH2Q2. Sharp absorptions are due to vibrational overtones from the solvent. The near-IR spectra of all the mono- and dications contain a higher energy band (1040 - 2240 nm) and a lower energy band (2240 - 2350 nm). The spectrum of 52 + has the third broad absorption. Representative spectra for complexes 56* and 56 2 + are shown in Figure 11. The higher energy band is assigned to a Class II IVCT transition. The half-widths (2100 -4400 cm - 1) and intensities of the higher energy band are consistent with this assignment. 3 1 ' 3 8 ' 1 3 6 In addition, the absorption maximum of the band depends both on solvent and on the ancillary ligands around the ruthenium (vide infra), also consistent with the 61 band being an IVCT transition. The lower energy band is narrower (430 - 950 cm"1) and is of lower intensity than the higher energy band while the absorption maximum for the lower energy band is relatively insensitive to the changes in the ancillary ligands on the ruthenium center. The features of the lower energy band obtained in a limited number of solvents indicate that the absorption band is largely solvent-independent. Although it is difficult to unequivocally assign the lower energy band, it is clear that the properties of the lower energy absorption differ dramatically from those of the higher energy band. It is most likely that the lower-energy band is due to a d - d transition in Fe111 which becomes accessible in the mono- and dicationic species. The energy of such a band would be expected to be solvent-independent and to appear at approximately the same energy for all the mono- and dications.137 In addition, since the electronic transition is localized on the Fe m , it would be relatively independent of the changes in the ancillary ligands on the Ru center. Multiple bands in other mixed-valence complexes have been assigned to either ligand-field splitting 1 3 8 or spin-orbit coupling in the metal centers.139 The markedly different behavior of the two absorption bands in the series of complexes described here makes these explanations less likely. The lower energy absorption is not due to intermolecular charge transfer as the relative band intensities do not exhibit to be concentration-dependent, and the spectra are all taken at low concentrations (10"4 - 10"3 M). The higher energy near-IR absorption bands for the dications 53 - 57 (1040 - 1595 nm) are assigned to the IVCT transition shown in Scheme 13. In these complexes both iron Scheme 13 Fe i n -Ru n -Fe ! A in 62 centers are oxidized, and thus IVCT between the iron centers is not possible. The absorption maximum (v,^) for the IVCT band shifts to higher energy as the number of carbonyl ligands on the ruthenium increases ( V m a x : 53 2 + < 56 2 + * 57 2 + < 552+). The origin of this shift in v m a x may be rationalized using two overlapping potential energy curves (Figure 12a). In this diagram, v o p is the energy required to excite an electron from state A to state B. For 2+ 2+ complexes 53 - 57 v o p should vary proportionally with the difference in ground-state energies (A£°). Importantly, A£° is expected to be larger as the electron density at the ruthenium is decreased via electron-accepting ligands. A B C D E (a) (b) Figure 12. Potential energy diagrams for initial and final states for (a) states A and B (Scheme 13), and (b) states C, D and E (Scheme 14). It is not possible to calculate AE° exactly from the measured oxidation potentials since state B cannot be isolated; however, AE° can be estimated from the electrochemical data. The transition from state A to B involves oxidation of the Ru n and concomitant reduction of the F e m center; therefore, eq 13 can be derived from eqs 8 and 12. A plot of V m a x for the dications 63 vs £p,a(3) - £1/2(2) should be linear on the base of eq 13. This data is plotted in Figure 13a, and demonstrates an approximately linear correlation between v m a x and £ p, a(3) - £1/2(2). Vmax = Xi + Xo + AE'+D + (e/hc)[Ep,a(3)- Em(2)] (13) The model shown in Figure 12a may be extended to charge transfer in the monocations (Scheme 14). Here one must consider a three-state potential energy diagram (Figure 12b) in which v o p ' corresponds to the energy required to optically excite an electron 64 from state C to D. States C and E are isoenergetic, while state D has a higher ground-state energy. Scheme 14 Fe n -Ru n -Fe I U [Fe^Ri^-Fe 1 1]* C D* + F e m - R u n - F e n E c W [ F e F - R u r a - F e n f F ^ - R ^ - F e 1 1 1 Figure 14. Relative energy diagrams for states A - D. [Fe11—Ru111—Fe11] * 'op F e n - R u n - F e m The absorption maximum for the IVCT band in the monocations is lower by 1000 -2000 cm - 1 than the corresponding transition in the spectra of the corresponding dications. The relative magnitude of the absorption maxima in the spectra of the mono- and dications may be predicted by considering the relative energies of the states involved (Figure 14). Both state's C and D are lower in energy than states A and B respectively, because C and D carry less total charge than A and B. Furthermore, oxidation of a Fe n center adjacent to a R u m (D to B) requires more energy than oxidation of a Fe11 adjacent to a Ru 1 1 (C to A). Thus v o p ' is expected 65 to be lower in energy than v o p . A transition involving long-range electron transfer between the two iron centers is also possible, but is expected to be of higher energy. A larger intermetallic distance often results in higher IVCT energy. The IVCT bands have been observed at 1800 nm for the monocation [Fc-Fc]+, 1560 nm for [Fc-C=C-Fc] + and 1180 nm for 61+ in which the Fe - Fe distance is still significantly shorter than in 53+ - 57+.43 A near-IR absorption is also observed in mixed-valence diferrocenylpolyenes;45 however, the intermetallic distance in these complexes in solution is not exactly known due to the nonrigidity of the polyene linker. An unusual band at 820 nm is present in the visible spectrum of 53+, and at 780 nm in 54+. It is possible that these bands arise from the direct Fen-to-Fen I charge transfer. Since states G and E are isoenergetic, state D can relax thermally to either C or E. Optical excitation to form state D followed by electron transfer to form E thus provides a pathway for charge transfer across the ruthenium bisacetylide bridge. The dependence of the v m a x of an IVCT band on solvent has been used as a diagnostic test of a Class II species. According to Hush model (eqs 8 and 9), the energy (vm a x) of an IVCT transition may be expected to vary linearly with (IIn - l/Ds) for a Class II system. The spectra of the dications are all found to be solvent-dependent. A plot of V n ^ as a function of (1/n2 - \IDS) for 532+ is shown in Figure 15, along with the best-fit line obtained from linear regression (R = 0.925). This behavior is consistent with complex 532+ behaving as a Class II partially delocalized system. 66 7000 6800 U S 6600 o 6400 \-6200 h 6000 \-i j i , . i • - T(g) ( e ) i X • / t ( d ) -- (a) I i ( C ) _ 1 S -i 1 i . i . i 0.1 0.2 0.3 0.4 l /« - l/D. 0.5 0.6 Figure 15. Plot of (near-IR) vs l/n2- l /D, for 532+ with the best-fit line (R = 0.925). (a) trichloroethylene; (b) chlorobenzene; (c) o-dichlorobenzene; (d) CICH2CH2CI; (e) CH2CI2; (f) nitrobenzene; (g) CH3COCH3; (h) C H 3 N 0 2 ; (i) CH 3 CN. The energy and intensities of the near-IR absorptions of mixed-valence complexes may be used to calculate a delocalization coefficient a (eq 7). The values calculated for a are shown in Tables 8 and 9. The value used for the Ru - Fe intermetallic distance d is obtained from the structural data for 52 (d = 6.20 A) 53 (d = 6.30 A) 5 3 and 55 (d = 6.15 A). Complex 54 is assumed to have the same intermetallic distance as 53; 56 and 57 are assumed 67 to have the same d as 55. As shown in Figure 13b the calculated values of a 2 for the dications 532+- 572+ in CH 2C1 2 are linearly correlated with 4,a(3) - Em(2) (R = 0.983), indicating that the complexes with smaller difference in the ground-state energies (A£°) have more charge delocalization between the iron and ruthenium centers. The monocations 55+- 57+ in CH 2C1 2 have similar calculated values of a 2 , while the values obtained for 53+ and 54+ are larger. The monocation 52+ has two broad TVCT absorptions which have similar a 2 values, both being lower than those calculated for 53+ - 57+. Comparison of the a 2 values with literature results for other molecules containing the same metal centers is instructive. The a 2 value for [(n5-C 5H 5)(PPh 3) 2RuC--CFc]+ is comparable (2.8 x 10~ 2), 1 3 6 while a 2 for [(NH3)5RuN=CFc]+ is less (2.3 x 10" 3). 1 4 0 The smaller values of a 2 for 52+ and 522+ are obtained even though the extent of delocalization observed from the C V of 52 is comparable to those for 56 and 57. Complex 52 contains a coordinated Cu 1 which can interact with the delocalized system, and this may decrease the effective distance between the donor and acceptor wavefunctions. 2.4 Discussion The cyclic voltammetric and spectroscopic data for 52 - 57 and their oxidized derivatives may be used to compare the electronic delocalization within this series of complexes. Complex 52 demonstrates the effect of varying the geometry at the ruthenium on the electronic delocalization. For 52, A£[a is 0.14 V, smaller than that observed for 53 (0.22 V). There are two factors which may influence ts£\a here; the geometry around the ruthenium center and the presence of the coordinated Cul. In complex 53 the electronic interaction between the terminal redox groups occurs via the TC bonds of the cyclopentadienyl and alkynyl moieties and the d-orbitals of the ruthenium. When the acetylide ligands are trans, as in 53, 68 the same d-orbitals on the metal are involved in backbonding with both acetylide ligands (dxz -» n and d x y -» n). This should enhance the interaction between the ferrocenyl groups in 53 relative to 52, where the acetylide ligands are cis, and backbonding involves three different d-orbitals, only one of which is common to both acetylide ligands (d x y —> n, dxz —» TC* for one acetylide, d x y —> n*, dyz -> n* for the other). The coordinated Cul in 52 could act either to enhance the electronic interaction between the ferrocenyl groups by acting as another bridge between the two C=C bonds, or to reduce the interaction through the ruthenium center by reducing conjugation between the ferrocenyl group and the ruthenium. It is difficult to predict which effect is more significant without comparing the electrochemical behavior of the analog of 52 in which the Cul is absent; however, this is impossible due to the instability of that analog. In the related complex [(Ti5-C5H4SiMe3)2Ti(C=CSiMe3)2]CuOTf129-132 the coordinated CuOTf group lowers the energy of the C=C absorption in the IR relative to the free ligand. This has been interpreted as the result of both backbonding from the Cu 1 to a n* orbital on the ligand and electron-donation from the n bond to the Cu 1. These results suggest that electronic interaction via the coordinated Cu 1 in 52 is possible, but the magnitude of this effect relative to interaction through the Ru center is difficult to predict. The behavior of the series of trans bisacetylide complexes 53 - 57 may be directly compared as the basic structures of these complexes are identical. Cyclic voltammetry shows that AE\a decreases as 54 > 53 > 56 « 57 > 55. Considering the properties of the ancillary ligands on the ruthenium found in this series of complexes, pyridine is a a-donor, the phosphines are strong a-donors and weak Ti-acceptors while carbonyl is a strong 7i-acceptor and a weak a-donor. The results obtained from the CVs of this series demonstrate that the 69 electronic interaction between the ferrocenyl groups depends primarily on the number of carbonyl ligands. Hence, the more carbonyl ligands present around the ruthenium, the smaller A£i /2 . This can be interpreted as resulting primarily from the strongly n-acidic nature of the carbonyl groups, which serve to withdraw the electron density from the ruthenium and thus decrease the electron density available for conjugation with the acetylide bonds. Two electron-donating dmpe ligands on the ruthenium in 54 result in a 0.09 V increase in AE\a compared with 53, while one carbonyl group decreases AEm by almost the same amount. Varying the number and nature of the non-carbonyl ligands has a somewhat smaller effect on AEia compared to that exerted by the carbonyl group. In addition to the trend in AEm, the potential measured for the Ru 1 1 1 1 1 oxidation wave in the C V increases as the number of carbonyl ligands on the ruthenium is increased. As electron density is lost from the metal to the carbonyl ligands, the ruthenium becomes increasingly difficult to oxidize. The trends in electron delocalization observed by C V are consistent with those observed by spectroscopic methods. In the visible region, the magnitude of the redshift observed in the L M C T band for 53+ - 57+ relative to 61+ increases as the number of donor ligands on the ruthenium is increased. Similarly, in the IR region both the C=C and C=0 absorptions correlate to the electrochemical results. Based on the C V data, 53 or 54 has a higher electron density at the ruthenium relative to the monocarbonyl complexes 56 and 57 and dicarbonyl 55. This results in a lower v<>c for 53 or 54 than for 56 or 57 since more backbonding is possible in 53 or 54. Complex 56 also has a donor ligand trans to the carbonyl group which allows for more backbonding with the lone carbonyl than in 55. 70 0.05 0.10 0.15 0.20 0.25 0.30 0.35 AEV2 (V) Figure 16. Plot of AE1/2 = £ 1 / 2 (2) - El/2(1) vs the difference in voe between the neutral and dicationic complexes (Avoc) with a best-fit line to the data (R = 0.999). Interestingly, the wavenumber difference between the acetylide absorptions of the neutral and dicationic species correlates with the degree of electronic derealization observed in the CV. The difference in the sequential ferrocenyl-oxidation potentials AEm is correlated to the difference in the C=C absorptions between the neutral and dicationic complexes (Avoc) (Figure 16). Complex 54 has the largest AE\a, and the difference in the absorption is the largest (93 cm - 1) between 54 and 542+; while 55 has the smallest AEj/2, and the difference is 71 smallest (36 cm - 1). This is a result of how much of the electron density at the ruthenium is available for backbonding with the acetylide ligands. Ancillary ligands such as carbonyls, which withdraw electron density, allow less participation of the ruthenium d-electrons in electron transfer along the acetylide backbone. Fc +-=-=-Fc Fc-=-=-Fc + Fc + -H -= -Fc Fc-s-j-=-Fc+ RuL/4 R u L 4 (a) (b) Figure 17. Potential diagrams for electron derealization in (a) 61 + and in a hypothetical molecule in which a RuL 4 is inserted into the central C - C bond in 61+. The charge derealization in these complexes is best explained by the potential-energy diagram shown in Figure 12b. The nature of the ancillary ligands on the ruthenium center affects the difference in ground-state energy (AE°). Donor ligands decrease AE°, thus facilitating charge transfer across the ruthenium bisacetylide bridge, while acceptor ligands increase AE°. Complexes 53 - 57 have similar structures, and comparison of the AEm within the series indicates a significant contribution from charge derealization to the magnitude of Mi /2 . Changes in the ligands on the Ru (II) result in increases in AE\a (0.13 V) from 0.09 V for 55 to 0.22 V for 53, and (0.22 V) to 0.31 V for 54. It is reasonable to conclude that this 72 increase in AEm is mainly due to the contribution from charge delocalization. For 53 and 54 this contribution exceeds the total stabilization energy in complex 61, which has AE\n = 0.10 V, thus there is less charge delocalization between the two ferrocenyl groups in complex 61 + than in 53 + and 54+. This result may be understood by considering the potential-energy diagram for 61 + (Figure 17a). Charge transfer between the two states occurs either in the ground states by overcoming the activation barrier (£„), or in the excited state via IVCT. Figure 17b shows the potential-energy diagram for a hypothetical molecule consisting of 61 + with a ruthenium center between the two ferrocenyl groups. Here the ruthenium facilitates electron delocalization by reducing the energy barrier in the ground state (E& )• The potential-energy diagrams for the mono- and dications of complexes 52 - 57 (Figure 12a and b) are slightly different since the distance between the two iron centers is greater than that shown in Figure 17b; however, the same principle is expected to hold. It is clearly demonstrated that the extent to which the energy barrier is lowered is influenced by the ancillary ligands on the ruthenium; therefore, the extent of delocalization is affected by these ligands. 2.5 Conclusions A series of ruthenium(II) bisferrocenylacetylide complexes are synthesized by coupling ruthenium halide complexes with FcC=CSn(«-Bu)3 in the presence of copper (I) halide. The amount of copper (I) halide used significantly affects the product formed. When a stoichiometric amount of Cul is used [c/s-Ru(dppm)2(C--CFc)2]CuI is obtained, while trans-. Ru(dppm)2(C=CFc)2 is obtained with catalytic Cul. Removal of coordinated Cul from [cis-Ru(dppm)2(C=CFc)2]CuI yields frms-Ru(dppm)2(C--CFc)2, while reaction of trans-73 Ru(dppm)2(C=CFc)2 with a stoichiometric amount of Cul yields [cis-Ru(dppm)2(C=CFc)2]CuI. The stability of the complexes in different oxidation states allows for studies of charge transfer between different metal centers in these complexes by electrochemical and spectroscopic characterization. Charge derealization between the iron and ruthenium centers in the oxidized species is observed, and the extent of derealization is evaluated. The ruthenium bisacetylide bridges facilitate electronic interactions between the terminal ferrocenyl groups. The interaction is enhanced when the ancillary ligands on the ruthenium center are electron donors and lessened when the ligands are acceptors. c 74 Chapter 3 Models for Conjugated Metal Acetylide Polymers: Ruthenium Oligothienylacetylide Complexes 3.1 Introduction The work described in Chapter 2 shows that the bridge in ruthenium bis(ferrocenylacetylide) complexes allows for derealization of charge between the two ferrocenyl groups, and that the derealization varies with the nature of the ancillary ligands siirrounding the ruthenium. These results suggest that linking conjugated organic groups with ruthenium bisacetylide bridges could result in extended conjugation involving both the metal center and the organic bridges. Lewis and coworkers have synthesized ruthenium bisacetylide polymers where the metal centers are linked by phenyl bridges; 7 1 ' 8 0 however, no electronic properties have been reported to date for these materials. Although these phenyl group-bridged polymers are readily accessible synthetically it may be necessary to link the metal centers in these polymers with organic fragments which are more electron-rich and have longer conjugation lengths in order to enhance derealization of charge. It has been demonstrated that block copolymers containing short oligothiophenes (> 4 thiophene units) are conductive. 1 0 ' 1 4 1 This suggests that ruthenium acetylide polymers containing oligothienyl bridges may be good" candidates for conducting materials. Toward this end, one approach is the preparation of electropolymerizable monomers consisting of oligothiophenes bridged by ruthenium bisacetylide groups (Figure 8). In this Chapter the preparation and characterization of the series of ruthenium mono(oligothienylacetylide) complexes 62a - c and bis(oligothienylacetylide) complexes 63a - c are described. The 75 interaction between the metal and organic fragments is probed. Electropolymerization of several of these complexes are also carried out. Ph 2 P^PPh 2 Cl-Ru—=====—R Ph 2 p(^PPh 2 62a- c R : Ph 2P x PPh 2 R = Ru = R P h 2 p ^ p p h 2 63a - c 3.2 Experimental General. Procedures were as described in Chapter 2 except as noted here. Diethyl ether and THF were dried by refluxing over sodium/benzophenone. Diisopropylamine was purified by distillation from K O H and was stored over molecular sieves (4 A). Trans-Pd(PPh 3 ) 2 Cl 2 , 1 4 2 , Ni(dppp)Cl 2, 1 4 3 lKtrimethylsilyl)-2-(2-thienyl)acetylene (64a),144 (2-thienyl)acetylene (65a),144 2,2'-bithiophene,144 5-bromo-2,2'-bithiophene,145 2,2':5',2"-terthiophene146 and 5-iodo-2,2':5',2"-terthiophene146 were all prepared using literature procedures. l-(Trime%lsilyl)-2-(5-(2,2'-bithienyl))acetylene (64b) and (5-(2,2'-bithienyl))acetylene (65b) were prepared by modification of the procedures for the preparation of 64a and 65a respectively.144 l-(Tributylstannyl)-2-(2-thienyl)acetylene (67a) and 1-(tributylstannyl)-2-(5-(2,2'-bithienyl))acetylene (67b) were prepared using the procedure for 1-(tributylstannyl)-2-(5-(2,2':5\2"-tertrnenyl))acetylene (67c), and used without further purification. The concentration of the complexes for electrochemical measurements was in the range 1.6 x 10 - 2.2 x 10 M unless otherwise noted. The visible and near-IR spectra of 76 62a+ - c+ and 63a+ - c+ in dry CH2CI2 solutions were collected on a Varian Cary 5 spectrometer, in which the sample cell was held at -17 °C under a nitrogen atmosphere. l-(Trimethylsilyl)-2-(5-(2,2,:5',2"-terthienyl))acetylene (64c). A suspension of 5-iodo-2,2':5',2"-terthiophene (0.64 g, 1.7 mmol), diisopropylamine (0.33 g, 3.3 mmol), trans-Pd(PPh3)2Cl2 (0.14 g, 0.20 mmol) and Cul (0.022 g, 0.12 mmol) in dry THF (60 mL) was degassed for 2 min with N2. Trimethylsilylacetylene (0.33 g, 3.4 mmol) was added to the suspension via syringe. After the solution was stirred overnight at room temperature the solution turned dark green. The reaction was quenched by adding distilled water (20 mL) to the solution. The organic layer was collected, and the aqueous layer was extracted with CH2CI2 (2 x 20 mL). The organic extracts were combined and washed with brine (30 mL) and distilled water, and then dried over anhydrous MgS04. The solvent was removed to obtain a brown solid, which was purified by flash chromatography (silica gel), using hexanes as eluant. Yield: 0.52 g (88%). ! H N M R (400 MHz, CDC13): 6 7.21 (dd, J = 5.1, 1.0 Hz, IH), 7.16 (dd, J = 3.6,1.0 Hz, IH), 7.12 (d, J = 3.8 Hz, IH), 7.05 (s, 2H), 7.00 (dd, J = 5.1, 3.6 Hz, IH), 6.97 (d, J = 3.8 Hz, IH), 0.30 (s, 9H). 1 3C{'H} N M R (75.429 MHz, CDCI3): 8 138.48, 136.85, 136.78, 135.22, 133.48, 127.86, 124.78, 124.66, 124.29, 123.84, 123.08, 121.72, 100.18,97.39, -0.17. Anal. Calcd for C i 7 Hi 6 S 3Si: C 59.25, H 4.68. Found: C 59.18, H 4.70. (5-(2,2':5',2"-Terthienyl))acetyIene (65c). To a solution of 64c (0.44 g, 1.3 mmol) in a mixture of CHCI3 and MeOH (60 mL, 1:3 v/v), was added anhydrous K2CO3 (0.36 g, 2.6 mmol). The suspension was stirred at room temperature for 2 h, after which time the solvent was removed, and the solid was extracted with CHCI3 (20 mL). The CHCI3 solution was then washed with distilled water (2 x 30 mL), and dried over anhydrous MgS04. The product was obtained as a yellow solid by removing the solvent and then purified by flash chromatography 77 (silica gel) using hexanes as eluant. Yield: 0.27 g (78%). ] H N M R (400 MHz, CDC13): 8 7.22 (dd, J = 5.1,1.1 Hz, IH), 7.15 - 7.17 (m, 2H), 7.05-7.07 (m, 2H), 6.99 - 7.02 (m, 2H), 3.40 (s, IH). 1 3C{'H} N M R (75.429 MHz, CDCI3): 8 138.85, 137.06, 136.79, 135.09, 133.98, 127.93, 124.99, 124.79, 124.35, 123.96, 123.09, 120.56, 82.56, 82.38. Anal. Calcd for Ci4H 8 S 3 : C 61.73, H 2.96. Found: C 62.04, H 2.96. Diisopropylaminotributylstannane. This compound was prepared by modification of a literature procedure.147 n-BuLi (1.6 M in hexanes; 31.7 mL, 50.7 mmol) was diluted with dry diethyl ether (30 mL), and cooled to -78 °C. Diisopropylamine (5.19 g, 51.3 mmol) was cooled to -78 °C, and was added dropwise via cannula while the rc-BuLi solution was being stirred. Upon completion of the addition, the mixture was allowed to warm to room temperature. A solution of tributyltin chloride (10.3 mL, 38.0 mmol) in dry diethyl ether (15 mL) was added to the stirred reaction mixture, causing an immediate color change to milky white. After being heated at reflux for 4 h, the solution was stirred at room temperature for 12 h, after which time the mixture became a pale yellow cloudy suspension. The lithium chloride was removed by suction filtration through Celite under N2. Removal of the solvent from the filtrate afforded an orange oil, which was distilled in vacuo to obtain a moisture-sensitive colorless oil (bp = 152 °C at 0.15 mmHg). Yield: 9.01 g (61%). *H N M R (200.132 MHz, CDCI3): 8 2.83 (septet, J = 6.3 Hz, 2H), 1.51-1.64 (m, 6H), 1.16 - 1.39 (m, 12H), 0.97 (d, J = 6.3 Hz, 12 H), 0.86 (t, J = 7.3 Hz, 9H). Anal. Calcd for CigftnNSn: C 55.40, H 10.59, N 3.59. Found: C 55.38, H 10.68, N 3.90. l-(TributyIstannyl)-2-(5-(2,2':5',2"-terthienyl))acetyIene (67c). This compound was prepared by modification of a published procedure.148 Diisopropylaminotributylstannane (0.72 g, 1.8 mmol) was added to a flask charged with 65c (0.50g, 1.8 mmol). Dry THF (20 78 mL) was added to the mixture, and the solution was stirred overnight in the absence of light at room temperature. The THF was removed, and the residual oil was held in vacuo at room temperature overnight to obtain a brown oil. Yield: 1.02 g (99%). ' H N M R (400 MHz, CDC13): 8 7.20 (dd, J = 5.1, 1.0 Hz, 1H), 7.15 (dd, J = 3.6, 1.0 Hz, 1H), 7.06 (d, J = 3.8 Hz, 1H), 7.04 (s, 2H), 7.00 (dd, J = 5.1, 3.6 Hz, 1H), 6.97 (d, J = 3.8 Hz, 1H), 1.57 - 1.64 (m, 6H), 1.32-1.41 (m, 6H), 1.07 (t, J = 8.1 Hz, 6H), 0.93 (t, J = 7.3 Hz, 9H). Anal. Calcd for C26H34S3Sn: C 55.62, H 6.10. Found: C 55.84, H 6.29. frans-{Ru=C=CHR(dppm)2(Cl)}[PF6] (R = 2-thienyI) (66a). To a solution of 59 (0.60 g, 0.64 mmol) and NaPF 6 (0.21 g, 1.3 mmol) in CH 2C1 2 (75 mL) was added 65a (0.15 g, 1.4 mmol). After being stirred at room temperature for 18 h, the red-brown solution was filtered through a filter-paper-tipped cannula. The solvent was removed in vacuo, and the residue was rinsed with diethyl ether to obtain a rust-colored solid. The solid was dried in vacuo at 90 °C for 6 days. Yield: 0.68 g (92%). "H N M R (400 MHz, CDC13): 8 7.15 - 7.45 (m, 40H), 6.87 (dd, J = 5.2, 0.9 Hz, 1H), 6.51 (dd, J = 5.2, 3.6 Hz, 1H), 5.30-5.40 (m, 2H), 5.01 - 5.10 (m, 3H), 3.22 - 3.25 (m, 1H). 3 IP{-H} N M R (81.015 MHz, CDC13): 8 -18.1 (s). Anal. Calcd for Cs^gClFePjRuS: C 58.06, H 4.18. Found: C 57.70, H 4.06. *raMs-Ru(dppm)2(Cl)(C--CR) (R = 2-thienyl) (62a). To a solution of 66a (0.62 g, 0.54 mmol) in CH 2 C1 2 (40 mL) was added l,8-diazabicyclo[5.4.0]undec-7-ene (DBU) (80 pL, 0.54 mmol) via syringe. The red-brown solution changed color quickly to yellow. After being stirred at room temperature for 2 h, the reaction mixture was filtered through a filter-paper-tipped cannula, and the solvent was removed in vacuo. The resulting dark yellow solid was taken up in a minimum of THF and transferred via cannula to a Schlenk filter charged with neutral alumina (Brockman, Activity I). The product was eluted with diethyl ether. 79 Removal of the solvent at reduced pressure, followed by rinsing with hexanes, yielded a yellow solid, which was recrystallized from layered chloroforrn/hexanes. The crystals were crushed and dried in vacuo at 90 °C for 6 days. Yield: 0.36 g (66%). ' H N M R (400 MHz, CDCI3): 6 7.38 - 7.50 (m, 16H), 7.20-7.30 (m, 8H), 7.15 (t, J - 7.6 Hz, 8H), 7.09 (t, J = 7.6 Hz, 8H), 6.57 - 6.60 (m, 2H), 5.68 (dd, J = 2.6, 1.9 Hz, IH), 4.88 (quintet, J = 4.2 Hz, 4H). 3 1P{'H} N M R (81.015 MHz, CDCI3): 8 -8.7 (s). Anal. Calcd. for C56H47CIP4RUS: C 66.43, H 4.68. Found: C 66.33, H 4.61. *ra/ts-Ru(dppm)2(Cl)(CsCR) (R = 5-(2,2'-bithienyl)) (62b). To a solution of 59 (0.31 g, 0.33 mmol) and NaPF 6 (0.16 g, 0.95 mmol) in CH 2C1 2 (30 mL) was added 65b (0.17 g, 0.89 mmol). The solution was stirred at room temperature for 7 h, turning dark-brown. The solution was then filtered through a filter-paper-tipped cannula, and a brown solid was obtained by removing the solvent. The solid was dissolved in CH2CI2, and filtered through a short Al203 column (Basic, Brockman Activity I) to give an orange solution. The volume of the solution was reduced to approximately 1 mL, and hexanes (40 mL) were added to the solution to induce the precipitation of an orange-yellow solid. The solid was dried overnight in vacuo at room temperature. Yield: 0.16 g (44%). *H N M R (400 MHz, CD 2C1 2): 8 7.50 -7.56 (m, 8H), 7.41 - 7.47 (m, 8H), 7.34 (t, J = 7.4 Hz, 4H), 7.30 (t, J = 7.4 Hz, 4H), 7.22 (t, J = 7.6 Hz, 8H), 7.16 (t, J = 7.6 Hz, 8H), 7.10 - 7.12 (m, IH), 6.96 - 6.98 (m, 2H), 6.70 (d, J = 3.4 Hz, IH), 5.54 (d, J = 3.4 Hz, IH), 4.86 - 4.99 (m, 4H). 3 1P{'H} N M R (81.015 MHz, CDC13): 8 -8.9 (s). Anal. Calcd for C6oH49ClP4RuS2: C 65.84, H 4.51. Found: C 65.92, H 4.52. *raws-Ru(dppm)2(Cl)(C>CR) (R = 5-(2,2':5',2"-terthienyl)) (62c). This complex was prepared as described for 62b. Yield: 53%. *H N M R (400 MHz, CD 2C1 2): 8 7.50 - 7.56 (m, 8H), 7.41 - 7.48 (m, 8H), 7.12 - 7.38 (m, 26H), 7.03 - 7.06 (m, 2H), 6.87 (broad, IH), 80 6.72 (d, J = 3.7 Hz, 1H), 5.54 (broad, 1H), 4.86 - 4.98 (m, 4H). 3 1P{'H} N M R (81.015 MHz, CDC13): 8 -8.9 (s). Anal. Calcd. for QaHsiCU^Rt^: C 65.33, H 4.37. Found: C 65.49, H 4.42. fraHs-Ru(dppm)2(C--CR)2 (R = 2-thienyl) (63a). To a deaerated solution of 59 (0.43 g, 0.46 mmol) and Cul (7 mg, 0.04 mmol) in chlorobenzene (25 mL) was added 67a (0.73 g, 1.8 mmol). The solution was heated at reflux overnight, and then cooled to room temperature. Chlorobenzene was removed, and the residual solid was dissolved in CH2CI2. The CH2CI2 solution was filtered through Celite to remove Cul. The filtrate was concentrated to approximately 10 mL, and hexanes were added. The solution was then cooled to -4 °C overnight, resulting in the precipitation of 63a. Yield: 0.30 g (60%). ! H N M R (400 MHz, CD2CI2): 8 7.47 - 7.55 (m, 16H), 7.28 (t, J = 7.4 Hz, 8H), 7.15 (t, J = 7.6 Hz, 16H), 6.63 -6.66 (m, 4H), 5.87 (t, J.= 2.3 Hz, 2H), 4.85 (quintet, J = 4.2 Hz, 4H). 3 1P{'H} N M R (81.015 MHz, CD2CI2): 8 -4.4 (s). Anal. Calcd for C62H50P3RUS2: C 68.69, H 4.65. Found: C 68.43, H4.53. frans-Ru(dppm)2(C--CR)2 (R = 5-(2,2'-bithienyl)) (63b). This complex was prepared as described for 63a. Yield: 80%. ' H N M R (400 MHz, CDC13): 8 7.42 - 7.48 (m, 16H), 7.26 (t, J = 7.4 Hz, 8H), 7.15 (t, J = 7.6 Hz, 16H), 7.07 (dd, J = 5.0, 1.2 Hz, 2H), 6.93 -6.99 (m, 4H), 6.74 (d, J = 3.7 Hz, 2H), 5.69 (d, J = 3.7 Hz, 2H), 4.81 (quintet, J = 4.1 Hz, 4H). 3 1P{'H} N M R (81.015 MHz, CDC13): 8 -6.3 (s). Anal. Calcd for C7oH54P4RuS4: C 67.35, H 4.36. Found: C 67.01, H 4.47. i>fl«s-Ru(dppm)2(C--CR)2 (R = 5-(2,2':5'2"-terthienyl)) (63c). The complex was prepared as described for 63a. Yield: 84%. ] H N M R (400 MHz, CDC13): 8 7.41 - 7.47 (m, 16H), 7.27 (t, J = 7.4 Hz, 8H), 7.11 - 7.19 (m, 20H), 7.02 (d, J = 3.6 Hz, 2H), 7.00 (dd, J = 81 5.1, 3.6,2H), 6.87 (d, J = 3.7 Hz, 2H), 6.75 (d, J = 3.7 Hz, 2H), 5.69 (d, J = 3.7 Hz, 2H), 4.82 (quintet, J = 4.0 Hz, 4H). 3 1P{'H} N M R (81.015 MHz, CDC13): 8 -6.3 (s). Anal. Calcd for C78H58P4RuS6: C 66.32, H 4.14. Found: C 65.98, H 4.04. Crystallographic Study. Data collection and structure determination were performed by the late Dr. Steven Rettig in this department. An orange irregular crystal of dimensions 0.40 x 0.35 x 0.20 mm was used. Data were obtained on a Rigaku/ADSC CCD area detector with graphite monochromated Mo-Ka radiation (A, = 0.71069 A). The data were collected at a temperature of-93° C. A total of 15312 independent reflections were measured (29 < 61°), of which 8814 had I > 3CT(I) and were considered to be observed. The data were corrected for Lorentz and polarization factors. Crystal data for 63c: C78H58P4RUS6, M r = 1412.64, triclinic, space group PI (#2), a = 13.1663(9) A, b = 15.5443(8) A, c = 17.757(2) A, a= 70.834(3)°, fi = 83.6388(10)°, / = 75.9874(6)°, V= 3328.7(4) A 3 , Z = 2,D (calc) = 1.409 g cm - 3 , ft (Mo Ka) = 5.65 cm - 1 , £(000) = 1452. The structure was solved by direct methods and expanded using Fourier techniques. Both terminal thiophene moieties are disordered with respect to a 180° rotation about the bond to the adjacent ring. For the ring containing S(3) and the minor component S(3a) the disorder was modeled by split-atom refinement with bond length constraints. A similar treatment of the more nearly ordered ring containing S(6) was not successful. As a result of the disorder, both resolved and unresolved, the geometry of the terminal thiophene rings is subject to errors larger than expected from the least-squares standard deviations. A l l other non-hydrogen atoms were refined anisotropically. Hydrogen atoms were fixed in idealized positions with C - H = 0.98 A and thermal parameters 1.2 times those of the parent atoms. The final cycle of full-matrix least-squares refinement was based on the observed data and 809 variable 82 parameters and converged with R\ = 0.037 and wR2 = 0.079. The maximum and minimum peaks on the final difference Fourier map were corresponded to 1.92 and -1.28 e A - 3 respectively. A l l calculations were performed using the teXsan crystallographic software package of Molecular Structure Corporation. 3.3 Results and Interpretation 3.3.1 Syntheses and Structure Complexes 62a - c are synthesized by coupling cw-Ruf^dppm^Cfe (59) with oligothienylacetylenes 65a - c in the presence of excess NaPFg, yielding ruthenium vinylidene complexes 66a - c (Scheme 15). Complexes 66a - c are converted to 62a - c either by passing through a basic AI2O3 column or by reaction with l,8-diazabicyclo[5.4.0.]undec-7-ene (DBU). Scheme 15 f ^ P h 2 P h o P ^ P P h , - | + P F 6 " P h P ^ P P h o Ph P r i 2 r N / A 1 1 2 T T •• r n 2 * \ / r n 2 >^<S + R - = - H C I - R U = C = C : " C l - R u ' - ^ - R Ph 2P I C l n , ' R / \ 2 U P P h 2 P V V P P h 2 P h 2 P v P P h 2 59 65a - c 66a - c 62a - c i : NaPF 6 , CH 2 C1 2 ; i i : DBU/neutral A1 2 0 3 or basic A 1 2 0 3 Scheme 16 PPPh 2 P h 2 P ^ P P h 2 P h 2 P - R u - C 1 + R - ^ - S n ( « - B u ) 3 C u I > R - ^ - W - S - R Ph 2 P^ I " C l C 6 H 5 C1 P h p / N p p h U P P h 2 A x j n 2 r ' v J J J J h 2 59 67a - c 63a - c 83 Bisacetylide complexes 63a - c are prepared via the coupling of 59 with the corresponding l-(tributylstannyl)-2-(oligothienyl)acetylenes 67a - c in the presence of catalytic Cul (Scheme 16). The reactions are carried out in the absence of light in chlorobenzene heated at reflux. Complexes 62a - c and 63a - c are fully characterized using 1 31 H and P NMR, IR and UV-vis spectroscopies, and the elemental analyses are all in the expected range. In solution, all the complexes decompose slowly when exposed to ambient light. The ' H N M R spectrum of 62c shows broadened peaks assigned to the protons close to> the ruthenium, and a single broadened peak is observed in the 3 1 P N M R spectrum. The broadened peaks sharpen when diisopropylamine is added to the N M R solution, and no new resonances are observed in the 3 1 P N M R spectrum, consistent with the broadening being a result of partial oxidation to the paramagnetic Rum species, and in situ reduction with diisopropylamine. It is possible that 62c is oxidized slowly in solution when the solution is exposed to air, or that a trace amount of the R u m species is formed on the basic alumina in the last step of the synthesis. This is the only compound in the series for which peak broadening is consistently observed in the N M R spectra. Complex 63c is crystallized from layered CtkCh/hexanes at 4°C to obtain orange irregular crystals whose structure is determined by single-crystal X-ray diffraction (Figure 18). Selected bond lengths and angles are listed in Table 10. The ruthenium center is in a distorted octahedral environment with the two terthienylacetylide ligands in a trans orientation around the Ru. The dihedral angles between the six thiophene rings in the structure are of interest since the extent of conjugation depends on the coplanarity of the rings. The thiophene rings are labeled 1 - 6 in Figure 18. The dihedral angle between the two innermost thiophene rings 84 (1 and 4) is 47.2°. The dihedral angles between rings 1 - 2 and 4 - 5 are small, 9.3° and 2.6° respectively, indicating that these rings are held nearly coplanar in the solid-state molecular structure. Both terminal thiophene rings (3 and 6) are disordered with respect to 180° rotation about the bond to the adjacent ring and have dihedral angles with the adjacent rings of 17.8° (2 -3 ) and 27.6° (5-6) . Table 10. Selected Bond Lengths (A) and Angles (deg) for 63c Ru(l) - P ( l ) 2.3306(8) Ru(l) - C ( l ) 2.059(3) Ru(l) -P(2) 2.3298(7) Ru(l) -C(15) 2.067(3) Ru(l) -P(3) 2.3243(8) C ( l ) - C(2) 1.195(4) Ru(l) -P(4) 2.3340(7) C(15) -C(16) 1.200(4) P ( l ) - Ru(l) -P(2) 72.65(3) P(3)- Ru( l ) -C( l ) 92.09(8) P ( l ) - Ru(l) -P(3) 178.63(3) P(3)- Ru(l)-C(15) 90.92(9) P ( l ) - Ru(l) -P(4) 108.70(3) P(4)- Ru( l ) -C( l ) 92.20(7) P(2)- Ru(l) -P(3) 106.12(3) P(4)- Ru(l)-C(15) 88.38(7) P(2)- Ru(l) -P(4) 178.42(3) C ( l ) - Ru(l)-C(15) 176.97(12) P(3)- Ru(l) -P(4) 72.52(3) Ru(l) - C ( l ) - C ( 2 ) 176.4(3) P ( l ) - Ru(l) - C ( l ) 87.26(8) Ru(l) -C(15)-C(16) 175.5(3) P ( l ) - Ru(l) -C(15) 89.74(9) C ( l ) - C(2)-C(3) 177.2(3) P(2)- Ru(l) - C ( l ) 87.03(7) C(15) -C(16)-C(17) 174.8(4) P(2)- Ru(l) -C(15) 92.46(7) 85 C 2 7 C 1 3 3 Figure 18. Solid-state molecular structure of 63c. 3.3.2 Electrochemistry Complexes 62a - c all have two oxidation waves in their cyclic voltammograms (Figure 19) in the range 0.0 to 1.6 V vs SCE. Wave A is assigned to a Ru11™ redox process since it appears at similar potentials in all three complexes (Table 11). Related complexes /ra/i5-Ru(dppm)2(Cl)(C=CR') (R' = Ph and 4-nitrophenyl) have oxidation waves due to the Ru1 1 7 1 1 1 process with £ p , a = 0.41 V vs SCE for R' = Ph and 0.59 V vs SCE for R' = 4-nitrophenyl.149 Inspection of the first oxidation waves (A) for 62a - c reveals features diagnostic of chemical irreversibility. For all these complexes plots of iPA to vm (v = scan rate) 86 deviate from linearity, and the ratio iPJC/ip fiis less than 1 (Table 11), consistent with some of the oxidized species undergoing following chemical reactions.150 These results suggest that isolation of the R u m complexes may be difficult, due to the instability of the oxidized species in solution. The ip>Q/ip,a ratio for 62c is closest to 1.0, so the oxidized form of this complex is most likely to be sufficiently stable to isolate. Attempts to isolate the R u m complex by oxidation of 62c using FcPF6 obtain a deeply colored gray-green solid; however, this complex is found to be impure. The IR spectrum of this solid contains an unassigned band at 1920 cm"1 in addition to the expected peak at 1976 c m - 1 . 1 3 6 In addition, electrochemical reduction of the oxidized 62c yields other products in addition to 62c, as indicated by new resonances in the 3 1 P N M R spectrum. Although all attempts to isolate the Rum complexes as pure solids are unsuccessful, these species can be spectroscopically characterized in situ (vide infra). The second oxidation wave (B) in the C V of 62a is clearly chemically irreversible. For complex 62b, wave B becomes more reversible only at high scan rates (> 100 mV/s), but is irreversible at slower scan rates. Wave B appears most chemically reversible for 62c. The oxidation potential of wave B decreases significantly as the length of the oligothienyl ligand increases (Table 11). Based on this behavior, the second oxidation wave (B) is assigned to a thiophene-based oxidation. Electrochemical studies on oligothiophenes have shown that longer oligomers have lower oxidation potentials.21'22 The potential of wave B in 62c is lower and more reversible than the first oxidation wave in the C V of the organic oligomer 64c (Ep;a = 1.07 V vs SCE). At the potential of wave B the ruthenium in 62c is already oxidized; yet, is still able to stabilize the ligand-based oxidation relative to the trimethylsilyl derivative 64c. The CVs of complexes 63a - c all contain multiple waves. Complexes 63a and 63b both show four distinct waves in the range 0 - 1.4 V vs SCE (Figure 20a and b). The lowest 87 potential waves (C for 63a; G for 63b) are assigned to a R u I I / m redox couple. Wave G is more reversible than wave C as shown by /PjC/zP,a measurements, and wave C becomes more reversible at higher scan rates. The three irreversible waves (D, E and F for 63a; H , I and J for 63b) observed at higher potentials are assigned to ligand-based oxidations, as well as oxidation of products resulting from decomposition during the electrochemical experiment. It is very clear that the first ones of these multiple waves (D and H) are ligand-based and are likely due to the same process which gives rise to wave B in the CVs of 62a - c. For complex 63b repeated scans over the range 0 - 1.4 V result in an increase in the current upon each subsequent scan (Figure 21a). This observation is consistent with deposition of conducting material on the electrode. Inspection of the electrode surface after multiple scans reveals the presence of an insoluble red film. During growth of the film, wave G decreases in intensity and eventually disappears, presumably due to the film blocking further monomer from reaching the electrode surface. This indicates that the film is insulating at 0.3 V vs SCE since a conducting film would allow oxidation of monomer in solution at this potential even if the monomer cannot penetrate to the electrode surface. Bisacetylide 63c shows four waves in its CV, all of which appear relatively reversible (Figure 20c). The Ru11™ wave (K) appears very close to where it is observed for 63a and 63b. The three waves (L, M and N) at higher potentials are due to ligand-based oxidation processes, by analogy with the results observed for 63a and 63b. When the concentration of the sample is increased, the shapes of the waves in the C V change as a result of deposition of conductive material on the electrode during the scan (Figure 21c, first scan). Multiple scans show clear evidence for this growth (Figure 21c), analogous to the observations for 63b. Extended scanning of solutions containing lower concentrations of 63c (3.5 x 10"4 M , as in 88 Figure 20c) also results in deposition of conducting material, albeit more slowly. Since the stability of the oxidized species formed under these conditions is in doubt, the exact nature of the conducting material formed in these experiments is unclear. It is possible that the material contains coordinated metal centers in addition to dimerized or polymerized oligothienyl ligands, or alternatively that decomposition of the complexes via ligand loss results in polythiophene-like conducting polymer films. No films form when 62a - c and 63a are scanned repeatedly under the same conditions. -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Voltage (V vs SCE) Figure 19. Cyclic voltammograms of (a) 62a (2.2 x 10"3 M), (b) 62b (1.6 x 10~3 M) and (c) 62c (1.8 x 10"3 M) in CH 2C1 2 containing 0.1 M [(«-Bu)4N]PF6. Scan rate = 100 mV/s; 89 Figure 20. Cyclic voltarnmograrns of (a) 63a, (b) 63b (4.0 x io~* M) and (c) 63c (3.5 x io~* M) in CH 2 C1 2 containing 0.1 M [(n-Bu)4N]PF6. The scan rate = 100 mV/s. The dotted lines show the cyclic voltammograms in the range of -0.2 - 0.6 V vs SCE. 90 l 1 i 1 i i • i • i • i • i 1 r _ i i i i i i i i i t . i . i . t -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Voltage (V vs SCE) Figure 21. Multiple scan cyclic voltammograms of (a) 63b (4.0 * 10"* M) and (b) 63c (1.8 x 10"3 M) in CH 2 C1 2 containing 0.1 M [(«-Bu)4N]PF6 Scan rate = 100 mV/s. 91 Table 11. Spectroscopic and Electrochemical Data for 62a - c Complex UV-vis-near-IR Mnm) (s ( I v T W ) ) 0 IR (KBr) vr>c (cm - 1) (Ruu/mrb £ P ,a(2f* 62a 328 (14000) 2063 0.36 0.50 1.25 62b 406 (22000) 2056 0.32 0.89 0.99 62c 334 (sh) (11000), 450 (33000) 2053 0.30 0.94 0.84 62c+ 375 (13000), 580 (29000), 665 (sh) (17000), 1090 (29000) 63a 338 (25000) 2050 0.33 0.59 0.98 63b 420 (53000) 2050 0.30 0.95 0.85 63c 340 (sh) (24000), 460 (75000) 2047 0.30 0.95 0.76 63c+ 400 (45000), 595 (28000), 1330 (23000), 1610 (sh) (16000) 64c 384 (34000) 2139 1.07^  a CH 2C1 2 , 20 °C except for 62c+ and 63c+ which were obtained at -17 °C. b V vs SCE, Pt working electrode, 20 °C. c R u n / m wave; scan rate =100 mV/s. d first oxidation wave. 3.3.3 Spectroscopic Characterization The frequencies of the infrared absorption bands for the C=C groups in the complexes are collected in Table 11. The absorptions appear at similar energies for the whole series, and are significantly lower in energy than the corresponding absorption for l-(trimethylsilyl)-2-(5-(2,2':5',2"-terthienyl))acetylene (64c) demonstrating the extent of backbonding of the Ru 1 1 center with the acetylide. 92 50000 40000 h '20000 10000 1 1 1 1 1 i 1 i 1 i i 1 \ \\ \\ 62c \ -ti \i \ \ \ \ \ \ \\ 62b . -\ \ 62a • \ \ 7 \ \ \ _ ^ ' : \ ' •' \ > : \ \ ^ _ i . i \ x i — ^ i i " •" ~ -> 0 250 300 350 400 450 500 Wavelength (nm) 550 600 Figure 22. UV-vis spectra of 62a (—), 62b (—) and 62c (•••) in CH 2 C1 2 The electronic spectra of 62a - c and 63a - c are shown in Figures 22 and 23 respectively. A l l the complexes exhibit strong ligand-based (dppm) absorption bands above 270 nm. Intense absorption bands with Xmax in the range 328 - 460 nm are assigned to thiophene-based rc - TI* transitions. These bands are absent in the spectra of analogous complexes which do not contain oligothienyl ligands, such as trans-Ru(dppm)2(Cl)(C--CFf),151 and the extinction coefficients are in the range expected for such transitions.152 These absorption bands are approximately twice as intense in the spectra of the diacetylides 63a - c as the corresponding bands for the monoacetylides 62a - c, and the 93 energy of the transition shifts to lower energy as the length of the oligothienyl ligand is increased. The energy of the n - n* absorption is 450 nm for 62c and 460 nm for 63c, while the absorption maximum for compound 64c appears at 384 nm and 355 nm for 2,2':5',2"-terthiophene,14 mdicating that the electron-donating Ru 1 1 reduces the n - n* energy. The differences in Xmax are small between the monoacetylide and diacetylide containing the same ligand, from which it is concluded that the n - n* transitions in the diacetylides are largely localized on each ligand rather than delocalized over the whole complex. 100000 80000 63c ^ 60000 K ~ 40000 h \ 20000 0 250 300 350 400 450 500 550 600 Wavelength (nm) Figure 23. UV-vis spectra of 63a (—), 63b (—) and 63c (•••) in CH 2C1 2 . 94 50000 I 1 1 1 1 1 1 1 . 1 . r 0 I I I I 1 I I I I I . ' " 1- • • . • - •!• • .'....l.-., 400 600 800 1000 1200 1400 1600 1800 2000 Wavelength (nm) Figure 24. Vis-near-ffi. spectra of 62c+ (•••) and 63c+ (—) in CH 2 C1 2 at -17 °C. Due to the instability of the Ru f f l complexes (vide supra) electronic spectra of these species cannot be obtained at room temperature; however, the solutions of the oxidized complexes can be prepared by the addition of one equivalent of a freshly prepared solution of FcPF6 to a solution of the neutral complex in dry CH 2C1 2 at -20 °C. The solutions are transferred to a spectrophotometric cell held at -17 °C, and the electronic spectra are obtained. The oxidation reactions are assumed to be complete when no further increase in the new 95 absorption bands is observed (within 30 min). After this time, the oxidized species 62c+ and 63c+ are stable in solution for at least 30 rnin with only slight changes in their absorption spectra. Complexes 62a+ - b+, 63a+ - b + still decompose at this temperature as evidenced by changes in the color of the solutions, and the spectra of these complexes are not reproducible. For these complexes oxidation and decomposition of the R u m species occur at a comparable rate, resulting in mixtures of complexes in solution. The electronic spectra of 62c+ and 63c+ at -17 °C are shown in Figure 24, and the data summarized in Table 11. In these spectra, three sets of bands (I - III) are observed in the visible and near-IR regions. The ferrocenium used in the preparation is reduced to ferrocene, which has a d - d transition in this region; however, it is very weak (441 nm, s = 91 M~ 'cn f ' ) 1 5 3 and does not interfere with the observed spectra. Band I is assigned to the n - n* absorption of the terthienyl group. This absorption shifts to higher energy in the Ru f f l species, since the metal center is less electron-donating than in the corresponding Ru 1 1 complexes. Bands II and HI consist of multiple, intense absorptions in the visible and near-IR region and are entirely absent in the spectra of the Ru n analogs. R u m complexes frequently exhibit ligand-to-metal charge-transfer (LMCT) absorptions,36'137 particularly with reducing-type ligands. Based on the intensities and energies of bands II and III these bands are assigned as L M C T bands from the terthienyl ligand to the R u m . Splitting of both ligand-donor orbitals and metal-acceptor orbitals would give rise to the multiple bands which are observed; however, it is not possible to assign these low-energy bands to specific transitions at this time. 96 3.4 Discussion In complexes 62a - c and 63a - c the length of the oligothienyl group has a dramatic effect on the oxidation potential of this group; however, it does not significantly affect the Ru11™ oxidation potential. This is similar to the effects observed in bimetallic Fe11 ferrocenylacetylide complexes 8a and 8b.56 The oxidation potential of the (Cp)(PP)Fe-moiety is -0.47 V vs Fc+/Fc in 8a (PP = dppm), and -0.84 V (PP = dmpe) in 8b, while the oxidation potential of the ferrocenyl group is relatively constant at +0.12 V vs Fc+/Fc in 8a and +0.08 V in 8b. Although the length of the oligothienyl ligand has little effect on the potential of the Ru11™ oxidation, the complexes containing the longer ligands are found to have more reversible Ru11™ oxidation waves. This is due to resonance delocalization of the positive charge in the R u r a species onto the oligothienyl group. This resonance stabilization is greater for more conjugated ligands, and nnnimizes further chemical reactions of the oxidized complexes. This is consistent with the complete irreversibility of the Ru11™ wave in trans-Ru(dppm)2(Cl)(C=CH),151 in which no resonance stabilization is possible. Charge delocalization in the monocations 62c+ and 63c+ may be evaluated by analysis of their electrochemical and spectroscopic data using Hush theory, which has been previously applied to L M C T processes.34"36 The terthienyl-to-Rum L M C T in these oxidized species can be represented by the potential energy diagram in Figure 7b, and the energy of the L M C T correlates to the oxidation potential difference between the metal and oligothienyl groups (eqs 8 and 12). The difference in the oxidation potentials (AE) of the terthienyl ligand and the Ru center is 0.54 V in 62c and 0.46 V in 63c, and the complex with the smaller AE (63c) also has the lower energy L M C T band. The lowest energy absorptions are observed at 1090 nm in 62c+, and 1330 nm with a shoulder at 1610 nm in 63c . Similarly, the difference in the 97 oxidation potentials for the two Fe11™ couples of 8a (0.62 V) and 8b (0.80 V) correlates with the energy of the IVCT band.5 6 The monocation 8a+ has an intense, broad IVCT band at 1595 nm, while 8b+ has a band at 1295 nm. 3.5 Conclusions The results in this Chapter support the conclusion that the rc system of the conjugated oligothienyl ligands interacts electronically with the Ru center. The electron-donating Ru n group decreases the energy of the TC - it* transition in the oligothienyl ligands, and the energy of this transition increases when the Ru n is oxidized. The reversibility of the Ru11™ oxidation wave improves as the length of the oligothienyl ligand is increased, indicating stabilization of the R u m species occurs via delocalization of positive charge from the metal to the oligothienyl ligand. The presence of low-energy charge-transfer transitions in the spectra of 62c+ and 63c+ is consistent with this delocalization. These results suggest that polymers containing Ru centers bridged by oligothienyl linkers could be conductive; however, it is clear that polymers having these groups will require complete stability of the Ru center in both oxidation states, otherwise decomposition may result in the formation of material of ill-defined composition. 98 Chapter 4 Charge Derealization in (Ferrocenylethynyl)oIigothiophene Complexes 4.1 Introduction The extent of charge derealization in metal complexes will depend on the magnitude of the energy barrier to charge transfer between donor and acceptor (£th as shown in Figure 7). As described in Chapter 2, the higher degree of charge derealization in the ruthenium(II) bis(ferrocenylacetylide) complexes is observed when the difference in oxidation potentials between the terminal groups and the Ru center is smallest. The same trend is also observed for 62c+ and 63c+, which show oligothienyl-to-Rum L M C T bands in the near-IR region. These L M C T absorption bands have very similar features to the F /CT bands described in Chapter 2. The electronic and optical properties of hybrid polymers which consist of metal and organic conjugated groups in the polymer backbone depend on charge derealization along the polymer backbone. Insight into the extent of derealization along the polymer backbone may be obtained by examining the L M C T processes in model complexes such as Fc +/Fc-=-R and Fc + /Fc-=-R-s-Fc/Fc + (R = conjugated organic groups). This Chapter describes the spectroscopic and electrochemical characterization of a series of compounds in which a metal center (ferrocene) is conjugated to oligothiophenes of varying oxidation potentials (68a - e and 69a - e). Analysis of this data allows important conclusions to be made regarding charge derealization in these complexes. 99 4.2 Experimental General. Procedures were the same as described in previous Chapters except as noted here. 5,5'-Dibromo-2,2'-bithiophene146 and 5,5"-dibromo-2,2':5',2"-terthiophene146 were prepared using literature procedures. 3,4-Ethylenedioxythiophene was a gift from Bayer Ltd. UV-vis-near-IR spectra of solutions of 68a + - e+ and 6 9 a 2 + - e 2 + in CH2CI2 containing 0.10 M [(n-Bu)4N]PF6 were obtained on a Varian Cary 5 spectrometer. Electrochemistry. Cyclic voltammograms were obtained under nitrogen at room temperature in a CH2CI2 solution containing 4 - 6 x 10~3 M complex and 0.5 M [(n-Bu4)N]PF6. Before addition of dry solvent the cells containing the electrodes and electrolyte were dried in vacuo at 90 °C overnight. Electrolyses of 68a - e and 69a - e were conducted in a CH2CI2 solution containing 0.10 M [(n-Bu)4N]PF6 using Pt mesh electrodes and a Ag/AgN0 3 (CH 3CN) reference electrode (0.39 V vs SCE). 100 2,5-Dibromo-3,4-ethylenedioxythiophene. This complex was prepared by a modification of the literature method.1 5 4 3,4-Ethylenedioxythiophene (1.42 g, 10.0 mmol) was dissolved in THF/CH 3 COOH (40 mL, 1:1 v/v), and 7v"-bromosuccinimide (3.74 g, 21.0 mmol) was added. After the solution was stirred at room temperature for 2 h, distilled water (100 mL) was added resulting in the precipitation of a silver-white crystalline solid. The solid was isolated by filtration, and dried over P2O5. Yield: 2.6 g (87%). 3,,4,-EthyIenedioxy-2,2':5',2"-terthiophene. Mg foil (0.60 g, 25 mmol) and one grain of I2 were added to a dry flask charged with dry diethyl ether (80 mL). The suspension was stirred for 10 rnin, and a solution of 2-bromothiophene (3.60 g, 22.0 mmol) in diethyl ether (20 mL) was then added dropwise over 30 min via a pressure-equalized addition funnel. After addition, the solution was heated at reflux for 2 h, and then cooled to room temperature. This solution was added to a solution of 2,5-dibromo-3,4-ethylenedioxythiophene (3.0 g, 10 mmol) and Ni(dppp)Cl2 (0.27 g, 0.50 mmol) in diethyl ether (50 mL) via a filter paper-tipped cannula. The mixture was then heated at reflux overnight. After the solution was cooled to room temperature, the reaction was quenched by addition of 1 M aq HC1 (50 mL). The organic layer was collected, washed with distilled water (2 x 50 mL) and dried over anhydrous MgSCM. After filtration, removal of the solvent resulted in an orange oil, which was purified by flash chromatography on silica gel using hexanes/diethyl ether (9:1 v/v) as eluant. The product was used without further purification in the next step. Yield: 3.02 g. lU N M R (400 MHz, de-benzene): 8 7.27 (dd, J = 3.6,1.1 Hz, 2H), 6.81 (dd, 7 = 5.1,1.1 Hz, 2H), 6.75 (dd, J = 5.1,3.6 Hz, 2H),3.35(s,4H). 5-Bromo-3',4,-ethylenedioxy-2,2,:5,,2"-terthiophene. 3',4-Ethylenedioxy-2,2':5',2"-terthiophene (1.00 g, 3.03 mmol) was dissolved in THF/CH3COOH (40 mL, 1:1 101 v/v). TV-Bromosuccinimide (0.55 g, 3.1 mmol) was added to the solution while the solution was stirred rapidly. After addition, the solution was stirred for another 1 h, then distilled water (50 mL) was added to induce the precipitation of a yellow solid which was collected by filtration. The crude yellow solid was dried over P2O5, and the product was purified by chromatography on silica gel using CJ-LCbThexanes (1:4 v/v) as eluant. Yield: 0.45 g (39%). *H N M R (400 MHz, de-benzene): 5 7.26 (d, J = 2.9 Hz, 1H), 6.81 (d, J= 4.9 Hz, 1H), 6.73 -6.76 (m, 2H), 6.64 (d, 7= 3.9 Hz, 1 H), 3.21 - 3.35 (m, 4H). Anal. Calcd C ^ a B r O ^ : C 43.64, H 2.35. Found: C 43.03, H 2.27. 5,5"-Dibromo-3',4,-ethylenedioxy-2,2':5',2"-terthiophene. S'^'-Ethylenedioxy-2,2':5',2"-terthiophene (1.00 g, 3.03 mmol) was dissolved in THF/CH3COOH (40 mL, 1:1 v/v). N-Bromosuccinimide (1.13 g, 6.36 mmol) was added portionwise while the solution was stirred rapidly. After addition the solution was stirred for 2 h, during which time a yellow solid formed. Distilled water (50 mL) was added to induce the precipitation of more yellow solid, which was collected by filtration. The yellow solid was dissolved in CH2CI2 (50 mL), and the solution was washed with distilled water (50 mL) and dried over anhydrous MgSCV After filtration, removal of the solvent yielded a yellow solid, which was purified by recrystallization from toluene. Yield: 1.04 g (71%). ' H N M R (400 MHz, de-benzene): 8 6.72 (d, J = 3.9 Hz, 2H), 6.65 (d, J = 3.9 Hz, 2H), 3.22 (s, 4H). Anal. Calcd Ci^gBrzChSs: C 36.22, H 1.74. Found: C 35.93, H 1.65. 2-Ferrocenylethynylthiophene (68a). Ethynylferrocene (0.38 g, 1.8 mmol) was added to a solution of 2-bromothiophene (0.25 g, 1.5 mmol), trans-Pd(PPh3)2Ci2 (53 mg, 0.076 mmol), Cul (15 mg, 0.077 mmol) and diisopropylamine (0.30 g, 3.0 mmol) in dry THF (50 mL). The solution was stirred and heated at reflux overnight. After the solution was 102 cooled to room temperature, distilled water (50 mL) was added to quench the reaction. The solution was extracted with CH2CI2 (2 x 40 mL). The organic layers were combined, washed with brine (50 mL) and distilled water (50 mL), and dried over anhydrous MgSCU. Removal of the solvent resulted in a brown solid, which was purified by flash chromatography on silica gel using hexanes as eluant. Yield: 0.30 g (68%). *H N M R (400 MHz, CD 2C1 2): 8 7.27 (dd, J = 1.1, 5.2 Hz, 1H), 7.21 (dd, J= 1.1, 3.6, 1H), 7.00 (dd, J= 3.6, 5.2, 1H), 4.50 (t, J= 1.8 Hz, 2H), 4.27 (t, J= 1.8 Hz, 2H), 4.25 (s, 5H). Anal. Calcd QeHnFeS: C 65.77, H 4.14. Found: C 66.14, H 4.27. 5-Ferrocenylethynyl-2,2'-bithiophene (68b). This complex was prepared as described for 68a using 5-bromo-2,2'-bithiophene. The product was purified by flash chromatography on silica gel using CH2Cl2/hexanes (1:20 v/v) as eluant. Yield: 74%. ! H N M R (400 MHz, CD2C12): 8 7.27 (dd, J= 1.0, 5.1 Hz, 1H), 7.21 (dd, J = 1.1, 3.6, 1H), 7.11 (d, J= 3.8 Hz, 1H), 7.06 (d, J= 3.8 Hz, 1H), 7.04 (dd, / = 3.6, 5.1, 1H), 4.51 (t, 7= 1.7,2H), 4.29 (t, J= 1.7,2H), 4.26 (s, 5H). Anal. Calcd C 2oHi 4S 2Fe: C 64.18, H 3.77. Found: C 64.24, H3.78. 5-Ferrocenylethynyl-2,2':5',2"-terthiophene (68c). This complex was prepared as described for 68a using 5-iodo-2,2':5',2"-terthiophene. The product was purified by flash chromatography on silica gel using C^Ch/hexanes (1:3 v/v) as eluant. Yield: 25%. ! H N M R (400 MHz, CDC13): 8 7.21 (dd, J= 1.1,5.1 Hz, 1H), 7.17 (dd,J= 1.1,3.6,1H), 7.08 (d, J = 3.8 Hz, 1H), 7.06 (s, 2H), 7.00 - 7.02 (m, 2H), 4.49 (t, J= 1.8, 2H), 4.24 - 4.25 (m, 7H) Anal. Calcd C24Hi6S3Fe: C 63.16, H 3.53. Found: C 62.85, H 3.43. 5-FerrocenylethynyI-3,4-ethylenedioxythiophene (68d). A solution of n-BuLi (4.0 mL, 1.6 M in hexanes, 6.4 mmol) was cooled to -78 °C and added dropwise to a solution of 103 2,5-dibromo-3,4-emylenedioxythiophene (1.8 g, 6.0 mmol) in dry THF (40 mL). The solution was then warmed up to room temperature, and the reaction was quenched by adding 1 M aq HC1 (30 mL). The solution was extracted with CH2CI2 (2 x 30 mL), and the organic layers were combined and dried over anhydrous MgSCV Removal of the solvent yielded 1.0 g of pale yellow 2-bromo-3,4-ethylenedioxythiophene, which was used without purification to prepare 68d, following the same procedure described for 68a. Pure 68d was obtained by flash chromatography on silica gel using CF^C^/hexanes (1:2.5 v/v) as eluant. Yield: 18%. *H N M R (400 MHz, CD 2C1 2): 5 6.27 (s, IH), 4.48 (t, J= 1.8 Hz, 2H), 4.28 - 4.30 (m, 2H), 4.26 (t, J= 1.8 Hz, 2H), 4.24 (s, 5H), 4.19 - 4.21 (m, 2H). Anal. Calcd Ci 8 Hi 4 0 2 SFe: C 61.73, H 4.03. Found: C 61.43, H 3.98. 5-Ferrocenylethynyl-3*,4,-ethyIenedioxy-2,2' :5',2"-terthiophene (68e). This complex was prepared as described for 68a using 5-bromo-3',4'-ethylenedioxy-2,2':5',2"-terthiophene. The product was purified by flash chromatography on silica gel using CH2Cl2/hexanes (1:3 v/v) as eluant. Yield: 62%. *H N M R (400 MHz, CD 2C1 2): 8 7.25 - 7.27 (m, 2H), 7.12 (d, J= 3.9 Hz, IH), 7.09 (d, ./= 3.9 Hz, IH), 7.04 (dd, J= 3.8,5.0 Hz, IH), 4.50 (t, J = 1.8 Hz, 2H), 4.40 (s, 4H), 4.28 (t, J - 1.8 Hz, 2H), 4.26 (s, 5H). Anal. Calcd C26Hi802S3Fe: C 60.70, H 3.53. Found: C 60.57, H 3.56. 2,5-Bis(ferrocenyIethynyI)thiophene (69a). Ethynylferrocene (0.52 g, 2.5 mmol) was added to a solution of 2,5-dibromothiophene (0.24 g, 1.0 mmol), Pd(PPh3)2Ci2 (70 mg, 0.10 mmol), Cul (20 mg, 0.10 mmol) and diisopropylamine (0.30 g, 3.0 mmol) in dry THF (50 mL). The solution was stirred and heated at reflux overnight. After the solution was cooled to room temperature, distilled water (30 mL) was added to quench the reaction. The solution was extracted with CH2CI2 (2 x 40 mL). The organic layers were collected and 104 combined, washed with brine (50 mL) and distilled water (50 mL), and then dried over anhydrous MgSC^. Removal of the solvent yielded a brown solid, which was purified by flash chromatography on silica gel using CH2Cl2/hexanes (1:4 v/v) as eluant. Yield: 0.41 g (82%). ! H N M R (400 MHz, CDCI3): 5 7.04 (s, 2H), 4.49 (t, J = 1.8 Hz, 4H), 4.25 (t, J= 1.8 Hz, 4H), 4.24 (s, 10H). Anal. Calcd C28H2oS2Fe2: C 67.23, H 4.03. Found: C 67.29, H 3.91. 5,5'-Bis(ferrocenylethynyl)-2,2'-bithiophene (69b). This complex was prepared as described for 69a using 5,5'-dibromo-2,2'-bitl-iophene. The product was purified by flash chromatography on silica gel using CH2Cl2/hexanes (1:3 v/v) as eluant. Yield: 80%. ! H N M R (400 MHz, CDC13): 8 7.09 (d, / = 3.8 Hz, 2H), 7.02 (d, 3.8 Hz, 2H), 4.49 (t, J= 1.8 Hz, 4H), 4.25 (t, J= 1.8 Hz, 4H), 4.24 (s, 10H). Anal. Calcd C 3 2 H 2 2 S 2 Fe 2 : C 66.00, H 3.81. Found: C 66.11,H 3.70. ' 5,5"-Bis(ferrocenylethynyl)-2,2':5,,2"-terthiophene (69c). This complex was prepared as described for 69a using 5,5n-dibromo-2,2':5\2"-terfhiophene. The product was purified by flash chromatography on silica gel using CH2Cl2/hexanes (3:7 v/v) as eluant. Yield: 65%. lH N M R (400 MHz, CDC13): 8 7.09 (d, J= 3.8 Hz, 2H), 7.06 (s, 2H), 7.02 (d, J = 3.8, 2H), 4.49 (t, J = 1.8, 4H), 4.25 (t, J = 1.8 Hz, 4H), 4.24 (s, 10H). Anal. Calcd C36H 2 4S 3Fe 2: C 65.07, H 3.64. Found: C 65.38, H 3.68. 2,5-Bis(ferrocenylethynyl)-3,4-ethyIenedioxythiophene (69d). This complex was prepared as described for 69a using 2,5-dibromo-3,4-emylenedioxythiophene. The product was purified by recrystallization from layered hexanes/CH2Cl2 solution, following by flash chromatography on silica gel using diethyl ether/hexanes (1:2 v/v) as eluant. Yield: 34%. ' H N M R (400 MHz, CD 2C1 2): 8 4.50 (t, J= 1.8 Hz, 4H), 4.31 (s, 4H), 4.28 (t, J= 1.8 Hz, 4H), 4.25 (s, 10H). Anal. Calcd C 3 0 H 2 2 O 2 SFe 2 : C 64.55, H 3.97. Found: C 63.94, H 3.98. 105 5,5"-Bis(ferrocenylethynyl)-3S4'-ethylenedioxy-2,2,:5S2"-terthiophene (69e). This complex was prepared as described for 69a using 5,5"-dibromo-3',4'-ethylenedioxy-2,2':5^2"-terthiophene. The product was purified by recrystallization from a layered CH2Cl2/hexanes solution, following by flash chromatography on silica gel using diethyl ether/hexanes (1:2 v/v) as eluant. Yield: 36%. ! H N M R (400 MHz, CD 2C1 2): 8 7.09 - 7.14 (m, 4H), 4.50 (t, J= 1.8 Hz, 4H), 4.42 (s, 4H), 4.29 (t, J= 1.8 Hz, 4H), 4.26 (s, 10H). Anal. Calcd C38H 2 60 2S 3Fe 2: C 63.17, H 3.63. Found: C 62.46, H 3.62. 4.3 Results and Interpretation 4.3.1 Syntheses Compounds 68a and 69a are prepared by coupling ethynylferrocene with mono- and dibromothiophene, respectively, using trans-Pd(PPh3)2Cl2 and Cul catalysts (Scheme 17), and 68b -e and 69b - e are obtained using the same procedures. The compounds are purified either by chromatography on silica gel or by recrystallization from CH2Cl2/hexanes or toluene, and characterized using ! H N M R and UV-vis spectroscopies and C, H elemental analyses. Scheme 17 Br trans-Pd(PPh3)2Cl2, Cul THF, diisopropylamine A 68a 2 Fe + Br S Br trans-Pd(PPh3)2Cl2, Cul THF, diisopropylamine A 69a 106 4.3.2 Electrochemistry Complexes 68a - e and 69a - e all contain ferrocenyl and oligothiophene groups which are both expected to show redox activity. The reversibility and relative oxidation potentials of the redox processes in these compounds are deteirnined by C V in CH2CI2 solution containing 0.5 M [(rt-Bu)4N]PF6. A l l the compounds have two oxidation waves in the range 0 - 1.8 V vs SCE (Figures 25 and 26), and the potentials for these waves are collected in Table 12. Table 12. Electrochemical and UV-vis Spectroscopic Data for 68a - e and 69a - e Complex £1/2(1) ±0.01 (V) a £p,a(2) ±0.01 (V) a AE ± 0.02 (V) 3* A^ax (nm) (s (IVr'cm ')) c 68a 0.55 1.67 1.12 305 (14000), 445 (610) 68b 0.55 1.40 0.85 350(25000), 445 (1600) (sh) 68c 0.55 1.13 0.58 392(35000) 68d 0.53 1.47 0.94 310(13000), 445(610) 68e 0.53 0.91 0.38 406 (39000) 69a 0.56 1.68 1.12 342 (26000), 446 (3300) (sh) 69b 0.55 1.42 0.87 390 (38000) 69c 0.55 1.23 0.68 416(44000) 69d 0.56 1.50 0.94 352 (27000), 448 (3700) (sh) 69e 0.54 1.04 0.50 433 (52000) 0 Volts vs SCE, Pt working electrode, CH2C12 containing 0.5 M [(n-Bu)4N]PF6,20 °C. b AE = [£p,a(2) -£1/2(1)]. cCFf 2Cl 2solution. 107 0.(3 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Voltage (V vs SCE) Figure 25. Cyclic voltammograms of (a) 68a, (b) 68d, (c) 68b, (d) 68c and (e) 68e in CH 2C1 2 containing 0.5 M [(/i-Bu4)N]PF6. Scan rate = 100 mV/s. 108 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Voltage (V vs SCE) Figure 26. Cyclic voltammograms of (a) 69a, (b) 69d, (c) 69b, (d) 69c and (e) 69e in CH 2C1 2 containing 0.5 M [(w-Bu)4N]PF6. Scan rate = 100 mV/s. The first oxidation wave is reversible and occurs at a potential very close to that of the Fe11™ couple in ethynylferrocene (E\a - 0.57 V vs SCE); therefore, this wave is assigned to the Fe™11 redox couple. The compounds containing two ferrocenyl groups (69a - e) show 109 only a single-oxidation wave for these centers, indicating that there is little ground-state interaction between the metal centers over the conjugated oligothiophene bridge. In the series of all-trans compounds Fc-(CH=CH) n-Fc (n = 1 - 6), peak separations are observed only for n < 3, 4 5 while for Fc-CH=CHCeFi4CH=CH-Fc no peak separation is observed.44 Thus, it is not surprising that no separation is observed in 2a - e, in which the Fe - Fe distance is considerably longer. The second oxidation wave in the cyclic voltammograms of all the compounds is irreversible (Figures 25 and 26), and the oxidation potential depends strongly on the nature of the oligothiophene group. Both longer conjugation length and the presence of electron-donating ethylenedioxy substituents (in 68d, 68e, 69d and 69e) result in a decrease in the potential of this wave. On the basis of these observations the second wave is assigned to an oligothiophene-based oxidation, the potential of which varies from 0.91 to 1.68 V vs SCE. The peak current of the second wave varies due to the differences in the stabilities of the resulting dications. Complexes 68a - e all contain a terminal thiophene ring with an unsubstituted a position, and these compounds are consequently good candidates to undergo a - a coupling when oxidized. The second oxidation waves of 68a and 68d are completely irreversible, and no new reduction waves which would be evidence of dimerization are observed (Figure 25a and 25b). On the other hand, the cyclic voltammograms of 68b, 68c and 68e all contain a number of features which indicate dimerization is occurring (Figure 25c - e). For all three compounds a new reduction wave appears after the first scan past the thiophene-oxidation wave. In subsequent scans, these reduction waves have corresponding oxidation features, and when solutions of these compounds are cycled repeatedly past the thiophene-oxidation wave 110 (0.7 - 1.5 V vs SCE for 68b, 0 - 1.3 V for 68c and 0 - 1.1 V for 68e), an electrochromic film deposits on the electrode surface. The films are orange-red in the neutral state, and black when fully oxidized. Characterization of these films by C V in monomer-free solution reveals a reversible wave at ~ 0.5 V vs SCE due to the Fe117111 couple as well as other waves assigned to oxidation of the oligothiophene groups (Figure 27). The electrodeposited films are proposed to consist of dimers 70,71 and 72. o o o o The oxidation wave at 1.01 V in 70 (Figure 27a) is quasireversible and occurs at a lower potential than the oxidation of the terthiophene group in 69c. This is consistent with the presence of a longer tetrathiophene bridge in this dimer. The new reduction wave at 0.63 V vs SCE may be due to a similar process to that observed in the CVs of 69c and 69e (vide infra). The C V of 71 (Figure 27b) contains two oxidation waves with Em at 0.51 and 0.91 V vs SCE. The wave at 0.51 V has a shoulder on the high potential side, which may be due to overlap between the Fe11™ oxidation and an oligothiophene oxidation wave. In 71, the bridge consists of a sexithiophene group. These have been shown to support two reversible one-I l l electron oxidations, for instance, in didodecylsexithiophene the first oxidation occurs at 0.34 V, and the second occurs at 0.54 V vs Fc + /Fc. 2 3 _i i I i I i I i I i i i i i I 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Voltage (V vs SCE) i Figure 27. Cyclic voltammograms of (a) 70, (b) 71 and (c) 72 in CH 2 C1 2 containing 0.5 M [(n-Bu)4N]PF6. Scan rate = 100 mV/s. 112 The C V of the dimer 72 shows similar features, with the waves having E\a at 0.35 and 0.82 V vs SCE assigned as sequential, reversible oxidations of the substituted sexithiophene bridge. The electron-donating ethylenedioxy groups in 72 result in lower potential waves for the bridge compared to 71. In 72 the bridge is oxidized at a lower potential than the ferrocenyl groups, and it is possible that an increase in the conductivity of the molecular wire linking the two terminal groups could lead to splitting in the potentials of the ferrocenyl waves; however, no splitting in this case is observed, possibly due to poor orbital overlap between the oligothiophene and the ethynyl linkage. For complexes 69a, 69b and 69d the oligothiophene-oxidation waves are irreversible, and no new reduction waves are observed; however, for 69c and 69e the second oxidation waves become more reversible at higher scan rates, and new reduction features appear (Figure 26). For example, at a scan rate of 100 mV/s a small reduction wave is observed at 0.96 V for 69c and at 0.64 V vs SCE for 69e. This wave is likely due to the reduction of the product resulting from coupling of 69c or 69e upon oxidation past the oligothiophene-oxidation potential, analogous to the behavior of 68c, 68d and 68e. Although the disubstituted complexes do not have unsubstituted a positions available, it is possible that P - P coupling occurs in these compounds. 4.3.3 Spectroscopic Characterization The UV-vis spectra of 68a - e and 69a - e all contain very strong absorption bands with Xmax between 305 and 433 nm, and these bands are assigned to an oligothiophene rc — TC* transition (Table 12). As expected, the absorption red-shifts and becomes more intense with an increase in the length of the oligothiophene group. The presence of the electron-donating 113 ethylenedioxy groups results in a slight decrease in the absorption maximum relative to the unsubstituted compound with an equally long oligothiophene group. Several of the compounds also have a weaker absorption band at ~ 446 nm, assigned to a Fe n d - d transition.135 This band is not observed in the spectrum of the complex containing a longer oligothiophene group because of overlap from the strong, broad TC - TC* transition of the oligothiophene group. Stable solutions of the oxidized species 68a+- e+ and 69a 2 +- e2+ in CH2CI2 containing 0.10 M [(n-Bu)4N]PF6 are prepared by constant-potential electrolysis at a potential 0.25 V above Em of the ferrocenyl wave, except for 68e which is electrolyzed 0.15 V above the iron-oxidation potential to preclude oxidation of the oligothiophene group. The UV-vis-near-ER spectra of these solutions are measured, and the data are collected in Table 13. A l l the complexes have a strong absorption band with Xm^ between 285 and 425 nm, with a low-energy shoulder between 370 - 520 nm. The high-energy absorption band appears at a similar wavelength, and is of comparable intensity, to that observed in the corresponding neutral complex, and this band is therefore assigned as TC - TC* transitions in the oligothiophene groups. The shoulder observed between 370 and 520 nm is due to a Cp —>• F e m ligand-to-metal charge-transfer (LMCT) transition, and has been observed previously at similar energies in the oxidized species of 52 - 57 discussed in Chapter 2. The spectra of 68a+ - e+ and 69a2+ - e2+ all contain broad, low-energy absorption bands with Xmax between 875 and 1290 nm (Figures 28 and 29). These absorption bands are assigned to oligothiophene -> F e m L M C T transitions. The energy maxima of these L M C T transitions correlate to the length of the oligothiophene groups, with the absorption maxima shifting to lower energy with increased conjugation. The ethylenedioxy substituent in 114 compounds with an identical conjugation length (for example 68c vs 68e) results in a shift to lower energy and an increase in the intensity of the absorption band. Table 13. UV-vis-Near-IR Spectroscopic Data for 68a+ - e+ and 69a2+ - e' Complex (nm) (± 5 nm) (e ( T V T W 1 ) ) 3 Avi/2 (cm-1)* ±100 / ( x 10"3)* (cm"1) 68a+ 280 (18000), 370 (5000) (sh), 875 (630 ± 20) 3460 1 0 ± 2 68b+ 340 (22000), 445 (8800) (sh), 1015 (1390 ± 50) 3680 24 ± 2 68c+ 385 (26000), 485(11000) (sh), 1090 (1550 ± 70) 3800 27 ± 2 68d+ 285 (23000), 410 (56000) (sh), 995 (1190 ± 50) 3240 1 8 ± 2 68e+ 400 (28000), 505 (12000) (sh), 1290 (2090 ± 90) 3750 3 6 ± 2 69a2 + 285 (29000), 425 (12000) (sh), 885 (1550 ± 60) 3805 27 ± 2 69b2 + 375 (28000), 475 (19000), 1010 (2600 ± 100) 3890 4 6 ± 2 69c 2 + 410 (33000) 475 (29000) (sh), 1080 (3600 ± 160) 3760 62 ± 3 69d 2 + 290 (28000), 320 (24000) (sh), 450 (14000) (sh), 4120 4 6 ± 2 985 (2400 ±90) 69e2+ 425 (30000), 520 (25000) (sh), 1255(4000 ± 150) 4300 79 ± 3 a CH 2C1 2 containing 0.10 M [(«-Bu)4N]PF6,20 °C. b lowest energy bands Wavelength (nm) Figure 28. Vis-near-IR spectra of 68a + - e +in CH 2C1 2 containing 0.1 M [(«-Bu)4N]PF6. 116 600 800 1000 1200 1400 1600 1800 2000 Wavelength (nm) Figure 29. Vis-near-IR spectra of 69a' - e" in CH 2C1 2 containing 0.1 M [(n-Bu)4N]PF6. 4.4 Discussion > The electrochemical and spectroscopic results are interpreted using the classical electron-transfer model of Hush. The absorption maximum (vm a x) of an IVCT band has been related to the difference in electrochemical potentials (AE) between two metal centers (eqs 8 and 12). A similar relationship exists for the L M C T transition band in Ru f f l complexes (eq 14) 36,155 j - e r e j) corrects for the difference between the Ru11™ potential with an oxidized 117 ligand attached and the measured Ru1 1 7 1 1 1 oxidation potential, and % is the sum of the inner and outer reorganizational parameters. V r ™ = Xi + X o + £ + (e//*c)A£ (14) A plot of Vmax vs AE will be linear if D and % are constant for a series of complexes in which only the oxidation potential of the ligand varies. These data are shown for 68a+- 68e+ and 69a2+- 69e2+in Figures 30 and 31, and in both plots the data are linearly correlated with R = 0.975 and 0.990 respectively. These correlations indicate that in both series a smaller oxidation-potential difference between the donor and the acceptor results in a lower optical-transition energy. The extinction coefficient of the L M C T transition in the oxidized complexes clearly increases with smaller AE. Assuming a Gaussian-peak shape, the oscillator strength,/ of the band can be calculated from eq 6. A plot of/"as a function of AE (Figure 32) shows a linear correlation within the series of monoferrocenyl cations, as well as within the diferrocenyl cations. The oscillator strengths of the dications are approximately twice those of the monocations due to the presence of twice the number of chromophores per molecule in the dications. From this plot it is clear that a smaller difference in oxidation potentials between donor and acceptor gives a greater charge-transfer transition oscillator strength. The significance of the correlations observed in Figures 30 - 32 lies in the relationship between the intensity and shape of a charge-transfer band and the extent of charge delocalization. For a one-electron system, the / and Vmax of an absorption band are related to the charge-transfer dipole moment Mby eq 15, which can be derived from eq 4. M2 = (9.22x IO"12) e 2 / /vmax G (15) 118 From eq 15 and the correlations in Figures 30 - 32 it is clear that in the series of compounds studied here M2 also correlates with AE. A smaller oxidation-potential difference between the donor and the acceptor results in a larger charge-transfer dipole moment. The dipole moment is related to a 2 and the transition dipole length d by eq 5. 12000 Figure 30. Absorption maxima Vmax (near-IR) vs the oxidation-potential difference AE £ p > a(2) - ExdX) for 68a + - e+. From eq 5 it is clear that for complexes in which d is approximately the same, a larger dipole moment correlates to a larger a 2 . Although the value of d is not known here, it is 119 reasonable that d does not vary significantly within the four pairs of complexes in the series with the same length oligothiophene group but with the different substitutents (for example 68a+and 68d+, or 69c 2 + and 69e2+). In these pairs, a direct correlation between a 2 and M2, and therefore AE, can be made. In each pair, the complex with the smaller value of AE has the larger / and lower Vmax for the L M C T transition. Therefore, in each pair, a decrease in the oxidation-potential gap between donor and acceptor results in greater charge derealization. 1 1 ' 1 1 1 — 1 i • i 11000 / . _ 10000 69bV^ ^6 9 d 2 + ^ u a s > 9000 • 6 9 c 2 + / V -8000 I . I . I . i . i 0-2 0.4 0.6 0.8 1.0 1.2 A£(V) Figure 31. Absorption maxima Vmax (near-IR) vs the oxidation-potential difference AE -£p;a(2) - £ I / 2 (1) for 69a 2 + - e2+. 120 80 £ 10 . 1 1 1 i | i %>9e2+ 1 i 1 i - ~ T — r - I ' 1 1 1 ^69c 2 + -68e+ 69b2+ 69d2+ -68c+ 68bt 69a2+^ _ 68d+ I . I . I . I . i . i 68a+ " i 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 AE(V) Figure 32. Oscillator strength/(near-IR) vs the oxidation-potential difference AE = £ p, a(2) -E\niX) for 68a+-e+ and 69a2+-e2+. As the conjugation length is increased (e.g. 68a+ - c*) the / increases, and the v,,,^ decreases, together giving a larger M2 (eq 15). Here, it is reasonable that d is not the same for all the complexes, so only the product (M2) of ct and d2 increases with an increase in the length of the oligothiophene ligand. Thus, either more charge is delocalized (if d does not change significantly) or a similar amount of charge is delocalized further along the rigid oligothiophene ligand (if a 2 does not change significantly). Either situation is desirable for a 121 polymer in which charge is more delocalized. Choosing a longer oligothiophene bridge which has an oxidation potential closer to that of the metal group will either result in a greater extent of delocalization of charge, or in charge being delocalized over a greater length along the conjugated backbone. ^ 4.5 Conclusions Oxidation of the Fe11 centers in (ferrocenylethynyl)oligothiophene complexes results in the appearance of an L M C T transition from the oligothienyl group to the F e m center in the near-IR region. The band maxima and oscillator strengths of these low-energy transitions both correlate linearly with the difference AE in the oxidation potentials between the metal and oligothienyl groups. The complexes with similar charge-transfer transition dipole lengths show an increase in the extent of charge delocalization with a smaller AE. Comparisons between complexes with different length oligothienyl groups show that a reduction in AE results either in greater delocalization of charge or in charge being delocalized further along the rigid oligothienyl group. The results described in this Chapter have important implications for the design of conjugated polymers which consist of redox-active metal and organic conjugated groups in the polymer backbone. It is clearly desirable to match the oxidation potentials of the metal and organic groups as closely as possible in order to achieve the highest degree of charge delocalization. 122 Chapter 5 Charge Delocalization in Oligothienylferrocene Monomers and Polymers 5.1 Introduction The C V of 2,5-bis(ferrocenylethynyl)thiophene (69a) described in Chapter 4 shows no ground-state interaction between the two ferrocenyl termini; however, a similar study of 2,5-bisferrocenylthiophene shows two separated Fe11™ waves with a potential difference of 0.14 V, indicating significant ground-state interaction.156 The presence of the C=C group in 69a appears to decrease the degree of charge delocalization relative to the complex without this group. The results described in Chapter 4 show that charge delocalization in complexes containing metal centers linked via oligothienyl groups may be enhanced by matching the oxidation potentials of the metal and oligothienyl groups. Therefore, one approach to obtaining metal-oligothiophene polymers with extensive delocalization is to construct monomers containing two or more linked thienyl groups. Coupling between such monomers should yield oligomers or polymers containing bridges between adjacent metal centers of at least four thiophene units. In this Chapter the preparation and electropolymerization of l,l'-bis(5-(2,2'-bithienyl))ferrocene (50) and l,l,-bis(5-(2,2,:5^2,,-terthienyl))ferrocene (73), and the spectroscopic and electrochemical characterization of these compounds and electropolymerized films are described. Higgins and coworkers have also published a similar study while this thesis is in preparation.117 123 5.2 Experimental General. Procedures were as described in previous Chapters except as noted here. TMEDA was purified by distillation from K O H and was stored over molecular sieves (4 A). l,r-Bis(2-thienyl)ferrocene (40) was prepared using the literature procedure. 6 9 ' 1 5 7 l,l'-Bis(5-(2,2'-bithienyl))ferrocene (50). /z-BuLi (3.5 mL, 1.6 M in hexanes, 5.6 mmol) was added via syringe at 0 °C to a solution of 40 (0.78 g, 2.3 mmol) and T M E D A (1.2 mL, 8.0 mmol) in a mixture of dry THF (20 mL) and hexanes (10 mL). The solution was heated at reflux for 2 h, then cooled to room temperature. A solution of anhydrous ZnCk (0.76 g, 0.56 mmol) in dry THF (20 mL) was then added via cannula, and the mixture was stirred for another 2 h. The mixture was added to a solution of 2-bromothiophene (0.86 g, 5.3 mmol) and Pd(PPh3)4 (60 mg, 0.052 mmol) in dry THF (20 mL) via cannula, and the rnixture was heated at 50 °C for 2 days while the solution was stirred. After the reaction mixture was cooled down to room temperature, a solution of 1 M aq HC1 (30 mL) was added to the reaction solution. The organic layer was collected, and the aqueous layer was extracted using V. methylene chloride (2 x 20 mL). The organic portions were combined and washed with distilled water (2 x 30 mL). The organic solution was dried over anhydrous MgS0 4 , and collected by filtration. Removal of the solvent yielded a brownish red solid, which was purified by flash chromatography on silica gel using hexanes/Q^Ck (3:2 v/v) as eluant. Pure 50 was also obtained by crystallization from the hexanes/CH2Ci2 solution at -4 °C overnight. Yield: 0.39 g (34%). ' H N M R (400 MHz, CDCI3): 5 7.15 (dd, J = 5.0, 1.0 Hz, 2H), 7.05 (dd, J = 3.5, 1.0 Hz, 2H), 6.96 (dd, J = 5.0, 3.5 Hz, 2H), 6.87 (d, J = 3.7 Hz, 2H), 6.73 (d, J = 3.7 Hz, 2H), 4.47 (t, J = 1.8 Hz, 4H), 4.24 (t, J = 1.8 Hz, 4H). Anal. Calcd C 26Hi 8S 4Fe: C 60.69, H 3.53. Found: C 60.74, H 3.42. 124 l,r-Bis(5-(2,2':5',2"-terthienyI))ferrocene (73). This complex was prepared as described for 50 using 5-bromo-(2,2'-bithiophene). The product was obtained in > 95% purity by crystallization either from hot toluene or chlorobenzene solution. Yield: 53%. *H N M R (400 MHz, CDCI3): 8 6.98 (s, broad, 2H), 6.95 (dd, J = 3.6,1.1 Hz, 2H), 6.80 - 6.83(m, 2H), 6.78 (d, J = 3.7,2H), 6.71 (dd, J = 5.1,1.1 Hz, 2H), 6.63 (dd, J = 5.1,3.6 Hz, 2H), 6.53 (d, J = 3.7 Hz, 2H), 4.37 (t, J = 1.8 Hz, 4H), 4.01 (t, J = 1.8 Hz, 4H). Anal. Calcd C34H2 S6Fe: C 60.16, H 3.27. Found: C 60.20, H 3.15. Electrochemistry. The working electrode was a Pt disc electrode with a diameter of 1.0 mm. A l l electrochemical polymerizations were carried out under nitrogen, and the cells containing the electrodes and [(«-Bu)4]NPF6 were dried in vacuo at 90 °C overnight before addition of dry solvent. The polymer film was grown on the electrode surface by scanning over a suitable potential range (0 - 1.5 V vs SCE for 50, and 0 - 1.2 V vs SCE for 73). The electrochemical polymerization of 50 was carried out at 20 °C in a CH2CI2 solution with an electrolyte concentration of 0.6 M , and the complex concentration was in a range of 5 x 10~3 -1 x 10~2 M . The same experiments for 73 were carried out with a saturated solution of 73 in 1,2-dichloroethane with 0.6 M electrolyte at 70 °C. Spectroelectrochemistry. A homemade spectroelectrochemical cell using a Pyrex cuvette into which an ITO working electrode could be inserted under inert atmosphere was used. The polymer films were deposited on the ITO electrodes from the same deposition solutions as used with the Pt electrodes, by scarining at 25 mV/s over a similar potential range (0 - 1.7 V vs SCE for 50, and -0.2 - 1.3 V vs SCE for 73). The poly-50 film was prepared at 20 °C, and the poly-73 film at 55 °C. After the film was reduced at 0 V , the films were removed from the solution, rinsed with CH2CI2 and dried in vacuo for 1 h. The film-coated 125 electrodes were then loaded into the previously dried cell containing [(n-Bu^NPFe, along with a Pt counter electrode and Ag wire reference electrode. The cell was dried in vacuo overnight at room temperature, and dry C H 2 Q 2 was added until the electrolyte concentration was 1.3 M . The vis-near-IR spectrum was obtained at 20 °C after the film was held at a specific potential until no further change in current was observed (~ 1 min), to allow the entire film to reach the potential. 5.3 Results and Discussion 5.3.1 Syntheses Complexes 50 and 73 are synthesized by the coupling of 40 with 2-bromothiophene and 5-bromo-2,2'-bithiophene respectively (Scheme 18). Complex 40 has been previously synthesized,69 while 50 and 73 are new compounds. Complex 50 is obtained as an orange solid which is soluble in CH2CI2 and is purified by chromatography on silica gel. The solubility of 73 at room temperature in most organic solvents is poor; however, its solubility increases at slightly elevated temperatures so it can be purified by recrystallization from hot toluene or chlorobenzene. 126 Scheme 18 Reagents: (a) BuLi, TMEDA, THF/hexanes. (b) ZnCl 2 , THF. (c) 2-bromothiophene, Pd(PPh3)4, THF. (d) 5-bromo-2,2'-bithiophene, Pd(PPh3)4, THF. 5.3.2 Electrochemistry Complexes 40, 50 and 73 are expected to show electroactivity both due to the Fe11™ couple and the thiophene groups, and the C V results are collected in Table 14. The C V of 40 at 20 °C in CH2CI2 containing 0.1 M [(«-Bu)4N]PF6 shows only one reversible oxidation wave with Em - 0.46 V vs SCE, due to the Fe11™ redox couple. Previous attempts by others to electropolymerize this monomer are also unsuccessful.1 1 7.1 5 7 The C V of 50 at 20 °C in CH2C1 2 containing 0.6 M [(n-Bu)4N]PF6 contains two oxidation waves in the range 0 - 1.5 V vs SCE. The lower potential wave with Em = 0.45 V vs SCE is reversible and assigned to the Fe117111 redox couple (Figure 33a). The higher potential wave at 1.33 V vs SCE is 127 electrochemically irreversible and is coupled to a reduction feature at 1.05 V (Figure 33b); this wave is assigned to a thiophene-based oxidation. Similar electrochemical behavior has been observed in a series of [CpRu(r|5-oligothiophene)]PF6 complexes.158 In the absence of water, scanning several times over the potential range 0 to 1.5 V results in the appearance of an electrochromic film on the electrode surface. The film is golden-red and stable in the neutral state, but becomes black when oxidized. Scanning from 0 to 0.7 V , where only the ferrocenyl groups are oxidized, does not result in any deposition. The C V of the electropolymerization process (Figure 33b) is consistent with the growth of a conducting film on the electrode since the current of both the ferrocenyl-oxidation wave and a new reduction wave at 0.8 V increases with scan number. Oxidative coupling of 50 results in a polymer (poly-50) which contains ferrocenyl groups linked by tetrathienyl bridges, and deposits on the electrode due to poor solubility. The new reduction wave is associated with the tetrathienyl groups, and is expected to occur at a lower potential than reduction of the bithienyl group due to more extended conjugation.21 The C V of the poly-50 film in monomer-free solution shows a broad oxidation wave at 0.6 V which is assigned as the Fe11™ oxidation (Figure 34). The associated reduction at 0.35 V is sharper, but is approximately equal in area to the oxidation wave. The C V also shows a broad feature at higher potentials which is due to oxidation of coupled oligothiophene moieties in the film. The film is stable to repeated cycling over the potential range 0 to 0.8 V , with only slight changes in the voltammogram during this process. 128 Table 14. UV-vis-near-IR Spectroscopic and Electrochemical Data for 40,50 and 73 Complex UV-vis-near-IR £i/2(l)* £ P >a(2)" A.(nm)(E(M _ 1cm- 1)) a V v s S C E V v s S C E 40 460 (830), 306 (17000) 0.46 -50 466 (2800, sh), 344 (35000) 0.45 1.33 73 384 (55000) 0.37 0.93 2,2'-bithiophenec 302 (12470) - 1.31 2,2,:5,,2"-terthiophenec 355 (25050) - 1.05 40+ 275 (18000), 365 (9000), 480 (3800), 815 (900, sh), 945 (1200) 50+ 345 (21000), 450 (9600), 560 (3700, sh), 1175(2300) 73+ 385 (51000), 505 (13000, sh), 585 (7800, sh), 1305 (3500) a CH2CI2,20 °C. b Conditions: Pt working electrode; 20 °C; CH 2C1 2; scan rate = 50 mV/s, except for 73: 70 °C; C1CH2CH2C1. c UV-v i s : ref. 1 4; electrochemical: ref. 2 1 129 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Voltage (V vs SCE) Figure 33. Cyclic voltammogram of 50 at 20 °C in CH 2C1 2 containing 0.6 M [(«-Bu) 4N]PF 6 on a Pt working electrode (a) between 0 - 0.8 V and (b) multiple scans between 0-1.6 V. Scan rate = 50 mV/s. 130 Figure 34. Cyclic voltammogram of poly-50 on a Pt working electrode at 20 °C in CH2CI2 containing 0.6 M [(«-Bu)4N]PF6. Scan rate = 50 mV/s. Due to the poor solubility of 73 in CH2CI2 at room temperature the cyclic voltammogram of this compound is obtained in CICH2CH2CI containing 0.6 M [(rc-Bu)4N]PF6 at 70 °C. The cyclic voltammogram contains multiple waves in the range 0 - 1.2 V vs SCE, including a reversible Fe11™ wave with Em = 0.37 V (Figure 35a), and two irreversible waves at 0.93 and 1.03 V due to thiophene-based oxidations. The Fe 1 1 1 1 1 waves of 40, 50 and 73 all appear at very similar potentials, indicating that those redox couples are not substantially 131 affected by the length of the oligothiophene group. Similar behavior is also observed in 62a -c and 63a - c. Increasing the length of the oligothienyl group from 50 to 73 results in a decrease in the second-oxidation potential by 0.40 V . In the absence of water, scanning over the potential range 0 to 1.2 V results in the growth of a golden-red film (poly-73) on the electrode surface (Figure 35b). The C V of the film in monomer-free solution shows similar features to those observed for poly-50 (Figure 34). This cyclic voltammogram (Figure 36) is recorded after the film is scanned ten times over the complete potential range. During these scans the intensity of the waves increase until it stabilizes as shown in Figure 36. After this treatment, the film is stable to repeated cycling over the potential range 0 to 1.5 V. The reasons for this unusual increase in intensity are not known. The thicknesses of the poly-50 and poly-73 films are estimated from integration of the cyclic voltammograms. The number of monomer units in the films is determined from the area under the ferrocenyl-redox waves. Assuming a density of 1.1 g/cm 3, 1 0 the same as that of electropolymerized polythiophene, the poly-50 film in Figure 34 and the poly-73 film in Figure 36 both have a thickness of 0.4 um. It is very important that electropolymerization of 50 and 73 be carried out under rigorous exclusion of water. The presence of even small amounts of water results in the formation of brittle black deposits which stop growing after 2 - 3 scans. The Fe1™ wave disappears after the first scan, which suggests that water reacts to form an insulating film which prevents the Fe11™ redox couple in dissolved monomer from being observed. In a related study, Higgins and coworkers have also attempted to electropolymerize 50 and 51, and have obtained films in which the ferrocenyl groups are electrochemically inaccessible.117 132 This different result is likely related to the amount of water present in the electropolymerization media. T 1 1 1 1 1 1 1 1 • r ' i i i i i i i i i i i i I 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Voltage (V vs SCE) Figure 35. Cyclic voltammogram of 73 at 70 °C in C1CH2CH2C1 containing 0.6 M [(«-Bu)4N]PF6 at a Pt working electrode (a) between 0 - 0.7 V and (b) multiple scans between 0 -1.3 V. Scan rate = 50 mV/s. 133 l — 1 — i — 1 — i — > — i — 1 — i — 1 — i — 1 — i — • — r I i i i i i i i i i ' • i • ' • i 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Voltage (V vs SCE) Figure 36. Cyclic voltammogram of poly-73 on a Pt working electrode at 20 °C in CH2CI2 containing 0.6 M [(n-Bu)4N]PF6. Scan rate = 50 mV/s. 5.3.3 Spectroscopic Characterization The spectroscopic data of both the neutral complexes and the monocations are collected in Table 14. The UV-vis spectra of 40, 50 and 73 in the neutral state contain a strong absorption band assigned to a n - n transition in the oligothienyl group. This band shifts to lower energy and becomes more intense as the conjugation length is increased. The weaker bands at 460 nm in the spectrum of 40 and the shoulder at 466 nm for 50 are assigned 134 as Fe11 d - d transitions, and appear very close to the ligand-field transition for ferrocene at 441 nm. 1 3 5 Solutions of the monocations 40+, 50+, and 73 + are prepared by electrochemical oxidation in CH2CI2 containing 0.13 M [(«-Bu)4N]PF6, and the vis-near-IR spectra are obtained (Figure 37). The spectra all contain several intense transitions between 300 - 800 nm, and the strongest ones are assigned to rc - n transitions in the oligothienyl groups. These transitions appear at similar energy and have similar intensity to those of the corresponding neutral complexes. The other bands between 300 - 800 nm are likely due to Cp —> F e m L M C T transitions and are of similar intensity and energy as those in ferrocenium.153 50000 h 40000 h -> r 5000 4000 3000 r \ 'E < 2000 -; CO 1000 ( c ) , - ; - j b ) 600 800 1000 1200 1400 1600 1800 2000 Wavelength (nm) i T J i • i 400 600 800 1000 1200 1400 1600 1800 2000 Wavelength (nm) Figure 37. Vis-near-IR spectra of (a) 40+, (b) 50+ and (c) 73 + in CH 2 C1 2 containing 0.13 M [(n-Bu)4N]PF6. 135 The spectra of 40 + , 5 0 + and 73 + contain a broad low-energy band with A^ax between 945 and 1305 nm having features very similar to those observed in the complexes described in Chapter 4. The absorption maximum and intensity of this band are dependent on the length of the conjugation in the ligand. The band appears at the highest energy (945 nm) and is the least intense (e = 1200 IVT'cm-1) for 40 + , and appears at the lowest energy (1345 nm) and is the most intense (e = 3500 IVT'cm -1) for 73 + . On the basis of this behavior, this low-energy band is assigned to an oligothiophene -» Fe m L M C T transition. The difference in oxidation potential between the oligothienyl group and the metal center correlates to the energy of this L M C T transition, as well as to the oscillator strength (f= 2.1 x IO - 2 cm - 1 for 40 + , 3.3 x IO - 2 cm - 1 for 5 0 + and 5.5 x 10 - 2 cm - 1 for 73"*). Increased conjugation in the oligothienyl group narrows the oxidation potential difference between the metal and the oligothienyl group, and thus enhances donor-accepting coupling between them. The energy and intensity of the L M C T transition in 40 + , 5 0 + and 7 3 + are consistent with significant delocalization of charge in these complexes, and this delocalization is the greatest for the complex with the most conjugated oligothienyl group (73 +). 5.3.4 Spectroelectrochemistry of E lec tropolymerized F i l m s In order to examine the effect of oxidation on the spectral characteristics of the electropolymerized films, in situ spectroelectrochemistry on poly-50 and poly-73 films on transparent indium tin oxide (ITO) electrodes is carried out. The cyclic voltammograms of these films are similar to those obtained on Pt electrodes, although the positions of all the peaks slightly shift due to the resistance of the ITO layer. In order to minimize IR drop in the 136 solution due to the relatively large separations between the electrodes in the spectroelectrochemical cell, high electrolyte concentrations are used (1.3 M). Optical spectra of the films are collected at three potentials: reduced (—0.1 - 0 V), with the ferrocene oxidized (~ 0.7 - 0.8 V) and fully oxidized ( -1 .5-1.7 V). These spectra over the range 300 - 1600 nm are shown for poly-50 and poly-73, respectively, in Figures 38 and 39. The spectra all show sharp absorptions at > 1100 nm due to vibrational overtones from the electrolyte and solvent. The reduced films (—0.1 - 0 V vs SCE) have a single broad band at 445 nm (poly-50) and 420 (poly-73) which are assigned to a Tt - Tr* transition of the oligothienyl group in the backbone. The absorption maxima for both films are close to those observed for polythiophene (between 418 - 480 nm depending on the method of preparation). The absorption maximum for poly-73 is expected at lower energy than for poly-50 due to the relative conjugation lengths in the two materials; however, the opposite is observed. This may be due to differences in the conditions (temperature and solvent) under which the two films are prepared. The conductivity of electropolymerized polythiophene films has been shown to be dependent on the conditions under which the films are prepared;159 similar effects may influence the absorption maxima for the neutral poly-50 and poly-73 films. Upon oxidation of the films to a potential at which the ferrocenyl groups are oxidized » several changes are observed in the spectra. In the case of poly-50 two bands are observed, at 495 and 1395 nm. The broad, higher energy band is predominantly due to the n - TC* transition with contributions from Cp -» F e m L M C T also possible, and the band with at 1395 nm is assigned as a charge-transfer band from the oligothiophene group to the Fe r a . It is interesting to compare this spectrum to that of 50+ in solution (Figure 37b). The lowest energy band shifts from 1175 nm in 50+ to 1395 nm in poly-50 oxidized at 0.8 V , consistent with the 137 charge transfer originating from a more conjugated moiety in the polymer. Coupling of 50 should result in a tetrathienyl bridge between ferrocenyl groups, which is expected to result in a charge-transfer band slightly lower in energy than that observed for 73 + which contains a terthienyl group. Figure 38. Spectroelectrochemistry of poly-50 on an ITO electrode at the oxidation potentials (a) -0.1 V , (b) 0.8 V and (c) 1.7 V vs SCE in CH 2C1 2 contauiing 1.3 M [(n-Bu)4N]PF6. 138 Figure 39. Spectroelectrochemistry of poly-73 on an ITO electrode at the oxidation potentials (a) 0 V , (b) 0.7 V and (c) 1.5 V vs SCE in CH 2C1 2 containing 1.3 M [(«-Bu)4N]PF6. In the spectrum of poly-73 oxidized at 0.7 V , the rc - n* transition does not shift substantially from neutral poly-73, although a broad low-energy shoulder does appear, and a very weak, broad band in the near-IR is present. Based on the spectra of 50 + and 73+, as well as the spectrum of poly-50 at 0.8 V, this absorption of poly-73 at 0.7 V is expected to be more intense than observed. It is possible that the ferrocenyl centers are not completely oxidized at 0.7 V resulting in a weaker absorption. Cycling the films twice between 0 and 0.7 - 0.8 V 139 demonstrates that the spectral changes are completely reversible, consistent with the stability of the films over this potential range as shown in the cyclic voltammograms of the films on the Pt electrodes. Oxidation of both poly-50 and poly-73 films to 1.5 - 1.7 V vs SCE results in dramatic changes in the spectra. For both poly-50 and poly-73 a very broad absorption between 400 and 1600 nm appears. In contrast to the reversibility of the spectral changes upon oxidation to 0.7 - 0.8 V , these broad absorptions do not completely disappear when the oxidized film is reduced back to -0.1 - 0 V indicating oxidation to 1.5 - 1.7 V results in some irreversible changes to the polymers. The appearance and positions of these broad bands suggest that they arise due to transitions to intergap states in the oxidized polymers. Oxidation or doping of polythiophene causes the interband absorption at 480 nm to decrease with the concomitant appearance of lower energy bands due to transitions between the valence band and localized levels in the band gap. 1 6 0 5.4 Conclusions The stability of ferrocene in both Fe11 and F e m oxidation states allows for the electropolymerization of oligothienylferrocene complexes without significant decomposition. The growth of electroactive films on the electrode surface is consistent with the formation of a conductive material during the electropolymerization process. The reversible electrochromic behavior of these films is similar to that observed for other conducting polymers such as polythiophene. Oxidation of the ferrocenyl groups in both the monomers and in the electropolymerized films results in the appearance of a low-energy transition. This band is 140 I in due to an oligothienyl group to Fe L M C T transition and is indicative of significant charge derealization, which is enhanced with increased conjugation in the oligothienyl group. 141 Chapter 6 Copper(I) Halide Catalyzed Trans - Cis Isomerization of RuCI2(dppm)2 6.1 Introduction Coupling reactions of ruthenium halide complexes with organostannane derivatives in the presence of a copper(I) halide are an important synthetic route to ruthenium-containing conjugated polymers. 6 4 ' 7 1 ' 8 0 ' 8 2 This route has been successfully used for the preparations of new complexes described in Chapters 2 and 3. In one of these syntheses the amount of copper(I) halide used results in different products. fra/w-Ru(dppm)2(C=CFc)2 (53) is obtained from the coupling reaction of Ru(dppm)2Cl2 with FcC=CSn(n-Bu)3 when catalytic Cul is used, while [cw-Ru(dppm)2(CsCFc)2]CuI (52) is isolated when excess Cul is used. The function of Cul in these coupling reactions is not very well understood, although it is likely that transmetalation of the acetylide group from RC-"CSnR'3 (R' = Bu, Me) to Cul is involved in the catalysis. This is related to the use of copper© compounds as cocatalysts in Stille reactions, in which organic electrophiles are coupled with organostannanes using Pd(0) catalysts. 1 6 1- 1 6 2 Besides the transmetalation, the copper(I) halide likely also plays another role in the coupling reactions which influences the structure of the product as described above for 52 and 53. In order to investigate this question, the reactivity of CuCI or Cul with trans-RuCl2(dppm)2 (19) or os-RuCl2(dppm)2 (59) is examined. Meyer has previously reported that the isomerization of 59 to 19 occurs by photochemical means or by oxidation and subsequent reduction, and that the reverse process (19 -» 59) occurs thermally in 1,2-dichloroethane at reflux (83 °C) in 10 h . 1 2 0 In this Chapter the catalytic behavior of CuCI and Cul for the isomerization of 19 to 59 is described, as well as the isolation and structural characterization of 142 a halide-bridged heterotrimetallic complex 74 that forms when 19 or 59 reacts with excess CuCl. 6.2 Experimental Copper(I) halide catalyzed conversion of 19 to 59. CuCl (0.7 mg, 0.007 mmol, 8 mol %) was added to a solution of 19 (87 mg, 0.093 mmol) dissolved in CH 2 C1 2 (20 mL). The solution was stirred for 24 h at room temperature in the dark, during which time the solution turned bright greenish yellow. The solution was then washed with with 2 M aq HC1 and distilled water to remove residual copper salts. The CH 2 C1 2 layer was collected and dried over MgS04. The solution was reduced in volume to approximately 2 mL and was then poured into hexanes (50 mL) to offord 74 mg of a yellow powder, which was 95% 59 and 5% 19. Pure 59 was obtained when the reaction was allowed to proceed for 6 days. When Cul was used as the catalyst under identical conditions, the same results were obtained. [{cw-RuCl2(dppm)2}2Cu][CuCI2] (74). CuCl (17 mg, 0.17 mmol) was added to a solution of 19 (96 mg, 0.10 mmol) in CH 2C1 2 (20 mL). The suspension was stirred at room temperature for 2 h, during which time it turned from orange-yellow to green-yellow. The solution was filtered through Celite 545, and the filtrate was reduced in volume to approximately 2 mL. The concentrated solution was poured into hexanes (100 mL), yielding a yellow powder. The solid was recrystallized from layered CH2Cl2/hexanes to obtain yellow needles, which were dried at 90 °C in vacuo for 2 days. Yield: 94 mg (89%). Anal. Calcd for CiooH88Cl6P8Cu2Ru2: C 57.70; H 4.26. Found: C 57.79; H 4.13. Crystallographic Study. Data collection and structure deterrnination were carried out by Dr. Glenn Yap (Department of Chemistry and Biochemistry, University of Windsor, 143 Ontario). A suitable single crystal of 74, grown by slow diffusion of hexanes into a 1,2-dichloroethane solution, was selected and mounted inside a thin-walled glass capillary which contained mother liquor. Unit-cell parameters were calculated from reflections obtained from 60 data frames collected at different sections of the Ewald sphere. The systematic absences in the diffraction data and the determined unit-cell parameters were consistent for space groups P2/c and Pc. The E statistics strongly suggested the centric option, which yielded chemically i reasonable and computationally stable results. A trial application of a semiempirical absorption correction based on redundant data at varying effective azimuthal angles yielded Trnax/Trnin at unity and was ignored. Table 15. Crystallographic Data for 74-solvent Empirical formula Ci07.13H88Cl6CU2P8RU2 V,A3 5787.0(3) Formula weight 2165.02 T , K 298 (2) Crystal system Monoclinic Radiation M o K a (0.71073 A) Space group P2/c Density (calcd) 1.242 g e m - 3 a, A 11.6059(4) Z 2 b,A 13.8415(5) R(F),a% 5.38 c, A 36.163(1) Rw(F2),a% 16.47 ,3 deg 95.03(1) GOF on F2 1.066 a Quantity rninimized = RwfF2) = Z[w(F02 - F2)2]; R = Z\F0-Fji /Z(F0); GOF = Goodness-of-fit. The [{c/s-RuCi2(dppm)2}2Cu]+ cation is located on a 2-fold axis. A [CuCk] - anion is located at an inversion center. Attempts to model several peaks of significant electron density, located away from the compound molecules, as a chemically recognizable, cocrystallized 144 solvent molecule were not successful. These peaks were assigned arbitrary carbon atom identities with refined partial site occupancies. A l l non-hydrogen atoms were refined with anisotropic displacement coefficients except those on the apparent solvent molecules, which were refined isotropically. Phenyl groups were refined as idealized, flat, rigid bodies. A l l hydrogen atoms were treated as idealized contributions except those on the apparent solvent molecules, which were ignored. The structure was solved by direct methods, completed by subsequent Fourier syntheses, and refined with full-matrix least-squares methods. A l l scattering factors and anomalous dispersion coefficients are contained in the SHELXTL 5.03 program library. 1 2 3 Table 15 contains the details of the structure determination. 6.3 Results and Interpretation 6.3.1 Catalytic Isomerization When a solution of 19 in CH2CI2 is stirred in the presence of CuCl or Cul (5 -10 mol %) at room temperature for 24 h, 59 forms in high yield (90 - 95%) (Scheme 19). If the reaction is allowed to proceed for 6 days at room temperature in the absence of light, 19 is quantitatively converted to 59. No 59 forms from 19 at room temperature in the absence of catalyst. Identical results are obtained from the catalytic reaction in the presence and absence of ambient laboratory light for 24 h. Stirring a solution of 59 without added coppenT) halide under ambient laboratory light results in the conversion to 60% 19 after 24 h; however, with added CuCl or Cul under ambient light, only a small amount (< 5%) of 19 is formed. These results suggest that the presence of the copper(I) halide prevents formation of significant amounts of 19, presumably by reconversion of photochemically formed 19 to 59. 145 Scheme 19 Ph 2p x PPh 2 C u C l o r C u I P h X ^ P h r i C l - R u - C l ™ 2 l ^ R < U P h 2 P ^ P P h 2 C H 2 C 1 2 ^ p p n f 19 59 6.3.2 Synthesis and Structure of 74 A small amount (< 5%) of a new phosphoms-containing product is also observed by 3 1 P N M R prior to workup when CuCI is used as the catalyst in Scheme 19. This product is obtained in high yield when a CH2CI2 solution of 19 or 59 is stirred with excess CuCI at room temperature for 2 h. This complex is isolated and characterized as 74. _ + / ^ P P h 2 P h 2 P ^ 1 P h 2 p , . I ' u . . c i , C u . c u I r u . P h 2 p P h 2 P ^ I ^ C f " C l " I "Ph 2 P V^PPh 2 P h 2 P ^ / 74 Diffusion of hexanes into a 1,2-dichloroethane solution of complex 74 produces yellow crystals whose structure is determined by single-crystal X-ray diffraction (Figure 40). The structure shows that the cation of 74 is a halide-bridged heterotrimetallic cationic complex in which the copper center has four chloride ligands bridging to the two ruthenium atoms. Although halide-bridging ligands are common in coordination chemistry, halide-bridged heterobimetallic complexes are uncommon, and only several have been structurally characterized. 1 6 3 - 1 6 6 Complex 74 is the first example of a complex containing a dihalide bridge between ruthenium and copper centers. A previous study proposed a single bridging chloride on the basis of N M R evidence for [(Ti3:n3-CioHi6)Cl2Ru(p-Cl)Cu(PCy3)].165 In 146 complex 74, the Cu atom is in a severely distorted tetrahedral environment with four equal Cu-Cl bond lengths of 2.398(2) A. In comparison, the Cu-Cl bonds in [{(C5H10NO)2Mo(p.2-S)2Cu(u2-CI)}2] are 2.375(1) and 2.341(2) A in length, while the Cu centers in this complex are much closer to a tetrahedral geometry than those in 74 . 1 6 7 The ruthenium centers in complex 74 are in a slightly twisted trigonal-antiprismatic coordination, with Cl(l), P(3), and P(2) forming one face and Cl(2), P(4) and P(l) forming the opposite face. The Ru-Cl bonds in 74 (2.469(2), 2.470(2) A ) are slightly longer than the corresponding bonds in 59 (2.440(2), 2.451(13) A), while the C l - R u - C l angle is smaller in 74 (82.03(7)) than in 59 (84.1(5)).168 The [CuCl2]~ counterion of 74 is linear with a Cu-Cl bond length of 2.101(4) A, similar to the Cu-Cl bond length in [(w-Bu)4N][CuCl2] (2.107(1) A ) . 1 6 9 Figure 40. ORTEP diagram of the solid-state molecular structure of 74-solvent. The solvent molecules and the phenyl groups, except the ipso carbon atoms, are omitted for clarity. The thermal ellipsoids are depicted at 30% probability. 147 Table 16. Selected Bond Lengths in 74-solvent (A) R u - P ( l ) 2.303(2) Cu(l) -Cl(2) 2.398(2) P(2)- C(36) 1.828(6) Ru-P(2) 2.363(2) Cu(l) - C1(2A) 2.398(2) P(2)- C(46) 1.820(4) Ru-P(3) 2.312(2) Cu(2) - Cl(3) 2.101(4) P(3)- C(2) 1.856(8) Ru - P(4) 2.351(2) Cu(2) -C1(3A) 2.101(4) P(3)- C(56) 1.832(5) R u - C l ( l ) 2.470(2) P ( l ) - C(l) 1.856(9) P(3)- C(66) 1.844(5) Ru-Cl(2) 2.469(2) P ( l ) - C(16) 1.842(5) P(4)- C(2) 1.822(9) C u ( l ) - C l ( l ) 2.398(2) P ( l ) - C(26) 1.845(5) P(4)- C(76) 1.80(1) Cu(l)-C1(1A) 2.398(2) m- C(l) 1.853(8) P(4)- C(86) 1.820(5) Table 17. Selected Bond Angles in 74-solvent (deg) P(l) - R u - P(2) 72.01(8) Cl ( l ) -Cu( l ) -C1(2A) 113.79(8) P(l) - R u - P(3) 95.16(8) C1(1A) - Cu(l) - C1(2A) 85.04(7) P(l) - R u - P(4) 100.38(8) . Cl(2)-Cu(l)-C1(2A) 134.9(1) P(2) - R u - P(3) 102.30(8) Cl(3)-Cu(2)-C1(3A) 180.0 P(2) - R u - P(4) 170.41(8) C ( l ) - P ( l ) - R u 96.6(2) P(3) - R u - P(4) 72.09(8) C ( 1 6 ) - P ( l ) - R u 126.1(2) P(l) - R u - Cl(l) 164.71(8) C ( 2 6 ) - P ( l ) - R u 119.3(2) P(l) - R u - Cl(2) 93.61(8) C ( l ) - P ( 2 ) - R u 94.6(3) P(2) - R u - Cl(l) 93.45(7) C(36)-P(2)-Ru 123.2(2) P(2) - R u - Cl(2) 92.34(8) C(46)-P(2)-Ru 123.1(2) 148 P ( 3 ) - R u - C l ( l ) 92.51(7) C(2)- P(3)-Ru 95.0(3) P(3) -Ru-Cl (2) 164.69(8) C(56) - P ( 3 ) - R u 127.1(2) P ( 4 ) - R u - C l ( l ) 94.56(7) C(66) - P ( 3 ) - R u 120.4(2) P(4) -Ru-Cl (2) 93.99(8) C(2)- P(4)-Ru 94.7(3) C l ( l ) - R u - C l ( 2 ) 82.03(7) C(76) - P ( 4 ) - R u 122.8(4) C l ( l ) - C u ( l ) - C l ( 2 ) 85.03(7) C(86) - P ( 4 ) - R u 121.8(2) C1(1A)-Cu(l)-Cl(2) 113.80(8) Cu(l) - C l ( l ) - R u 96.42(7) Cl ( l ) -Cu( l ) -C1(1A) 131.2(1) Cu(l) - C l ( 2 ) - R u 96.44(8) 6.3.3 NMR Studies When pure crystalline 74 is dissolved in CD2CI2, the 3 1 P N M R spectrum obtained suggests that several complexes are present in solution (Figure 41). Complex 74 is poorly soluble in most other weak donor solvents; however, the 3 1 P N M R spectrum of 74 in mixtures of CD 2 C1 2 and acetone or THF is similar to that in pure CD 2C1 2 . Complex 74 reacts with stronger donor solvents such as C H 3 C N or CH3NO2. In CD2CL2, the 3 1 P N M R spectrum consists of two major triplets (8 -0.5 and -28.2) and two smaller triplets (8 -0.4 and -25.3). In addition, there are broad peaks at 8 0 and -25 which overlap the triplets. The four sharp triplets are assigned to two cis-substituted ruthenium complexes present in an approximately 4:1 ratio when 74 dissolves. Neither of these complexes is 59, and the small chemical shift differences between the sets of peaks indicate that they are structurally similar. These two major compounds are proposed to be 74 and the intermediate 75 (Scheme 20). 149 tfbt/i+toy* D******** I — i — j — i — i — i — r -" T — i — i — i — i I i i— i— i—p -10 -20 -30 ppm Figure 41. 3 1P N M R spectrum of 74 in CD 2C1 2 . The molar conductivity of an acetone solution of 74 is 207 Q _ 1 cm 2 mol - 1 , indicating that the major species in solution is the 1:1 electrolyte 74. The broad peaks in the 3 1P N M R spectrum may be due to species 76, which would result if one of the Cu-Cl bonds in 75 breaks. The broad peaks can result from inequivalency of the phosphorus atoms in 76 or from chemical exchange with 59. Since 74 is a strong electrolyte, the relative concentrations of the species in Scheme 20 are expected to be independent of the amount of 74 added. A 10-fold increase in the amount of 74 dissolved does not change the ratio of 74 to the sum of 75 and 76 (integrated together in the P N M R spectrum), consistent with the equilibria in Scheme 20. 150 Addition of excess [(n-Bu)4N]Cl or 2 M aq H C l to a solution of 74 in C H 2 C 1 2 results in complete conversion of all species present to 59. The added C P presumably complexes the copper as C u C l 2 _ , leaving 59 as the only mtherrium-containing species. Subsequent addition of excess CuCI to a solution of 74 which had been treated with excess [(«-Bu)4N]Cl results in the disappearance of peaks due to 59 and the reappearance of sharp peaks due to 74 and 75 as well as the broad peaks 8 0 and -25 in the 3 1 P N M R spectrum. These experiments demonstrate that 59 and 74 - 76 may be interconverted in the presence of CuCI and Cl", and that the relative concentrations of these species in solution depend on the amount of CuCI and C l - present. Scheme 20 / ^ P P h 2 P h 2 P ^ l ~ i + P h 2 P ^ J ^ C l _ . C k J . P h 2 P „ „ . -1 .Ru. , C u . ^Ru^ 1 C u C l 2 P h 2 P ^ | " C l " C l I Ph 2 P \ ^ P P h 2 P h 2 P ^ / 74 Ph,P- • C L l 2 l ^ R u ' ^ ; C u - C l P h 2 P ^ | ^ C l U P P h 2 75 P P h 2 P h 2 P ^ J ^ C l z Ru P h 2 P ^ | ^ C U U- P P h 2 <X 76 C l The progress of the reaction of 19 with one equivalent of CuCI in C D 2 C 1 2 is followed by P N M R . Since CuCI is poorly soluble, only a small amount dissolves initially, and the reaction conditions are identical to the catalytic experiments. Upon addition of CuCI into the tube, the singlet in the 3 1 P N M R at 8 -7.7 due to 19 broadens. In the early stages of the 151 reaction, sharp resonances due to 59 begin to appear, and eventually the resonances due to 74 and 75 also become apparent. When an identical experiment is carried out with 59 and stoichiometric CuCI, reaction is very rapid in the N M R tube with complete disappearance of 59 and formation of the equilibrium mixture of74 - 76 within 5 min. . Complex 19 isomerizes to 59 when a small amount of 74 is added to the solution. A solution of 19 and 4 mol % 74 in CH 2C1 2 is stirred at room temperature for 24 h, affording a solution which contained 90% 59 and 10% 19 by 3 1 P NMR. Addition of a small amount of complex 74 to the isomerization reaction is analogous to adding catalytic CuCI. Intermediates 75 and 76 are formed directly in solution from 74, introducing coordinated CuCI, which acts as a catalyst for the isomerization. The UV-vis absorption spectrum obtained when 74 is dissolved in CH 2 C1 2 is similar to that of 59 (Figure 42). It is likely that the bands in the spectrum of 74 are due to the same transitions observed for 59, with slight distortions due to the change in the ligand field around ruthenium upon coordination of the copper center. Since both 74 and 75 have very similar ligand environments at the ruthenium center, it is reasonable that both complexes exhibit similar visible transitions to 59. Solutions of 74 are stable under ambient laboratory light, consistent with the cis chloride ligands being locked in place on the ruthenium center by the copper, preventing photochemical isomerization. 152 5000 r-4000 -3000 -"9 • '§ 2000 -w 1000 -0 I 1 • 1 i I i I •' • " " ^ ~ - l 300 350 400 450 500 550 Wavelength (nm) Figure 42. Absorption spectra of 59 (...) and 74 (—) in CH2CI2. 6.4 Discussion The following mechanism is proposed for the CuCl-catalyzed isomerization of 19 to 59. As the CuCl dissolves, it coordinates to one of the chloride ligands in 19, yielding halide-bridged intermediate 77 (Scheme 21). Complex 77 is not observed in the 3 'P N M R spectrum, but a broadening of the singlet at 8 -7.7 is observed. This broaderiing is specific to this peak, and is not due to nonspecific broadening as would be caused by paramagnetic Cu 1 1 impurities. It is possible that the broadening is caused by exchange between 77 and 19. The ruthenium center in intermediate 77 is still trans substituted. Coordination of the CuCl is expected to 153 weaken the Ru-Cl bond, facilitating isomerization to the more stable cis orientation at the ruthenium center. The Ru-Cl bond is lengthened in 74 compared to 59, consistent with a weakening of the Ru-Cl bond upon coordination of Cu1. Thermodynamically, RuCl2(dppm)2 prefers the cis orientation. Meyer argued that this is because 59 is stabilized by rc-donation from the Cl through the Ru to a trans P, whereas this favorable interaction is not possible in the trans isomer. 1 2 0 Scheme 21 Ph 2 pf>Ph 2 + C u C 1 C l - R u - C l . -Ph2P^Vph2 - CuCl 19 p P P h 2 P h 2 P x ^ C l + C u ^ 1 ^ P ' J ^ C l - C u C l U P P h 2 59 P h 2 P ^ P P h 2 C l - R u - C l - C u - C l P h 2 P ^ P P h 2 77 isomerization / ^ P P h 2 P h ^ R u ^ C u - C l P h 2 P " | ^ C f U P P « 2 75 It is proposed here that 77 isomerizes to 75, in which the ruthenium center has a cis geometry, and the CuCl is coordinated to both chloride ligands. The intermediate 75 can then lose CuCl, presumably via 76, to yield 59. This pathway is likely under catalytic conditions when there is a large excess of ruthenium species relative to CuCl. Under these conditions, any 75 which forms probably reacts with 19, yielding 59 and 77. The equilibrium between 77 and 75 is expected to favor 75 due to both the favorable geometry of the ruthenium center in this complex, and the stabilizing effect of coordination of the copper with the two bridging 154 chlorides. The effect of the CuCI is therefore to weaken the Ru-Cl bond in 77 and to stabilize intermediate 75. When stoichiometric amounts of CuCI are available, sufficient copper is present to complex all the available ruthenium species. Since the reaction between 59 and CuCI is rapid, it is reasonable that 74 and 75 are the final products observed in the reaction between 19 and stoichiometric CuCI. Because CuCI is poorly soluble, 59 is observed first in this reaction followed by 74 and 75 as 19 is consumed, and 59 is formed. The coppeifl) halide catalyzed isomerization of 19 to 59 provides an alternative, low-temperature route to the cis complex. Although a previous report indicated that complexes structurally related to 19 such as i/ra«5-[RuCl2((o-C6H4(PMePh)2)2] can be isomerized to the corresponding cis isomer using excess triemylaluminum, this procedure involves air-sensitive reagents and proceeds via hydrido intermediates which must be hydrolyzed to regenerate the product.170 The method described here may allow for the convenient preparation of cis isomers of similar complexes which are difficult to isomerize thermally. The results in this Chapter are important for understanding copper(I) halide catalyzed reactions involving 19 and 59. It is likely that coupling reactions of ruthenium halide complexes with organostannane involve intermediates such as 74 - 77 in which the copper coordinates to the ruthenium halide complex. When the starting material is 19, copper-catalyzed isomerization may generate cis-substituted intermediates in which the chloride ligands are more labile, facilitating reaction with the organostannane. In addition, since 74 is a soluble complex, it may be useful to use this directly as a reagent with organostannanes to make bisacetylide complexes and polymers. 155 6.5 Conclusions Copper(I) halides are studied as catalysts for the trans - cis isomerization of RuCi2(dppm)2. When RuCi2(dppm)2 reacts with excess CuCI, a halide-bridged heterotrimetallic complex is isolated. An N M R study on the reaction of RuCi2(dppm)2 with CuCI provides some insight into the isomerization mechanism. Chapter 7 Suggestions for Future Work 156 The work in this thesis opens many avenues for future exploration of metal-containing oligomers and polymers. Some of these are summarized in this Chapter. Ruthenium oligothienylacetylide polymers may have significant conductivities when doped because oxidation of the ruthenium(II) centers in 62c and 63c results in very strong low-energy absorptions in the near-IR region. Measurement of the conductivities of the electropolymerized films of those materials would be interesting. Chemical coupling may also be used to produce such polymers. Preliminary results have showed that coupling of 58 with 78 gives an insoluble brown-red solid, which is presumably the polymer 79 (Scheme 22). Further characterization of this solid and preparation of analogous polymers which are soluble would be useful. The solubility of this polymer is expected to increase as alkyl groups are incorporated into either the oligothienyl groups or phosphine ligands, but care must be taken to consider the steric effects of these groups, which may reduce the efficiency of the coupling. Scheme 22 r~\ / - A Electropolymerization of 63b and 63c results in the deposition of electrochromic films on the electrodes, but the nature of the films is not clear because of the poor stability of 157 oxidized 63b and 63c in solution at room temperature. The stability of these species could be enhanced through the use of more electron-rich phosphine ligands on the metal and oligothienyl groups with a lower oxidation potential. The latter could be achieved by using oligothienyl groups with a longer conjugation length or with electron-donating substituents. Preliminary conductivity measurements on films prepared by electropolymerization of 50 and 73, and electrodimerization of 68c and 68e are attempted on interdigitated microelectrodes with an interelectrode spacing of 5 um. This method, developed by Wrighton and coworkers, allows the conductivity to be measured as a function of applied potential.171" 1 7 3 This method requires that the film must fill the interelectrode gap. In all cases, it is impossible to obtain sufficiently thick films to bridge the gap between the electrodes, and it is clear that interdigitated electrodes with a smaller gap (around 1 um) are needed. Murray and coworkers have measured the conductivity of polymer films using a sandwich-electrode arrangement in which a polymer film is deposited on a Pt surface, with a gold film evaporated on top of the polymer film.174 A modified method is attempted to obtain conductivity measurements of films of the dimers 71 and 72. In these experiments, a Pt electrode coated with a film of the dimer is dipped into a mercury pool, and the resistance between the Pt and mercury electrodes measured. Using this method the conductivity of the neutral film of the dimer 71 is obtained (3 x 10~8 S/cm), but consistent results can not be obtained for the oxidized film. It is impossible to measure the conductivity of films of 72 in this way because they crack when dry resulting in shorts between the electrodes. Further investigations are needed to obtain satisfactory and useful data from those measurements. 158 References (1) Feynman, R. Eng. Sci. 1960,23,22. (2) Aviram, A. ; Ratner, M . Molecular Electronics: Science and Technology; The New York Academy of Sciences:, 1998; Vol. 852. (3) Reed, M . A. Proc. IEEE 1999,87,652-658. (4) Reinerth, W. A.; Jones, L.; Burgin, T. P.; Zhou, C. W.; Muller, C. J.; Deshpande, M . R.; Reed, M . A.; Tour, J. M . Nanotechnology 1998,9,246-250. (5) Munn, R. W. Chem. Br. 1984, 518-524. (6) Mirkin, C. A.; Ratner, M . A. Annu. Rev. Phys. Chem. 1992,43,719-754. (7) Kelly, T. R.; Silva, H. D.; Silva, R. A. Nature 1999,401,150-152. (8) Koumura, N . ; Zijlstra, R. W.; Delden, R. A. V.; Harada, N . ; Feringa, B. L. Nature 1999,401,152-154. (9) Shirakawa, H.; Louis, E. J.; MacDiarmid, A. G.; Chiang, C. K.; Heeger, A. J. J. Chem. Soc, Chem. Commun. 1977,578-580. (10) Skotheim, T. A . Handbook of Conducting Polymers; 1st ed.; Marcel Dekker: New York and Basel, 1986; Vol. 1-2. (11) Skotheim, T. A.; Elsenbaumer, R. L.; Reynolds, J. R. Handbook of Conducting Polymers; 2nd ed.; Marcel Dekker: New York, 1998. (12) Naarmann, H.; Theophilou, N . Synth. Met. 1987,22,1-8. (13) Tsukamoto, J.; Takahashi, A.; Kawasaki, K. Jpn. J. Appl. Phys. 1990,29,125-130. (14) Pham, C. V.; Burkhardt, A.; Shabana, R.; Cunningham, D. D.; Mark, H. B. J.; Zimmer, H. Phosphorus, Sulfur, and Silicon 1989,46,153-168. 159 (15) Salzner, U. ; Lagowski, J. B.; Pickup, P. G.; Poirier, R. A . Synth. Met. 1998, 96, 177-189. (16) Hotta, S.; Soga, M . ; Sonoda, N . Synth. Met. 1988,26,267-269. (17) Kossmehl, G.; Chatzitheodorou, G. Mol. Cryst. Liq. Cryst. 1982,53,291-296. (18) Berlin, A.; Pagani, G. A.; Sannicolo, F. J. Chem. Soc, Chem. Commun. 1986, 1663-1664. (19) Tourillon, G.; Gamier, J. A . J. Electroanal. Chem. 1984,161, 51-58. (20) Waltman, R. J.; Bargon, J.; Diaz, A. F. J. Phys. Chem. 1983,1459-1463. (21) Diaz, A . F.; Crowley, J.; Bargon, J.; Gardini, G. P.; Torrance, J. B. J. Electroanal. Chem. 1981,121, 355-361. (22) Martinez, F.; Voelkel, R.; Naegele, D.; Naarmann, H. Mol. Cryst. Liq. Cryst. 1989, 167,227-232. (23) Bauerle, P.; Segelbacher, U . ; Gaudl, K. -U. ; Huttenlocher, D.; Mehring, M . Angew. Chem. Int. Ed. Engl. 1993,32,76-78. (24) Fichou, D.; Demanze, F.; Horowitz, G.; Hajlaoui, R.; Constant, M . ; Ganier, F. Synth. Met. 1997,85,1309-1312. (25) Zotti, G.; Schiavon, G.; Berlin, A.; Pagani, G. Chem. Mater. 1993,5,430-436. (26) Kagan, J.; Arora, S. K. Heterocycles 1983,20,1937-1940. (27) Horowitz, G.; Hajlaoui, R.; Kouki, F. Eur. Phys. J. Appl. Phys. 1998, / , 361-367. (28) Horowitz, G. Adv. Mater. 1998,10,365-377. (29) Torsi, L.; Dodabalapur, A.; Rothberg, L. J.; Fung, A . W. P.; Katz, H . E. Phys. Rev. B 1998,57,2271-2275. (30) Fichou, D.; Horowitz, G.; Xu, B.; Gamier, F. Synth. Met. 1990,39,243-260. 160 Hush, N . S. Prog. Inorg. Chem. 1967,8,391-444. Nelson, S. F.; Tran, H. Q.; Nagy, M . A. J. Am. Chem. Soc. 1998,120,298-304. Lambert, C ; N611, G. J. Am. Chem. Soc. 1999,121,8434-8442. Desjardins, P.; Yap, G. P. A. ; Crutchley, R. J. Inorg. Chem. 1999,38,5901-5905. Creutz, C.; Newton, M . D.; Sutin, N . J. Photochem. Photobiol. A: Chem. 1994,82,47-59. Evans, C. E. B.; Ducharme, D.; Naklicki, M . L.; Crutchley, R. J. Inorg. Chem. 1995, 34,1350-154. Creutz, C. Prog. Inorg. Chem. 1983,30,1-73. Crutchley, R. J. Adv. Inorg. Chem. 1994,41,273-325. Creutz, C ; Taube, H. J. Am. Chem. Soc. 1973,95,1086-1094. Richardson, D. E.; Taube, H. Coord. Chem. Rev. 1984, 60,107-128. Rosa, R. D. L.; Chang, P. J.; Salaymeh, F.; Curtis, J. C. Inorg. Chem. 1985, 24, 4229-4231. Robin, M . B.; Day, P. Adv. Inorg. Chem. Radiochem. 1967,10,247-422. Levanda, G ; Bechgaard, K.; Cowan, D. O. J. Org. Chem. 1976,41,2700-2704. Morrison, W. H.; Krogsrud, J., S.; Hendrickson, D. S. Inorg. Chem. 1973,12, 1998-2004. Ribou, A . - C ; Launay, J.-P.; Sachtleben, M . L.; L i , H.; Spangler, C. W. Inorg. Chem. 1996,35,3735-3740. Coat, F.; Lapinte, C. Organometallics 1996,15, A11A19. Belanzoni, P.; Re, N . ; Sgamellotti, A. ; Floriani, C. J. Chem. Soc, Dalton Trans. 1998, 1825-1835. 161 (48) Brady, M . ; Weng, W.; Zhou, Y. ; Seyler, J. W.; Amoroso, A . J.; Arif, A . M . ; Bohme, M . ; Frenking, G.; A., Gladysz. J. A . J. Am. Chem. Soc. 1997,119, 775-788. (49) Guillemot, M . ; Toupet, L.; Lapinte, C. Organometallics 1998, / 7,1928-1930. (50) Bunz, U . H. Angew. Chem. Int. Ed. Engl. 1996,35,969-971. (51) Bartik, T.; Bartik, B.; Brady, M . ; Dembinski, R.; Gladysz, J. A . Angew. Chem. Int. Ed. Engl. 1996, 35,414-417. (52) Colbert, M . C. B.; Lewis, J.; Long, N . J.; Raithby, P. R.; Younus, M . ; White, A . J. P.; Williams, D., J.; Payne, N . N . ; Yellowlees, L.; Beljonne, D.; Chawdhury, N . ; Friend, R. H. Organometallics 1998,17,3034-3043. (53) Jones, N . D.; Wolf, M . O.; Giaquinta, D. M . Organometallics 1997,16,1352-1354. (54) Colbert, M . C. B.; Lewis, J.; Long, N . J.; Raithby, P. R.; White, A . J. P.; Williams, D., J. J. Chem. Soc, Dalton Trans. 1997,99-104. (55) Osella, D.; Gambino, O.; Nervi, C ; Ravera, M . ; Russo, M . V. ; Infante, G. Inorg. Chim. Acta 1994,225,35-40. (56) Sato, M . ; Hayashi, Y. ; Kumakura, S.; Shimizu, N . ; Katada, M . ; Kawata, S. Organometallics 1996,15,721-728. (57) Hanack, M . ; Lang, M . Adv. Mater. 1994, 6,819-833. (58) Takahashi, S.; Kariya, M . ; Yatake, T.; Sonogshira, K.; Hagihara, N . Macromolecules 1978, / / , 1063-1066. (59) Sonogashira, K.; Kataoka, S.; Takahashi, S.; Hagihara, N . J. Organomet. Chem. 1978, 750,319-327. (60) Takahashi, S.; Ohyama, Y. ; Murata, E.; Sonogashira, K.; Hagihara, N . J. Polym. Sci., Polym. Chem. Ed. 1980,18,349-353. 162 (61) Takahashi, S.; Murata, E.; Sonogashira, K.; Hagihara, N . J. Polym. Sci., Polym. Chem. Ed. 1980,18,661-669. (62) Hagihara, N . ; Sonogashira, K.; Takahashi, S. Adv. Polym. Sci. 1981,41,149-179. (63) Takahashi, S.; Morimoto, H.; Murata, E.; Kataoka, S.; Sonogashira, K.; Hagihara, N . J. Polym. Sci., Polym. Chem. Ed. 1982,20, 565-573. (64) Johnson, B. F. G.; Kakkar, A . K.; Khan, M . S.; Lewis, J. J. Organomet. Chem. 1991, 409, C12-C14. (65) Khan, M . S.; Schwartz, N . A.; Kakkar, A. K.; Lin, B.; Raithby, P. R.; Lewis, J. Z. Anorg. Allg. Chem. 1992, o7<5,121-124. (66) Khan, M . S.; Davies, S. J.; Kakkar, A . K.; Schwartz, D.; Lin, B.; Johnson, B. F. G.; Lewis, J. J. Organomet. Chem. 1992,424, ill-91. (67) Lewis, J.; Khan, M . S.; Kakkar, A. K.; Johnson, B. F. G. J. Organomet. Chem. 1992, 425,165-176. (68) Sun, Y. ; Taylor, N . J.; Carty, A. J. Organometallics 1992, / / , 4293-4300. (69) Gamier, F.; Yassar, A. , U.S. Patent 5272238,1993 (70) Frapper, G.; Kertesz, M . Inorg. Chem. 1993,32,732-740. (71) Faulkner, C. W.; Ingham, S. L.; Khan, M . S.; Lewis, J.; Long, N . J.; Raithby, P. R. J. Organomet. Chem. 1994,482,139-145. (72) Markwell, R., D.; Butler, I. S.; Kakkar, A . K.; Khan, M . S.; Al-Zakwani, Z. H.; Lewis, J. Organometallics 1996,15,2331-2337. (73) Irwin, M . J.; Vital , J. J.; Puddephatt, R. J. Organometallics 1997,16,3541-3547. (74) Russo, M . V.; Furlani, A.; Altamura, P.; Fratoddi, I.; Polzonetti, G. Polymer 1997, 38, 3677-3690. (75) Chawdhury, N . ; Kohler, A.; Friend, R. H.; Wong, W. Y . ; Lewis, J.; Younus, M . ; Raithby, P. R.; Corcoran, T. C ; Al-Mandhary, M . R. A.; Khan, M . S. J. Chem. Phys. 1999,10,4963-4970. (76) Kingsborough, R. P.; Swager, T. M . Prog. Inorg. Chem. 1999,48,123-231. (77) Porter, P. L.; Guha, S.; Kang, K ; Frazier, C. C. Polymer 1991,32,1756-1760. (78) Dray, A . E.; Wittmann, F.; Friend, R. H.; Donald, A . M . ; Khan, M . S.; Lewis, J.; Johnson, B. F. G. Synth. Met. 1991,41-43,871-874. (79) Messier, J.; Kajzar, F.; Prasad, ^.Organic Molecules for Nonlinear Optics and Photonics; Kluwer Academic Publishers: Dordrecht; Boston, 1991. (80) Atherton, Z.; Faulkner, C. W.; Ingham, S. L.; Kakkar, A . K ; Khan, M . S.; Lewis, J.; Long, N . J.; Raithby, P. R. J. Organomet. Chem. 1993,462,265-270. (81) Chawdhury, N . ; Kohlor, A.; Friend, R. H.; Younus, M . ; Long, N . J.; Raithby, P. R.; Lewis, J. Macromolecules 1998,31, 722-727. (82) Davies, S. J.; Johnson, B. F. G.; Lewis, J.; Raithby, P. R. J. Organomet. Chem. 1991, 414, C51-C53. (83) Khan, M . S.; Kakkar, A . K.; Ingham, S. L.; Raithby, Paul R.; Lewis, J. J. Organomet. Chem. 1994,472,247-255. (84) Khan, M . S.; Kakkar, A. K.; Long, N . J.; Lewis, J.; Raithby, P. R.; Nguyen, P.; Marder, F.; Wittmann, F.; Friend, R. H. J. Mater. Chem. 1994,4,1227-1232. (85) Lewis, J.; Long, N . J.; Raithby, P. R.; Shields, G. P.; Wong, W.-Y.; Younus, M . J. Chem. Soc, Dalton Trans. 1997,4283-4288. (86) Lavastre, O.; Even, M . ; Dixneuf, P. H.; Pacreau, A. ; Vairon, J. P. Organometallics 1996,15,1530-1531. 164 (87) Bilow, N . ; Landis, A. ; Rosenberg, H. / . Polym. Sci., Polym. Chem. Ed. 1969, 7,2719-2736. (88) Cowan, D. O.; Park, J.; Pittma, C. U . J.; Sasaki, Y. ; Mukherjee, T. K.; Diamond, N . A . J. Am. Chem. Soc. 1972,94, 5110-5112. (89) Pittman, C. U . J.; Sasaki, Y. ; Mukherjee, T. K. Chem. Lett. 1975,383-386. (90) Neuse, E. W.; Bednarik, L. Macromolecules 1979,12,187-195. (91) Yamamoto, T.; Sanechika, K.-I.; Yamamoto, A. ; Katada, M . ; Motoyama, I.; Sano, H. Inorg. Chim. Acta 1983, 73, 75-82. (92) Neuse, E. W.; Trifan, D. S. J. Am. Chem. Soc. 1963,1952-1958. (93) Neuse, E. W.; Khan, F. B. D. Macromolecules 1986,19,269-272. (94) Nelson, J. M . ; Rengel, H.; Manners, I. J. Am. Chem. Soc. 1993,115,7035-7036. (95) Barlow, S.; O'Hare, D. Chem. Rev. 1997,97, 637-669. (96) Buretea, M . ; Tilley, T. D. Organometallics 1997,16,1507-1510. (97) Arnold, R.; Matchett, S. A.; Rosenblum, M . Organometallics 1988, 7,2261-2266. (98) Nugent, H. M . ; Rosenblum, M . / . Am. Chem. Soc. 1993,115,3848-3849. (99) Funaki, H.; Aramaki, K.; Nishihara, H. Synth. Met. 1995, 74, 59-64. (100) Yamamoto, T.; Morikita, T.; Maruyama, T.; Kubota, k.; Katada, M . Macromolecules 1997,30,5390-5396. (101) Knapp, R.; Velten, U. ; Rehahn, M . Polymer 1998,39,5827-5838. (102) Morikita, T.; Maruyama, T.; T., Y. ; Kubota, K ; Katada, M . Inorg. Chim. Acta 1998, 269,310-312. (103) Foucher, D. A.; Honeyman, C. H.; Nelson, C. H.; Tang, B. Z.; Manners, I. Angew. Chem. Int. Ed. Engl. 1993,32,1702-1711. 165 (104) Foucher, D. A. ; Tang, B. Z.; Manners, I. J. Am. Chem. Soc. 1992,114,6246-6248. (105) Brandt, P. F.; Rauchuss, T. B. J. Am. Chem. Soc. 1992,114,1926-1927. (106) Manners, I. Adv. Mater. 1994,6,68-71. (107) Manners, I. Adv. Organomet. Chem. 1995,37,131-168. (108) Hirao, T.; Kurashina, M . ; Aramaki, K.; Nishihara, H. J. Chem. Soc, Dalton Trans. 1996,2929-2933. (109) Foucher, D. A. ; Ziembinski, R.; Tang, B. Z.; Macdonald, P. M . ; Massey, J.; Jaeger, R.; Vancso, G. J.; Manners, I. Macromolecules 1993,26,2878-2884. (110) Hmyene, M . ; Yassar, A.; Escorne, M . ; Percheron-Guegan, A.; Gamier, F. Adv. Mater. 1994,6,564-568. (111) Ingham, S. L.; Khan, M . S.; Lewis, J.; Long, N . J.; Raithby, P. R. J. Organomet. Chem. 1994,470,153-159. (112) Wrighton, M . S. Science 1986,231,32-37. (113) Segawa, H.; Wu, F.-P.; Nakayama, N . ; Maruyama, H.; Sagisaka, S.; Higuchi, N . ; Fujitsuka, M . ; Shimidzu, T. Synth. Met. 1995, 71,2151-2154. (114) Zhu, S. S.; Kingsborough, R. P.; Swager, T. M . J. Mater. Chem. 1999,9,2123-2131. (115) Kingsborough, R. P.; Swager, T. M . Adv. Mater. 1998,10,1100-1104. (116) Kingsborough, R. P.; Swager, T, M . J. Am. Chem. Soc. 1999,121,8825-8834. (117) Higgins, S. J.; Jones, C. L.; Francis, S. M . Synth. Met. 1999,98,211-214. (118) Manna, J.; John, K. D.; Hopkins, M . D. Adv. Organomet. Chem. 1995,38,79-154. (119) Chart, J.; Hayter, R. G. J. Chem. Soc. 1961,896-904. (120) Sullivan, B. P.; Meyer, T. J. Inorg. Chem. 1982,21,1037-1040. (121) Smart, J. C ; Pinsky, B. L. J. Am. Chem. Soc. 1980,102,1009-1014. 166 (122) Polin, J.; S'chottenberger, H. Org. Synth. 1996, 73,262-269. (123) Sheldrick, G. M . , Siemens XRD, Madison, WI. (124) Warn, V . W.-W.; Choi, S. W.-K.; Chan, C.-L.; Cheung, K . -K. J. Chem. Soc, Chem. Commun. 1996,2067-2068. (125) Yamazaki, S.; Deeming, A . J. J. Chem. Soc, Dalton Trans. 1993,3051-3057. (126) Abu Salah, O. M . ; Bruce, M . I.; Churchill, M . R.; Bezman, S. A. J. Chem. Soc, Chem. Commun. 1972, 858-859. (127) Abu Salah, O. M . ; Bruce, M . I.; Redhouse, A . D. J. Chem. Soc, Chem. Commun. 1974,855-856. (128) Churchill, M . R.; Bezman, S. A. Inorg. Chem. 1974,13,1418-1426. (129) Janssen, M . D.; Herres, M . ; Zsolnai, L.; Grove, D. M . ; Spek, A . L.; Lang, H.; van Koten, G. Organometallics 1995,14,1098-1100. (130) Janssen, M . D.; Herres, M . ; Zsolnai, L.; Spek, A. L.; Grove, D. M . ; Lang, H.; van Koten, G. Inorg. Chem. 1996,35,2476-2483. (131) Janssen, M . D.; Herres, M . ; Spek, A . L.; Grove, D. M . ; Lang, H. ; van Koten, G. J. Chem. Soc, Chem. Commun. 1995,925-926. (132) Janssen, M . D.; Kohler, K.; Herres, M . ; Dedieu, A.; Smeets, W. J. J.; Spek, A . L.; Grove, D. M . ; Lang, H.; van Koten, G. J. Am. Chem. Soc. 1996,118,4817-4829. (133) Colbert, M . C. B.; Ingham, S. L.; Lewis, J.; Long, N . J.; Raithby, P. R. J. Chem. Soc, Dalton Trans. 1994,2215-2216. (134) Evans, D. F. J. Chem. Soc. 1959,2003-2005. (135) Geoffroy, G. L. ; Wrighton, M . S. Organometllic Photochemistry; Academic Press: New York, 1979. 167 (136) Sato, M . ; Shintate, H.; Kawata, Y. ; Sekino, M . ; Katada, M . ; Kawata, S. Organometallics 1994,13,1956-1962. (137) Lever, A . B. P. Inorganic Electronic Spectroscopy; 2nd ed.; Elsevier: New York, 1984. (138) Creutz, C. Inorg. Chem. 1978,17,3723-3725. (139) Kober, E. M . ; Goldsby, K. A.; Narayana, D. N . S.; Meyer, T. J. J. Am. Chem. Soc. 1983,105,4303-4309. (140) Dowling, N . ; Henry, P. M . ; Lewis, N . A.; Taube, H. Inorg. Chem. 1981, 20, 2345-2348. (141) Hong, Y. ; Miller, L. L. Chem. Mater. 1995, 7,1999-2000. (142) Itatani, H.; Bailar, J. C. J. J. Am. Oil Chem. Soc. 1967,44,147-151. (143) Van Hecke, G. R.; Horrocks, W. D. J. Inorg. Chem. 1966,5,1968-1974. (144) Wu, R.; Schumm, J. B.; Pearson, D. L.; Tour, J. M . J. Org. Chem. 1996, 61, 6906-6921. (145) Bauerle, P.; Wurthner, F.; Gotz, G.; Effenberger, F. Synthesis 1993,1099-1103. (146) MacEachern, A. ; Soucy, C ; Leitch, L.; Arnason, J. T.; Morand, P. Tetrahedron 1988, 44,2403-2412. (147) Jones, K.; Lappert, M . F. J. Chem. Soc. 1965,1944-1951. (148) Viola, E.; Sterzo, C. L.; Crescenzi, R.; Frachey, G. J. Organomet. Chem. 1995, 493, C9-C13. (149) These complexes were prepared previously but were not characterized by CV. see: Hodge, A . J.; Ingham, S. L.; Kakkar, A . K ; Klan, M . D.; Lewis, J.; Long, N . J.; Parker, D. G.; Raithby, P. R. J. Organomet. Chem. 1995,488,205-210. 168 r (150) Rossiter, B. W.; Hamilton, J. F. Physical Methods of Chemistry; Wiley: New York, 1986; Vol. 2. (151) This complex was prepared previously but was not characterized by C V and UV-vis spectroscopy, see: Touchard, D.; Haquette, P.; Pirio, N . ; Toupet, L.; Dixneuf, P. H. Organometallics 1993,12,3132-3139. (152) Miller, L. L.; Yu, Y . J. Org. Chem. 1995, 60, 6813-6819. (153) Sohn, Y . S.; Hendrickson, D. N . ; Gray, H. B. J. Am. Chem. Soc. 1971,93,3603-3612. (154) Sotzing, G. A.; Reynolds, J. R.; Steel, P. J. Chem. Mater. 1996,8, 882-889. (155) Curtis, J. C ; Sullivan, B. P.; Meyer, T. J. Inorg. Chem. 1983,22,224-236. (156) Sato, M . ; Shiraogawa, M . CACSForum (Japanese) 1993,13, 30-32. (157) Wolfgang, H.; Andrejewski, D., German Patent DE3703771,1988. (158) Graf, D. D.; Mann, K. R. Inorg. Chem. 1997,36,150-157. (159) Hotta, S.; Hosaka, T.; Shimotsuma, W. Synth. Met. 1983, 6,69-71. (160) Patil, A . O.; Heeger, A. J.; Wudl, F. Chem. Rev. 1988,88,183-200. (161) Liebeskind, L. S.; Fengl, R. W. J. Org. Chem. 1990,55,6359-5364. (162) Stille, J. K. Angew. Chem. Int. Ed. Engl. 1986,25,508-524. (163) Baker, R. W.; Braithwaite, M . J.; Nyholm, R. S. J. Chem. Soc, Dalton Trans. 1972, 1924-1928. (164) Briant, C. E.; Rowland, K. A.; Webber, C. T.; Mingos, D. M . P. J. Chem. Soc, Dalton Trans. 1981,1515-1519. (165) Severin, K.; Polbom, K.; Beck, W. Inorg, Chim. Acta 1995,240,339-346. (166) Sekutowski, D.; Jungst, R.; Stucky, G. D. Inorg. Chem. 1978,17,1848-1855. (167) Bristow, S.; Gamer, C. D.; Clegg, W. Inorg. Chim. Acta 1983, 76, L261-L262. 169 (168) Chakravarty, A . R.; Cotton, F. A. ; Schwotzer, W. Inorg. Chim. Acta 1984, 84, 179-185. (169) Asplund, M . ; Jagner, S.; Nilsson, M . Acta Chem. Scand. (A) 1983,37,57-62. (170) Grocott, S. C ; Wild, S. B. Inorg. Chem. 1982,21, 3526-3534. (171) Kittlesen, G. P.; White, H. S.; Wrighton, M . S. J. Am. Chem. Soc. 1984,106, 7389-7396. (172) Paul, E. W.; Ricco, A. J.; Wrighton, M . S. J. Phys. Chem. 1985,89,1441-1447. (173) Thackeray, J. W.; White, H. S.; Wrighton, M . S. J. Phys. Chem. 1985,89,5133-5140. (174) Pickup, P. G.; Kutner, W.; Leidner, C. R.; Murray, R. W. J. Am. Chem. Soc. 1984, 106,1991-1998. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061484/manifest

Comment

Related Items