New Directions in the Coordination Chemistry of
Verdazyl Radicals
by Stephen D.J. McKinnon M.Sc., University of Victoria, 2005 B.Sc., University of Victoria, 2001 A Dissertation Submitted in Partial Fulfillmentof the Requirements for the Degree of Doctor of Philosophy
in the Department of Chemistry
Stephen D.J. McKinnon, 2010 University of Victoria
All rights reserved. This thesis may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.
M.Sc., University of Victoria, 2005 B.Sc., University of Victoria, 2001
Supervisory Committee
Dr. Robin G. Hicks, (Department of Chemistry) Supervisor
Dr. Thomas M. Fyles, (Department of Chemistry) Departmental Member
Dr. David J. Berg, (Department of Chemistry) Departmental Member
Dr. Byoung C. Choi, (Department of Physics) Outside Member
Dr. Michael O. Wolf, (Department of Chemistry, UBC) External Member
Abstract
A series of palladium and platinum complexes of verdazyl radicals were prepared to study the intermolecular magnetic exchange coupling. Reaction of bidentate verdazyl radicals with (RCN)2MCl2 (R = Me or Ph; M = Pd or Pt) yielded square planar
(verdazyl)MCl2 complexes. The isolated complexes crystallized in either an infinite 1D array or as loosely associated π-stacked dimer pairs. Molecules stacked with either M–M or M–N(verdazyl) close contacts. Molecules that stacked with a M-M close contact exhibited weak antiferromagnetic coupling. Molecules that stacked with a M– N(verdazyl) close contact had coupling that was an order-of-magnitude weaker, but the type of exchange was also metal dependent. While the palladium complex exhibited weak antiferromagnetic coupling, the exchange in the analogous platinum complex was ferromagnetic. The difference between the two was attributed to increased spin leakage onto the platinum centre relative to palladium. The differing electronic behaviour of the two metals was evident in the data from EPR and UV/vis spectroscopies.
Ruthenium complexes of a verdazyl radical were prepared by the reaction of a bidentate verdazyl with Ru(L)2(MeCN)2 (L = acac or hfac). The complexes were isolated
in two or more oxidation states and all characterized by FT-IR, UV/vis/NIR, and EPR spectroscopies, and their structures determined by X-ray crystallography. Experimental data was further explained and supported with time-dependant DFT calculations which were performed by Dr. A. B. P. Lever at York University, Toronto, Ontario. When the complex contained an electron-rich metal fragment, Ru(acac)2, noninnocent behaviour
was observed. There was a large degree of orbital mixing, so that the spin distribution was approximately 39% metal and 56% ligand. The contrasting complex with the electron-poor fragment, Ru(hfac)2, behaved more innocently, the majority of charge was
localized and the spin was ligand-based.
Verdazyl-bridged diruthenium complexes were prepared from a bisbidentate verdazyl and Ru(L)2(MeCN)2 (L = acac or hfac) to study the effect of a neutral radical
bridge on mixed-valence properties. Structural data from X-ray crystallography, spectroscopic data from EPR, FT-IR, and UV/vis/NIR spectroscopies, and comparison to
states.
Supervisory Committee
Dr. Robin G. Hicks, (Department of Chemistry)
Supervisor
Dr. Thomas M. Fyles, (Department of Chemistry)
Departmental Member
Dr. David J. Berg, (Department of Chemistry)
Departmental Member
Dr. Byoung C. Choi, (Department of Physics)
Outside Member
Dr. Michael O. Wolf, (Department of Chemistry, UBC)
Table of Contents
Supervisory Committee ... ii
Abstract ... iii
Table of Contents ... v
List of Tables ... viii
List of Figures ... x
List of Schemes ... xiv
List of Numbered Compounds ... xvi
List of Abbreviations ... xxii
Acknowledgments ... xxvi Dedication ... xxvii Chapter 1 - Introduction ... 1 1.1 Stable radicals ... 1 1.2 Metal-radical complexes ... 3 1.2.1 Phenoxyl radicals ... 4 1.2.2 Nitroxide radicals ... 6 1.2.3 Aminyl radicals ... 9 1.2.4 Thiazyl radicals ... 10 1.2.5 Verdazyl radicals ... 11 1.3 Thesis objectives ... 13
Chapter 2 - Palladium and Platinum Complexes of Verdazyl Radicals ... 15
2.1 Intermolecular magnetic coupling ... 15
2.1.1 Molecular crystal ferromagnets ... 15
2.1.2 Solid-state magnetism of verdazyl radicals ... 18
2.2 Synthesis and characterization of palladium-verdazyl complexes ... 20
2.2.1 Synthesis ... 20
2.2.2 EPR spectroscopy ... 22
3.1 Redox-active ligands ... 40
3.2 Synthesis and characterization of ruthenium-verdazyl complexes ... 44
3.2.1 Synthesis ... 44 3.2.2 Crystal structures ... 48 3.2.3 Vibrational spectroscopy ... 55 3.2.4 EPR spectroscopy ... 56 3.2.5 Electronic spectroscopy ... 57 3.2.6 Cyclic voltammetry ... 60 3.3 Discussion ... 63 3.4 Experimental ... 65
3.4.1 Methods and materials ... 65
3.4.2 Syntheses... 65
Chapter 4 - Synthesis and Properties of Diruthenium-verdazyl Complexes ... 71
4.1 Mixed valence ... 71
4.2 Synthesis and characterization of diruthenium-verdazyl complexes ... 74
4.2.1 Synthesis ... 74 4.2.2 Crystal structures ... 79 4.2.3 Vibrational spectroscopy ... 87 4.2.4 EPR spectroscopy ... 88 4.2.5 Electronic spectroscopy ... 90 4.2.6 Cyclic Voltammetry ... 95 4.3 Discussion ... 98
4.4 Experimental ... 101
4.4.1 Methods and materials ... 101
4.4.2 Syntheses... 101
Chapter 5 - Summary and Future Directions ... 107
Appendix A: Equations ... 119
Appendix B: Calculated UV/vis/NIR spectra ... 120
Appendix C: Crystallographic parameters ... 122
Table 2.4 - Selected bond lengths and angles for structure 2.11 (estimated standard deviations in parentheses). ... 28 Table 2.5 - Selected bond lengths and angles for structure 2.14 (estimated standard deviations in parentheses). ... 33 Table 2.6 - Selected bond lengths and angles for structure 2.15 (estimated standard deviations in parentheses). ... 35 Table 3.1 - Selected bond lengths and angles for the two independent molecules of 3.12 (estimated standard deviations in parentheses). ... 49 Table 3.2 - Selected bond lengths and angles for structure 3.13 (estimated standard deviations in parentheses). ... 50 Table 3.3 - Selected bond lengths and angles for the two independent molecules of 3.12+ (estimated standard deviations in parentheses). ... 52 Table 3.4 - Selected bond lengths and angles for the two independent molecules of 3.13+ (estimated standard deviations in parentheses). ... 53 Table 3.5 - Selected bond lengths and angles for structure 3.13–––– (estimated standard
deviations in parentheses). ... 54 Table 3.6 - Carbonyl stretching frequencies for verdazyls 1.30 and 3.11, and ruthenium-verdazyl complexes. ... 55 Table 3.7 - EPR hyperfine coupling constants and g-values of the neutral ruthenium-verdazyl complexes. ... 57 Table 3.8 - Redox potentials in V vs. Fc/Fc+ for ligands 1.30 and 3.11 and the ruthenium complexes. ... 62
Table 4.1 - Selected bond lengths and angles for 4.8 (estimated standard deviations in parentheses). ... 80 Table 4.2 - Selected bond lengths and angles for 4.9 (estimated standard deviations in parentheses). ... 81 Table 4.3 - Selected bond lengths and angles for 4.8+ (estimated standard deviations in parentheses). ... 82 Table 4.4 - Selected bond lengths and angles for 4.9+ (estimated standard deviations in parentheses). ... 84 Table 4.5 - Selected bond lengths and angles for 4.82+ (estimated standard deviations in parentheses). ... 85 Table 4.6 - Selected bond lengths and angles for 4.9– (estimated standard deviations in
parentheses). ... 86 Table 4.7 - Carbonyl stretching frequencies for verdazyl 3.11 and ruthenium-verdazyl complexes. ... 87 Table 4.8 - EPR hyperfine coupling constants and g-values of ligand 3.11 and the
diruthenium-verdazyl complexes. ... 90 Table 4.9 - Redox potentials (in V vs. Fc/Fc+) for ligand 3.11 and diruthenium
Figure 2.2 - Radical-radical contacts in 2.6, Cu(hfac2)(1.31). Hydrogen atoms and CF3
groups removed for clarity. Thermal ellipsoids are at the 50% probability level. ... 19 Figure 2.3 - UV-visible spectra of palladium complex 2.10 and ligand 1.30 in CH2Cl2. . 21
Figure 2.4 - Room temperature EPR spectrum of 2.9 in CH2Cl2 (top) and simulated
spectrum (bottom). ... 22 Figure 2.5 - Room temperature EPR spectrum of 2.10 in CH2Cl2 (top) and simulated
spectrum (bottom). ... 23 Figure 2.6 - Room temperature EPR spectrum of 2.11 in CH2Cl2 (top) and simulated
spectrum (bottom). ... 23 Figure 2.7 - Molecular structure of 2.9 (left). Intermolecular contacts within the 1D chains of 2.9 (right). Thermal ellipsoids are at the 50% probability level. ... 25 Figure 2.8 - χ vs. T and χT vs. T magnetic data for complex 2.9. Experimental () and fit
data (line). ... 26 Figure 2.9 - Molecular structure of complex 2.10. Intermolecular contacts within the dimers (right).Thermal ellipsoids are at the 50% probability level. ... 26 Figure 2.10 - χ vs. T (left) and χT vs. T (right) magnetic data for complex 2.10.
Experimental () and fit data (line). ... 27
Figure 2.11 - Molecular structure of complex 2.11 (left). Intermolecular contacts within the dimers (right). Thermal ellipsoids are at the 50% probability level. ... 28 Figure 2.12 - χ vs. T (left) and χT vs. T (right) magnetic data for complex 2.11.
Experimental () and fit data (line). ... 29
Figure 2.14 - Room temperature EPR spectrum of 2.14 in CH2Cl2 (top) with spectral
simulation (bottom). The arrow denotes the presence of a free ligand signal. ... 32 Figure 2.15 - Solid state structure of complex 2.14. Intermolecular contacts within the dimers (right).Thermal ellipsoids are at the 50% probability level. ... 33 Figure 2.16 - χ vs. T and χT vs. T magnetic data for complex 2.14. Experimental () and
fit data (line). ... 34 Figure 2.17 - Molecular structure of complex 2.15 (left). Intermolecular contacts within the dimers (right). Thermal ellipsoids are at the 50% probability level. ... 34 Figure 2.18 - χ vs. T (left) and χT vs. T (right) magnetic data for complex 2.15.
Experimental () and fit data (line). ... 35
Figure 3.1 - Molecular structure of complex 3.12 (left). Alternate view showing distortion in verdazyl ring (right). Hydrogen atoms omitted for clarity. Thermal ellipsoids are at the 50% probability level. ... 49 Figure 3.2 - Molecular structure of complex 3.13. Hydrogen atoms omitted for clarity. Thermal ellipsoids are at the 50% probability level. ... 50 Figure 3.3 - Molecular structure of 3.12+. Hydrogen atoms and BPh4–––– counterion omitted
for clarity. Thermal ellipsoids are at the 50% probability level. ... 51 Figure 3.4 - Molecular structure of 3.13++++. Hydrogen atoms and BF
4–––– counterion omitted
for clarity. Thermal ellipsoids are at the 50% probability level. ... 53 Figure 3.5 - Molecular structure of 3.13–––– .Hydrogen atoms and cobaltocenium cation
omitted for clarity. Thermal ellipsoids are at the 50% probability level. ... 54 Figure 3.6 - Room temperature EPR spectra of 3.12 (left) and 3.14 (right) in CH2Cl2.
Arrows denote presence of verdazyl impurity. ... 56 Figure 3.7 - Room temperature EPR spectrum of 3.13 in CH2Cl2 (top) with spectral
simulation (bottom). ... 57 Figure 3.8 - UV/Vis/NIR absorption spectra of 3.12, 3.13, and verdazyl 1.30 in CH2Cl2.
... 58 Figure 3.9 - UV/Vis/NIR absorption spectra of 3.12+ and 3.13+ in CH2Cl2. ... 59
Figure 3.10 - UV/Vis/NIR absorption spectrum of 3.13– in CH
2Cl2. ... 60
Figure 4.4 - Molecular structure of 4.9+. Hydrogen atoms, CF3 groups, PF6– counterion,
and solvent molecules omitted for clarity. Thermal ellipsoids are at the 50% probability level. ... 83 Figure 4.5 - Molecular structure of 4.82+. Hydrogen atoms and PF6– counterions omitted
for clarity. Thermal ellipsoids are at the 50% probability level. ... 85 Figure 4.6 - Molecular structure of 4.9–. Hydrogen atoms and CoCp
2+ counterion omitted
for clarity. Thermal ellipsoids are at the 50% probability level. ... 86 Figure 4.7 - EPR spectrum of 3.11 in CH2Cl2 at room temperature (top) with spectral
simulation (bottom). ... 88 Figure 4.8 - EPR spectrum of 4.8 recorded in CH2Cl2 at room temperature. ... 89
Figure 4.9 - EPR spectrum of 4.9 recorded in CH2Cl2 at room temperature. ... 89
Figure 4.10 - UV/vis/NIR spectra of 4.8 and mononuclear complex 3.12 recorded in CH2Cl2... 91
Figure 4.11 - UV/vis/NIR spectra of 4.9 and mononuclear complex 3.12 recorded in CH2Cl2... 91
Figure 4.12 - UV/vis/NIR spectra of (4.8+)PF6 and (3.12+)PF6 recorded in CH2Cl2. ... 92
Figure 4.13 - UV/vis/NIR spectrum of (4.9+)PF6 and mononuclear complex
3.13+ recorded in CH2Cl2. ... 93
Figure 4.14 - UV/vis/NIR spectrum of (4.82+)(PF6)2 recorded in CH2Cl2. ... 94
Figure 4.15 - UV/vis/NIR spectrum of CoCp2(4.9–)and mononuclear complex
Figure 4.16 - CV of 4.8 in CH3CN / 0.1M Bu4NBF4 electrolyte. ... 96
Figure 4.17 - CV of 4.10 in CH3CN / 0.1M Bu4NBF4 electrolyte. ... 96
Figure 4.18 - CV of 4.9 in CH3CN / 0.1M Bu4NBF4 electrolyte. ... 97
Scheme 1.4 - Redox activity in the verdazyl radical. ... 14
Scheme 2.1 - Synthesis of palladium(II)verdazyl complexes. ... 20
Scheme 2.2 - Synthesis of platinum(II)verdazyl complexes. ... 30
Scheme 3.1 - Three possible resonance structures used to describe the electronic structure of bis(dithiolene) metal complexes 3.1 (M = Ni, Pd, or Pt)... 40
Scheme 3.2 - The three redox states of dioxolenes. ... 41
Scheme 3.3 - Valence tautomerism in cobalt dioxolene complex 3.3. ... 41
Scheme 3.4 - Oxidation states of ruthenium-verdazyl complexes and their possible electron configurations. ... 44
Scheme 3.5 - Synthesis of verdazyl 3.11. ... 45
Scheme 3.6 - Synthesis of complexes 3.12 and 3.13. ... 46
Scheme 3.7 - Synthesis of complexes 3.14 and 3.15. ... 46
Scheme 3.8 - Synthesis of cationic complexes 3.12+ and 3.13+. ... 47
Scheme 3.9 - Synthesis of anionic complex CoCp2(3.13–). ... 47
Scheme 3.10 - Two limiting resonance forms of 3.12. ... 64
Scheme 4.1 - Intramolecular CT transitions in a localized (a) or delocalised (b) mixed-valent complex. ... 72
Scheme 4.2 - Structure of complex 4.2 (a) and the two resonance structures that contribute to the overall electronic structure of 4.2 (b). ... 73
Scheme 4.3 - Synthesis of diruthenium complexes 4.8 and 4.9. ... 75
Scheme 4.4 - Synthesis of cationic diruthenium complexes 4.8+ and 4.9+. ... 76
Scheme 4.5 - Synthesis of dication 4.82+. ... 76
Scheme 4.6 - Attempted synthesis of anion 4.8– and synthesis of anion 4.9–. ... 77
Scheme 4.7 - Synthesis of diruthenium complex 4.102+ and 4.10. ... 78
Scheme 4.8 - Synthesis of diruthenium complex 4.11. ... 78
Scheme 4.9 - Contributing resonance structures for the various oxidation states of verdazyl bridged diruthenium complexes, anion (a), neutral (b, boxed), cation (c), and dication (d). ... 100
Cu O O N N N 1.5 Cu O N N O S S 1.6 Ph Ph Ni O N N O 1.7 R NR O N N O O R N N O R 1.8 1.9 1.10 M O N O O CF3 O O F3C
1.11, M = Cu, V=O, Mn(TEMPO) CF3
F3C
1.12, M = Mn, Co, Ni, Cu, Zn
O M O O O CF3 CF3 F3C F3C O O N N N N O O R R 1.13 1.14 N N N O O N N N O N N O O 1.16 N NH N N O O 1.15 N NH
M N O N O N N O O N N N N O O N N M N N O O N N M N O O N M O N N O 1.17 N N N O O O 1.18 Mn Mn Mn O O O O O O O O O OO O O O = hfac N Rh(bipy) = 1.19 Ir 1.20 N N N S S N N S N S S S N NS S R R R R R R 1.21 1.22 1.23 1.24 S N S N S N Cu O O O O Cu O O O O S N S N S N 1.26 O O = hfac N N S S N M O O O O F3C CF3 F3C CF3
1.25, M = Mn, Fe, Ni, Co, Cu
N N N N R' R R N N N N R' R R X 1.28 (X = O, S) X H 1.27 (X = H, alkyl) N N NMe N N 1.29, R = Me 1.30, R = iPr 1.31, R = Me1.32, R = iPr O OH N N N N N N N N N N N N N N N N O O O O R R R R Me Me Me Me 1.33 1.34 N N N N N O Me Me N 1.35
N N Me Me O 1.39 NC CN 2.1 NO2 N O 2.2 OH N N O O 2.3 N N O O 2.4 OH N S Se N N Se S Et Cl 2.5 N N N O M Cl Cl 2.7, M = Pd 2.8, M = Pt N N N N N N Me Me O Cu O O O O F3C CF3 F3C CF3 2.6 Me N N N N N O Pd Cl Cl N N N N N O Pd Cl Cl N N N N N N O Pd Cl Cl 2.9 2.10 2.11 N N Pd Cl Cl 2.12 N N N N N O Pt Cl Cl 2.13 N N N N N O Pt Cl Cl 2.14 N N N N N N O Pt Cl Cl 2.15 S R R S M R S S R S R R S M R S S R S R R S M R S S R 3.1 (M = Ni, Pd, or Pt)
O O O O O O -e -e -e++++ -e++++ 3.2 3.2 3.22222 N N Co O O O O tBu tBu tBu tBu 3.3 N N O Zr O L L L 3.4 L = THF O N Ir CF3 3.5 N N N Ar Ar 3.6 N N N Ar Ar 3.7 Fe N2 N2 N N N Ar Ar 3.8 Co N2 N N NH N N HN O NH2 N N NH2 O 2HCl 3.9 2HCl 3.10 N N N N N N O 3.11 N N N N N iPr iPr O Ru O O O O R R R R R = CH3, 3.12 R = CF3, 3.13 N N N N N N iPr iPr O Ru O O O O R R R R R = CH3, 3.14 R = CF3, 3.15 BPh4 N N N N N iPr iPr O Ru O O O O Me Me Me Me (3.12+)BPh 4− BF4 N N N N N iPr iPr O Ru O O O O F3C CF3 F3C CF3 (3.13+)BF 4 N N N N N iPr iPr O Ru O O O O Me Me Me Me CoCp*2+(3.12 ) Co
4.2 O P Ph2 Ru N N Ru N N N N N N 4.6 R E N N E R 4.7 (E = O, NH) O O O O O O R R E E 4.5 (E = O, NH) 4.4 4.3 R N N N N N N iPr iPr O Ru O O O O R R R R Ru O O O O R R R R R = CH3, 4.8 R = CF3, 4.9 N N N N N N iPr iPr O Ru O O O O R R R R Ru O O O O R R R R R = CH3, (4.8+)PF6 R = CF3, (4.9+)PF 6 PF6 N N N N N N iPr iPr O Ru O O O O Me Me Me Me Ru O O O O Me Me Me Me (PF6)2 (4.82+)(PF6)2 2 N N N N N N iPr iPr O Ru O O O O Me Me Me Me Ru O O O O Me Me Me Me Co CoCp*2(4.8 ) N N N N N N iPr iPr O Ru O O O O F3C CF3 F3C CF3 Ru O O O O CF3 F3C CF3 F3C Co CoCp2(4.9 )
N N N N Ru O O O O Me Me Me Me Ru O O O O Me Me Me Me (4.102+)(PF 6)2 (PF6)2 N N N N Ru O O O O Me Me Me Me Ru O O O O Me Me Me Me 4.10 2+ N N N N Ru O O O O F3C CF3 F3C CF3 Ru O O O O CF3 F3C CF3 F3C 4.11 N N N N N iPr iPr O Ru CO CO CO CO N N N N N iPr iPr O Mo CO CO CO CO 5.1 5.2 N N N N N N iPr iPr O Ru L L N N N N N N iPr iPr O Ru L L N N N N N N iPr iPr O Ru L L N N N N N N iPr iPr O Ru L L Ru L L 5.3
bipy 2,2’-dipyridyl
BL bridging ligand
br broad (NMR and IR peak descriptor)
Bu butyl °C degrees Celsius C Curie constant cm centimeter cm-1 wavenumber Cp cyclopentadienyl Cp* pentamethylcyclopentadienyl CT charge-transfer d doublet DPPH 1,1-diphenyl-2-picrylhydrazyl EI electron impact
emu electromagnetic units (cm3)
EPR electron paramagnetic resonance
Ered reduction potential
Et ethyl
EtOH ethanol
Fc/Fc+ ferrocene/ferrocenium
FM ferromagnetic
g g-factor (or Landé factor)
G Gauss
GAO galactose oxidase
GHz gigahertz
h Planck’s constant (6.6260755 × 10-34 J s)
h hour(s)
hfac 1,1,1,5,5,5-hexafluoroacetylacetonate
HOMO highest occupied molecular orbital
Hz hertz
iPr isopropyl
IR infrared
J coupling constant (NMR) or magnetic exchange coupling constant
k Boltzmann constant (1.3806580 × 10-23 J K-1) K Kelvin KBr potassium bromide Kc Comproportionation constant L litres LMCT ligand-to-metal charge-transfer
LUMO lowest unoccupied molecular orbital
m multiplet M molarity Me methyl mg milligram min minute(s) MHz megahertz mL millilitre MLCT metal-to-ligand charge-transfer mp melting point MO molecular orbital(s) mol mole mmol millimole
Ph phenyl
ppm parts per million
q quartet
R agreement factor
R2 quality of fit
RT room temperature
s singlet (NMR), strong (IR) or seconds
S spin multiplicity
SOMO singly occupied molecular orbital
SQUID superconducting quantum interference device
t triplet
TLC thin layer chromatography
UV ultraviolet V volt vd• verdazyl vis visible vw very weak β Bohr magneton (9.27401549 × 10-24 J T-1)
δ parts per million
∆ heat
ε molecular extinction coefficient
µ magnetic moment or denoting bridging ligand
µeff effective magnetic moment
θ Weiss constant
ρ density
I would also like thank the members of the Hicks group past and present for all of your support, personal and professional, and for sharing all of your knowledge and experience in synthetic chemistry. Many thanks to Dr. Marty Lemaire, Dr. Greg Patenaude, Dr. M’hamed Chahma, Dr. Rajsapan Jain, Dr. Khayrul Kabir, Dr. Peter Otieno, Dr. Bryan Koivisto, Dr. Dan Myles, Dr. Joe Gilroy, Dr. Tyler Trefz, Dr. Kevin Anderson, Graeme Nawn, Bartoz Nowak, Cooper Johnson, Kate Waldie, and Derek Mandel.
Lastly, I would like to thank all of the faculty and staff at in chemistry department work so hard to ensure that this is such a great place to work and learn. The secretaries and office staff have all been there to clarify university procedures and make sure deadlines are met. Thanks to Bob Dean and Terry Wiley for keeping the large number of computers and vital instruments in working order. I would especially like to thank my teaching supervisors Kelly Fawkes and Dave Berry for helping make teaching such an enjoyable and rewarding experience.
Dedication
triphenylmethyl, or trityl, radical (1.1) was the product of an attempted synthesis of hexaphenylethane. Although it is not stable or isolable, the trityl radical is a persistent species, existing indefinitely in dilute deoxygenated solutions. Almost immediately after Gomberg’s report of an organic radical, the first example of an isolated stable organic radical, the nitroxide porphyrexide 1.2, was reported.2 Since this discovery, many stable radicals have been reported and studied extensively to better understand the interesting chemistry that arises with the presence of an unpaired electron.
Radicals possess properties that are rare or do not exist in closed-shell molecules. For this reason, stable radicals have been particularly attractive in material science applications. One such property associated with unpaired electrons is magnetism. A large number of stable radical derivatives have been studied to better understand the interactions of radicals in the solid-state, and how these interactions affect the observed magnetic properties.3 This topic will be further expanded on in Chapter 2. Stable radicals have also been used in technologically important applications. Nitroxide radical-containing polymeric materials have been incorporated into the so-called “organic radical batteries”.4 This new generation of battery offers a number of distinct advantages owing
1.1 N N O NH HN 1.2
to their organic nature; they are lightweight, flexible, fast to charge, and free of heavy metals which are a major component of most conventional batteries.
Stable radicals have been used in a number of contexts outside of the advanced materials applications mentioned above. Stable radicals have been used extensively to study biological systems through electron paramagnetic resonance (EPR) spectroscopy, a technique first introduced in the 1960s by McConnell and coworkers.5 The “spin label”,
or “spin probe”, is a radical species which is attached to another molecule or macromolecule such as a protein. The EPR spectra of a labelled species can used to gain essential information about the environment and dynamics surrounding the label, or to probe the different molecular environments within a complex system.
Some stable radicals have also found utility in probing the kinetics and mechanisms of radical-induced polymerizations6, radical rearrangements7, and homolytic dissociations8 as a so-called radical trap. A stable radical is added to the reaction under study to trap any reactive radical intermediates that are formed (Scheme 1.1). Pairing of the odd electrons of the radical species results in a closed-shell species composed of the radical trap bound to the radical intermediate fragment. This allows for detection and identification of radical species (•R) that would otherwise be too short-lived or in too low a concentration to observe.
Another use of stable radicals of significant technological importance is living radical polymerization. Unlike the conventional radical polymerizations, this technique utilizes a stable radical to control propagation of the growing polymer chain. This gives polymers with low polydispersity, controlled molecular weights, and defined chain ends, allowing the properties of the material to be more easily tuned.9,10 As Scheme 1.2 shows,
a stable radical can reversibly bond to the end of the growing polymer chain. The labile bonding of the stable radical to the growing polymer effectively reduces the concentration of the reactive chain end, which minimizes irreversible termination reactions. Therefore, the polymer grows in a more controlled, predictable manner.
N O + R N O R
1.2 Metal-radical complexes
The coordination chemistry of stable radicals has also been the focus of research interest over the past few decades. As discussed previously, the properties of radicals lend themselves to a range of potential applications. Combination of these properties with those of transition elements opens up an even wider range of possibilities. In the realm of magnetic materials, the metal-radical approach to molecular magnetism11 was introduced to take advantage of these properties. Magnetic materials might also be produced incorporating other optical or redox properties. Metal complexes of some radicals have also been employed as stoichiometric reagents or catalysts. The synergy between a particular metal and radical ligand has the potential to create species with reactivity not observed in either the metal or radical alone.
There are some requirements with respect to the radicals that are used as ligands that are specific to their intended use. Firstly, the radical must have a lone pair capable of bonding to a metal, in addition to the unpaired electron. The unpaired electron of the radical should also be proximal to the metal centre for there to be any sizable interaction between the radical and metal.
The following sections comprise an overview of neutral radicals that have been investigated as ligands. Absent are discussions of complexes of O2 and NO, both of
which are radicals. Though many complexes of O2 and NO exist,12-14 these molecules are
Scheme 1.2 - General mechanism for living radical polymerization, where R is a stable radical
generally treated as substrates. Also, these radicals are not able to be structurally modified, unlike the radicals in the following sections, which can be to alter the properties of the radical ligand. The lack of tunability has meant that they have not been studied for their properties based on metal-radical interactions. Therefore, the following sections are focused on neutral organic radicals that can be derivatized.
1.2.1 Phenoxyl radicals
Phenoxyls radicals are formed by the oxidation of phenols and had been proposed as far back as the 1920’s. The first example of a stable phenoxyl (1.3) was discovered independently by Cook15 and Muller16 in 1953. Some polyhalophenoxyl radicals, such as
1.4, are also stable.17 The stability of structures 1.3 and 1.4 is the result of the large
substituents present ortho and para to the phenolic oxygen. The steric protection is needed to prevent decomposition through C-O or C-C bond forming reactions.
Phenoxyl radicals are of particular importance in biological systems. Tyrosine, an amino acid essential to a large number of organisms, contains a phenol moiety. The tyrosyl radical has been shown to be a key component for a number of enzymatic redox processes. Galactose oxidase (GAO), a fungal enzyme, contains a copper ion coordinated to two tyrosine residues. One of these tyrosines exists as a phenoxyl radical, which has an
ortho-thioether bond to a cysteine residue (Figure 1.1a). This enzyme functions to
catalyze the two-electron oxidation of D-galactose to D-galacto-hexodialdose, which is coupled to a reduction of O2 to H2O2 (Figure 1.1b).18 This reactivity has possible
applications in chemical synthesis, so many complexes of phenoxyl radicals have been prepared to model the active site of this or related enzymes.19
O 1.3 O 1.4 Cl Cl Cl
The first model compound of the active site of GAO was synthesised in 1993 by Tolman and coworkers.20 The Zn2+ and Cu2+ complexes that were investigated utilised a 1,4,7-triazacyclononane ligand framework with two appended phenolates. Electrochemical or chemical oxidation of the neutral species gave the phenoxyl radical complexes 1.5. The complex of the diamagnetic Zn2+ was used to observe the coordinated phenoxyl without the added complication of a paramagnetic metal centre.
Since Tolman’s report of a GAO active site model, several groups attempted to make complexes to mimic the catalytic properties of this enzyme. For example, Stack and Wang reported a complex capable of oxidizing primary alcohols to aldehydes.21 The neutral (disalicylidene)diimine complex has a distorted square planar geometry about copper (II) centre. Oxidation with 1.1 equivalents of tris(4-bromophenyl)aminium hexachloroantimonate converts the complex to its phenoxyl form 1.6.
Cu O O N N N 1.5
A limited number of phenoxyl radical complexes have been studied so far that feature metal centres other than copper or zinc. Yamauchi and coworkers have reported a Ni(II) (disalicylidene)diimine complex (1.7) that has been shown to undergo temperature dependant valence tautomerism.22 Because of the mixing between the metal and ligand frontier orbitals, the structure can be interconverted between the Ni(II)-phenoxyl and Ni(III)-phenolate forms by changing temperature.
1.2.2 Nitroxide radicals
The first stable organic nitroxide radical was synthesized in 1901 by Piloty and Schwerin.2 Since then, the nitroxides (1.8) have become the most well studied class of radicals owing to their ease of synthesis and the excellent stability of many derivatives. The nitroxide functional group has been incorporated in to other radical derivatives, such as the nitronyl nitroxides (1.9) and imino nitroxides (1.10). Like the nitroxides, the nitronyl nitroxides and imino nitroxides display excellent stability.
Ni O N N O 1.7 R NR O N N O O R N N O R 1.8 1.9 1.10 Cu O N N O S S 1.6 Ph Ph
In complexes like 1.11 and 1.12, the monodentate oxygen donors make nitroxides poor ligands; their coordination geometries are difficult to predict and metal substrates tend to be limited to very Lewis acidic metals. To combat this, the chelate effect has been exploited to give nitroxide derivatives with stronger binding and more well-defined coordination geometries. There are many examples in which pyridine substituted nitroxide derivatives 1.13 and 1.14 have been used to form complexes with transition metals.31-34
Bridging chelating nitroxide ligands have also been used to prepare extended structures. Rey and coworkers reported a one-dimensional manganese coordination polymer utilising imidazole substituted radical 1.15.35 Bis(bidentate) ligand 1.15, or its
M O N O O CF3 O O F3C
1.11, M = Cu, V=O, Mn(TEMPO) CF3
F3C
1.12, M = Mn, Co, Ni, Cu, Zn
O M O O O CF3 CF3 F3C F3C O O N N N O R R 1.13 1.14 N N N O O N N N O
benzannulated derivative 1.16, have further been used in the preparation of two-dimensional manganese coordination arrays (1.17).36
Inoue and coworkers have used a different approach to extended structures using a nitroxide triradical.37 Combination of a tris(monodentate) nitroxide radical and Mn(hfac)2 gave a three-dimensional coordination network 1.18. Each manganese centre is
octahedral with two nitroxides coordinating in a trans orientation. The resulting material orders magnetically, albeit only at cryogenic temperatures.
Chemical applications for the nitroxide radicals have also been developed in conjunction with metal ions. Simple nitroxides, e.g. TEMPO, have been shown to catalyze the oxidation of alcohols to carbonyl compounds in the presence of a stoichiometric co-oxidant.38 TEMPO has also been used in conjunction with transition metals, including copper and ruthenium, for similar transformations.39,40 The benefit of these transition metal-based systems is that atmospheric O2 is used as the stoichiometric
oxidant. It is still unknown how the reaction actually proceeds, though several mechanisms have been postulated. Scheme 1.3 shows a possible catalytic cycle for a
M N O N O N N O O N N N N O O N N M N N O O N N M N O O N M O N N O 1.17 N N O O 1.16 N NH N N O OH 1.15 N NH N N N O O O 1.18 Mn Mn Mn O O O O O O O O O OO O O O = hfac
1.2.3 Aminyl radicals
Aminyls (•NR2)are a family of nitrogen centred radicals derived from loss of a
hydrogen atom from the corresponding amine HNR2. These radicals tend to be very
unstable and are reactive intermediates in a variety of chemical reactions.42,43 The few examples of aminyls that have been isolated have a large amount of steric protection and further stabilisation by either heteroatoms (O or S) or by coordination to a metal. Because of issues of stability, the first complex of an aminyl radical was not isolated until 2005. Grützmacher and co-workers synthesised a five-coordinate rhodium (I) complex 1.19, which features two bulky 5-H-dibenzo[a,d]cyclohepten-5-yl substituents on the aminyl nitrogen.44 These substituents provide both steric bulk to shroud the radical and two η2 -alkenes to coordinate the rhodium centre. A small number of other aminyl radical complexes have since been reported.45
N O Cu(II) N HO R1 R2 H OH R1 R2 H O Cu(II) 1/2O2
Scheme 1.3 - Proposed catalytic cycle for the oxidation of alcohols with a copper-nitroxide catalyst
Complexes of aminyl radicals have shown promise for use as stoichiometric or catalytic reagents. Recently, Grützmacher and co-workers synthesized an iridium analogue of complex 1.20, which, through hydrogen abstraction, has been used to form a variety of disilanes, disulfides and disiloxanes.46 The same complex is also a capable
catalyst for the dehydration of aromatic and aliphatic alcohols to give aldehydes.47
1.2.4 Thiazyl radicals
Thiazyls are a class of stable radicals, the simplest examples of which are the thioaminyls, with the general structure R–SN•–R'. The R and R' groups tend to be large in
order to provide steric protection of the unpaired electron. Several examples of heterocyclic thiazyl derivatives have also been reported which have been studied extensively for their magnetic and conducting properties.48,49
The coordination chemistry of thiazyl radicals is currently in its early stages of development.50 The most common thiazyl derivatives to be used as ligands are based on the 1,2,3,5-dithiadiazolyls 1.21. Because of the presence of both sulfur and nitrogen heteroatoms in the ring, complexes of 1,2,3,5-dithiadiazolyls 1.21 have been reported
N Rh(bipy) = 1.19 Ir 1.20 N N = N S S N N S N S S S N N SS R R R R R R 1.21 1.22 1.23 1.24
1.2.5 Verdazyl radicals
Verdazyls were discovered accidentally by Kuhn and Trischmann in 1963 during studies on the alkylation of aryl formazans.54 The first examples of verdazyl radicals
(1.27) have a saturated carbon atom in the C6 position of the tetrazine ring and aryl substituents at the 1- and 5-positions. A number of derivatives have been prepared in which the substituent at the 3-position are alkyl, aryl or other groups. Verdazyls have also been prepared which have a carbonyl or thiocarbonyl group at the C6 position of the tetrazine ring (1.28).55 6-oxoverdazyls most commonly have alkyl substituents (methyl, benzyl, or isopropyl) on the nitrogen of the ring, though a few examples feature aryl R-group substituents. The synthetic pathway allows for a large number of C3 substituents to be used and many derivatives have been made with alkyl or aryl groups at the 3-position.
N N N N R' R R N N N N R' R R X 1.28 (X = O, S) X H 1.27 (X = H, alkyl) 5 4 3 2 1 6 R R R'
Figure 1.2 - Two types of verdazyl radicals with ring numbering scheme at centre. S N S N Cu O O O O Cu O O O O S N S N S N 1.26 O O = hfac N N N M O O O F3C CF3
Although verdazyl radicals have been known for nearly 50 years, their coordination chemistry has only been studied over the last 13 years.56 A number of verdazyl derivatives have been synthesised with donor atom containing substituents and been used successfully as ligands. Nearly all this work has been based on derivatives of 6-oxoverdazyls 1.28; there is but one example that utilises a derivative of verdazyl 1.27 as a ligand.57 Structures 1.29 - 1.35 are representative examples of verdazyl radicals that
have been utilised as ligands. A number of other more complex derivatives have also been synthesized for the purpose of forming extended coordination networks.58
The coordination chemistry of the verdazyls has been limited to the study of their magnetic properties. Most of the complexes have been synthesised with the mid to late 1st
row transition elements, though some complexes of heavier metals are known. Simple mononuclear complexes (1.36), featuring verdazyl ligand 1.30 and hfac ancillary ligands, were investigated by Hicks and coworkers.59,60 The magnetic properties of the different complexes have helped understand the mechanism of magnetic exchange between the spin on the verdazyl and metal. Dinuclear complexes 1.37, where the verdazyl serves to bridge two metal centres, have also been investigated.61 In the context of molecular magnetism, these complexes are of value to assess whether the verdazyl might be a viable magnetic coupler in the formation of extended coordination networks.
N N NMe N N 1.29, R = Me 1.30, R = iPr 1.31, R = Me1.32, R = iPr O OH N N N N N N N N N N N N N N N N O O O O R R R R Me Me Me Me 1.33 1.34 N N N N N O Me Me N 1.35 1.37, M = Ni, Mn N N N N N N Me Me O M O O O O F3C CF3 F3C CF3 M O O O O CF3 F3C CF3 F3C N N N N N Me Me O M O O O O F3C CF3 F3C CF3 1.36, M = Ni, Mn, Cu
1.3 Thesis objectives
Over the past decade and a half, work on complexes of verdazyl radicals has been primarily focused on understanding intermolecular magnetic interactions between the radical and metal. This has provided important information on the nature of intramolecular exchange interactions, but little has been done to explore other possible properties of these radicals when coordinated. Although intramolecular magnetic interactions tend to be much stronger, in the solid-state intermolecular interactions between molecules are important too. A variety of molecular solids have been investigated for their magnetic properties. In such materials, these properties depend on intermolecular contacts in the solid state. It is of interest to attempt to elucidate patterns or trends in discrete complexes of verdazyl radicals. Chapter 2 is concerned with intermolecular exchange interactions in verdazyl complexes with the diamagnetic metals, palladium and platinum.
Earlier studies had suggested verdazyls are redox-active and more recent work in the Hicks group studied this in detail.63 Verdazyl radicals 1.28 can be oxidized by one
N N N N N N N N Me Me O Me Me O Cu Cu X X 1.39 N N N N N N N N Me Me O Me Me O 1.38
electron to 1.28+ or reduced by one electron to 1.28-. But, as of yet, no work has been published investigating the redox behaviour of the radical coordinated to a metal ion. Metal complexes of phenoxyl, nitroxide, and aminyl radicals show a wealth of interesting chemical properties which arise from metal-radical interactions in which both the metal and ligand are redox-active. It is of interest to speculate on possible physical or chemical properties of redox-active complexes of the verdazyl; of particular interest is the combination of a redox active ligand with a redox active metal. In this context, the major thrust of this thesis is focused on verdazyl complexes of ruthenium. Chapter 3 is focused on mononuclear complexes, which feature a verdazyl bound to a single ruthenium ion. Chapter 4 presents dinuclear complexes, in which the verdazyl serves as a bridging ligand between two metal centres. Chapter 5 describes preliminary studies on mono- and dinuclear ruthenium complexes of a verdazyl diradical.
R' N N N N O R R R' N N N N O R R R' N N N N O R R -e -+e --e -+e -1.28 1.28+ 1.28
conventional magnet counterparts. For these materials in general, the bulk magnetic properties are a function of the magnetic interactions within the solid which can be through bond and/or through space. So far, there have been several types of materials investigated for their magnetic properties, the most simple of which being composed solely of radicals or radical ions. Others have prepared polyradicals containing two or more unpaired electrons.64 The metal-radical approach utilizes the combination of radical ligands and paramagnetic metals covalently bound.
Unlike intramolecular exchange, intermolecular magnetic exchange coupling is through space and relies on close contacts between molecules in the solid state. For this to occur, atoms in close contact have to bear some spin density and have to be close enough to have some overlap of their orbitals. Though usually weak, the possibility of intermolecular interactions propagated through one, two, or three dimensions in the solid state make them significant and important to understand. It is this type of interactions that lead to the long range ordering observed in organic molecular crystal65 and
charge-transfer ferromagnets66, examples of which will be discussed in the following sections.
2.1.1 Molecular crystal ferromagnets
An important class of molecular materials studied for their magnetic properties are radical ion salts. The first example of this type to exhibit bulk ferromagnetic ordering was [FeIII(C5Me5)2]•+[TCNE]•- 2.1 (TCNE = tetracyanoethylene) which orders
ferromagnetically at 4.8 K.67,68 In the solid-state, 2.1 forms infinite 1-D chains with alternating donor-acceptor moieties (•••D•+A•-D•+A•-•••) and strong intrachain exchange interactions. There are also interchain interactions which lead to 3-D ordering. These
interactions tend to be very weak and the magnetism can be thought of as 1-D except at very low temperatures. Since this discovery, there have been many radical ion salts of this type, featuring a paramagnetic metallocenium cation and a radical anion, exhibiting a range of magnetic properties. Among these, the highest ordering temperature reported so far has been for [MnIII(C5Me5)2]•+[TCNQ]•- (TCNQ = tetracyanoquinodimethane), which
becomes a ferromagnet below 6.2 K.69
The first example of a purely organic molecular crystal ferromagnet was reported in 1991 by Tamura and coworkers.70 The para-nitrophenyl nitronyl nitroxide radical 2.2 was found to crystallize in four different phases. The orthorhombic β-phase has a transition to a ferromagnetically ordered state with a Tc of 0.6 K. Below this temperature,
magnetization versus field experiments confirm the ferromagnetism of 2.2, indicating three-dimensional ordering. Based on the crystal structure, at least two kinds of interactions exist, one between molecules in a two-dimensional sheet and a second, weaker interaction between sheets.
Following the discovery of ferromagnetic ordering in nitroxide radical 2.2 in the early 90s, there was increased interest in the study of the structure-property relationships in nitronyl nitroxide radicals. Because it is almost impossible to predict how radicals pack in the solid-state, it has been very difficult to study these relationships systematically. In an attempt to control solid-state organisation, some have taken a crystal engineering approach materials in which so-called ‘crystal design elements’, such as hydrogen bonding substituents, are used to give predetermined molecular arrangements. This approach had been used by Veciana and coworkers in the study of a series of
NO2 N N O O 2.2 Fe NC CN NC CN 2.1
a more general sense, Veciana, Novoa and coworkers had taken a statistical approach to correlate crystal packing and magnetism.73 A library of crystalline nitronyl nitroxides
were selected in which the radicals had either dominant ferromagnetic or dominant antiferromagnetic interactions. Unfortunately, analysis of the crystal packing has not revealed any obvious correlations between solid-state packing and the magnitude or sign of magnetic interactions.
Other systems have been studied to further understand the mechanisms of magnetic exchange and to improve on the low ordering temperatures observed for previous organic ferromagnets. The thiazyls are a class of stable radicals that, due to their planarity and propensity to form stacked structures in the solid-state, are well suited for this research. Oakley and coworkers had recently reported a bis(thiaselenazolyl) radical,
2.5, that displays bulk ferromagnetism with a Tc of 12.3 K.74 The crystal structure reveals
that in the solid state molecules of 2.5 are organized in slipped π-stack arrays, linked by intermolecular Se--Se’ contacts.
Oakley and coworkers have developed a systematic approach to understanding and predicting the conducting and magnetic behaviour in some classes of thiazyl radicals.75 Thiazyl radicals tend to crystallize in π-stacked arrays. Calculations involving variations in the slippage of the stacks with a fixed intermolecular contact distance have
N S Se N N Se S Et Cl 2.5
been used to create energy surface plots to give a qualitative understanding of the solid-state behaviour of these radicals. Such plots may aid in the design of materials with particular magnetic or conducting properties.
2.1.2 Solid-state magnetism of verdazyl radicals
Besides understanding intramolecular interactions with metal ions, there have been several examples of verdazyl radicals studied for their intermolecular interactions in the solid-state.76 The majority of studies have been on 1,3,5-triaryl substituted verdazyls with the general structure 1.27. Generally, magnetic interactions in these radicals have been weak, partially as a result of the bulky aromatic substituents which prevent close contacts between neighbouring verdazyl rings.
Hicks and coworkers have reported very strong magnetic exchange in a pyridine-substituted 6-oxoverdazyl 1.29:hydroquinone molecular solid.77 The radicals crystallize into head-to-tail π-stacked arrays (Figure 2.1). The 1D radical stacks are linked together by hydrogen bonding between the pyridyl nitrogen of the verdazyl and hydroquinone. The result is strong antiferromagnetic exchange between radicals within the π-stacks. Because of the head-to-tail arrangement, the exchange was initially assumed to be mediated by the pyridine substituents. In 2006, the magnetic properties of 1.29 were reinvestigated with the aid of high level DFT calculations.78 The conclusion of this study
was that the main path of magnetic exchange involves intramolecular overlap of verdazyl rings at the C3 position, a site which does not contribute to the SOMO, but possesses some negative spin density via spin polarization.
A number of metal complexes of verdazyl radicals have been reported in which π-stacking of the radical ligands occurs in the solid-state.56,79 This has been observed more
N N N N N O N N N N N O N N N N N O
Figure 2.1 - Head-to-tail π-stacking observed in verdazyl radical 1.29 with dominant magnetic exchange pathway (---).
To date, there has been no investigation of π-stacking in verdazyl metal complexes in which the metal is nonmagnetic. To maximize the possibility of π-stacking interactions in the solid-state, the target complexes should be planar. The d8 metals tend
to form square planar complexes and, when combined with planar ligands, have the propensity to form stacked structures in the solid state. So far, there has been little work reported in which radicals have been coordinated to palladium or platinum metals. Most have had ligands coordinated so that the spin bearing atoms are far from and have little or no interaction with the metal centre. For example, Ouahab, Sutter, and coworkers have reported imino nitroxide radical complexes 2.7 and 2.8 in which a the ligand is bonded through the spin-bearing imino nitrogen.34 Both 2.7 and 2.8 crystallize into 1D π-stacks in the solid-state. Though the unpaired electron in imino nitroxide radicals is somewhat delocalised, the majority of spin density is on the N-O unit, away from the metal centre. The focus of this chapter is on the solid-state and magnetic properties of Pd(II) and Pt(II) complexes of verdazyl radicals.
N N N O M Cl Cl 2.7, M = Pd 2.8, M = Pt
Figure 2.2 - Radical-radical contacts in 2.6, Cu(hfac2)(1.31). Hydrogen atoms and CF3
2.2 Synthesis and characterization of palladium-verdazyl complexes
2.2.1 Synthesis
The verdazyl radicals chosen for this work were bidentate chelates that mimic the coordination of bipyridine. Ligands 1.2980, 1.3081, and 1.3263 have been reported previously. Palladium (II) complexes 2.9, 2.10, and 2.11 were all synthesized by adding the appropriate verdazyl to a hot solution of a bis(nitrile)palladium dichloride (Scheme 2.1). The complexes were all isolated as dark reddish crystalline solids in moderate yields. In all cases, slow cooling of the reaction mixtures resulted in formation of crystalline materials. All of the complexes were characterized by FT-IR, UV-vis, and EPR spectroscopies and elemental analysis. 2.9 was found to be only sparingly soluble in some polar organic solvents. 2.10 and 2.11 were generally more soluble, dissolving in the same solvents as 2.9, and in chlorinated solvents. All of the palladium complexes investigated are indefinitely stable in both solution and the solid-state.
N N N N N O PdCl Cl N N N N N O Pd Cl Cl N N N N N N O Pd Cl Cl N N N N N O N N N N N O N N N N N N O PdCl2 CH3CN / ∆∆∆∆ PdCl2 CH3CN / ∆∆∆∆ (PhCN)2PdCl2 Acetone / ∆∆∆∆ 1.29 1.30 1.32 2.9 2.10 2.11 Scheme 2.1 - Synthesis of palladium(II)verdazyl complexes.
compared to free ligand 1.29 (1683 cm-1). In complex 2.10, the difference is quite small, with v(C=O) at 1689 cm-1, compared to ligand 1.30 at 1686 cm-1. Complex 2.11 has a carbonyl stretching frequency of 1697 cm-1, only slightly higher than that of free ligand
1.32 (1689 cm-1).
The electronic spectra of palladium complexes 2.9, 2.10, and 2.11 all bear a resemblance to the electronic spectra of the uncoordinated radicals. The spectra of complex 2.10 and ligand 1.30 are shown in Figure 2.3 as a representative example. Generally, the absorptions of the metal complexes are red-shifted by approximately 50 nm and more intensely absorbing relative to the free ligands. Coordination also results in the emergence of a new absorption at the edge of the visible region (λ = 360 nm (ε = 2500) for 2.9, 360 nm (ε = 2500) for 2.10, and 360 nm (ε = 2500) for 2.11).
2.2.2 EPR spectroscopy
The room temperature EPR spectra of the palladium complexes were all recorded in dichloromethane. Hyperfine coupling constants determined by spectral simulation are provided in Table 2.1.
The spectrum of compound 2.9 (Figure 2.4) consists of broad 13-line pattern that arises from coupling of the unpaired electron to the four nitrogen atoms of the tetrazine ring and six N-methyl protons. The lack of any obvious coupling to the spin active isotope of palladium (105Pd = 22.3 %, S = 5/2), and the fact that the g-value of 2.0077 is close to that of the free ligand 1.29 (g = 2.0043), suggest that the spin density is primarily located on the ligand.
The spectra of compounds 2.10 and 2.11 show similar hyperfine coupling to the tetrazine nitrogens (aN) as observed in compound 2.9, but differ in that there is only weak coupling to the two isopropyl methine protons (Figure 2.5 and Figure 2.6). The weak coupling, a(H), is consistent with that observed in the free ligands. The total spectral width is narrower and there is some hyperfine coupling to the palladium centre apparent at the edges of the spectrum. Like 2.9, the g-values for 2.10 (2.0065) and 2.11 (2.0074) both suggest the spin is primarily ligand-based in these complexes.
Figure 2.4 - Room temperature EPR spectrum of 2.9 in CH2Cl2 (top) and simulated
Table 2.1 - EPR hyperfine coupling constants and g-values for palladium(II)verdazyl complexes. Compound a(N) a(N) a(N) a(N) a(H) a(H) g-value 1.2980 6.5a 5.3b 5.3c 2.0037 2.9 8.4 6.8 6.1 5.8 5.8 4.8 2.0077 1.3081 6.5a 5.3b 1.5d 2.0043 2.10 8.4 6.5 5.6 4.8 1.8 0.3 2.0065 1.32 6.6a 5.3b 1.2d 2.0047 2.11 8.7 6.8 5.5 4.6 2.0 0.2 2.0074
a Coupling to two nitrogens a(N2, N4), b Coupling to two nitrogens a(N1, N3), c Coupling to six methyl hydrogens a(CH
3), d Coupling to two methine hydrogens a(CH)
Figure 2.6 - Room temperature EPR spectrum of 2.11 in CH2Cl2 (top) and simulated
spectrum (bottom).
Figure 2.5 - Room temperature EPR spectrum of 2.10 in CH2Cl2 (top) and simulated
2.2.3 Structural and magnetic properties
The molecular structure of 2.9 is shown in Figure 2.7 with selected bond lengths and angles in Table 2.2. The geometry around the palladium centre is square planar, as expected for a d8 metal. Minor deviations from the ideal square planar geometry are the result of the smaller bite angle of the verdazyl ligand. Unlike the bipyridine in 2.12, the ligand 1.29 is not perfectly planar upon coordination (torsion angle defined by N2-C1-C11-N10 = 3.1°) and does not bond coplanar with the PdCl2 plane. The longer
metal-verdazyl bond, Pd-N(2) [2.0700(18) Å], reflects the weaker donor ability of the metal-verdazyl compared to the pyridine, Pd-N(10) [2.0251(18) Å], observed in previous complexes. Typically, upon coordination there are only minor changes to the verdazyl ring relative to the free ligand. In 2.9, there is a pronounced loss of symmetry, the N(2)-C(1) [1.355(3) Å] bond is longer and the N(4)-C(1) [1.314(3) Å] bond slightly shorter. The C-N bonds in free ligand 1.29 are 1.32 Å. The N-N bonds also change so that N(1)-N(2) [1.366(3) Å] is longer than N(3)-N(4) [1.347(3) Å], but they do not differ significantly from the free ligand (1.36 Å).
In the solid state, molecules of 2.9 associate into 1D columns; molecules stack in a head-to-toe arrangement with alignment of the palladium centre of one with N(2) of its neighbours (Figure 2.7). The intermolecular Pd-N(2) contacts within the chains are 3.49 Å and 3.54 Å. The type of stacking in 2.9 differs from that observed for the bipyridine analogue 2.12 in which the palladium centres of neighbouring molecules are aligned within the π-stacks with a separation of 3.46 Å.82
N N Pd Cl Cl 2.12
Table 2.2 - Selected bond lengths and angles for structure 2.9 (estimated standard deviations in parentheses).
Atom Length (Å) Atom Angle (deg)
O-C(2) 1.207(3) Cl(1)-Pd-Cl(2) 87.38(2) N(1)-N(2) 1.366(3) Cl(1)-Pd-N(10) 92.12(6) N(1)-C(2) 1.385(3) Cl(2)-Pd-N(2) 100.62(5) N(2)-C(1) 1.355(3) N(2)-Pd-N(11) 79.78(7) N(3)-N(4) 1.347(3) Pd-N(2)-C(1) 112.51(14) N(3)-C(2) 1.378(3) N(2)-C(1)-C(11) 115.70(19) N(4)-C(1) 1.314(3) N(10)-C(11)-C(1) 114.88(19) Pd-N(2) 2.0700(18) Pd-N(10)-C(11) 114.70(14) Pd-N(10) 2.0251(18) N(3)-N(4)-C(1) 115.9(2) Pd-Cl(1) 2.2762(6) N(1)-N(2)-C(1) 114.76(18) Pd-Cl(2) 2.2868(6)
The magnetic data for complex 2.9 is shown as a plot of χ vs. T and χT vs. T in Figure 2.8. In the plot of χ vs. T, there is a maximum at 3 K, below which χ decreases suggesting antiferromagnetic behaviour. This is shown more clearly in the χT plot, where the moment appears to approach 0 emu•K•mol-1 as the temperature decreases to 2 K. The
room temperature moment is 0.37 emu•K•mol-1, as expected for an S = ½ system. Fitting
the χ vs. T data using the Bonner-Fisher linear chain model83 (For equation see Appendix A), with g = 2.0077 (from EPR) fixed; gave J = -1.76 ± 0.01 cm-1 and a quality of fit of
R2 = 0.9996. The χT vs. T data fit gave J = -1.79 ± 0.01 cm-1 with R2 = 0.9977.
Figure 2.7 - Molecular structure of 2.9 (left). Intermolecular contacts within the 1D chains of 2.9 (right). Thermal ellipsoids are at the 50% probability level.
