A Study of Nitronyl and Imino Nitroxide Radicals Attached to Heterocyclic Cores. High Spin Building Blocks Towards Organic Magnets. Dissertation zur Erlangung des Grades „Doktor der Naturwissenschaften“ am Fachbereich Chemie und Pharmazie der Johnnes Gutenberg-Universität in Mainz vorgelegt von Giorgio Zoppellaro geb. in Cassano Magnago, Italien Mainz, 19.11.2004 Dekan: Herr Prof. Dr. R. Zentel 1. Berichterstatter: Herr Prof. Dr. K. Müllen 2. Berichterstatter: Herr Prof. Dr. H. Meier Tag der mündlichen Prüfung: 19.11. 2004 Die vorliegende Arbeit wurde in der Zeit von November 2001 bis September 2004 am Max-Planck-Institut für Polymerforschung in Mainz unter Anleitung von Herrn P.D. Dr. M. Baumgarten und Herrn Prof. Dr. K. Müllen durchgeführt. Non chiederci la parola che squadri da ogni lato l'animo nostro informe, e a lettere di fuoco lo dichiari e risplenda come un croco perduto in mezzo a un polveroso prato. Ah l'uomo che se ne va sicuro, agli altri ed a se stesso amico, e l'ombra sua non cura che la canicola stampa sopra uno scalcinato muro! Non domandarci la formula che mondi possa aprirti, sì qualche storta sillaba e secca come un ramo. Codesto solo oggi possiamo dirti, ciò che non siamo, ciò che non vogliamo. Eugenio Montale, Ossi di sepia, 1923 1975 Nobel Laureate in Literature Zusammenfassung __________________________________________________________________________ Stabile organische Radikale mit zusätzlichen Funktionalitäten wie Donor/Akzepotor Eigenschaften und Ligandeneignung für Übergangsmetallkomplexierung repräsentieren eine synthetische Herausforderung beim Streben nach der Konstruktion hochdimensionaler heterospin Strukturen. In diesem Hinblick wurden acht neue Hochspinbiradikal-Moleküle zusammen mit ihren Monoradikal- Pendants in dieser Arbeit hergestellt. Die Wahl der Liganden als organische Distanzhalter der Radikaleinheiten wurde auf stickstoffhaltige Heterozyklen (Pyridin und Pyrazol) gelenkt. Diese wurden weiterhin mit den stabilen Spinträgern Nitronylnitroxid- (NN) und Iminonitroxidfragmenten (IN) dekoriert. Ihre Synthese beinhaltete mehrstufige Umsetzungen (Brominierung, Iodierung, N- und Carbaldehyd Schutzgruppen, Stille-Kupplung, Grignard Reaktion, etc.) um die Mono- und Dicarbaldehyd- heterocyclenderivate als Schlüsselvorläufer der Radikaleinheiten zu gewinnen. Die Carbaldehyd-Zwischenstufen wurden Kondensationsreaktionen mit 2,3-Dimethyl-2,3- bis(hydroxylamino)-butan unterworfen (üblicherweise in Dioxan unter Argon für ~ 7 Tage), gefolgt von der Oxidation der Bis-hydroxylimidazolidin-Vorläufer unter Phasentransferkatalyse (NaIO4/H2O). Die Radikalmoleküle wurden mit verschiedenen spektroskopischen Methoden untersucht (FT/IR, UV/Vis/ EPR etc.) und ihre Einkristalle mit Röntgenstrahlbeugung gemessen. Die UV/VIS- Lösungsspektren zeigten in einem breiten Bereich verschiedener Lösungsmittelpolaritäten keine spezifische Wechselwirkung zwischen Lösungsmittel und Radikaleinheit, während ihre Stabilitäten in protischen Lösunsgmitteln wie MeOH stark abnahmen. Als Pulver konnten sie jedoch im Kühlschrank an der Luft für eine Jahr gelagert werden, ohne sich zu zersetzen. Die spektroskopischen Fingerabdrücke der Radikale wurden eindeutig identifiziert and erschienen stark abhängig vom Typ des π- Ringsystems an das die Spinträger gekoppelt wurden. Basierend auf diesen Informationen wurde ein schnelles Protokoll etabliert, das eine direkte Zuordnung der Art der Radikale und ihrer Anzahl ermöglicht, sowie ihre Reinheit und Verunreinigungen zu definieren. In Lösung bestätigte die Analyse der EPR Spektren der Biradikale die starke Austauschwechselwirkung J zwischen den Radikalfragmenten über die Kopplungseinheiten (J >> an, an ist die Stickstoffhyperfeinkopplungskonstante). Dies wurde weiter unterstützt durch die Beobachtungen in gefrorener Lösung über die Nullfeldaufspaltungen und verbotenen Halbfeldübergänge (∆ms = 2). Die Temperaturabhängigkeiten der ∆ms = 2 - EPR Signale wurden bis herunter auf 4 K gemessen und das exakte Vorzeichen und die Größe von J ermittelt. Diese Arbeit unterstreicht die Möglichkeit über synthetische Chemie eine Feineinstellung der „through bond“ Austauschwechselwirkung zwischen verwandten π- und σ- konjugierten Heterozyklen zu erreichen, in denen der S = 1 Grundzustand angenommen wird. Zusätzlich zeigten diese Resultate, dass die Übertragung der Spinpolarisation durch verschiedene Koppler sehr effektiv war. Abstract __________________________________________________________________________ Stable organic radicals with additional functionalities such as donor/acceptor properties and ligand capabilities to transition metal ions represent a synthetic challenge in the quest of constructing highly dimensional heterospin structures. In this frame, eight novel high spin biradical systems were prepared in this work, together with their monoradical counterparts. The choice of the ligands, as organic spacer for the radical entities, was directed towards nitrogen containing heterocycles (pyridine and pyrazole). Those were further decorated with the stable spin carrier’s nitronyl nitroxide (NN) and imino nitroxide fragments. Their synthesis involved multi-step procedures (bromination, iodination, N- and carbaldehyde protecting groups, Stille coupling, Grignard reaction, etc.) in order to assemble the mono and biscarbaldehyde hetero-derivatives, the key precursors for the radical entities. Such intermediates were subjected to condensation reaction with 2,3-dimethyl-2,3- bis(hydroxylamino)-butane (generally in dioxane under argon for ~ 7 days), followed by oxidation of the bis-hydroxylimidazolidyn precursor under phase transfer conditions (NaIO4 / H2O). The radical molecules were studied using different spectroscopic techniques (FT/IR, UV/Vis, EPR) and on the single crystals, by X-ray diffractions. The UV/Vis solution spectra witnessed no specific interaction between solvent and radical moieties, in a broad range of solvent polarities, while their stabilities strongly decreased in protic solvents, especially in THF and methanol. As powders, however, they could be stored in cold under air for a year without decomposition. The spectroscopic fingerprints of the radical entities were unambiguously identified and appeared strongly dependent on the type of π-ring system in which the spin carriers were attached to. Based on these information’s, a quick protocol was established that allowed to assign straightforwardly the type of radical and their numbers (monoradical, biradical), to define their purities and the nature of the contaminants. In solution, the EPR analysis in the biradical systems confirmed the strong exchange interaction, J, between the radical fragments through the couplers (J >> aN, with aN the nitrogen hyperfine interaction), further supported by the observation in frozen state of both zero-field-splitting (zfs) and forbidden half-field transitions (∆ms = 2). These findings were consistent with dipolar couplings accounting for S = 1 state species. The temperature dependencies of the ∆ms = 2 EPR signals were followed down to cryogenic temperature (4 K), and the exact sign and magnitude of J were derived. This work underlined the opportunity via synthetic chemistry to fine tune the through-bond exchange interaction among closely related π- and σ-conjugated hetero-systems, in which the S = 1 ground state were preferentially adopted. In addition, these results showed that the propagation of the spin polarization were effective through different couplers. Therefore those constitute unprecedented findings as compared with the limited number of similar systems known in literature in which no comparable effects were observed. ___________________________________________________________________________________________ Table of Contents Chapter 1- Prelude p.1 1.1. Organic molecular magnetism. p.1 1.2. Anticipated properties of organic magnetic materials, pros and cons. p.2 1.3. Classes of organic magnetic molecules, their design and thesis objectives. p.2 References. p.6 Chapter 2- Synthesis of Pyridine and Pyrazole containing Radicals p.11 2.1. Pyridine containing radicals. p.12 2.2. Pyrazolylpyridine containing radicals. p.25 References. p.46 Chapter 3- The Radical´s Optical Properties p.49 3.1. The UV/Vis absorption spectra of the radical systems. p.49 3.2. Theoretical predictions of the UV/Vis absorption spectra in the p.53 biradical systems. 3.3. The IR absorption spectra of the radicals. p.58 Chapter 4- EPR Analysis p.62 4.1. The EPR analysis of monoradical systems in solution. p.62 4.2. The EPR of biradical systems in solution. p.69 4.2.1 The spectral EPR analysis of the biradical systems in solution. p.72 4.2.2 The determination of thermally activated spin-states p.74 in the biradical systems. 4.3. The observed EPR spectra of the π- conjugated biradicals in solution. p.75 4.4. The EPR spectra of the σ - conjugated biradicals in solution. p.82 4.5. The EPR of mono and biradical systems in frozen solutions. p.85 4.6.1 The observed EPR spectra of the monoradical systems in frozen solutions . p.87 4.6.2 The observed EPR spectra of the biradical systems in frozen solutions. p.88 4.7. The EPR saturation behaviour of mono and biradical systems p.92 in frozen solutions. 4.8. Determination of the electronic ground state in the magnetically dilute biradical systems. p.96 4.8.1. The terpyridine based biradicals. p.96 4.8.2. The bispyrazolylpyridine based biradicals. p.99 i Table of Contents ___________________________________________________________________________ 4.8.3. The pyrazolylbipyridine based biradicals. p.101 4.8.4. The σ - conjugated biradicals. p.102 4.9. Conclusion. p.102 References. p.108 Chapter 5- Crystal Structures of the Radicals p.111 5.1. The structure of 5,5"-bis(1-oxyl-3-oxo-4,4,5,5-tetramethyl-imidazolidin-2-yl)2,2':6',2"- terpyridine (12) and the supramolecular π-stacking chain formation. p.111 5.2. The structure of 2,6-bis[4'-(3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazol-1'- yl]-pyridine (26) and the zig-zag chain formation. p.114 5.3. The structure of 4'',5'-bis[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl]-6-(pyrazol- 1''yl)-2,2'-bipyridine (33) and the dimers formation. p.116 5.4. The structure of 6-bromo-5'[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'- bipyridine (8) and the dimers formation. p.117 5.5. Conclusion. p.118 Chapter 6- Summary and Outlook p.120 Chapter 7- Experimental Session p.126 7.1. Materials and Methods. p.126 7.2. Data treatment. p.127 7.2.1. Synthesis of 6-bromo-3-pyridinecarbaldehyde (1). p.128 7.2.2..Synthesis of 2-bromo-5-[1,3]dioxolan-2-yl-pyridine (2). p.129 7.2.3. Synthesis of 2-tributylstannyl-5-[1,3]dioxolan-2-yl-pyridine (3). p.130 7.2.4. Synthesis of 2-tributylstannyl-6-bromopyridine (4). p.130 7.2.5. Synthesis of 6’-bromo-[2,2]’-dipyridinyl-5’-carbaldehyde (5). p.131 7.2.6. Synthesis of 2,3-dimethyl-2,3-bis(hydroxylamino)-butane (6). p.132 7.2.7. Synthesis of 6-bromo-5'[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]- 2,2'-bipyridine (7). p.133 7.2.8. Synthesis of 6-bromo-5'[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]- 2,2'-bipyridine (8). p.134 7.2.9. Synthesis of 6-bromo-5'[1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]- p.134 2,2'-bipyridine (9). 7.2.10. Synthesis of 5, 5"-diformyl-2,2':6',2" terpyridine (10). p.135 7.2.11.Synthesis of 5,5"-bis(1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"- terpyridine (11). p.136 ii Table of Contents ___________________________________________________________________________ 7.2.12. Synthesis of 5,5"-bis(1-oxyl-3-oxo-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"- terpyridine (12). p.137 7.2.13. Synthesis of 5,5"-bis(1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"- terpyridine (13). p.138 7.2.14. Synthesis of triformylmethane (14). p.139 7.2.15. Synthesis of 4-formyl-1(H)-pyrazole (15)(Method A). p.139 Synthesis of 4-formyl-1(H)-pyrazole (Method B). p.140 7.2.16. Synthesis of 4-iodo-pyrazole (16). p.140 7.2.17. Synthesis of 1-(1-ethoxyethyl)-4-Iodo-pyrazole (17). p.141 7.2.18. Synthesis of 4-formyl-1(H)-pyrazole (18). p.141 7.2.19. Synthesis of 2,6-bis(4'-formylpyrazol-1'-yl)-pyridine (19). p.142 7.2.20. Synthesis of 2,6-bis-pyrazol-1-yl-pyridine (20). p.143 7.2.21. Synthesis of 2,6-bis-(4-iIodo-pyrazol-1-yl)-pyridine (21). p.144 7.2.22. Synthesis of 2,6-bis-(4-bromo-pyrazol-1-yl)-pyridine (22). p.145 7.2.23. Synthesis of 2,6-bis-(4-trimethylsilamylethynyl-pyrazol-1-yl)-pyridine (23). p.146 7.2.24. Synthesis of 2,6-bis-(4-ethynyl-pyrazol-1-yl)-pyridine (24). p.147 (See 7.2.19) Synthesis of 2,6-bis(4'-formylpyrazol-1'-yl)-pyridine by Grignard reaction. p.148 7.2.25. Synthesis of 2,6-bis[4'-(1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl)pyrazol-1'- yl]-pyridine (25). p.148 7.2.26. Synthesis of 2,6-bis[4'-(3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazol-1'-yl]- pyridine (26). p.149 7.2.27. Synthesis of 2,6-bis[4-(1-hydroxy-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]- pyridine (27). p.149 7.2.28. Synthesis of 2,6-bis[4-(1-oxyl-3-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]- pyridine (28). p.150 7.2.29. Synthesis of 2-bromo-6-hydrazinopyridine (29). p.151 7.2.30. Synthesis of 6-bromo-2-[4'-formylpyrazol-1'-yl]-pyridine (30). p.151 7.2.31. Synthesis of 6'-(4-formyl-pyrazol-1-yl)-[2,2']-bipyridinyl-5-carbaldeyde (31). p.152 7.2.32. Synthesis of 4'',5'-bis[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-6-(pyrazol-1''- yl)-2,2'-bipyridine (32) . p.153 7.2.33. Synthesis of 4'',5'-bis[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl]-6-(pyrazol-1''- yl)-2,2'-bipyridine (33). p.154 7.2.34. Synthesis of 4'',5'-bis[-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl]-6-(pyrazol-1''-yl)-2,2'- bipyridine (34). p.155 7.2.35. Synthesis of 2-(4-formylpyrazolyl)pyridine (35). p.155 7.2.36. Synthesis of 2[4-(1-hydroxy-3-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]- pyridine (36). p.156 iii Table of Contents ___________________________________________________________________________ 7.2.37. Synthesis of 2[4-(1-oxide-3-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)-pyrazolyl]- pyridine (37). p.157 7.2.38. Synthesis of 2[4-(1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]- pyridine (38). p.157 7.2.39. Synthesis of 2,6-bis(4-formylpyrazolylmethyl)pyridine (39). p.158 7.2.40. Synthesis of 2,6-bis[4-(1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl)-pyrazolyl- methyl]-pyridine (40). p.159 7.2.41. Synthesis of 2,6-bis[4-(1-oxyl-3-oxide-4,4,5,5-tetramethylimidazolin-2-yl)- pyrazolylmethyl]-pyridine (41). p.159 7.2.42. Synthesis of 2,6-bis[4-(1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolylmethyl]- pyridine (42). p.160 Acknowledgements p.161 Curriculum Vitae p.162 iv Table of Contents ___________________________________________________________________________ List of Symbols _________________________________________________________________ │ψ(0)│2 – Unpaired electron density at the nucleus a - Hyperfine coupling A - Hyperfine splitting tensor aH - Proton hyperfine splitting aiso - Isotropic hyperfine splitting (Fermi contact aN - Nitrogen hyperfine splitting term) c - Concentration C - Curie constant d - Distance D - Zero-field-splitting DI- Double Integrated EPR signal intensity g - g tensor ge - Electron g factor gn - Nuclear g factor h - Plancks constant H - magnetic field strength Ĥ- Spin Hamiltonian Î - Nuclear spin operator I - Nuclear spin quantum number J - Spin-spin exchange coupling energy J - Total angular momentum KB - Boltzmann constant L/G - Lorentzian- Gaussian ratio m - Distance vector M - Magnetization MI - Nuclear spin quantum number Ms - Electron spin quantum number NA - Avogadro´s number r - Distance between two atoms Ŝ - Electron spin operator S - Spin quantum number T - temperature Tc - Curie temperature βe - Bohr Magneton βn - Nuclear Magneton ∆BPP - Difference in peak to peak width ∆E - Difference in energy ∆EST - Singlet-Triplet energy difference ∆ν - Frequency shift in wavenumber ε - Extinction coefficient Θ - Torsional angle Θ - Weiss constant µeff - Effective magnetic moment ν - Frequency ρ - Spin density of electron χ - Molar susceptibility χ − Susceptibility M ____________________________________________________________________________ List of Abbreviations ___________________________________________________________________________ Å - Armstrong AF – Antiferromagnetic coupler AO - Atomic orbital b.p. - Boiling point cm - Centimeter CW - Continuous wave CHCl3 - Chloroform CH2Cl2 - Dichloromethane DMF - Dimethylformamide ENDOR – Electron nuclear double resonance EPR - Electron paramagnetic resonance FC - Ferromagnetic coupler g - gram h - Hour hfc - Hyperfine coupling IN - Iminonitroxide IR - Infrared m.p. - Melting point MeOH - Methanol min - Minute MO - Molecular orbital mol - Mole mW - MilliWatt NN - Nitronylnitroxide nm - Nanometer NMR - Nuclear magnetic resonance RT - Room temperature THF - Tetrahydrofuran TLC - Thin layer chromatography UV/Vis - Ultraviolet/visible zfs - Zero-field-splitting v Table of Contents ___________________________________________________________________________ Table I Conversion factors (EPR) dB µWatt √ µWatt dB µWatt √ µWatt 40 20.1 4.4833 23 1010 31.7805 39 25.3 5.02991 22 1270 35.63706 38 31.9 5.64801 21 1600 40 37 40 6.32456 20 2000 44.72136 36 50.4 7.0993 19 2530 50.29911 35 63.6 7.97496 18 3180 56.39149 34 80 8.94427 17 4000 63.24555 33 101 10.04988 16 5040 70.99296 32 127 11.26943 15 6350 79.68689 31 160 12.64911 14 8010 89.4986 30 201 14.17745 13 10100 100.49876 29 250 15.81139 12 1270 35.63706 28 320 17.88854 11 16000 126.49111 27 400 20 10 20100 141.77447 26 510 22.58318 9 25300 159.05974 25 640 25.29822 8 31800 178.32555 24 800 28.28427 7 40100 200.24984 -- 6 50400 224.49944 dB µWatt √ µWatt dB µWatt √ µWatt Table II Conversion factors for Energy Units cm-1 MHz aJ eV kJ/mol kcal/mol K (Kelvin) cm-1 1 2.997925 1.986447 × 1.239842 1.196266 2.85914 × 1.438769 × 104 10-5 × 10-4 × 10-2 10-3 MHz 3.33564 × 1 6.626076 × 4.135669 3.990313 9.53708 × 4.79922 × 10-5 10-10 × 10-9 × 10-7 10-8 10-5 aJ 50341.1 1.509189 1 6.241506 602.2137 143.9325 7.24292 × × 109 104 eV 8065.54 2.417988 0.1602177 1 96.4853 23.0605 1.16045 × × 108 104 kJ/mol 83.5935 2.506069 1.660540 × 1.036427 1 0.239006 120.272 × 106 10-3 × 10-2 kcal/mol 349.755 1.048539 6.947700 × 4.336411 4.184 1 503.217 × 107 10-3 × 10-2 K 0.695039 2.08367 × 1.380658 × 8.61738 × 8.31451 × 1.98722 × 1 (Kelvin) 104 10-5 10-5 10-3 10-3 vi ___________________________________________________________________________________________ Chapter 1 - Prelude It was in 1856 when W. H. Perkin discovered his famous colorant “mauvein” serendipitously [1], and since then organic compounds have played an important role in industrial and material chemistry [2], especially in the field of dyes [3] and pigments [4]. However, it was only after 1950 that in some organic compounds normally having insulating properties, electrons have been found to have conducting properties [5], i.e., several kinds of organic semi-conductors, conductors and even superconductors have been discovered and developed [6]. Compared with the advances in organic conducting materials, the more recent development of organomagnetic materials is rather similar in the sense that even the most sophisticated properties can be rationally designed by a systematic modification of the organic molecular structures. The notion of organomagnetic materials showing metallic properties, such as electron conductivity and ferromagnetism [7], began several decades ago. The goal was to create an assembly of organic molecules or macromolecules containing only light elements (C,H,N,O,S, etc.), and yet possessing the electron/hole mobility or spin alignment that is inherent to metals or their oxides. 1.1. Organic molecular magnetism. The progresses towards obtaining organomagnetic materials have been triggered by the advance in the chemistry of organic radicals [8]. With the notation of “organic radical”, we refer to molecules that contain “unpaired” electrons, each one formally associated with different atomic centers in the isolated molecular unit. In order to achieve magnetic properties in these organic molecules, it is necessary to obtain cooperative interactions among their unpaired electrons, i.e. to keep the unpaired spins paralel to each other. However, the control of such long-range spin interactions turned out to be extremely hard to accomplish [9]. It is consequently not surprising that only a decade or so has passed since high-spin molecules with parallel spin alignment in the bulk have been developed, although the theoretical possibility of organic based magnetic materials was suggested as early as 1963 [10], when McConnell proposed through-space models for building up bulk ferromagnetic interactions among spin carriers. As originally postulated, the assemble of such long range interactions among spin carriers consist of the following sequence of events: (1) the design of the isolated molecular unit, with one, two, or more already interacting entities; (2) a large molecule with several interacting entities; (3) mesoscopic-size molecules with added complexities; and (4) assembly of molecules to supramolecular clusters, monomolecular layers, or bulk solids. The goals of such an approach through rational design and synthesis of molecules and molecular 1 Chapter 1 – Prelude ___________________________________________________________________________ assemblies, are to prepare either materials with superior properties compared to their existing “natural” or artificial counterparts, or to gain better insight to more complex systems [11]. From the preparative organic chemist’s perspective, the first point is crucial. This implies the research of novel spin-containing building blocks (i.e. free radicals and radical ions) and the definition of their exchange coupling. These issues are addressed in this work. 1.2. Anticipated properties of organic magnetic materials, pros and cons. The first application would be the replacement of existing bulk magnets or magnetic recording devices. Values of the saturation magnetisation, Ms, for molecular organic based magnets are comparable to metallic magnets on a molar basis. The inherently large molecular weight (per magnetic moment) of organomagnetic material and their low density, however, result in a smaller saturation magnetisation on either a volume or a mass basis, i.e. the spin concentration is low. As a consequence, they feature lower magnetisations, small exchange energies (because of large interspin distances), and low temperature of transition (Tc, the Curie-temperature) to the ferromagnetic state. This means that organic magnets are unlikely to compare with existing magnets. Further disadvantages are the inherent chemical instability of many organic materials and their “aging” with time. They might offer, however, several advantages with respect to their inorganic counterparts: (1) organomagnetic materials are transparent in nature, therefore a variety of optical properties may be expected, i.e. photomagnetic switches and polarized light manipulation in integrated optical devices; (2) the biocompatibility of such materials may lead to application in magnetic imaging and transducers for medical implants; (3) tuning properties via organic chemistry, processability, low environmental contamination; (4) electronic/magnetic molecular devices for applications in modern computer technology. Other than commercial opportunities, the future realization of the potential of organic ferromagnetism is significant from a basic point of view, in which extended ferromagnetic exchange interactions through s and p orbitals may provide a better insight into the phenomenon of magnetism, far beyond of being fully understood. 1.3. Classes of organic magnetic molecules, their design and thesis objectives. Several spin carrying units are currently in active use towards building organic based magnets. Some examples are the nitronylnitroxide, NN, iminonitroxide, IN, tbutylnitroxide, NO, verdazyl radicals, VZ, carbenes, nitrenes, phenoxides, ArO, ketyl radicals, and triphenyl methyl radical, TPM. These spin units are shown together in Figure 1.1. However, NN, IN, NO, VZ constitute one of the rare classes of radicals capable being handled under ordinary 2 Chapter 1 – Prelude ___________________________________________________________________________ conditions on the laboratory bench, allowing their isolation and purification as stable substances. The choice of the spin carriers in this thesis was directed towards the NN and IN radical’s type. They offer multiple opportunities such as H-bond formation, π-stacking, and even extension to mixed organic-inorganic hybrid structures, thanks to the chelating properties of both N and NO moieties that readily provide coordination sites for metal acetylacetonates. X R R N N + N N N N O N N N X= S;O-O O O R R R R NN IN NO VZ O R R O Ar N R C C C R R Ar Ar R ArO nitrene `trityl` radicalcarbene ketyl TPM Figure 1.1: The most often encountered radical units used for organic molecular magnets. For assembling a high spin molecule, the spin alignment in the organic compound must be arranged in such a way that interactions between the spins are allowed to be parallel (ferromagnetic, S ≥ 1). This appears to work against nature, because organic compounds with unpaired electrons normally have a strong tendency to bind together with antiparralel spins (antiferromagnetic, S=0) or they do not interact at all, behaving as independent spin units. It is therefore necessary to use a "ferromagnetic coupler” (FC) or spacer, which still allows interactions between two or more unpaired electrons but prevents their pairing [12]. FC S = 1/2 S = 1/2 S = 1 3 Chapter 1 – Prelude ___________________________________________________________________________ Conceptually, the interactions among unpaired spins can be regarded as the fine balance of three physical mechanisms [13]: (i) direct coupling; (ii) indirect coupling; (iii) spin polarisation. The design of the molecular backbone allows to decrease the effect of direct coupling, and to tune the indirect coupling (so called through-bond interaction), upon choosing the type of bonds involved (connectivity) and molecular topology. Then, the nature of the radical defines the strength of the spin polarisation contribution [14]. The endeavour of chemists via synthetic design would be, in principle, to gain full control of these three factors. In reality, the most used synthetic tool to accomplish a high spin molecule relies on shaping at least the through-bond interaction between/among unpaired spins, leading to so called non-Kekulé structures, as in m-xylene, where no double bond between the unpaired electrons can possibly be formed (Figure 1.2A). The p- and o- xylene, in contrast, allow spin pairing toward the more stable quinoid structures in the Kekulé forms as shown in Figure.1.2 B, and therefore the low spin state (S = 0) might be expected. non-Kekule´ Kekule´ . . . . * * . * . .* . S = 0 n*-n = 0 * S = 1 * n*-n = 2 * * * * * (A) * *(B) Figure 1.2. non-Kekulé versus Kekulé structures. In case of m-xylene (non-Kekulé), the ground state is thought to be a triplet (S = 1). This is due to the presence of degenerate non-bonding molecular orbitals (NBMOs) of a non- disjoint type. In 1950, Longuet-Higgins [16a] proposed a rule, the topological model, to predict the ground spin state in a π-conjugated molecule on the basis of the Hund´s rule. This relation is given by equation (1), 4 Chapter 1 – Prelude ___________________________________________________________________________ [nNBMO = (N-2T); S = 0.5 (N-2T)] (1) Where, N is the number of π-centres and T the number of double bonds. Later, Ovchinnikov [16c] using valence bond theory and a Pariser-Parr-Pople Hamiltonian, proposed a different model (spin polarization). In this case, if the conjugated carbon framework can be divided in alternant “starred” n* and “unstarred” n (spin polarized) centres, in such a way that each starred atom faces contacts with only unstarred ones, then through equation (2), and by counting half of the difference between n* and n, the value of the net spin S in the system can be anticipated. S = 0.5 (n*-n) (2) Because the spin multiplicity is expressed as 2S+1, the Ovchinnikov rule also predicts triplet ground state for non-Kekulé molecules. Following these approaches many high spin molecules connected by different FC units have been designed and synthesized [15]. Such theoretical guidelines [16], on the other hand, cannot be straightforwardly extended neither to hetero-systems containing radicals (like in NN and IN) and couplers, nor clearly to non- alternant systems (e.g. five member rings). In these cases, several factors such as the specific molecular geometry of the molecule [17], the nature and position of the substituents in the organic spacer [18], and the heteroatom influence [19] seem to influence strongly the ground spin state. To put this thesis work on molecular magnetism into perspective, and to address the other points raised in the introduction, namely how to arrange each single-unit into a supramolecular network [20], the choice of both radicals and type of couplers are therefore determinant. One possible approach to achieve such a goal was based on intermolecular interaction via H- bonding, or stacking among pure organic carriers, and this led to the attainment of bulk ferromagnetism (Tc) where the p-nitrophenyl-nitronylnitroxide made in the Kinoshita group [21] and the 1,3,5,7-tetramethyl-2,6-diazaadamantane-N,N′-dioxyl by Rassat [22] represent two among other outstanding examples [23] (Figure 1.3 A and 1.3 B, respectively). 5 Chapter 1 – Prelude ___________________________________________________________________________ (A) (B) (C) (D) S H3C CH + N N 3 S S N N -O O N N N N O N F F NO F F NO2 Cl CN Tc = 0.65 K Tc = 1.48 K Tc = 0.67 K Tc = 36 K 1991 1993 1995 2000 Ref. 22 Ref. 21 Ref. 23f Ref. 23g Figure 1.3: Examples of pure organic based magnets. Another approach relies on the combination between paramagnetic metal ions and pure organic radicals, leading to a new class of hybrid organic-inorganic materials [24]. Two of such examples are shown in Figure 1.4 and 1.5. N N Mn(hfac) O O Mn(hfac) 2 2 O O Mn(hfac)2 = 2+Mn N O Mn(hfac) O O2 Figure 1.4: The Inoue and Iwamura’s 2D honey-comb metallo-organic higher dimensional compound [Ref.24g]. Figure 1.5: The Mathevet and Luneau’s 3D polymeric metal-radical network [Ref.24h]. 6 Chapter 1 – Prelude ___________________________________________________________________________ A note of caution should be made, nevertheless, about the prospects of purely organic molecular magnets. So far, despite much research effort, the highest Tc recorded is 36 K. There appears to be some barrier to high Tcs, and it is therefore likely that further progress can be made with a high-spin metal organic hybrid complex crystal or polymer. Therefore, the hybrid approach has been pursued within this research work, by making accessible novel synthetic routes towards chelating coupling unit cores (terpyridine, bipyridine, pyrazolylpyridine), followed by their functionalisation with stable nitronyl- and iminonitroxide radical moieties. Through this thesis, the reader will find in Chapter 2 the synthetic strategies towards such molecular entities, which in turn set up flexible routes for further extension on analogous heterosystems. Besides synthesis, the full characterization of the radical systems and their intermediates are needed. Therefore, a comprehensive study of their optical properties (absorption and IR spectroscopy) is given in Chapter 3. Some practical tools for assessing the precursor/radical fingerprints and their purities are provided. The Chapter 4 presents the whole EPR analysis of the monoradical and biradical systems. When more than one spin unit (i.e. biradicals) is connected to the organic core, the theoretical and experimental steps necessary to define their molecular ground state spin multiplicity are described in detail. Finally, in Chapter 5, the molecular structures of one monoradical and three biradical systems are analysed, together with a perspective of their supramolecular assemblies. 7 Chapter 1 – Prelude ___________________________________________________________________________ References [1] W.H. Perkin, Br. Pat., 1984, 1856. [2] K. Ziegler and G. Natta, Nobel Lecture, December 12, 1963. [3] J. F. W. A. von Baeyer (1905), Nobel Lecture, Chemistry, 1901-1921, Elsevier Publishing Group, Amsterdam 1966. [4] R. M. Willstätter (1915), Nobel Lecture, Chemistry, 1901-1921, Elsevier Publishing Group, Amsterdam 1966. [5] (a) H. Shirakawa, E.J. Louis, A.G. MacDiarmid, C.K. Chiang, A.J. Heeger, Chem. Commun., 1977, 578. (b) C.K. Chiang, C.R. Jr. Fincher, Y. W. Park, A.J. Heeger, H. Shirakawa, E.J. Louis, Phys. Rev Lett., 1977, 39, 1098. (c) B. Råndy, In Cojugated Polymers and Related Materials: The Interconnection of Chemical and Electronic Structures, W.R Salaneck, I. Lündström, B. Råndy, Eds., Oxford University Press: Oxford, UK, 1993, Chapter 3. (d) A.J. Heeger, J. Phys. Chem. B, 2001, 36, 8475. [6] M. R. Bryce, Chem. Soc. Rev., 1991, 20, 355. (b) J. M. williams, A.J. Schultz, U. Geiser, K. D. Carlson, A. M. Kini, H.H. wang, W.-K. Kwok, M.-H. Whangbo, J. E. Screiber, Science, 1991, 252, 1501. [7] (a) H.K.J., Bushow, E.P. Wohlfart, Ferromagnetic Materials, Eds., North Holland: Amsterdam, 1980-1990, Vols. 1-5. (b) K.H. Fisher, J.A. Hertz, Spin Glasses, Cambridge University Press: Cambridge, 1991. (c) D.R. Tilley, J. Tilley, Superfluidity and Superconductivity, Hilger, Bristol, 1986. [8] (a) G. Herzberg, Nobel Lecture, December 11, 1971. (b) A. R. Forrester, J. M. Hay, R. H. Thomson, Organic Chemistry of Stable Free Radicals, Academic Press, New York, 1968.(c) E. G. Rozantsev, Free Nitroxyl Radicals, Plenum Press, New York, 1970. (d) E. G. Rozantsev, V. D. Sholle, Synthesis 1971, 190, 401.(e) . Sayre, J. Am. Chem. Soc. 1955, 77, 6689. (f)M. Lamchen, T. W. Mittag, J. Chem. Soc. 1966, 2300. (g) E. F. Ullman, J. H. Osiecki, D.G.B.Boocock, J. Am. Chem. Soc. 1972, 94, 7049. (h) J. F. W. Keana, Chem. Rev. 1978, 78, 37; (i) L. B. Vordarsky, Imidazoline Nitroxides, CRC Press, Boca Raton, Florida, 1988, Vol.I-II. (l) H. G. Aulich in Nitrones, Nitronates and Nitroxides, S. Patai and Z. Rappoport (Eds.) John Wiley and Sons, New York, 1989, p. 313. (m) M.-E. Brik, Heterocycles 1995, 41, 2827. [9] (a) O. Kahn, Magnetism: A Supramolecular Function, Eds., Kluwer, Dordrecht, 1996. (b) J. S. Miller, M. Drillon, Magnetism: Molecules to Materials III, Wiley-VCH, Weinheim, 2001. (c) J. S. Miller, M. Drillon, Magnetism: Molecules to Materials I, II, IV, Wiley-VCH, Weinheim, 2003. [10] H. M. McConnell, J. Chem. Phys. 1963, 39, 1916. [11] ](a) J. S. Miller and A. Epstein, Angew. Chem. Int. Ed. 1994, 106, 399. (b) S. Nakatsuji and H. Anzai, J. Mat. Chem. 1997, 7, 2161. (c) J. A. Crayston, J. N. Devine, J. C. Walton, Tetrahedron 2000, 56, 7829. (d) U. Hartmann, Ann. Rev. Mat. Sci., 1999, 29, 53. (e) R.J. Bushby, J.-P. Paillaud, Introduction to Molecular Electronics; M.C. Petty, M.R. Brice, D. Bloor, 8 Chapter 1 – Prelude ___________________________________________________________________________ Eds., Edward Arnold: London, 1995, Chapter 4. (f) J.A. Crayston, J. N. Devine, J.C. Walton, Tetrahedron, 2000, 56, 7829. [12] P. Lafenete, J.J. Nova, M.J. Bearpark, P. Celani, M. Olivucci, M.A. Robb, Theor. Chem. Acc. 1999, 102, 309. [13] (a) Y. Molin, K. M. Salikhov, K. I. Zamaraev, Spin Exchange, Springer Verlag, Berlin, 1980, p.11. (b) E. Coronado, B. S. Tsukerblat, R. Georges, in: E. Coronado et al. (Eds), Molecular Magnetism: From Molecular Assemblies to the Device, NATO ASI Series E, vol. 321, 1996, p. 65. [14] J. E. Wertz, J. R. Bolton, Electron Spin Resonance, Elementary Theory and Practical Applications, Chapman & Hall, 1986. [15] (a) A. Calder, A.R. Forrester, P.G. James, G.R. Luckhurst, J.Am.Chem.Soc. 1969, 91, 3724. (b) K. Inoue, H. Iwamura, Angew.Chem.Int.Ed. 1995, 34, 927. (c) T. Ishida, H. Iwamura, J.Am.Chem.Soc. 1991, 113, 4238. (d) F. Kanno, K. Inoue, N. Koga, H. Iwamura, J.Phys. Chem. 1993, 97, 13267. (e) K. Inoue, H. Iwamura, J.Am.Chem.Soc. 1994, 116, 3173. (f) T. Itoh, K. Matsuda , H. Iwamura, Angew.Chem. 1999, 111, 1886. (g) T. Itoh, K. Matsuda ; H. Iwamura, K. Hori, J.Am.Chem.Soc. 2000, 122, 2567. (h) D. Shiomi, M. Tamura, H. Sawa, R. Kato, M. Kinoshita, Synt.Met. 1993, 56, 3279. (i) L. Catala, P. Turek, J. Le Moigne, A. De Cian, N. Kyrisakas, Tetrahedron Lett. 2000, 41, 1015. (j) F. Mathevet, D. Luneau, J.Am.Chem.Soc. 2001, 123, 7465 - 7466. [16] (a) H. C. Longuet-Higgins, J. Chem. Phys. 1950, 18, 265. (b) W. T. Borden, E. R. Davidson, J. Am. Chem. Soc. 1977, 99, 4587. (c) A. A. Ovchinnikov, Theor. Chim. Acta 1978, 47, 297. (d) P. M. Lahti, Magnetic Properties of Organic Materials, Marcel Dekker, New York, 1999. [17] K. Okada, T. Imakura, M. Oda, M. Baumgarten, J. Am. Chem. Soc. 1996, 118, 3047 [18] (a) M. Dvolaitzki, R. Chiarelli, A. Rassat, Angew. Chem. Int. Ed. Engl. 1992, 31, 180. (b) F. Kanno, K. Inoue, N. Koga, H. Iwamura, J. Am. Chem. Soc. 1993, 115, 847. [19] (a) A.P. Jr. West, S.K. Silverman, D.A. Dougherty, J. Am. Chem. Soc. 1996, 118, 1452. (b) S.V. Chapyshev, R. Walton, J.A. Sanborn, P.M Lahti, J. Am. Chem. Soc. 2000, 122, 1580- 1588. (c) M. Rule, A.R. Matlin, D.E. Seeger, E.F. Hilinski, D.A. Dougherty, J.A. Berson, Tetrahedron. 1982, 38, 787. (d) Y. Liao, C. Xie, P.M. Lahti, R.T. Weber, J. Jiang, D.P. Barr, J. Org. Chem. 1999, 64 (14), 5176-5182. [20] (a) J. S. Miller, A. Epstein, W. M. Reiff, Chem. Rev. 1988, 88, 201. (b) A. Rajca, Chem. Rev. 1994, 94, 871. (c) J. S. Miller, Inorg. Chem. 2000, 39,4392.(d) D. Luneau, Curr. Op. in Sol. State Mat. Sci. 2001, 5, 123. (e) for a collection of the recent progreesses in the field see Proceedings of the 8th International Conference on Molecule-Based Magnets (ICMM 2002), Polyhedron 2003, 22, 1725-2584. 9 Chapter 1 – Prelude ___________________________________________________________________________ [21] M. Tamura, Y. Nakazawa, D. Shiomi, K. Nozawa, Y. Hosokoshi, M. Ishikawa, M. Takahashi, M. Kinoshita, Chem. Phys. Lett. 1991, 186, 401. [22] R. Chiarelli, M. A. Novak, A. Rassat and J. L. Tholence, Nature (London) 1993, 363, 147. [23] (a) T. Sugawara, M. M. Matsushita, A. Izuoka, N. Wada, N. Takeda, M. Ishikawa, J. Chem. Soc. Chem. Commun. 1994, 1723. (b) J. Cirujeda, M. Mas, E. Molins, F. L. dePhantou, J. Laugier, J. G. Park, C. Paulsen, P. Rey, C. Rovira, J. Veciana, J. Chem. Soc. Chem. Commun. 1995, 709. (c) J. Veciana, J. Cirujeda, C. Rovira, J. Vidal-Gancedo, Adv. Mater. 1995, 7, 221. (d) A. Caneschi, F. Ferraro, D. Gatteschi, A. leLirzin, M. A. Novak, E. Rentschler, R. Sessoli, Adv. Mater. 1995, 7, 476. (e) Y. Pey, O. Kahn, M. A. Aebersold, L. Ouahab, F. LeBerre, L. Pardi, J. L. Tholence, Adv. Mater. 1994, 6, 681. (f) K. Mukai, K. Konishi, K. Nedaki, K. Takeda, J. Magn. Mater., 1995, 140, 1449. (g) P. Carretta, D. Gatteschi, A. Lascialfari, Physica B. 2000, pp. 94-105. [24] (a) J.Cirujeda, L. E. Ochanko, J. M. Amigo, G. Rovira, J. Rius, J. Veciana, Angew. Chem. 1995, 107, 99-102; Angew. Chem. Int. Ed. Engl. 1995, 34, 55-57. (b) D. A. Shultz, S. H. Bodnar, K. E. Vostrikova, J. W. Kampf, Inorg. Chem. 2000, 39, 6091-6093. (c) H. O. Stumpf, L. Ouahab, Y. Pei, D. Grandjean, O. Kahn, Science 1993, 261, 447-449. (d) K.Inoue, T. Hayamizu, H. Iwamura, D. Hashizume, Y. Ohashi, J. Am. Chem. Soc. 1996, 118, 1803. (d) G. Ballester, E. Coronado, C. Giménez-Saiz, F. Romero, Angew. Chem. 2001, 113, 814. Angew. Chem. Int. Ed. Engl. 2001, 40, 792. (e) K. Fegy, D. Luneau, T. Ohm, C. Paulsen, P. Rey, Angew. Chem. 1998, 110, 1331. Angew Chem. Int. Ed. Engl. 1998, 37, 1270. (f) C. Benelli, D. Gatteschi, Chem. Rev. 2002, 102, 2369. (g) K. Inoue, H. Iwamura, J.Am.Chem.Soc. 1994, 116,3173-3174. (h) F. Mathevet, D. Luneau, J. Am. Chem. Soc. 2001, 123, 7465. 10 ___________________________________________________________________________________________ Chapter 2 - Synthesis of Pyridine and Pyrazole containing Radicals Nitronyl (NN) and imino nitroxide (IN) free radicals were described in the 1970´s by Ullman in his pioneering work [1-3], and since then they have attracted much attention with a revival in the 1990´s. This increasing interest mainly stems from the use of these paramagnetic species as building blocks for designing molecular magnetic materials [4]. Successful achievements in this field such as purely organic ferromagnetically ordered solids and metal-organic exchange-coupled complexes that exhibit versatile magnetic properties have triggered the synthesis of hundreds of these free radicals [5]. The synthesis of nitronyl- and imino nitroxides relies almost exclusively [6] on the condensation of 2,3- bishydroxylamino-2,3-dimethylbutane [7a] with an aldehyde, and oxidation of the condensation product to afford the radical derivatives. However, from either the preparative synthetic and physical chemist’s perspectives, the challenge is dual [8]. It consists not only on building carbaldehyde derivatives anywhere, but also to topologically control the position such that it allows ferromagnetic intramolecular interactions among the spin carriers through the organic backbone (coupling unit, CU). The synthetic choice for the coupling unit cores in this thesis were directed towards the synthesis of hetero-ligands based on terpyridines and pyrazolylpyridines. Both are known for their rich coordination chemistry [9] and have been used widely CU in supramolecular assemblies [10], in molecular biology [11] optical devices [12] spin-crossover R R compounds [13] and even in photochemistry [14]. Unfortunately substituted terpyridines and O bispyrazolylpyridines [15] were limited by tedious NNN = multistep synthesis. In addition, carbaldehyde- + N functionalised pyridines and pyrazoles are overall the R - O less known derivatives. In this 2nd Chapter are O described the synthetic pathways that led to novel N protocols for the nitronyl- and imino nitroxide radicals IN = (R) attached to terpyridine, bispyrazolylpyridine and N pyrazolylbispyridine cores as heterocyclic functional Figure 2.1: Schematic sketch of coupling unit (CU)(Figure 2.1). the heterocyclic functional coupling unit core (CU) functionalized with radical moieties (R). 11 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ 2.1. Pyridine containing radicals Only three terpyridine ligands bearing Ullman radicals were synthesized so far [16]. Those included functionalisations either in the positions 6-6" of the terminal pyridine rings or in the 4' position of the central pyridine ring. 4' 5' 3' 6' N 2' 1' 5'' N N 5 6'' 6 None of them showed to encompass a triplet ground state. Our target was to combine into the terpyridine core two Ullman radicals in position 5-5" namely 5,5"-bis(1-oxyl-3-oxo-4,4,5,5- tetramethylimidazolin-2-yl)2,2':6',2"-terpyridine (12) and 5,5"-bis(1-oxyl-4,4,5,5- tetramethylimidazolin-2-yl)2,2':6',2"-terpyridine (13). In these positions the spin carriers would feature intramolecular ferromagnetic interaction with formation of triplet (S=1) ground state. O N O N N N N N N N N N N + 12 + N 13N N O - NN-Terpy - O O IN-Terpy O Two recent reports described the bis-carbaldehyde terpyridine 10 functionalised in 5-5" [17], and both afforded the product in low amount (from 45 to 290 mg per run) after a multistep synthesis. We therefore searched for a different route that could leave open further synthetic extensions. The synthetic path is described in Scheme 2.1. 12 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ O O Br 1) nBuLi / Et2O H HO OH O Br 2) DMF reflux / C HN Br N 6 6 Br N - 78 °C 1 2 67 % 87 % 1) nBuLi 2) (But)3 SnCl O O Prepared and used (But)3 Sn N in situ 3 Br N Br 1) nBuLi / Et O2O 1 2) (But)3 SnCl H Br N N Br Prepared and used N in situ H (But)3 Sn N Br 5% Pd(PPh3)2Cl2 5 10% PPh 4 3 O 58 % 4' 3' 5' N 2'' Pd(II)/ PPh3 N N H HCl 6N 2' N 6' 5 + 3 O H N N 5'' H T 5 O O 6 1O 10 1'' 6'' O 73 % HOHN 6 HOHN NN-Terpy 13 Excess NaIO4 / H2O 16 % HO N OH N N N N NaIO4 / H2O 11 IN-Terpy 12 N 86 % N 28 % OH HO Scheme 2.1 13 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ The first step relies on building the 6-bromo-3-pyridinecarbaldehyde (1), starting from the commercially available 2,5-dibromo-pyridine. This bromo-derivative can be selectively mono-lithiated in position 5 upon working at -78°C using ether as reaction solvent (Scheme 2.2). The reaction in ether is heterogeneous, due to the poor solubility of the 2,5-dibromo- pyridine in this solvent at low temperature (< 40°C). Nevertheless the lithium derivative is completely soluble at -78°. As reported in the short description found in literature [18], the two bromine groups in position 5 and 2 feature different reactivity towards lithium exchange. In particular the bromine in position 5 seems kinetically more reactive, while the bromine in position 2 is thermodynamically more stable. Thus, long aging of the reaction mixture with the lithiating agent should lead to substitution in 2, and short aging (< 60 min) should give major substitution in 5. As pointed out by the authors [18], also the solvent plays a role. Polar solvents (e.g. toluene) favor substitution in position 2, while apolar solvents favor substitution in position 5. Even though the authors suggested either ether or THF, several trials made by using THF or other solvents like glyme in which the starting material is far more soluble, led to much less overall yield. Also the concentration of the starting pyridine halide is important for achieving a successful reaction, and should be kept within 0.1-0.2 M. The quenching of the 5- lithio-2-bromo-pyridine with N-N-dimethylformamide (DMF) gave straightforwardly the corresponding 6-bromo-3-pyridinecarbaldehyde (1) (yield 67%) [19], after hydrolysis in NH4Cl and chromatographic separation on silica column. Br 1) nBuLi CHO Br N 2) DMF Br N 1 Scheme 2.2 The second step consisted in the protection of the carbaldehyde group in 1 (Scheme 2.3). It was easily achieved by formation of the correspondent dioxolane 2-bromo-5- [1,3]dioxolan-2-yl-pyridine (2) using toluene-4-sulfonic-acid as catalyst (CH3C6H4SO3H × H2O, 8% mol) with ethyleneglycole as protecting group. The benzene proved to be the best solvent for this reaction. The product 2 was obtained as yellowish oil that became solid on standing (yield 87%). O CHO HO OH O Br N Reflux Br N 1 2 Scheme 2.3 14 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ Lehn and co-workers [19] suggested a different procedure for this reaction, using Amberlist-15 as acid catalyst (yield 82%), but without providing any synthetic explanation. In a later report they recommended a different protecting group based on propanediol and PPTS (C5H5N+HtsO-, yield 80%) [16b]. Our procedure might offer an easier alternative route. CHO Br N (1) O O Br N (2) Figure 2.2: 13C-NMR (250 MHz, r.t.) spectra of 1 and 2 recorded in CDCl3. Further, the lithium exchange of the bromine in position 2 [20] followed by quenching with Bu3SnCl afforded 2-tributylstannyl-5-[1,3]dioxolan-2-yl-pyridine (3) (Scheme 2.4). The compound 3 was used later for Stille coupling reaction with derivative 5 (6-bromo-2,2’- dipyridine-5’-carbaldehyde). The next synthetic part illustrates the preparation of the monostannyl-derivative starting from commercially available 2,6-dibromopyridine. The 2,6- dibromopyridine was initially mono-lithiated working at -40°C in ether (Scheme 2.5). The subsequent quenching with tributyltin-chloride (Bu3SnCl) was achieved by lowering the temperature at -78°C and yielded 2-tributylstannyl-6-bromopyridine 4 as pale yellowish oil. 15 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ O O O 1) nBuLi O 2) (Bu)3SnClBr N Sn N 2 3 Scheme 2.4 1) nBuLi Br N Br 2) (Bu)3SnCl Sn N Br 4 Scheme 2.5 The stannyl-compounds 3 and 4, apart from the known toxicity of similar compounds [21], cannot be chromatographed either in alumina and silica column without substantial decomposition. Although they seemed quite stable at room temperature, they cannot be stored for long time even under argon. Thus they were prepared and used in situ. The Stille coupling between 4 and 6-bromo-3-pyridinecarbaldehyde (1) (Scheme 2.6) in anhydrous toluene under argon gave the first precursor for the synthesis of the biscarbaldehyde terpyridine, 6-bromo-[2,2’]-dipyridinyl-5’-carbaldehyde (5). The optimized conditions obtained for the coupling reaction were achieved by using an excess of 1 (2.4 eq) with respect to 4 (1 eq), together with Pd(PPh3)2Cl2 (5% mol) and PPh3 (10% mol) as catalyst, without addition of the additive CuI (yield 58%). Trials made in the presence of CuI in catalytic amount (<5% with respect to 4) did not improve the yield, while higher amounts (up to 10% with respect to 4) led to a drastic decrease on the whole yield (< 30%). Although it is reported that CuI can generally raise the reaction rate (>102) due to its free ligand scavenging ability, strong ligands in solution (e.g. polypyridines) may compete and inhibit the rate-limiting transmetalation step [22]. CHO + Pd(II) N Br Sn N Br Br N Toluene, reflux N OHC 4 1 5 Scheme 2.6 A side product obtained from the Stille reaction consisted in the homo-coupling of 1 to afford the [2,2']bipyridinyl-5,5'-dicarbaldehyde (A) (Scheme 2.7). 16 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ CHO Pd(II) OHC CHO Br N Toluene, reflux N N 1 A Scheme 2.7 d c Compound A OHC CHOa b N N Figure 2.3: 1H-NMR (250 MHz, r.t.) spectrum recorded in DMSO-d6; note that the signals (d) and (b) are further splitted (doublet-doublet) by through space interaction with the carbaldehyde proton (a). This reaction, whose mechanism is not known, was substantiated by using 1 (2 eq.), Pd(PPh3)2Cl2 (5% mol) and PPh3 (10% mol) as catalyst and tributyltin-chloride (Bu3SnCl, 0.3 eq.), without coupling partner. After 60 hours of reaction the compound A was obtained in an unexpectedly good yield (60%) after chromatographic purification on silica column (Scheme 2.8). CHO 1) Pd(II) / (Bu)3SnCl OHC CHO Br N Toluene, reflux N N 1 A Scheme 2.8 It is worth to notice that similar reports on homocoupling under Stille conditions can be found in literature [15i]. In addition one report based on homocoupling induced by hexa-n- butyldistannane compound (e.g. on 2,5-dibromopyridine leading to 5,5’dibromo-bipyridine) 17 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ recently appeared [23]. Although deeper investigation on A was not conducted, the reaction 2.8 may represent an alternative synthesis of A with respect to that one reported [24] based on Swern oxidation of the alcohol 5,5’-hydroxymethyl-[2,2’]bipyridine as shown in Scheme 2.9. It is also interesting to point out that the more classical nickel (II) catalyzed homocoupling largely employed for aryl halides has never been tested in the case of aldehyde substituted pyridine halides [25]. O Cl R Cl Cl Cl O - CO + OH O O -COO O 2 S H C CH +S 3 3 HO N N H3C CH3 S+ H C CH Et3N3 3 H3C CH3 S+ R O H Et3N OHC CHO O N N H R A SH3C CH2 O S+ H3C CH3 Scheme 2.9 Finally, the Stille-coupling reaction between 3 and 5 gave the 5,5"-diformyl-2,2':6',2"- terpyridine (10) (yield 73%)(Scheme 2.10). This reaction occurred in degassed and dry toluene under argon, in a similar way as found for 1 and 4, by heating the reaction mixture to reflux for 60 hours; further hydrolysis of the dioxolane T in the presence of HCl (6 N) followed by basification provided the dialdehyde 10. O HCl O Pd(II)N Br T reflux+ N (Bu)3Sn N N N NOHC OHC CHO 3 5 10 Scheme 2.10 Some comments may be useful at this point. A major consideration in working up reaction mixtures from Stille coupling is the removal of the tin byproducts. While trimethyltin chloride is water soluble and rather volatile, tributyltin chloride has a low volatility and is 18 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ soluble in most organic solvents. Separation on silica gel is difficult due to the tendency of tributyltin to elute even under non polar conditions, and to streak on the column. However, with basic compounds, e.g. oligopyridines, the workup of the reaction mixture is somehow easy. While oligopyridines are soluble in concentrated hydrochloric acid, the tin-byproducts can be removed by extraction with dichloromethane. Neutralisation of the acid phase gives the free oligopyridine ligands. In presence of sensitive groups like carbaldehydes, the basification needs to be carried out with sodium or potassium carbonate rather than sodium hydroxide in order to prevent Cannizzaro reactions that lead to the formation of the diacid. This derivative, very soluble in water, was on the other hand, not completely characterized. The organic compounds which contain no trace of tin byproducts can be extracted with CH2Cl2 and purified by chromatography. The diformyl derivative 10 (1 eq) was finally subjected to Ullman coupling with 2,3-bishydroxylamino-2,3-dimethyl-butane (6, 3 eq) using a mixture of 1,4-dioxane, trichloro-methane and methanol (4/3/3) as solvents for 7 days, under argon at room temperature to afford the white precipitate 5,5"-bis(1,3-dihydroxy-4,4,5,5- tetramethylimidazolidin-2-yl)2,2':6',2" terpyridine (11) as radical precursor (yield 86%)(Scheme 2.11). 6 NaOH HN NH H2SO4 THF / 4°C HN NH OH OH OH OH 50% O O HO N OHN + 6 N N MeOH / CHCl N N N N 3 OHC CHO 11 10 N N OH HO Scheme 2.11 Heating of the reaction mixture (50°C) in order to speed up the condensation induced fast decomposition either of 6 and 11, and no precipitate was obtained. Only an oily mixture was recovered (the solution turned into pale orange after 1 day) that afforded a very small amount of the biradical 12 or 13 after oxidation with NaIO4 (vide infra). It is worth to notice that when part of the tin-biproducts were left in the dialdehyde 10, they catalyzed fast decomposition (oxidation) of the 2,3-bishydroxylamino-2,3-dimethylbutane 6. A very small amount of the radical precursor 11 was obtained after the condensation reaction, indicative for the presence of two competitive mechanisms. A pale rose colour of the reaction mixture was observed in this case. The decomposition of 6 occurs slowly in presence of oxygen to acetone 19 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ oxyme (pale rose, see Experimental Session) according to a mechanism similar to the synchronous trans-elimination [7b] (Scheme 2.12). Since the presence of O2 in the medium is ruled out and the condensation reaction between 6 and 10 appeared rather slow, any tin- residues left should act somehow faster on 6 with a mechanism similar to the hydrogen abstraction induced by dioxygen. Therefore only very pure 10 and careful exclusion of O2 could lead to successful formation of 11. OH NHOH NH OH N OOH- + O +2 OH NH NHOH OH NH+ OH 2 N + H2O2 Scheme 2.12 The compound 11 was oxidized at room temperature (Scheme 2.13) working under phase transfer conditions (CHCl3/CH2Cl2/H2O), by using slight excess of NaIO4 (2.5 eq) with respect to 11 (1.0 eq) for 30 min using argon saturated solution. After column chromatography 5,5"-bis(1-oxyl-3-oxo-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"-terpyridine (12) was obtained as deep-green powder (yield 16%). Similarly but using an excess of NaIO4 (4 eq) with respect to 11 (1.0 eq) and warming the mixture up to 40 °C, the 5,5"-bis(1-oxyl-4,4,5,5- tetramethylimidazolidin-2-yl)2,2':6',2"-terpyridine (13) was obtained as orange-red powder (yield 28%). O N O N N N N 2 eq. NaIO N + 124 + N HO N OH O - - O N N N N N 11 N OH HO Excess NaIO4 N N N N N N 13 N O O Scheme 2.13 One of the by products identified after prolonged oxidation carried under argon at room temperature (2 hours) with stoichiometric amount of NaIO4 with respect to 11 (2 eq./1 eq.) was 5,5''-bis-(4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazol-2-yl)-[2,2';6',2'']terpyridine. This 20 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ terpyridine derivative could only be eluted through alumina column upon using MeOH as polar solvent. Therefore the major problems encountered in obtaining compounds 12 and 13 were avoiding the loss of one or both oxygen groups respectively (dehydratation) by balancing the amount of oxidizing agent and the reaction time. The nitronylnitroxide derivative 12 was obtained always in smaller amount with respect to the imino radical 13. This result arises by the fact that the oxidation reaction with NaIO4 occurs at the interface between H2O and CHCl3, leading to a statistical mixture of monoxidized, fully oxidized and overoxidised/decomposed 11. The use of another oxidizing agent largely employed in the case of precursors for nitroxide radicals (PbO2) showed to be far less effective in controlling side reactions. The 6- bromo-[2,2’]-dipyridinyl-5’-carbaldehyde (5) was used for the synthesis of the monoradicals 6- bromo-5'[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'-bipyridine (8) and 6-bromo- 5'[1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'-bipyridine (9) according to the Scheme 2.14. 6 HN NH HO N Br N Br OH OH N N N 7 OHC 5 dioxane / CHCl3 4 days in argon NOH NaIO4 Excess NaIO4 O N Br O N Br N N 9 N N 8 N N+ -O Scheme 2.14 The condensation with 2,3-bishydroxylamino-2,3-dimethyl-butane (6) afforded the radical precursor 6-bromo-5'[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'-bipyridine (7) (yield 85%). In Figure 2.4 and 2.5 are shown the 1H-NMR spectra of 5 and 7. Subsequent oxidation of 7 with sodium periodate afforded either 8 (yield 25%) or 9 (yield 28%) depending on the amount of oxidant used. As previously mentioned in the case of 11, heating the reaction mixture should be avoided. In fact, while in principle we might increase the amount of at least the imino compound 9, in practice the dehydratation process of 7 made it hard to control the radical oxidation with NaIO4 and almost no imino nitroxide radical was recovered after this step. Also in this case, the use of the other types of oxidazing agents (PbO2) that allowed to avoid the phase transfer conditions, gave a smaller overall yield for both 8 and 9 (< 15%). In the 1H-NMR spectra, the precursors of the Ullman radicals showed always characteristic peaks associated with the imidazolyl moiety. As reported for compound 7 in 21 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ Figure 2.5, and compound 11 in Figure 2.6C, the C-H proton of the imidazolyl ring featured a well defined resonance around 4.5- 4.6 ppm, while the two cis/trans methyl carbons –CH3 gave two singlet at ~1.0 ppm. In solution all the pyridine-based radicals were very sensitive in presence of traces of acids. The following order of stability 9 >8 >13 >12 (from more stable to less stable) can be drawn. These have been defined by monitoring the decrease of the double integration of their EPR signals at room temperature versus time. In solvents like toluene or hexane they all showed high stability (up to a year), and unexpectedly also in ethylacetate and 2-propanol. They should not be kept for long time in acetone, CH2Cl2 and CHCl3 although they feature high solubility in these solvents. Other solvents like MeOH and EtOH or ethereal solvents like THF destroyed the biradical 12 and 13 very fast (~ one day, see Table 2.1 at the end of this Chapter) while the monoradicals 8 and 9 were completely lost in approximately one week. The solvent effect on the radical stabilities could be easily monitored by observing a fading in the blue or red colour being accompanied with the absence of EPR signal either in solution or in frozen state. The purifications of these radicals were always carried out on neutral alumina (Al2O3) since they showed same trend of decomposition when silica (SiO2) was used. 22 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ Compound 5 b e a d c N Br N OHC g f Figure 2.4: 1H-NMR (250 MHz, r.t.) spectrum recorded in CDCl3. Note that the signals (c) and (f) are further splitted in doublets by (g) as found previously in compound A (see its 1H-NMR spectrum in Figure 2.3). Compound b 7 e ad c HO N Br N g N N f h OH Figure 2.5: 1H-NMR (250 MHz, r.t.) spectrum recorded in DMSO-d6. 23 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ Compound T a c (A) d hg N b N N l OHC O f e i m O a b Compound d 10 c N N N OHC e CHOf (B) Compound b d 11 e c HO N OH (C) N g N N N h N f N OH HO a Figure 2.6: 1H-NMR (250 MHz, r.t.) spectra of the terpyridine derivatives recorded in DMSO- d6. 24 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ 2.2. Pyrazolylpyridine containing radicals In order to extend the work from the terpyridine to a similar hetero-system, the 2,6- bispyrazolylpyridine core was then considered. Pz2Py can be regarded as a terpyridine- analogue, since it reproduces the tridentate nitrogen binding motif of the terpyridine core. The final target was to attach in the positions 4’,4’’ two NN or IN radical fragments. Pz2PY 4 3 5 2' 2" N 2 6 N 3' N N N1' 1 1" 3" 4' 5' 5" 4" Such design led to the novel symmetric biradical derivatives 2,6-bis[4'-(3-oxide-1-oxyl- 4,4,5,5-tetramethylimidazolin-2-yl)pyrazol-1'-yl]-pyridine (26) and 2,6-bis[4-(1-oxyl-3-4,4,5,5- tetramethylimidazolin-2-yl)pyrazolyl]pyridine (28). N N NN N N N N N N O 26 O N N N 28 N+ N + N N N -O O O O- Unfortunately we did not find in literature either estabilished synthetic routes towards the 4’,4’’-biscarbaldehyde functionality, that represents the key precursor for the Ullman radicals, nor any type of radical units appended anywhere on the 2,6-bispyrazolylpyridine backbone. In order to justify the synthetic effort the following considerations have been taken into account. (1) The synthetic development of novel functionalities in position 4’,4’’, besides the radical units, might in principle enable easier metal complexation without the hindrance induced by the usually encountered substituents for the terminal pyrazoles (methyl, phenyl, etc.) in the positions 3’,3” or 5’,5”. In order to attain coordination of the metal in 3’,3” or 5’,5 bispyrazolylpyridine derivatives often are required prolonged reaction time with the metal salt and high temperatures. (2) Novel functionalizations of 4’,4” substituted bispyrazolylpyridines are promising precursors for spin-crossover complexes and optoelectronics. (3) The presence of the two pyrazolyl fragments renders the systems 26 and 28 non-alternant. 25 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ Therefore, the clear definition of the ground spin state multiplicity must be explored. Furthermore, they will offer valuable model systems for the synthetic development of other cores based on non-alternant heterocyclic unit. Based on these motivations, the synthetic strategy towards the biradicals 26 and 28 is outlined below as Scheme 2.15. CHO O OHC CH NH2-NH2 / MeOH H N N 10 - 45%CHO HCl 6N H 22 h, RT 14 15 4 3 5 1. K / Diglyme / 70°C H 5' 2 6 H 4' 15 N N N+ Br N Br 1' 45%2. 110°C / 72 h N 1O N3' 2' O19 HOHN NHOH OH OHN 19 N N N N 42% Dioxane N N N N 10 days / R.T. OH 25 OH NaIO4 CHCl3 / H2O 26 27% HOHN NHOH OH OHN 19 N N N N Dioxane 46%N N N N 7 days / 60 °C 27 NaIO4 CHCl3 / H2O 28 51% Scheme 2.15 The first synthetic step consisted on the preparation of the triformylmethane (14), according to a very brief description reported in literature [26](Scheme 2.16) using N,N dimethyl- formamide, phosphorus oxychloride and bromoacetic acid. H3C CHO N CHO + POCl3 + Br COOH HO CHO CHOH3C CHO CHO M 14 Scheme 2.16 26 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ The course of this reaction, as suggested by the authors, seems very complex where the triformylmethane obtained is in equilibrium with 2-hydroxymethylene-malonaldehyde (M) [26]. As underlined in the experimental section, the delicate point consisted on the pre-reaction between DMF and POCl3 (this is the formylating agent) and then, once the reaction with bromoacetic acid is completed, the decomposition (in ice) and basification of the very acid mixture needed to be performed fast, without reaching too basic environment (up to pH ~8). The product 14 (variable yield 10-45%) infact appeared very sensitive to oxidize to the acid under air (Cannizzaro type reaction) in basic environment. Since basification cannot be avoided, this represented the major draw back for such reaction that led to the variable yields as above reported. Then, the condensation of triformylmethane (14) (1 eq) with hydrazine-monohydrate (1 eq) was sucesfull only when the hydrazine was added very slowly, over 3 hours. It was additionally carried out in acidified alcoholic medium [27] (Scheme 2.17). The reaction occurred simply at room temperature, by stirring for 20 hours, followed by basification and separation of 15 over silica column. The pyrazol-4- carboxaldehyde 15 (yield 52%) was finally obtained as yellowish solid. Its 1H-NMR and 13C- NMR spectra are shown both in Figure 2.7A and 2.7B respectively. CHO HH2N NH2 CHO N CHO CHO MeOH / HCl N 15 Scheme 2.17 The 2,6-bis(4-formyl-pyrazolyl)-pyridine (19), that represents the key precursor towards the biradical systems, was synthesized in one step reaction according to Scheme 2.18, by condensation between the nucleophilic potassium-salt of 15 with 2,6-dibromopyridine in diethylene glycol dimethyl ether (diglyme) as reaction solvent (yield 45%). The choice of diglyme was mostly dictated by the fact that the first bromine group in the 2,6-dibromopyridine is usually replaced relatively easily (70°C, 48h) [15m], however substitution of the strongly inactivated second bromine group requires higher temperature and prolonged reaction time. H N K CHO Br N Br N N N OHC N CHO O O O 110°C / 72 h N N 15 19 70°C Scheme 2.18 27 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ As advantages the diglyme offers with respect to the other most often used ethereal solvents (e.g. THF, Et2O) the higher boiling point (162°C at 760 mmHg) and stability at high temperature. In addition, its ability to chelate cations (the potassium cation in this case) left the nucleophile much more active. The figure 2.8 shows the 1H-NMR spectra of the dialdehyde 19. -CHO (A) H -NH N CHO N (B) 2 0 0 1 8 0 1 6 0 1 4 0 1 2 0 1 0 0 8 0 6 0 4 0 2 0 0 ( p p m) Figure 2.7: (A) 1H-NMR (250 MHz, r.t.) spectrum for compound 15 recorded in DMSO-d6 and (B) its 13C-NMR (63 MHz, r.t.) spectrum recorded in CDCl3. 28 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ Compound d 19 e c N N OHC N N CHO a Nb Figure 2.8: 1H-NMR (250 MHz, r.t.) spectrum for compound 19 recorded in DMSO-d6. We explored two other possibilities in order to obtain compound 19. One relied on building the pyrazol-4-carboxaldehyde (15) starting from the commercially available pyrazole, and the second method consisted in synthesizing at first the bispyrazolylpyridine backbone followed by functionalisation of the terminal pyrazoles in positions 4. While pyrazole 3,5 disubstituted derivatives are readily generated via a Claisen condensation to form a 1,3- dicarbonyl followed by condensation with hydrazine [15m,28] as shown in scheme 2.19, those pyrazoles functionalised in position 4 are much less common and often required tedious multi- step approaches. H R1O O + + NaH ONa O H N R CH3 R1 OEt Et2O R R H N NH1 2 2 N R R1, R = , t-Bu , H , CH3 Scheme 2.19 Since only 4-substituted pyrazole were needed, one method would be the reaction of 4- lithiopyrazoles with electrophiles [29]. The 4-halogen substituted pyrazoles are readily generated by N-protection followed by electrophilic halogenation [30] (or nitration [31]). However, there are only few examples of pyrazole C-4 lithiation [32], all based on bromine- lithium exchange and obviously as mentioned above required a protecting group in position 1 (N-protection). Such reaction proceeds with low chemoselectivity due to competing 29 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ deprotonation at C-5 [32a] or isomerization of the 4-lithiopyrazole to the corresponding 5- lithiopyrazole [33] (Scheme 2.20). Obviously this problem can be avoided upon introduction of a second protecting group in the C-5 position, making this procedure even more tedious [32f]. 1 5N HO HO P OBr2 N P N 2 4 NN Br BrN 3 O O N N OH nBuLi HO-N PO Cl E NCOOH N 1) (E) electrophile Li E and N 2) strong Base or Acid HO-N Li andPO Cl N N N P = , trimethylbenzoylbromideBr trimethylsilylchloride/LDA Scheme 2.20 We decided to avoid the lithiation procedure. One report [34a] described the regiospecific monoiodination of 1-benzyloxypyrazole [34b] without protecting group at C-5 followed by magnesium-iodine exchange. Subsequent reaction with DMF produced the corresponding 4-formyl-1-benzyloxypyrazole in good yield (88%). Hydrolysis in acid environment gave the 1-hydroxypyrazole. Further deprotection of the nitrogen 1 was not described. This reaction is shown in Scheme 2.21. Br i 1) , NEt(Pr )2 2) IICl , K2CON OH 3 N O N 81% N 1) i-PrMgBr OHC 2) DMF N O N 88% Scheme 2.21 In attempts to avoid initial N-protection, a mixture of NaI/I2 in sodium acetate/water or KI/I2 was used as iodinating agents for the pyrazole. However, the product obtained in both 30 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ cases consisted of a mixture of all three isomers (3, 4 and 5 iodo-pyrazole). The following protocol appeared recently as alternative route by using a mixture of HIO3 /I2 in acetic acid [35]. On the other hand, attempts to reproduce the procedure as reported led also to a mixture of products plus unreacted starting material. While the pyrazole can be easily recovered (water soluble) separation of the isomers is very difficult to achieve. As pointed out in the experimental part, the key point consisted in the slow addition of the iodinating agent (drop by drop); kinetically only the position 4 is favoured, and since the 4-substituted iodo pyrazole (16) is not very soluble even in acetic acid, its precipitation drives the reaction to completion (yield 88%). As reported in Scheme 2.22, the subsequent step consisted in the N- protection of the pyrazolyl-nitrogen in position 1 using ethylvinylethere in slightly acidified medium (HCl) followed by cromatographic separation on alumina (Al2O3, yield > 94%). The product 1-(1-ethoxyethyl)-4-iodo-pyrazole (17) collected is a pale yellowish oil. The 1H-NMR spectrum of 17 is shown in Figure 2.9. H N HIO H3 N O N CH COOH I O N 3 N Benzene/HCl IN 16 17 1) EtMgBr / THF 2) DMF HN 3) dioxane / HCl CHO N 18 = 15 Scheme 2.22 C o m p o u n d 1 7 a c a b O N I d N b c d 10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 (ppm) Figure 2.9: 1H-NMR (250 MHz, r.t.) spectrum for compound 17 recorded in CDCl3. 31 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ The final step towards the synthesis of 15 consisted in the formation of the correspondent Grignard derivative. Reaction of 1-(1-ethoxyethyl)-4-iodo-pyrazole with magnesium wire was ineffective either in THF and glyme, even upon heating over 90°C and in presence of CuI [similar to Note 17 in Reference 34a]. As an alternative, the reaction with a stronger Grignard reagent (ethylmagnesium bromide) proceeded very well. The Grignard derivative of 17 was not soluble and at the end of the reaction a solid paste was formed. This intermediate is unstable and should be kept below 4°C. Nevertheless, slow addition of the electrophyle DMF led to the resolubilization of the solid. After hydrolysis and column chromatography (ethylacetate/hexane) the pyrazol-4-carboxaldehyde (18) is obtained in a very good yield (80%). Although this route is longer than that previously suggested (reactions 2.16 and 2.17) was always very successful. The second route for building the biscarbaldehyde derivative 19 relies on an even more simplified procedure, and further allows flexibility in developing novel functionalisations for the bispyrazolylpyridine core that can be used for different cross-coupling reactions. The pyrazolylpyridine core was easily built by nucleophilic substitution between the pyrazole anion prepared by reacting pyrazole (1 eq.) with potassium metal (1 eq.) and 2,6-dibromo-pyridine (Scheme 2.23). H N K Br N Br N N N N N + N Br N N N glyme 140°C 50°C 20 C Scheme 2.23 This reaction as discussed in Scheme 2.18 is rather slow (4 days) and afforded 20 in moderate yield (~50%) after chromatographic separation on silica (Rf = 0.1-0.5 in CHCl3/hexane/ethylacetate, 6/2/1). A side product consisted in the mono-substituted 2-bromo- 6-pyrazolyl-1-yl-pyridine (Compound C, Rf = 0.8) that was easily separated from 20. In Figure 2.10 and 2.11 are shown their respective 1H-NMR spectra. 32 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ a Br N N N b Figure 2.10: 1H-NMR (250 MHz, r.t.) spectrum for compound (C) recorded in CDCl3. a N N N N N b Figure 2.11: 1H-NMR (250 MHz, r.t.) spectrum for compound 20 recorded in CDCl3. 33 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ The nexth step consisted in the functionalisation of the terminal pyrazoles in position 4. An old report described some electrophilic reactions (e.g. bromination, chlorinations, nitrations) carried out on various 2-(pyrazolyl-1’-yl)-pyridines but we did not find any report of similar reactions on 20 [36]. As shown in Scheme 2.24 the symmetric 2,6-bis-(4-Iodo- pyrazolyl-1-yl)-pyridine (21) and 2,6-bis-(4-Bromo-pyrazolyl-1-yl)-pyridine (22) were obtained from very good (21, >80%) to good yield (22, 57%). This opened the way to access a large number of derivatives which are shown in Scheme 2.24. The Grignard exchange on 21 was very successful, and allowed the synthesis of the biscarbaldehyde 19 easily. The 13C-NMR spectra for compounds 21, 22, and 24 are reported in Figure 2.12 Br2 / acetic acid N N H2SO4 diluted N N N yield 57 % N N N N N 22Br Br 20 HIO3 / I2/ acetic acid H2SO4 diluted N N N N N argon 1) EtMgBr yield 80 % 21 2) DMF I I Pd(II) /CuI / P(Ph)3 N TMS, argon N N N N 19 N OHC CHO yield 91 % N N N N yield 72 % 23 MeOH / THF Si argonK2CO3 Si N N yield 95 % N N N 24 Scheme 2.24 34 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ N N N N N I I N N N N N Br Br N N N N N Figure 2.12: 13C-NMR (63 MHz, r.t.) spectra for compounds 21 and 22 recorded in CDCl3 and 24 in DMSO-d6. 35 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ The Scheme 2.25 illustrates the subsequent steps towards the nitronylnitroxide and iminonitroxide biradicals. The condensation between the diformyl-derivative 19 with 2,3- bishydroxylamino-2,3-dimethylbutane (6) in dioxan occurred while stirring at room temperature under argon over ten days, and afforded the radical precursor 2,6-bis[4-(1,3-dihydroxy- 4,4,5,5-tetramethylimidazolidin-2-yl)-pyrazolyl]-pyridine (25) as yellowish powder (yield 42%). Heating the reaction mixture increased the decomposition of the hydroxyl-derivative in solution, and led to the formation of compound 2,6-bis[4-(1-hydroxy-4,4,5,5- tetramethylimidazolin-2-yl)pyrazolyl]-pyridine (27). The sodium-periodate oxidation of 25 carried under phase transfer conditions (CHCl3/H2O) gave the crude biradical 2,6-bis[4'-(3- oxide-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazol-1'-yl]-pyridine (26) in the organic phase. The side products (i.e mono nitronyl nitroxide, mono imino nitroxide) were easily separated by column chromatography (silica gel, acetone/light petroleum ether, b.p. 30-40°C, 2/8) and highly pure 26 was collected (Rf = 0.34) and recrystallised from CHCl3 (blue crystals, yield 27%). N N N N N N N N N N HN NH OH OH 25 OH NaIOHO 4 O 26 O N N N N N+ +dioxane N N N 10 days in argon OH HO O- -O N N N N N 19 OHC CHO N N N HN NH N N N N N N N Excess OH OH 27 dioxane, 60°C N N NaIO4 N 28 N 7 days in argon N N OH HO N NO O Scheme 2.25 Similarly, the oxidation of the radical precursor 26 under phase transfer conditions (CHCl3/H2O) afforded the crude biradical 2,6-bis[4-(1-oxyl-3-4,4,5,5-tetramethylimidazolin-2- yl)pyrazolyl]pyridine (28) in the organic phase. The purification of the crude mixture was carried out by column chromatography (silica gel, acetone/light petroleum ether, b.p. 30-40°C, 2/8, Rf = 0.46), then recrystallisation from CH2Cl2 gave pure 28 (orange powder, yield 51%). As expected, the biradical 28 was obtained in larger yield with respect to 27. Once isolated, the biradicals are stable as powder over several months, while in protic solvents (e.g. CH2Cl2 or CHCl3), that may catalyze loss of water molecules, a clear decrease of their stability was observed upon prolonged storage. However, in aprotic solvents like toluene, as previously found for the terpyridine and the bispyridine based radicals, the compounds 26 and 28 could safely be kept for long time. Even when they were heated up to 60°C for several hours, no hint 36 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ of decompositions was observed, providing that the medium was maintained oxygen free. In Figure 2.13 are reported for comparison the 13C-NMR spectra of the biscarbaldehyde 19 and the radical precursor 25. Compound 19 N N N N N OHC CHO Compound 25 N N N N N HO OH N (A) N N OH NHO (B) (C) Figure 2.13: 13C-NMR (63 MHz, r.t.) spectra for compounds 19 and 25 recorded in DMSO-d6. The precursors of the Ullman base radicals feature always characteristic peaks associated with the imidazolyl moiety. As an example, the C-H carbon of the imidazolyl ring in compound 25, marked with (A) in Figure 2.13, exhibits a well defined resonance around 83 ppm, while the two cis/trans methyl carbons (C) fall at ~ 17 and ~ 24 ppm respectively. The quaternary carbon (B) instead gives constantly a sharp peak around 66 ppm. 37 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ The similar synthetic procedures used to assemble the biradical derivatives 26 and 28 have been employed also for the synthesis of the two pyrazolyl-based monoradical systems, the 2[4-(1-oxide-3-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]-pyridine (37) (NN) and 2[4-(1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]-pyridine (38) (IN). Their synthesis were necessary in order to obtain either suitable references for the pyrazole based biradical systems (26 and 28), and to probe up to which extent the electronic properties of the radical fragments (NN and IN) are influenced upon connection to different types of π-hetero rings (pyrazole based radicals versus pyridine based radicals). The Scheme 2.26 describes the reaction steps. The condensation between hydrazinopyridine with triformylmethane afforded in very good yield the 2-(4-formylpyrazolyl)- pyridine (35). Then, further condensation with 2,3-bishydroxylamino-2,3-dimethylbutane (6) in methanol/trichloromethane mixture under argon gave the air sensitive monoradical precursor 2[4-(1-hydroxy-3-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]-pyridine (36). The nitronyl nitroxide radical 37 was obtained by usual NaIO4 oxidation under phase transfer conditions (water/chloroform) carried on the precursors 36. However, excess of NaIO4 did not provide the correspondent imino nitroxide radical 38, even when the reaction mixture was heated up to 50°C. Thus, it has been used a much stronger oxidizing agent (NaNO2/HCl). Both radicals were obtained in a very good yield after purifications on silica column, and this represents a relevant difference with respect to the radicals directly attached to pyridine moieties that suffer of fast over-oxidation processes and decomposition, even within the purification in column when silica have been used. Furthermore, the overall yields of both mono- and biradical systems directly linked to pyrazole (26, 28, 37 and 38) are much higher with respect to those connected to pyridine units. CHO CHO HN NH CHO N OH OHN NH NH N N N N N 2 MeOH / HCl MeOH / CHCl - 3 35 36 O + CHO N 76% 30% N NaNO2 / HCl HONaIO4 N N N N N N 38 37 ON N 50% N 60% + O NO Scheme 2.26 38 - Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ A decrease in the π-conjugation between radical moieties in a biradical system should induce a decrease in their through-bond interaction. However, while such decrease is expected, it would be extremely difficult to asses quantitatively this effect by using theoretical predictions. Therefore, we decided to synthesize two model systems by taking as reference the biradicals 26 (NN) and 28 (IN), and introducing two σ-bonds between the pyrazolyl- imidazolidin moieties and the central pyridine ring, in order to attain their related σ-conjugated radicals 2,6-bis[4-(1-oxyl-3-oxide-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolylmethyl] pyridine (41) and 2,6-bis[4-(1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolylmethyl] pyridine (42). σ σ N N N N N N N N N N 26 (NN) 41 (NN) 28 (IN) 42 (IN) The synthetic steps are described below as Scheme 2.27, and provided the radicals 41 and 42 in a very good yield. 15 6 H N N HN NH CHO N N OH OH N NaH / THF N NBr Br MeOH / CHClN N 3 N N 60°C HO N N OH OHC 39 CHO N 40 N yield 54% N yield 62% N OH HO NaIO4 N NaNO2 N N N O N O HCl N N N O N N O N N41 N 42 N N+ yield 64% + N N yield 64% N O - - O Scheme 2.27 The 2,6-bis(4-formylpyrazolylmethyl)-pyridine (39) was made accessible by reacting the commercially available 2,6-dibromomethylpyridine with 4-formylpyrazole (15) and sodium hydride in THF. The nucleophilic substitution occurred very fast (3 hours), and gave the 39 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ biscarbaldehyde 39 in good yield (54%). Further condensation with 2,3-bishydroxylamino-2,3- dimethylbutane (6) in methanol/trichloromethane mixture under argon gave within 2 days the biradical precursor 2[4-(1-hydroxy-3-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]-pyridine (40). The NN radical 41 was obtained by usual NaIO4 oxidation under phase transfer conditions (water/chloroform), and separated as blue powder. However, as previously found for the monoradicals 38, using excess of NaIO4 in 40 did not provide the desired imino nitroxide biradical 42. Therefore both 40 and 41 appeared very resistant towards over- oxidation processes. Thus, the stronger oxidizing agent NaNO2/HCl was employed on the precursor 41, and the biradical 42 was cleanly obtained as orange-red powder. Also in this case, the purifications of the radicals have been performed easily on silica column, without showing any trend of decomposition. The 13C NMR spectra of 39 and the radical (NN) precursor 40 are shown in the Figure 2.14. Compound 39 Compound 40 Figure 2.14: 13C-NMR (63 MHz, r.t.) spectra for compounds 39 (CDCl3) and 40 (DMSO-d6). 40 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ In order to provide an intermediate case (radical distances vs exchange pathways) for the π-conjugated N-heterocyclic cores, which could stays in between the terpyridine and the bispyrazolylpyridine units, it was necessary to synthesize a non symmetric nitrogen-based ligand. The choice was straightforwardly directed to the pyrazolylbipyridine element (P). P N N N R R N N NN N N N N R R R R N Large radical distances Small R = nitronylnitroxide and/or iminonitroxide radical The successful synthetic path that led to the biradicals based on P, required the preparation of the derivative 4'',5'-diformyl-6-(pyrazol-1''-yl)-2,2'-bipyridine (31). This was obtained using the combined knowledge gained on the terpyridine and bispyrazolylpyridine systems. The necessary synthetic steps are outlined in the following section. N N N N OHC 31 CHO The coupling reaction between 6-bromo-2,2’-dipyridine-5’-carbaldehyde (5) as previously prepared with either the potassium or the sodium hydride salt of pyrazol-4- carboxaldehyde (18) in THF was almost ineffective, and led to a complex mixture of by- products, where the major loss of the carbaldehyde group in 5 was observed (Scheme 2.28). N Br N HN K or NaH OHC 5 CHO N THF THF / reflux x 18 Scheme 2.28 41 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ This failure seemed consistent with the poor reactivity of the bromine group in position 6 due to lack of activation towards nucleophilic substitutions. In addition the protection of the carbaldehyde group in 5 turned to be surprisingly difficult to achieve by using ethyleneglycole with either toluene-4-sulfonic-acid or Amberlist-15 as acid catalyst. This represents a substantial difference with the previous protection of the 6-bromo-3-pyridinecarbaldehyde (1). The successful reaction path for 31 was then carried out as follows: the 2,6-dibromopyridine was reacted in alcoholic medium (ButOH) with excess of hydrazine-monohydrate (solution in THF), and gave the 2-bromo-6-hydrazinopyridine (29) as yellowish crystals (yield 57% - 63%). The condensation between 29 and triformylmethane (14) was achieved in acidified alcoholic medium (MeOH/HCl), and afforded the 6-bromo-2-[4'-formylpyrazol-1'-yl]-pyridine (30). It was obtained as fine precipitate after neutralization with aqueous Na2CO3 solution and chromatographic separation on silica column (CHCl3/Hexane/Ethylacetate, 1/3/1, Rf = 0.67) (yield 70%). CHO CHO H2N NH2 CHO Br N N N Br N Br OH Br N NHNH2 MeOH / HCl 5 h 29 1 day 30 CHO r.t. Scheme 2.29 As an alternative, 30 can be obtained as well by reacting the 2,6-dibromopyridine (1.eq.) with the potassium salt of pyrazol-4-carboxaldehyde (18) (1.2 eq) in THF under argon, and heating the mixture for 60 h at 70° (yield 60%) followed by purification on silica column using the same procedure reported above. Both routes provided comparable yields but the one shown in Scheme 2.29 was somehow faster. Then, the Stille coupling reaction between 30 and 2-tributylstannyl-5-[1,3]dioxolan-2-yl-pyridine (3) was achieved in toluene under rigorous argon atmosphere for 60 hours, in presence of Pd(PPh3)2Cl2 (5%), PPh3 (10%) and copper iodide (CuI) as catalyst. The presence of CuI was essential, since the use of the Pd(II)/PPh3 catalyst itself led to a drastic decrease in the overall yield (from 46% to ~ 20%). This effect is exactly opposite to that one previously observed for the similar Stille coupling reaction between 5 and 3. Further hydrolysis of the dioxolane in presence of HCl (6 N) followed by basification gave 4'',5'-diformyl-6-(pyrazol-1''-yl)-2,2'-bipyridine (31) as shown in Scheme 2.30. 42 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ O O 1) Pd(II) N N N N Br N N N + Toluene,refluxSn N 2) HCl, reflux OHC 31 CHO 30 CHO 3 Scheme 2.30 The bisformyl-derivative 31 (1 eq) was then subjected to Ullman coupling with 2,3- bishydroxylamino-2,3-dimethylbutane (6, 3.5 eq) using a mixture of 1,4-dioxane and trichloromethane (1/1) as solvents for 7 days, under argon at room temperature as depicted in Scheme 2.31. 6 HN NH N N N N N N OH OH HO N N dioxane / CHCl3 N OHC 31 CHO 7 days in argon 32 OH N N OH NHO Excess NaIO NaIO4 4 N N N NO yield 18% N N O N N N yield 9% 34 ON NN N 33 O N+ + N O - N- O Scheme 2.31 The reaction mixture showed no hint of precipitate formation. This was consistent with the partial dehydratation of the bis(N,N')-hydroxy-imidazolidine to N-hydroxy-imidazolidine. The use of the FTIR technique proved to be very useful in this case. In fact, upon collecting aliquots of the reaction mixture at various times, we could monitor the proceeding of the condensation reaction just by observing the disappearance of the carbaldehyde peak (with ν at 1689 cm-1C=O ). The yellowish powder obtained, upon solvent removal, contained the crude 4'',5'-bis[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-6-(pyrazol-1''-yl)-2,2'-bipyridine (32) (crude yield 73%) and this powder was oxidized, working under phase transfer conditions (CHCl3/H2O) at room temperature, by using slight excess of NaIO4 (2.5 eq) with respect to 32 43 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ (1.0 eq) for 20 min. After collection of the organic phase side products (mono- nitronyl nitroxide and mono imino nitroxide radicals) were separated by column chromatography (aluminum-oxide, acetone/light petroleum ether) while the non-oxidized and/or decomposed hydroxy-imidazolidine 32 was not eluted at all, and remained strongly retained at the top of the column. Then, pure 4'',5'-bis[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl]-6-(pyrazol-1''- yl)-2,2'-bipyridine (33) was obtained as deep-blue powder. In a similar manner, the 4'',5'-bis[- 1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl]-6-(pyrazol-1''-yl)-2,2'-bipyridine (34) was synthesized by using excess of oxidant (NaIO4, 4.0 eq) with respect to 32 (1.0 eq) followed by chromatographic separation on neutral alumina, to afford 34 as orange-red powder. Unfortunately, was not possible to increase the yield of the nitronylnitroxide biradical 33 since, as reported previously, the NN radicals attached to pyridine moieties are far more sensitive towards over-oxidation processes with respect to those appended on the pyrazole rings. Finally, in Table 2.1, is shown a comparative outlook of the radical stabilities in solution for the nitronyl nitroxide (NN) series. Table 2.1. The NN radical stabilities in solution. NN radicals hexane toluene THF acetone CH2Cl2 MeOH CHCl3 12 Stable Stable Not stable Stable Stable Not stable Stable biradical (~ 1 year) (~1 year) (< day) (~6 months) (~3 months) (< day) (~ 1 month) ~[10-3 M] ~[10-3 M] >[10-3 M] >>[10-3 M] >>[10-3 M] >[10-3 M] >>[10-3 M] 8 Stable Stable Not stable Stable Stable Not stable Stable ~[10-3 monoradical M] (~1 year)* (~ week) (~8 months) (~4 months) (< week) (< 2 months) ~[10-3 M] >[10-3 M] >>[10-3 M] >>[10-3 M] >[10-3 M] >>[10-3 M] 26 Stable Stable Not stable Stable Stable Not stable Stable biradical (~ 1 year) (~1 year)* (~ 1 day) (~9 months) (~6 months) (< 1 day) (<4 months) < [10-3 M] ~[10-3 M] >[10-3 M] >>[10-3 M] >>[10-3 M] >[10-3 M] >>[10-3 M] 36 Stable Stable Not stable Stable Stable Not stable Stable monoradical (~ 1 year) (~1 year)* (~ week) (~ 1 year) (~7 months) (~ week) (<6 months) > [10-3 M] >[10-3 M] >[10-3 M] >>[10-3 M] >>[10-3 M] >[10-3 M] >>[10-3 M] 33 Stable Stable Not stable Stable Stable Not stable Stable biradical (~ 1 year) (~1 year) (<< day) (~6 months) (~2 months) (<< day) (~1 month) < [10-3 M] ~[10-3 M] >[10-3 M] >>[10-3 M] >>[10-3 M] >[10-3 M] >>[10-3 M] 41 Stable Stable Less stable Stable Stable Less stable Stable biradical (~ 1 year) (~1 year)* (> week)* (~ 1 year) (~7 months) (> week) (<7 months) > [10-3 M] >[10-3 M] >[10-3 M] >>[10-3 M] >>[10-3 M] >[10-3 M] >>[10-3 M] The asterisk (*) indicates that the radical can be heated up to 60°C without apparent decomposition for three hours even in presence of NaH under argon. The numbers in brackets [x] indicates the relative solubility in the specific solvent. The imino nitroxide radicals showed comparable or slightly higher stabilities in these solvent series. In addition, two other solvents (ethylacetate and 2-propanol) were recently tested in which the radicals featured similar stabilities as compared with, for example, acetone. All the solvents were purchased as spectrophotometric grade, and used as received with the exception of THF that was distilled in presence of sodium under argon. 44 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ Br N NHNH2 Compound 29 -NH2 -NH- Br N N N CHO Compound 30 N N N N OHC (A) CHO (B) Compound 31 Figure 2.15: 1H-NMR (250 MHz, r.t.) spectra for compounds 29 recorded in CDCl3, 30 and 31 recorded in DMSO-d6. 45 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ References [1] Osiecki J.H. and Ullman E.F., J. Am. Chem. Soc., 90 (4) 1968, 1078. [2] Ullman E.F., Call L., Osiecki J.H., J. Org. Chem. 35 (11), 1970, 3623. [3] Ullman E.F., Osiecki J.H, Boocock D.G.B., Darcy R., J. Am. Chem. Soc. 94 (20), 1972, 7049. [4] (a) O. Kahn, Magnetism: A Supramolecular Function, Eds., Kluwer, Dordrecht, 1996. (b) J. S. Miller, M. Drillon, Magnetism: Molecules to Materials III, Wiley-VCH, Weinheim, 2001. (c) J. S. Miller, M. Drillon, Magnetism: Molecules to Materials I, II, IV, Wiley-VCH, Weinheim, 2003. [5] For a collection of the recent progreesses in the field see Proceedings of the 8th International Conference on Molecule-Based Magnets (ICMM 2002), Polyhedron 2003, 22, 1725-2584. [6] Formation of 2-phenyl imidazoline from 2,3-diamino-2,3-dimethylbutane and phenyl- thioamide followed by oxidation with NaWO4/H2O2 has been reported by Aurich H.G., Czepluch H., Hahn K., Tetrahedron Lett. 50, 1977, 4373. [7] (a) Hirel C., Vostrikova E.K., Pécaut J., Ovcharenko V.I., Rey P., Chem. Eur. J. 7 (9), 2001, 2007. (b) G. V. Shustov, N. B. Tavakalyan, L. L. Shustova, A. P. Pleshkova, R. G. Kostianovskii, Bull. Acad. Sci. URSS, Div. Chem. Sci., 1982, 31 (Engl. transl.) [8](a) J. S. Miller and A. Epstein, Angew. Chem. Int. Ed., 106, 1994, 399. (b) S. Nakatsuji and H. Anzai, J. Mat. Chem. 7, 1997, 2161. (c) J. A. Crayston, J. N. Devine, J. C. Walton, Tetrahedron, 56, 2000, 7829. [9] (a) Chelucci G., Thummel R.P., Chem. Rev. 102, 2002, 3129. (b) Cargill Thompson A.M.W., Coord. Chem. Rev. 160, 1997, 1. (c) Lehmann U., Henze O., Schlüter A.D., Chem. Eur. J. 5, 1999, 854. (d) Romanenko G.V., El’tsov I.V., Ovcharenko V.I., J. of Struct. Chem., 43, 2002, 700. (e) Solanki N.K., McInnes E.J.L., Mabbs F.E., Radojevic S., McPartlin M., Feeder N., Davies J.E., Halcrow M.A., Angew. Chem. Int. Ed., 37, 1998, 2221. [10] (a) J. P. Sauvage, J. P.Collin, J. C. Chambron, S. Guillerez, C. Coudret, V. Balzani, F. Barigelletti, L. De Cola, L. Flamigni, Chem. Rev., 94, 1994, 993. (b) J.–M. Lehn, Supramolecular Chemistry. Concepts and Perspectives, Wiley VCH, Weinheim, Germany, 1995. (c) Drain C.M., Lehn J.-M., J. Chem. Soc. Chem. Commun., 1994, 2313. (d) Loi M, Hosseini M.W., Jouaiti A., De Cian A., Fisher J., Eur. J. Inorg. Chem., 1999, 1981. (e) Jouaiti A., Loï M., Hosseini M.W., De Cian A., J. Chem. Soc. Chem. Commun., 2000, 2085. [11] Trawick B.N., Daniher A.T., BashkinJ.K., Chem. Rev., 98, 1998, 939. [12] E. Coronado, P. Delhaés, D. Gatteschi, J. S. Miller, Eds., Molecular Magnetism: From Molecular Assemblies to the Devices, NATO ASI Series E 321, Kluwer Academic Publishers, Dordrecht 1996. 46 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ [13] Edwin C. Constable, Gerhard Baum, Eckhard Bill, Raylene Dyson, Rudi van Eldik, Dieter Fenske, Susan Kaderli, Darrell Morris, Anton Neubrand, Markus Neuburger, Diane R. Smith, Karl Wieghardt, Margareta Zehnder, Andreas D. Zuberbühler, Chem. Eur. J., 5 (2), 1999, 498. (b) Kremer S., Henke W., Reinen D., Inorg. Chem., 21, 1982, 3013. (c) Figgis B.N., Kucharski E.S., White A.W., Aus. J. Chem., 36, 1983, 1537. (d) Money V.A., Evans I.R., Halcrow M.A., Goeta A.E., Howard J.A.K., Chem.. Commun. 2003, 158. [14] A. Harriman, R. Ziessel, Coord. Chem. Rev., 171, 1998, 331. [15] Constable E.C., Ward M.D., Corr S., Inorg. Chim. Acta, 141, 1988, 201. (b) Dietrich- Buchecker C.O., Marnot P.A., Sauvage J.P., Tetrahedron Lett., 1983, 5291. (c) Bell, T.W., Firestone A.J., J. Org. Chem., 51, 1986, 764. (d) Schubert U.S., Eschbaumer C, Georg Hochwimmer, Synthesis, 5, 1999, 779. (e) Ziener U., Lehn J-M., Mourran A., Moller M, Chem Eur. J., 8 (4), 2002, 951. (f) Kelly TR, Lebedev R.L., J. Org. Chem. 67 (7), 2002, 2197. (g) El- Ghayoury A., Ziessel R., J. Org. Chem., 65, 2000, 7757. (h) Halcrow M.A., Brechin E.K., McInnes E.J.L., Mabbs F.E., Davies J.E., J. Chem. Soc., Dalton Trans., 1998, 2477. (i) Lehmann U., Henze O., Schlüter A.D., Chem. Eur. J., 5 (3), 1999, 854. (l) Constable E.C., Ward M.D., J. Chem. Soc. Dalton Trans., 1990, 1405. (m) Jameson L. D., Goldsby K.A., J. Org. Chem., 55, 1990, 4992. (n) Mukherjee R., Coord. Chem. Rev., 203, 2000, 151. [16](a) M. A. Halcrow, E. K. Brechin, E. J. L. McInnes, F. E. Mabbs, J. E. Davies, J. Chem. Soc., Dalton Trans., 1998, 2477. (b) C. Stroh, R. Ziessel, Tetrahedron Lett. 40, 1999, 4543. (c) C. Stroh, P. Turek, P. Rabu, R. Ziessel, Inorg. Chem. 40, 2001, 5334. [17] (a) Sasaki I., Daran J.C., Balavoine G.G.A., Synthesis, 5, 1999, 815. (b) Goral V., Nelen M.I., Eliseev A.V., Lehn J.-M., PNAS, 98 (4), 2001, 1347. [18] X. Wang, P. Rabbat, P. O`Shea, R. Tiller, E. J. J. Grabowski, P. J. Reider Tetrahedron Lett., 41, 2000, 4335. [19] F. J. Romero-Salguero, J.-M. Lehn Tetrahedron Lett., 40, 1999, 859. [20] D. Cai, D. L. Hughes, T. R. Verhoeven , Tetrahedron Lett., 37, 1996, 2537. [21] For the toxicity of stannyl-compounds see for example: A. G. Davies, P. J. Smith, Comprehensive Organometallic Chemistry (Ed. E. W. Abel) Pergamon, Oxford 2, 1982, 608. [22] See for example “On the nature of the “Copper Effect” in the Stille Cross-Coupling”. Farina V. Kapadia S. Krishnan B., Wang, C; Liebeskind L.S., J. Org. Chem., 59, 1994, 5905. [23] Schwab P.F.H., Fleischer F., Michl J., J. Org. Chem., 67 (2), 2002, 443. [24] J. de Mendoza, E. Mesa, Juan-Carlos Rodríguez-Ubis, P. Vázquez, F. Vögtle, Paul- Michael Windscheif, K. Rissanen, J.-M. Lehn, D. Lilienbaum, R. Ziessel, Angew. Chem. Int. Ed. (Engl.) 30 (10), 1991, 1331. [25] (a) Iyoda M., Otsuka H., Sato K., Nisato N., Oda M., “Homocoupling of aryl halides using Nickel(II) complex and Zinc in presence of Et4NI. An efficient method for the synthesis of 47 Chapter 2 – Synthesis of Pyridine and Pyrazole containing Radicals ___________________________________________________________________________ biaryls and bipyridines”, Bull. Chem. Soc. Jpn, 63, 1990, 80. (b) Jolly P.V., “Nickel catalyzed coupling of organic halides and related reactions” in “Comprehensive Organometallic Chemistry” Ed. G. Wilkinson, Pergamon Press., Oxford, Vol. 8, 1982, p.713. [26] Z. Arnold Z., Coll. Czech. Chem. Commun. 26, 1961, 3051. [27] Takagi K., Bajnati A., Hubert-Habart M., Bull. Soc. Chim. Fr., 127, 1990, 660. [28] (a) Sorrel T.N., Tetrahedron, 45, 1989,3.(b) Trofimenko S., Calabrese J.C., Thompson J.S., Inorg. Chem. 26, 1987, 1507. (c) Almirante N., Cerri A., Fedrizzi G., Marazzi G., Santagostino M., Tetrahedron Letters, 39, 1998, 3287. [29] Grimmet M., Iddon B. Heterocycles, 37, 1994, 2087. [30] (a) Hüttel, R., Schäffer O., Jochum P., Liebigs Ann. Chem., 593, 1955, 200. (b) Lipp M., Dallacker F., Munnes S., Liebigs Ann. Chem., 618, 1958, 110. [31] Hüttel R., Büchele F., Jochum P., Chem. Ber. 88, 1955, 1577. [32] (a) Hüttel, R., Schön M.E., Liebigs Ann. Chem., 625, 1959, 55. (b) Hahn M., Heinisch G., Holzer W., Schwarz H., J. Heterocycl. Chem. 28, 1991, 1189. (c) Iwata S., Qian C.-P, Tanaka K., Chem. Lett. 1992, 357. (d) Sakamoto T., Shiga F., Uchiyama D., Kondo Y., Yamanaka H., Heterocycles, 22, 1992, 813. (e) Elguero J., Jaramillo C., Pardo C., Synthesis, 1997, 563. (f) Balle T., Per Vedsǿ, Begtrup M., J. Org. Chem. 64 (15), 1999, 5366. [33] Tertov B.A., Morkovnik A.S., Chem. Heterocycl. Compd. (Engl. Transl.) 11, 1975, 343. [34] (a) Felding J, Kristensen J., Bjerregaard T., Sander L., Per Vedsǿ, Begtrup M., J. Org. Chem., 64 (11), 1999, 4196. (b) Begtrup M. Per Vedsǿ, J. Chem. Soc. Perkin Trans. 1 1995, 243. [35] Vasilevsky S.F., Klyatskaya S.V., Tretyakov E.V., Elguero J., Heterocycles, 60 (4), 2003, 879. [36] Khan M.A., Pinto A.A.A., J. Heterocyclic Chem., 18, 1981, 9. 48 ____________________________________________________________________________ Chapter 3 - The Radical`s Optical Properties The nitronyl nitroxide (NN) and imino nitroxide (IN) radical systems feature clear differences in their UV/Vis and IR optical properties. The spectroscopic fingerprints of the radical entities have been unambiguously identified and appeared strongly dependent on the type of π-ring system in which the radicals have been attached on. Based on these information’s, we established a quick protocol that allows to assign straightforwardly the type of radical and their numbers (monoradical, biradical), to define their purities and the nature of the contaminants. 3.1. The UV/Vis absorption spectra of the radical systems. The nitronyl nitroxide (NN) derivatives show light blue colour in solution (Figure 3.1), with a broad absorption band characterized by several vibronic components in the visible region of the spectra. These bands have been associated to the n → π* transitions of the aminoxyl-oxide residues (at ~ 600 nm) and their assignement were assessed by theoretical calculations on the absorption envelope (vide Section 3.2). Their intensities and spectral distributions did not show appreciable dependence upon changing solvent polarities (e.g for 12 in hexane λmax at 605 nm and in nitromethane λmax at 603 nm, see Figure 3.2.) but indeed strongly depend on which type of hetero-ring the spin carriers have been grafted on. The NN pyridine based radicals present always a weak but distinctive vibronic band around 740 nm (ε ≤ 70 M-1 × cm-1), that is missing when the radicals are connected to the pyrazolyl-moiety. In addition, within the pyridine based NN, the terpyridine biradical 12 shows a remarkable blue shift of the absorption envelope in the visible region with respect to the bipyridyl monoradical 8, being accompanied by a weakening in the molar extinction (ε). This hampering effect is not present in the pyrazole-based radical systems, in which for example the biradical 26 features just twice the molar extinction of the monoradical 37. The second distinctive absorption of NN falls in the UV region of the spectrum (<400 nm) and represents another fingerprint that allows to distinguish pyridine and pyrazole NN radicals. In the formers case this absorption is red- shifted over 380 nm, while in the latter is blue shifted (> 10 nm). This band is always very strong, and arises from π → π* transition of the aminoxyl-oxide residues. In the non- symmetric radical 33, in which pyrazole and pyridine moieties are both present, such effect is particularly clear. The transition at 373 nm (ε = 17510 M-1 × cm-1) originates from the radical connected to the pyrazole ring; in fact in the biradical system 26 this transition occur at 375 nm, ε = 16960 M-1 × cm-1 while in the monoradical 37 occur at 368 nm, ε = 11100 M-1 × cm-1. The similar transition at 392 nm (ε =10685 M-1× cm-1) originates from the radical connected to 49 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ the pyridine ring since in the monoradical 8 it occurs at 394 nm with ε = 8960 M-1 × cm-1 and in the biradical 12 occurs at 387 nm with ε = 13100 M-1 × cm-1. In the case of 41 this transition is strongly blue-shifted (~ 20 nm) due to the broken symmetry that results in less aromatic character of the molecule induced by the presence of σ(-CH2) bonds. However it appears as the most enhanced within the NN radical series. Figure 3.1: The absorption spectra of the NN radicals recorded in dilute toluene solutions at r.t. 50 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ The imino nitroxide (IN) derivatives feature a light orange-red colour in solution (Figure 3.3), with a broad absorption band around ~ 470 nm (n → π* transitions). Likewise as previously discussed for the NN radicals, also in the INs the intensities and spectral distributions of the n → π* transitions did not present appreciable dependence upon changing solvent polarities but strongly depend on the type of hetero-ring in which they have been connected (the pyridine rings hamper in a similar manner the radical optical properties as compared with those pyrazolyl-based). As shown in the case of 12 also the terpyridine biradical 13 shows a similar blue-shift on λmax. This effect arises from the increased aromatic character of the molecule once compared with its related monoradical 9. However, the absorption envelope of 13 in the visible is accompanied by a weakening in vibronic components, and a more remarkable quenching of the molar extinction (ε). The pyrazolyl- based radicals instead follow a somewhat simpler trend where the biradical 28 (λmax at 468 nm, ε = 1400 M-1 × cm-1) exhibits close to twice the extinction of the monoradical 38 (λmax at 469 nm, ε = 768 M-1 × cm-1), and the biradical 34 (λ -1 -1max at 467 nm, ε = 1082 M × cm ) in which both moieties are present shows an overall absorption very close to the superposition of the isolated monoradical envelopes 9 (λmax at 464 nm, ε = 321 M-1 × cm-1) and 38, in excellent agreement with the observed properties of the parent NN system 37. As previously observed in the case of 41, also in 42 the broad absorption in the visible appeared the most enhanced within this series of radical. The other distinct absorption of the radical moieties (π → π* transition of the aminoxyl-residues) is here strongly blue-shifted and cannot be discriminated from the π → π* transition of the organic backbone. Nevertheless the UV/Vis analysis represented a complementary and quick tool that allowed to identify the presence of different radical impurities (within the limit of 5%); in the IN based radical an absorption around 370 nm indicates the presence of left NN radical, vice versa an absorption in a solution of NN radical around 470 nm is consistent with the presence of an IN impurity. 51 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ Figure 3.2: The absorption spectra of the 12 as function of the solvent polarity recorded at r.t. Figure 3.3: The absorption spectra of the IN radicals recorded in dilute toluene solutions at r.t. 52 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ 3.2. Theoretical predictions of the UV/Vis absorption spectra in the biradical systems. The semiempirical approach ROHF/AM1 (Austin Model) including CIS (20,20) (20 electrons in 20 molecular orbitals) has been used to simulate the absorption spectra of three selected biradicals 12, 13, and 26. This would verify the applicability of the computational procedure for theoretical prediction of the absorption spectra of similar compounds, and furthermore allow assignment of the observed experimental transitions. It is worth to note that the simulations were based on the optimised geometry for the triplet-ground state. Therefore the nitronyl nitroxide biradical based on pyridine 12, the other one based on pyrazolylpyridine 26 and the imino nitroxide biradical 13 were selected. These radicals feature fairly large ∆EST gap (see Chapter 4). The calculated absorption spectra are summarised in Table 3.1 and in the Figure 3.4 and are expressed in function of the calculated oscillator strength. In order to explain the results reported in table 3.1 some points are therein clarified. An atom or molecule can be stimulated by light to change from one energy state to another. An atom or molecule in an excited energy state can also decay spontaneously to a lower state. The probability of an atom or molecule changing states depends on the nature of the initial and final state wavefunctions, how strongly light can interact with them, and on the intensity of any incident light. To a first approximation, transitions strengths or simply the probability that a certain type of transition is occurring, are governed by selection rules which determine whether a transition is allowed or disallowed. Practical measurements of transitions strengths are usually described in terms of the Einstein A and B coefficients or the oscillator strength (f). 1. Selection Rules 1. The parity of the initial and final wavefunctions must be different. 2. The spin can not change, ∆S = 0. 3. The change in orbital angular momentum can be ∆L = 0, ±1, but L=0 to L=0 transitions are not allowed. 4. The change in total angular momentum can be ∆J = 0, ±1, but J=0 to J=0 transitions are not allowed. 2. Transition Probability The transition probability is R2 with units of J (Joule) cm3, where R is the transition moment given by (a) R = < X | u | X > (a) 53 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ with u is the dipole moment operator and X the wavefunction. Basically what this equation indicates is that the strength of a transition is relative to how strongly the dipole moment of a resonance between energy states can couple to the electric field of a light wave. 3. Einstein coefficients For a two-level system (ground-state level i and upper level j), the rate of an upward stimulated transition (absorption, -dNi/dt or dNj/dt) is: -dNi /dt = Ni Bij Uv where Ni is the number density of atoms in the ground state, Uv is the light intensity, and the proportionality factor Bij is the Einstein B coefficient for absorption: B = 8 π3 R2ij / 3 h gi For stimulated emission the Einstein coefficient becomes: Bij = Bij gi /gj where gi and gj are the degeneracies of the ground and excited states, respectively. Atoms in the excited state can decay without the presence of an external light field due to stimulation due to "zero-point fluctuations." Zero-point fluctuations are the dynamic variations in the shape of an electronic orbital at any instant in time. These instantaneous orbitals can be described by a linear combination of the wavefunctions of the system, which provides the mechanism for transitions between different states of the system. The spontaneous decay rate: (-dNj/dt or dNi/dt) change into: -dNj/dt = Nj * Aji where Aji is the Einstein coefficient for spontaneous emission: Aji = 8 π h Bij gi / g 3j λm = 64 π4 R2 / 3 h gj λ 3m Since atoms in the upper level can decay by both spontaneous and stimulated emission, the total downward rate (-dNj/dt or dNi/dt) is given by: -dNi /dt = Ni (Aij + Bij Uv) 54 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ 4. Oscillator strength The oscillator strength of a transition is therefore a dimensionless number that is useful for comparing different transitions. It is defined as the ratio’s strength of an atomic or molecular transition with respect to the theoretical transition strength of a single electron, using a harmonic-oscillator model. For absorption: fij = 4 ε0 h c me Bij / e2 λm and for emission: fji = fij gi/gj Oscillator strengths can range from 0 to 1, or a small integer. A strong transition will have an f close to 1. Oscillator strengths greater than 1 result from the degeneracy of real electronic systems. Based on the computation, the predicted transition wavelengths correspond fairly well to the experimental values. The longest wavelength transitions in both 12 and 13 originate from the radical. In 13 this is an n→π* transition at 440 nm from the radical SOMO to an anti- bonding π* MO. In 12, however, there are two degenerate transitions (at 611 nm and 610 nm) between the radical HOMO and one of its anti-bonding σ*-type MOs. Unexpectedly, these two transitions are rather intensive. The second characteristic transition for nitronyl nitroxide (<415 nm) is reproduced as well. However, the predicted wavelength is higher than the experimental value of 387 nm. The calculated π→π* transitions arising from the terpyridine core follow a reasonable trend. The wavelength in 13 (286 nm) corresponds to the one measured for pure terpyridine (see experimental session). A bathochromic shift of ~ 70 nm is observed in 12 due to the enhanced conjugation. This is also in line with the experimentally measured spectra. In the case of the radical 26 a hypsochromic shift at 584 nm (HOMO (NN) → POMO (NN) and HOMO-1 (NN) → POMO (NN)) was calculated for the first fingerprint of the radical moiety. This is not more intense as compared with 12, and therefore does not reproduce the trend experimentally observed (pyridine hamper and pyrazole enhance) suggesting the presence of limits in this type of computation when singlet-triplet states are both thermally populated (see selection rules, point 2). The second fingerprint indeed is well reproduced (372 nm, π→π* transition from the core to the radical, observed 375 nm) and it is in line with the hypsochromic shift experimentally observed in all the pyrazolyl-based radicals. 55 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ Figure 3.4. Pictorial view of the calculated UV/Vis transitions in 12 (A), 13 (B) and 26 (C) by ROHF/AM1/CIS(20,20). The singlet-triplet gap on the optimised molecular geometry is collected in the small Table depicted on the right. 56 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ Table 3.1. UV/VIS spectra simulated by ROAM1 / CIS(20,20) for triplet ground state. Transitions Oscillator for 12 λ, nm strength MOs involved Comment π→σ* transition involving only 1 611 0.014 HOMO (NN) → LUMO+1 (NN) the spin bearing part of the radical π→σ* transition involving only 2 610 0.035 HOMO (NN) → LUMO+1 (NN) the spin bearing part of the radical 3 415 0.008 HOMO (NN) → LUMO+3 (terpy) π→π* transition from the radical to the terpyridine core 4 356 0.022 HOMO (terpy) → LUMO (terpy) π→π* transition from the HOMO-1 (terpy) → LUMO (terpy) terpyridine core 5 356 0.052 HOMO (terpy) → LUMO+1 (terpy) π→π* transition from the HOMO (terpy) → LUMO (terpy) terpyridine core 6 352 0.016 HOMO-* (terpy) → LUMO+* (terpy) π→π* transition from the terpyridine core 7 307 0.023 POMO (NN) → LUMO+2 (terpy) n→π* transition from the POMO (NN) → LUMO+3 (terpy) radical to the terpyridine core Transitions , nm Oscillator for 13 λ strength MOs involved Comment The n→π* transition involving 1 440 0.035 POMO (IN) → LUMO (IN) only the spin bearing part of the radical 2 313 0.170 HOMO (terpy) → LUMO+1 (terpy) π→π* transition from the HOMO-1 (terpy) → LUMO (terpy) terpyridine core 3 310 0.038 HOMO (terpy) → LUMO (terpy) π→π* transition from the HOMO-1 (terpy) → LUMO+1 (terpy) terpyridine core 4 306 0.013 HOMO-3 (IN) → POMO (IN) π→n transition from the HOMO-3 (IN) → POMO (IN) radical 5 286 0.669 HOMO (terpy) → LUMO+1 (terpy) π→π* transition from the HOMO-1 (terpy) → LUMO (terpy) terpyridine core 6 286 0.162 HOMO (terpy) → LUMO (terpy) π→π* transition from the HOMO-1 (terpy) → LUMO+1 (terpy) terpyridine core Transitions Oscillator for 26 λ, nm strength MOs involved Comment HOMO (NN) → POMO (NN) π→n transition involving only 1 584 0.002 HOMO-1 (NN) POMO (NN) the spin bearing part of the → radical 2 382 0.005 HOMO-3 (pyr) → LUMO +1 (pyr) π→π* transition involving only the pyrazolyl core 3 372 0.065 HOMO (pyr) → LUMO (NN) π→π* transition from the core HOMO-1 (pyr) → LUMO (NN) to the radical 4 372 0.042 HOMO (pyr) → LUMO (NN) π→π* transition from the core HOMO-1 (pyr) → LUMO (NN) to the radical 5 365 0.021 HOMO-1 (pyr) → LUMO+3 (pyr) π→π* transition from the HOMO (pyr) → LUMO (pyr) pyrazolyl core 6 320 0.009 POMO (NN) → LUMO (pyr) n→π* transition from the radical to the pyrazolyl core 7 320 0.011 POMO (NN) → LUMO (pyr) n→π* transition from the radical to the pyrazolyl core 8 319 0.005 HOMO-3 (NN) → LUMO +1 (pyr) π→π* transition from the radical to the core 57 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ 3.3. The IR absorption spectra of the radicals. Clear differences between NN and IN radicals were witnessed if one compares their respective IR absorption envelopes. This technique represents a quick complementary tool in order to monitor the process of condensation reactions between the different aldehydes with the 2,3-dimethyl-2,3-bis(hydroxylamino)-butane. It is worth to note that all the isolated radical precursors (e.g 7, 11, 26, 36 and 41) could not be purified either on alumina or silica columns due to their quick decomposition, but only by rapid washing with different solvents (see Experimental Session). The IR analyses was very helpful when, within the condensation reaction, the mixture showed no hint of precipitate formation and the radical precursor could not be isolated from the reaction medium due to its partial dehydratation from the bis-N,N'- hydroxy-imidazolidine to N-hydroxy-imidazolidine as for 32. In the Figure 3.4 is shown the spectrum of the isolated free base 2,3-dimethyl-2,3-bis(hydroxylamino)-butane. Figure 3.4: FT-IR spectrum of the free base recorded in KBr pellet at r.t. For comparison, in Figure 3.5 are reported the infrared spectra of 5, 5"-diformyl- 2,2':6',2" terpyridine (10) the precursor 5,5"-bis(1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin- 2-yl)2,2':6',2" terpyridine (11), the nitronyl nitroxide biradical 12 and the imino nitroxide biradical 13. The complete conversion of the dicarbaldeyde 10 into the radical precursors 11 clearly implied the disappearance of the strong signal at 1693 cm-1 (C=O) being accompanied with the formation of another strong and broad signal around 3252 cm-1 for 11 originating from the OH stretching-mode. After the oxidation process with NaIO4 and purifications of the radicals the OH signals disappeared in 12 and 13 being accompanied, in the case of 12, by an 58 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ unusually strong signal at 1351 cm-1. This was assumed to arise from the N-O stretching of the imidazolidin moiety. In the case of 13 the novel signal, absent in 11, is shifted to longer wavenumbers (1378 cm-1) and accompanied by the appearance of a strong signal at 1554 cm- 1 that may be tentatively attributed to the C=N stretching mode. In Figure 3.6 is reported a case in which isolation of the radical precursor 32 was not possible. Therefore the IR technique was applied in order to follow the condensation reaction between 31 and the 2,3- dimethyl-2,3-bis(hydroxylamino)-butane, by collecting aliquots of the reaction mixture at different times until complete disappearance of the carbaldehyde signal (,νC=O appeared at 1689 cm-1). Upon oxidation of the precursor 32, the difference between the nitronyl and imino- nitroxide radicals is visible in the shift at longer wavenumbers associated with the N-O stretching frequency (from 1352 cm-1 in NN 33 to 1371 cm-1 in IN 34) together with a clear decrease in the relative signal intensities. These differences were found consistent in all the radicals purified. The NN radicals always featured the N-O stretching frequency around 1350- 1360 cm-1 with a very strong signal, while in the IN radicals such resonance is shifted (around 1370 cm-1), much less intense and, in the case of 28 and 43, appeared embedded in the spectra and could not be unambiguously identified. In Figure 3.7 those traces are reported together to allow quick comparison. Figure 3.5: FT-IR spectra in KBr pellet recorded at r.t. for the terpyridine derivatives. 59 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ N O N N N N 34 O N N N N N N N N N HO N N N OHC 31 32 OHCHO N N OH NHO N N N NO N O N+ 33 + N O N- - O Figure 3.6: FT-IR spectra in KBr pellet recorded at r.t. Note the asymmetric peak at 1689 cm- 1 in 31 originating from the two slightly different carbaldehyde groups (one attached to the pyridine and the other to the pyrazole moiety). 60 Chapter 3- The Radical`s Optical Properties ___________________________________________________________________________ Figure 3.7: FT-IR spectra magnified in the region of the N-O stretching vibration (ν, cm-1) for NN and IN radicals, recorded in KBr pellet at r.t. 61 ____________________________________________________________________________ Chapter 4 - EPR Analysis 4.1. The EPR analysis of monoradical systems in solution. The EPR spectra in dilute (≤ 10-4 M) and oxygen free toluene solutions of the nitronyl nitroxide monoradicals 8 and 37 exhibit clear isotropic five line pattern at giso = 2.0066(1) and 2.0065(1) respectively. The spectra are shown in Figure 4.1 for 8 and 4.2 for 37. O O- N Br N N N N+ N N +N N 8 37 -O O Such five lines spectra originate from the interaction of the unpaired electron with the two equivalent nitrogen nuclei of the imidazolyl moiety. The relative intensity of each line follows the expected 1:2:3:2:1 ratio. The spin energy levels for the S=1/2 system, as for the radicals 8 and 37, are described by the spin-Hamiltonian Ĥ (1), in which are reported the contributions from the electron-Zeeman (ĤEZ) and the hyperfine interaction (ĤHF). Higher order terms such as nuclear Zeeman and Quadrupole are usually not considered [1]. Ĥ = ĤEZ + ĤHF + higher order terms (1) Ĥ = gβeŜiBo + ∑ aijŜi Îj (2) The electron-Zeeman term (ĤEZ = = gβeŜiBo) describes the interaction between the electron spin operator Ŝi and the applied external magnetic field tensor Bo, βe is the Bohr magneton (|eh/ 4 π me|= 9.2740 × 10-24 J/T, written also as µB). Here the g value is the observable, and any deviations from ge (the free electron in vacuum with resonance value at 2.0023) result from the so called spin-orbit coupling between the unpaired spin (Ŝi) and the orbital angular momentum (L). This coupling can be described by the Hamiltonian (3) Ĥ = λLŜi (3) Where λ represents the spin-orbit coupling constant (sometimes expressed as ξ =2Ŝiλ). The spin orbit coupling is influenced by the admixture of other orbitals, e.g. lone pair orbitals, to the singly occupied orbitals; λ is constant for a particular shell in a particular atom, and increases sharply with the atomic mass. For organic radicals, such as the nitronyl- and imino nitroxide objects of this thesis, the spin-orbit coupling constant is usually very small and therefore the 62 Chapter 4- EPR Analysis ___________________________________________________________________________ observed deviation of the g value from ge for both 8 and 37 is small (∆g = gobs – ge ~ 0.005). In the spin-Hamiltonian (2) the hyperfine term (ĤHF = ∑ aijŜi Îj) describes the interaction between the magnetic moment of the electron spin (Ŝi) and the magnetic moment of a nucleus in the vicinity (Îj), and the coefficient aij represents the hyperfine coupling constant (hfc). Thus, the best fitting for the observed EPR lines in 8 has been achieved by applying the spin- Hamiltonian (2), with the parameters aN = 0.748(2) mT, ∆Bpp = 0.105 mT, Lorentzian/Gaussian line-width ratio = 1/3 [1a, 1d], and giso = 2.0066 (1). The fitting is shown in Figure 4.1 as dashed-dotted line. The L/G ratio is taken into account because the intensity of the EPR transition is a lineshape-function according to equation (a) I (ν, B) = Cν P 2ij f [(ν-ν )20 , σv] (a) where C is a constant dependent on the type of microwave cavity and its temperature, ν0 is the frequency required for resonance at any particular magnetic field B (see the Zeeman term ν= gβeŜiBo / h), f is the line-shape function (Gaussian, Lorentzian or a mixture of both), σv is the line-width parameter (in frequency units) and Pij2 is the time-independent part of the transition probability between the state i and j. 1.0 0.8 toluene 2-propanol 0.5 0.4 0.0 0.0 -0.5 -0.4 -1.0 (A) -0.8 (B) 332 333 334 335 336 331 332 333 334 335 336 337 B / mT B / mT Figure 4.1: (A) EPR spectrum for the radicals 8 recorded in toluene solution (black lines) at 293 K with concentrations of 8 × 10-5 M. Experimental parameters: 9.39987 GHz, 100 kHz modulation frequency, 0.03 mT, 21 msec time constant, 42 sec sweep time, 2.0 mW microwave power, 105 gain, 4 scan were accumulated and averaged. The dashed-dotted line represents the computer simulations with parameters given in the text. (B) Comparison between the EPR spectrum of 8 recorded in toluene (solid line) and then in 2-propanol (dashed-line, ------) in which no clear solvent effect on aN has been observed. Experimental parameters as those reported in (A). 63 dχII / dB (a.u.) dχ II / dB (a.u.) Chapter 4- EPR Analysis ___________________________________________________________________________ Figure 4.2: EPR spectrum for the radicals 37 recorded in toluene (black line) and in 2-propanol (dotted blue line) solutions at 293 K with concentration of 10-4 M. Experimental parameters (toluene) 9.40021 GHz, (2-propanol) 9.40026 GHz, then 100 kHz modulation frequency, 0.03 mT, 21 msec time constant, 42 sec sweep time, 1.0 mW microwave power, 105 gain, 4 scan were accumulated and averaged. The two spectra show the solvent effect in which aN increases from ~ 0.755(3) mT in Toluene up to ~ 0.767(3) mT in 2- propanol. In the monoradical 8, in addition to the major five lines, the high resolution spectra reveals a more complex pattern (Figure 4.3, bold blue line) with at least 13 visible additional splittings overlapped on each N hfc. Figure 4.3: The high resolution EPR spectrum of 8 (blue line, 8 × 10-5 M in toluene) for the MI =+1 line, recorded with the following parameters: 9.39759 GHz, 100 kHz modulation frequency, 0.01 mT modulation amplitude, 41 msec time constant, 84 sec sweep time, 0.8 mW microwave power, 8×104 gain, temperature 293 K, 10 scan were accumulated and averaged. The other lines represent the computer simulations with parameters given in the text. These superimposed lines originate from the presence of additional couplings of the single unpaired electron (S = ½) with twelve hydrogen nuclei (I=½,) of the four methyl groups (aH = 0.022(1) mT), the 4’ (aH = 0.041(1) mT) and 6’ hydrogens (aH = 0.044(2) mT) of the pyridyl moiety, therefore experimentally demonstrate that non "zero spin density" resides also in the pyridine ring. Such effect is known as spin-polarization [Reference 1b, and Chapter 4 in 1c]. These lines rapidly saturate at relatively low powers (< 5 mW) at room temperature, while the major five lines hardly saturate even at 20 mW at room temperature. The parameters employed for the simulation of the H-hfc agree well with the estimated H-hfc found for similar systems in the literature. These references are collected in Figure 4.5. 64 Chapter 4- EPR Analysis ___________________________________________________________________________ Similarly to what has been shown previously in 8, also for the monoradical 37 in addition to the major five lines, the high resolution spectra of the central transition line (MI = 0) recorded in 2-propanol reveals a complex pattern (Figure 4.4A) with at least 17 visible additional splittings overlapped on each N hfc. These transitions originate again from extra couplings with the imidazolyl protons of the four methyl groups and the two hydrogens of the pyrazole ring. However, the spectrum is not symmetric with respect to the centre, once compared with the similar one for 8 reported in Figure 4.3. In particular, the third line (enclosed in a red circle in the figure 4.4A) is not well resolved. The identical effect has been observed in the nitronylnitroxide monoradical based on the pyrazole P (Figure 4.4B). Figure 4.4: (A) The high resolution EPR spectrum of 37 (7 × 10-5 M in 2-propanol) for the MI = 0 line, recorded with following parameters: 9.41021 GHz, 100 kHz modulation frequency, 0.01 mT modulation amplitude, 41 msec time constant, 84 sec sweep time, 0.5 mW microwave power, temperature 293 K. (B) The high resolution EPR spectrum of P for the MI = 0 line recorded under identical conditions as in 37. The small inset in (B) shows the total EPR spectrum of P. 65 Chapter 4- EPR Analysis ___________________________________________________________________________ Solution EPR Solvent aN aMe aHpyrazole aNpyrazole O N H2O 8.33 0.185 0.315 0.094N O-+ H2O/NaOH 8.45 0.205 0.279 0.064 hfc in Gauss T = 293 K N H N Ref. 2a - O Solvent aN aMe aHbenzene N+ N hexane 7.40 0.21 0.54 only ortho-protons N N hfc in Gauss T = 293 K O Ref. 2b - O Solvent aN a+ Me aHbenzene aOH N 0.207 OH CCl4 7.50 0.499 0.177 < 0.02 N ortho-H meta-H O hfc in Gauss T = 293 KRef. 2c O Solvent aN aMe aN thienoimidazole aN thienoimidazole aH, (N)H N N =N- -N-H S N N Benzene/ 7.15 0.20 0.47 0.12 0.16 H methanolO Benzene 7.18 0.20 0.47 0.12 Ref. 2d methanol-d1 hfc in Gauss T = 293 K - O H H 1 N + H-ENDOR Ref. 2e H 0.468 aN = 7.29 N Unit in Gauss - 0.206 O H H 0.519 -0.293 Observed in mineral oil at 290 K 1H-NMR (CDCl3 , RT) - ∆H aO H-ortho ∆H aH-meta ∆H aH-para ∆H aMe N+ Ref. 2f -3.31 + 0.446 +1.31 -0.177 -2.83 +0.382 +1.485 -0.201 N O - O N+ Ref. 2f -2.20 +0.296 +1.00 -0.135 -2.20 +0.296 +1.415 -0.196 N N +0.80 -0.108 O - O N+ Ref. 2f N -3.63 +0.490 +0.92 -0.125 +1.450 -0.191 N O - O N+ Ref. 2f -3.21 +0.435 +1.40 -0.189 -3.21 +0.435 +1.440 -0.194 N N O ∆H in KHz, a in Gauss ∆H = -a (γe γN) g βΗ/ 4kT Figure 4.5: The proton and nitrogen hyperfine coupling (aH, aN) constants obtained experimentally in nitronyl nitroxide radicals as found in the literature. 66 - + Chapter 4- EPR Analysis ___________________________________________________________________________ The EPR spectra in dilute and oxygen free toluene solutions of the imino nitroxide monoradicals 9 and 38 feature seven line patterns at giso = 2.0061(1) and 2.0060(1) respectively. N Br N N N N N 9 N NN 38 O O These lines originate from the interaction of the unpaired electron with the two non equivalent nitrogen nuclei of the imidazolyl moiety, with intensities ratio 1:1:2:1:2:1:1, since two of the N- hfc splittings overlap each other. The spectrum of 9 is shown in Figure 4.6, together with the best simulation (dashed-dotted line) obtained with parameters aN1 = 0.885(2) mT, aN2 = 0.430(3) mT, ∆Bpp = 0.106 mT, and assuming pure Lorentzian line. 1.0 Figure 4.6: EPR spectrum of Observed the radical 9 recorded in dilute Simulated (10-4 M) and oxygen free 0.5 toluene solution at 293 K. Experimental parameters: 9.400210 GHz, 100 kHz modulation frequency, 0.03 mT 0.0 modulation amplitude, 21 msec time constant, 42 sec sweep time, 2.0 mW microwave power -0.5 (0.8 mW for B), 105 gain, 4 scan were accumulated and averaged. -1.0 331 332 333 334 335 336 337 B / mT The nitronyl nitroxide and imino nitroxide monoradicals followed nicely the Curie-law in solution with linear increase of the signal intensities (DI) upon lowering the temperature according to equation (6). This is clearly expected for isolated S = ½ spin systems. DI = χEPR ∝ (Na/3kbT)× µ 2 2B g S(S+1) = C/T (6) In equation (6) Na represents the Avogadro number, kb Boltzman constant, µB the Bohr magneton and χEPR the magnetic susceptibility. Note that χ is actually the sum of χ(para) + χ(dia) + χ(TIP) where only the paramagnetic contribution χ(para) (large and positive) is temperature dependent, χ(dia) is small and negative (≤ 0.001×χ(para)), and χ(TIP) is small and positive. Although the double integration of the EPR absorption line discharges the 67 dχ II / dB (a.u.) Chapter 4- EPR Analysis ___________________________________________________________________________ diamagnetic part, in practice extrapolation of the linear fit in the Curie Plot for T→ ∞, should provide a negative intercept; this corresponds to the χ(dia) term. All the recorded solution EPR spectra for the monoradical systems 8, 9, 37 and 38 feature a very small increase in the peak- to-peak line-width upon cooling. This is shown in two examples for the radical 8 (NN) and 9 (IN) (Figure 4.7A and 4.7B respectively) along with the related Curie-behaviour for 8 (Figure 4.8A and 4.8B). Temperature 1.0(A) 298 K 288 K 278 K 0.5 268 K 258 K 248 K 238 K 0.0 228 K 218 K 208 K -0.5 -1.0 3315 3330 3345 3360 3375 B / Gauss 1.0 Temperature (B) 298 K 278 K 258 K 0.5 238 K 218 K 202 K 0.0 -0.5 -1.0 3320 3330 3340 3350 3360 3370 B / Gauss Figure 4.7: EPR spectra of the NN monoradical 8 (A) an the IN monoradical 9 (B), recorded as a function of the temperature. The samples were allowed to equilibrate in the cavity-cell at the fixed temperature for over 10 min. Each spectrum was both base-line and centre-field corrected. The error in the temperature setting was found to be within ± 1K. 68 DI (a.u.) DI (a.u.) Chapter 4- EPR Analysis ___________________________________________________________________________ 1.2 -4 4.8 10 M 10-3 M -3 4.4 1.1 7 x 10 M 4.0 1.0 3.6 (A) 0.9 (B) 3.2 0.8 0.0032 0.0036 0.0040 0.0044 0.0048 200 220 240 260 280 300 320 1 / T (K-1) Temperature (K) Figure 4.8: (A) The Curie-behaviour in solution of radical 8. The closed circles (●, DI) represent the double integration of the EPR signals recorded at various temperatures. All the spectra were base-line corrected prior to their double integration. The solid line shows the theoretical linear fitting according to the Curie equation (6) with the Curie constant, C, large and positive (C = + 1113.6, from the linear fitting) and the diamagnetic term small and negative (-0.6). In the Curie constant is also included the temperature independent paramagnetism χ(TIP). (B) The product DI × T vs T for 8 recorded with different radical concentrations. The light blue arrow underlines the suitable range in which the monoradical unit can be used as spin-standard. The red arrow shows the temperature range where a large deviation from the linearity in DI × T vs T is observed. This effect corresponds to a phase change in the fluid toluene solution. 4.2. The EPR of biradical systems in solution. The EPR spectra of the biradicals in solution feature a pattern very different from those observed in the simple S= ½ systems, and therefore the theoretical background is therein treated in few more details. The low lying excited states of a magnetic system are generally described in terms of a spin-Hamiltonian as earlier shown in the case of S=1/2 systems. For two interacting spins the phenomenological spin-Hamiltonian (1) is modified into (7): Ĥ = gβBoŜa,b – 2JŜaŜb + ∑ij aNij × (ŜaÎNij + ŜbÎNij) (7) The term 2JŜaŜb = Ĥexch represents the electron-exchange interaction, also termed Heisenberg-Dirac-van Vleck Hamiltonian, HDVV [3,4], Ŝa and Ŝb the electron-spin operators and J the isotropic electron-exchange coupling constant. When the EPR spectra are analyzed in terms of (7), the spin exchange parameter plays the role of numerical fitting parameter needed to reproduce the experimental data. At first approximation in (7) the contributions arising from slow molecular interconversion in solution are neglected. Those are responsible for the sometimes observed strong temperature dependence in the linewidth-broadening. In fact in exchange-coupled systems the ESR linewidth is determined by both the fluctuation of the (t)-dependent exchange interaction J(t) around its time averaged (the main exchange 69 D.I (a.u.) DI x T / [C] x Nav = Spin concentration standard Chapter 4- EPR Analysis ___________________________________________________________________________ term in equation (7)), while the resonance positions of each line are determined mainly by . Thus, if the time dependent term is considered, the exchange term in the spin-Hamiltonian (7) from Ĥ = - 2JŜaŜb needs to be changed into (8) Ĥ = 2 ∑a,b [Ja,b(t) - ] ŜaŜb (8) The theoretical and conceptual bases for understanding spin exchange interactions were laid out by Anderson [5], by Löwdin [6], by Nesbet [7], by Hai, Thibeault and Hoffmann [8], by Kahn and Briat [9], by Noodleman [10], Noodleman and Davidson [11]. A quantitative description of spin-exchange interaction from the theoretical perspective represents however a challenging task, since it requires state-of-the-art computational efforts on the basis of either configuration interaction wave functions (CI) [12-14] or density functional theory DFT [15-17]. Recently Illas et al. have provided a comprehensive review on the conceptual and theoretical issues concerning these quantitative approaches [18]. The HDVV model Hamiltonian 2JŜaŜb acts in spin space only and, hence, assumes that the spatial part of the wave functions involved in magnetic coupling is the same for all the neutral spin configurations. This treatment is discussed in details in some advanced books [20]. For a two electron spin system, the spin space has a dimension of four, and the basis for this space is simply given by the combination between α and β spin (α for spin up, β for spin down). This leads to |αα>, |ββ>, |αβ>, and |βα>. Since in the HDVV model, the total square spin operator S2, and its z-component S 2z , commute with each other, it is possible to find a set of eigenfunctions common to the operators. The eigenfunctions of S2 and S 2z are denoted |S, Ms> and lead to the description of the four spin states as |1,1> = |αα>, |1,-1> = |ββ>, |0,0> = |αβ>, and |1,0> |βα>. Therefore the spin states are combined in singlet |S> |0,0> and triplet |T> with its three Sz components. The singlet and triplet states are also egenfunctions of the HDVV Hamiltonian with energies (3/2) J and (-1/2) J [21,22]. Therefore the magnetic coupling for 2J < 0 1,+1> constant is given by the energy difference - J/2 1,0> 1,-1> between |S>-|T>=2J corresponding to the singlet- triplet energy gap (∆EST) as shown in Figure 4.9 Energy 2J = ∆E for a singlet ground state. ST This was the procedure that has been adopted when numerical diagonalization of the spin- + 3J/2 0,0> Hamiltonian (7) was used for the simulation of the entire solution EPR envelopes in the biradical Figure 4.9: The singlet-triplet energy gap systems. The exchange interaction J however for S = 1 system when 2J < 0. cannot be fully exploited for structural 70 Chapter 4- EPR Analysis ___________________________________________________________________________ investigation because it depends on several contributions, such as the radical distance and the number, nature, geometry and topology of the bonding involved between the radical entities. However, the contributions to J can be broadly classified into through-bond and through-space. The through-bond contribution is regarded as an electrostatic interaction, and decreases quite rapidly if the bondings are single σ bonds. A general expression for radical centers connected with n number of σ bonds was proposed several years ago [23] in the form: J [MHz] = (-1)n (3 × 10 6-n) (9) The through-bond interaction originates from spin-polarisation effects of the molecular bonding involved within the coupling unit. Such effect is expressed as a function of (ζ )= ρ↑ - ρ↓ / ρT, where the symbols indicate ρ↑ spin up densities (positive, α, spin up, parallel to the applied field) and spin down ρ↓ densities (negative, β, spin down, antiparallel), and ρT the total spin densities (e.g. in the isolated biradical molecule ρT = 1, and in the general case ρT = ST ). We might depict the spin-densities in the form (10) ρ (r) = ∑I ρI ψI (r) ψI*(r) = 1 (10) With J ∝ f ρ(r) = fρ (θ,r) (11) The exchange energy term J is linked with (ρ (r)) with a non-univocal direct correlation (vide infra). In equation (10) ρI (r) and ψI are the local spin-densities on the ith atoms, ψI its atomic wave-functions and (r) its distribution in the space (spin-densities distribution). The sum indicates that this distribution is normalized to one (over the entire molecule) and each atom will retain a fraction of it. By writing ψI = ∑J cij Φi in which the molecular function of the unpaired electron can be expressed as linear combination of atomic functions Φi through the coefficients cij we get the relation (12) [24-25]: ρ (r) = | c 2ij| (12) Thus ρI simply indicate the probability that the electron in the molecular orbital ψI resides in the atomic orbital Φj, and measures the unitless unpaired π-electron population ρI on the atom J when this atom bears only a single orbital occurring in ψI. 71 Chapter 4- EPR Analysis ___________________________________________________________________________ It is intuitive to suggest that a less efficient spin-polarization should lead to a decrease in the “communication” between radical sites through-bond, up to the limit in which no interaction occurs, and the radical entities behave as independent spins. In addition, increasing the dihedral θ angular torsions ( i.e. when θ ≠ 0°) between radical sites and coupling unit, as indicated in the general relation (11), leads to a decrease in the magnitude of J. Furthermore J may even change sign at angles θ ~ 80° [26]. The second contribution to J is due to the so called through-space interactions, which varies with the radical separation, . This is a magnetic interaction in the real sense, and it is based on the magnetic dipole moment generated by the unpaired spins (dipolar spin-spin interaction). Although the functional dependence on is also not well established, an exponential decrease with increasing is generally assumed according to equation (13) J = J0 exp [-α ( - d)] (13) Where d represents the minimum distance-approach, J0 and α are empirical values [27]. However an estimation of J0 can sometimes be obtained by applying Anderson’s theory [28a,b]. 4.2.1 The spectral EPR analysis of the biradical systems in solution. Considering the case of nitronyl and imino nitroxide biradical systems, according to the relative size of J with respect to , three cases could be encountered in the observed solution EPR envelopes: 1) Case |J| >> |a| the line positions in the observed spectrum are given by equations (14), (15) and (16) where MI (=mI (a) + mI(b)) represents the z-component of the total nuclear spin angular momentum for the halves (a) and (b) carrying the radical units (the imidazolyl moieties), ∆mI = |mI (a) – mI(b)|, and ϕ = arctan (aN∆mI / 4J) hν (T) = gβB0 + (aN/2) MI ± (aN∆mI / 2)tanϕ (14) hν (S) = gβB0 + (aN/2) MI ± (aN∆mI / 2)tanϕ ± 2J (15) with (T) = [1 / √ 2(α β + β α)] and (S) = [1 / √ 2(α β - β α)] (16) When J is large compared to a (e.g. for aN ~ 0.75 mT, J > 7.5 mT), ϕ → 0 and equation (14) reduced into (17) 72 Chapter 4- EPR Analysis ___________________________________________________________________________ hν (T) = gβB0 + aN/2 MI (17) Thus one would anticipate (2×4×1 + 1) = 9 lines pattern for the nitronyl nitroxide and (2×2×1 + 1) × (2×2×1 + 1) = 25 lines for the imino nitroxide biradical, with the observed aN half of that featured by the related monoradical. Since most of the transition lines in the imino nitroxide biradical will overlap each other, a 13 lines pattern is practically observed. Simulation of the solution spectra can provide the lower limiting value of J. The total spread in magnetic field (spectral-width) of the observed EPR spectrum would be the same as that observed in the monoradical case. 2) Case |J| << |a| In the spin-Hamiltonian (7) the exchange term almost vanishes leading to the limiting equation (J~0) (18) Ĥ = gβB0(Ŝa,b) + ∑ij aNij × (ŜaÎNij + ŜbÎNij) = gβeBoŜi + ∑ aijŜi Îj (18) Each radical centre behaves independently, and can be treated as uncorrelated spin-system (e.g. S=1/2 system). It follows that the total spread in the magnetic field of the observed EPR spectrum exactly matches that of a monoradical system with intensity doubled compared to the later. 3) Case |J| ~ |a| The mixing between (T) and (S) through hyperfine interaction is possible. In the EPR spectrum the numbers of observed lines are determined by the ratio aN∆mI / 4J. Therefore in the limit of J = a hν (T) = gβB 20 + (aN/2) MI ± (aN∆mI / 8) (19) The numbers of observed lines will be more than nine for the nitronyl nitroxide and more than thirtheen for the imino nitroxide biradicals. The total spectral-width in the observed EPR spectrum would be larger than that observed in the respective monoradical case. The simulation of the spectrum by EPR analyses studies would provide the exact size of |J| and not only the relative magnitude. 73 Chapter 4- EPR Analysis ___________________________________________________________________________ 4.2.2 The determination of thermally activated spin-states in the biradical systems. The double integration (DI) of the EPR envelope of paramagnetic species is directly proportional to the concentration ([CS]) of the unpaired electrons (spin-concentration) and, in absence of signal saturation DI is inversely proportional to the absolute temperature, according to the Curie-law (equation 6). Hence, comparison between the integrated intensities of the biradical system versus monoradical standards with known concentrations, recorded under identical conditions (filling factor, temperature, power, modulation amplitude, etc.) would provide the spin concentration of the biradical system. However, different spin-states contribute in different way according to the ratio |J/kbT|. We therefore can encounter the two limiting cases (A) and (B) in which the relation (20) and (21) can be obtained by combining the Curie and the Bleaney-Bowers equations: (A) |2J/kbT| >> 1 DI(S=1) = cNA × {(4∆E/3kbT) / [3 + exp –(2J/ kbT)]} (biradical) (20) DI( –(2J/ kbT)S=1/2) = cNA × {(2∆E/3kbT) / [3 + exp ]} (monoradical) (21) Here again NA represents the Avogadro number, kb the Boltzmann constant and DI the double integration of the signal intensities (for monoradical S = ½ or biradical S = 1); the coefficients written in bold in equation (20) and (21) simply indicate the total number of states (Ŝ) (e.g four states for biradical, two states for monoradical), under the frame of Boltzmann distribution (vide infra). The DI ratio between the biradical systems against monoradical standard is obtained upon applying the limit: J→ ∞ (biradical) in equation (20), J→ 0 (monoradical) in equation (21) Therefore: DI(S=1) / DI(S=1/2) = [cNA × (4∆E/9kbT)] / [cNA × (2∆E/12kbT)]= 2.66666 spin (22) Note that J is assumed large (∞) and positive. If J → - ∞, then [3 + exp –(2J/ kbT)] → ∞ and DI(S=1) = 0 at every temperature. (B) |2J/kbT| << 1 74 Chapter 4- EPR Analysis ___________________________________________________________________________ The thermal energy is much larger than the exchange energy, and singlet-triplet states are in equilibrium and follow the Boltzmann distribution. The ratio of molecules in triplet state: Ntriplet / cNA = ¾ × (exp–2J// kbT) = 75% = 0.75 (23) And those in the singlet: N –2J// kbTsinglet / cNA = ¼ × (exp ) = 25% = 0.25 (24) The double integrated signal intensities compared with a monoradical standard would thus account for two doublets (2.0 spins): [Ntriplet / cNA + Nsinglet / cNA]/ [NS=1/2 / cNA] = 0.75 × 2.6666 + 0.25 × 0 = 2.00 spin (25) Therefore only when the thermal energy becomes smaller than the exchange energy (kbT<׀J׀), the ground state starts to populate at the expense of the thermally excited state, until the equilibrium is fully shifted (T → 0) to give in case of triplet ground state (S=1) DI = 2.6666 DI mono, or in the case of singlet ground state (S=0) DI = 0. Consequently from the simulation of the diluted solution EPR spectrum, and the measure of the spin-concentration for a biradical system it is possible to obtain both lower and upper limits of the intramolecular exchange interaction J but not its sign. 4.3. The observed EPR spectra of the π- conjugated biradicals in solution. The EPR spectra of the terpyridine-based biradical 12 (NN) and 13 (IN) are better resolved in dilute toluene solution [1 × 10-4 M] rather than in chloroform or dichloromethane, and are shown in Figure 4.9 and 4.10 respectively. The nitronyl nitroxide biradical 12 exhibits a well resolved nine line pattern, hence demonstrating that the intramolecular exchange interaction between the two radical fragments is much larger than the hyperfine terms (2J/aN >>1, i.e. 2J >> 7 × 10-4 cm-1). The observed line spacing (apparent hyperfine interaction, aN) as pointed out in equation (17) would therefore correspond to half of that of the simple monoradical system (where aN = 0.748 mT). The observed spectrum (black line) is well reproduced by using the simple spin-Hamiltonian (2) with the apparent hyperfine interaction aN/2 = 0.374(1) mT over four equivalent nitrogen nuclei, an average g value, giso = 2.0066(1), pure Lorentzian line-shape for the single components and peak-to-peak line width, ∆Bpp = 0.135 mT. The simulation is shown with a magenta dotted line in Figure 4.10. An identical result is achieved by full diagonalization of the spin-Hamiltonian (7) with aN = 20.944 MHz = 75 Chapter 4- EPR Analysis ___________________________________________________________________________ 0.748 mT, |J/aN| > 55 (e.g. |2J/kb| = ∆EST ≥ 0.08 cm-1 (Gaussian line-width = 3.194 MHz, 0.840 MHz modulation amplitude, weight factor = 1 for the averaged molecular conformations under strong exchange limit, (rms ≤ 0.3105)). Figure 4.10. EPR spectrum of the nitronyl nitroxide biradical 12 (black line) recorded in dilute and oxygen free toluene solutions. Parameters: Frequency 9.40022 GHz, 100 KHz mod. frequency, 0.03 mT modulation amplitude, 21 msec time constant, 42 sec sweep time, 2.6 mW microwave power, 105 gain, temperature 260 K, 6 scan were accumulated and averaged. The dotted magenta line shows the spectrum simulation with parameters given in the text. Similarly, the strong exchange-interaction between the radical fragments is observed in the imino nitroxide biradical 13 (Figure 4.11). The spectrum can also be reproduced by assuming the strong exchange limit (J/aN>>1, i.e. J >> 4.2 × 10-4 cm-1) with the following parameters, aN1/2 = 0.430(1) mT, aN2/2 = 0.225(2) mT, giso = 2.0061(1), the Lorentzian/Gaussian line-shape of 1/3 and the peak-to-peak line width, ∆Bpp = 0.200 mT (red dotted line). The simple simulation according to the spin-Hamiltonian (2) however, although reproducing exactly the peak positions, is not completely satisfactory in reproducing the intensities of lines 1 and 13. A superior fitting can be achieved using again numerical diagonalization of the spin-Hamiltonian (7) with parameters 2׀J/kb0.065 ≤ ׀ cm-1 (rms ≤ 0.3544), |aN1|iso= 11.679 MHz, |aN2|iso = 25.573 MHz, Gaussian line-width = 2.216 MHz, 0.840 MHz modulation amplitude, weight factor = 0.97 for the averaged molecular conformations in strong exchange limit. This simulation corresponds to the blue dashed line in Figure 4.11. 76 Chapter 4- EPR Analysis ___________________________________________________________________________ Figure 4.11. EPR spectrum of the imino nitroxide biradical 13 (black line) recorded in dilute and oxygen free toluene solutions. Parameters: Frequency 9.40240 GHz, 100 KHz mod. frequency, 0.3 Gauss modulation amplitude, 21 msec time constant, 42 sec sweep time, 5.0 mW microwave power, 4 x 104 gain, temperature 260 K, 4 scan were accumulated and averaged. The different lines in the figure (red and blue dashed line) show the spectra simulations with parameters given in the text. Clearly the value of the exchange interaction |J| evaluated from the simulation based on the spin-Hamiltonian (7) does not provide the sign (either positive or negative), but indeed its lower limiting value is obtained. The solution EPR spectra for the biradical systems 26 (NN), 28 (IN), 33 (NN) and 34 (IN) are analysed in the same frame as for 12 and 13, since they all features |J/aN | >>1. Their spectra are shown in Figure 4.12 (26 and 28) and 4.13 (33 and 34) together with their simulations and estimated parameters. The EPR analysis for the biradical system 41 and 42 (with a methylene-bridge between the pyrazole moieties and the central pyridine ring) are treated separately later in the chapter. The total spectral width and number of observed lines do not appreciable change upon changing the solvents (e.g. CH2Cl2, CHCl3, acetone, THF, MeOH or toluene) for all these biradical. This implies that no specific interaction between the solvent and radical systems occurs. However, as mentioned earlier in Chapter 2 (Table 2.1) the radical stabilities in solution strongly decreased in protic solvents, and especially when THF and MeOH are used, they have been destroyed in a day. This effect is 77 Chapter 4- EPR Analysis ___________________________________________________________________________ shown in Figure 4.14 as one example for the case of the imino nitroxide biradical 28. The biradical systems 12 (NN), 13 (IN), 26 (NN), 28 (IN), in diluted solutions feature clear Curie- like behaviour upon cooling, with linear increase of the double-integrated signal intensities with lowering the temperature. Some broadening in the peak-to-peak line width is observed in all cases upon cooling, without severe loss of the original nine and thirteen lines pattern. The spin-concentration accounts for ~ 2.0 (± 0.1) uncorrelated spin, a indication of thermally activated spin states. The non-symmetric radicals 33 (NN) and 34 (IN) featured alternating line-width upon decreasing the temperature, together with small deviations from the linear Curie behaviour (within the 10% of the EPR limit in the double integration in the overall spectral resolution); for 33 this is particularly pronounced in a narrow temperature range (278 – 258 K), although recovering of the original nine line pattern is observed below 250 K. In 34 on the other hand, a homogeneous distortion of the thirteen lines spectrum recorded in the high temperature range occurs upon cooling, and leads to a broad dominating seven line pattern at 218 K. These spectra are shown in Figure 4.15 featuring both no solvent dependency and full reversibility (e.g. from low to high temperature both nine and thirteen line patterns are recovered even upon repeated cycles). The changes in the line-width are not easy to rationalize, but might be explained by the out-of phase rotation between the radicals attached on pyridine versus those connected to the pyrazole moiety. Although much work has been devoted to the analysis of the line-shape dependency versus temperature for monoradical systems, including nitronylnitroxide [29], very little is still known about line-shape effects on spin-coupled systems. At concentrations higher than 8 ×10-3 M, all the radicals synthesized show very poorly resolved patterns in solutions, due to line-broadening arising from the interactions among unpaired electrons on surrounding centres. Therefore, in order to probe the pure intramolecular interactions between the unpaired spins, magnetically diluted samples with concentrations << 8 ×10-3 M are needed. 78 Chapter 4- EPR Analysis ___________________________________________________________________________ O- O N+ N N N N N N 26 N +N O - O N N N N N N N 28 N N O O Figure 4.12. (A) EPR spectrum of 26 (solid line) recorded in dilute and oxygen free toluene solution at 298 K and its computer simulation (dashed-dotted line), with giso = 2.0065(1), 2׀J/kb ,cm-1 (rms ≤ 0.31), Gaussian line-width = 6.12 MHz, modulation amplitude = 0.84 MHz 0.07 ≤׀ MHz, weight factor 1.00, the other factors are kept the same as the 21.44 = ׀āiso׀ experimentally recorded spectrum. Experimental parameters: 9.40334 GHz, 100 kHz modulation frequency, 4.0 mW power, 0.03 mT modulation amplitude, 21 ms time constant, 42 s sweep time, 104 gain, 4 scan were accumulated and averaged. (B) EPR spectrum of 28 (solid line) recorded in diluted and oxygen free toluene solution at 298 K and its computer simulation (dashed-dotted line), with giso =2.0060(1), 2׀J/kb0.06 ≤ ׀ cm-1 (rms ≤ 0.3795), Lorentzian line-width = 4.78 MHz, modulation amplitude = 0.84 MHz, ׀āiso18.67 = ׀ MHz where MHz, weight factor ≥ 0.73 for 25.72 = ׀aN2׀ MHz and 11.62 = ׀aN1׀ with [2/ ׀aN׀+ ׀aN1׀] = ׀āiso׀ the averaged molecular conformations in strong exchange limit, the other factors are kept the same as the experimentally recorded spectrum. Experimental parameters: 9.40269 GHz, 100 kHz modulation frequency, 5.0 mW power, 0.03 mT modulation amplitude, 21 ms time constant, 42 s sweep time, 5×103 gain, 4 scan were accumulated and averaged. 79 Chapter 4- EPR Analysis ___________________________________________________________________________ 1.0 (A) 0.5 0.0 -0.5 -1.0 331 332 333 334 335 336 337 B / mT 1.0 (B) 0.5 0.0 -0.5 -1.0 331 332 333 334 335 336 337 B / mT Figure 4.13. (A) EPR spectra of 33 (solid line), and (B) 34 (solid line), recorded in dilute (10-4 M) toluene solutions at 293 K. Parameters for: (A) giso =2.0066(1) at 9.4002 GHz, (B) giso =2.0061(1) at 9.40021 GHz, and 100 kHz mod. frequency, 0.3 Gauss modulation amplitude, 21 msec time constant, 42 sec sweep time, 2.6 mW microwave power, 105 gain, 4 scan were accumulated and averaged. The parameters used for the simulations (dashed lines) according to the spin-Hamiltonian (7) are reported in Table 4.2. 80 II dχ II / dB (a.u.) dχ / dB (a.u.) Chapter 4- EPR Analysis ___________________________________________________________________________ Table 4.2: Simulation parameters for the biradical 33 and 34 of the r.t. solution EPR spectra (see Figure 4.13A for 33 and 4.13B for 34 obtained by numerical diagonalization of the spin- Hamiltonian Ĥ (7) Biradical giso 2׀J/kb׀ |aN1|iso |aN2|iso Gaussian Modulation rms line-width amplitude 33 2.0066(1) ≥ 0.069 6.876 × - 1.065 × 2.802 × ≤ 0.806 cm-1 10-4 10-4 cm-1 10-5 cm-1 cm-1 34 2.0061(1) ≥ 0.061 3.896 × 8.530 × 7.393 × 2.802 × ≤ 0.696 cm-1 10-4 10-4 10-5 cm-1 10-5 cm-1 cm-1 cm-1 Figure 4.14. EPR spectra showing the solvent and stability effects of the imino nitroxide biradical 28. Note that upon addition of MeOH (trace B), and after 1 hour of aging, approximately 20% of the biradical (A) is converted into monoradical. Upon addition of THF on (A) (see trace C), although the spectral resolution increased, the double integration of the signal intensities accounts for ~ 50 % of biradical left in solution. After one day no signal could be detected by EPR either in (B) and (C) as indication for the completely loss of the radical system. 81 Chapter 4- EPR Analysis ___________________________________________________________________________ Figure 4.15. The alternating line-width effect in the EPR solution (toluene) spectra for the radicals 33 (A) and 34 (B) observed upon cooling. The numbers on the right of each spectrum indicates the spin-concentration of the biradical system at that temperature with ± 1 K in temperature error (T/K is written on the left of each series of spectra). The temperature- dependent spin concentrations have been obtained by comparison with the related monoradical standards 8 and 9. 4.4. The EPR spectra of the σ - conjugated biradicals in solution. The introduction of two σ bonds within the coupling unit clearly induced loss of the π- conjugation between the radical carriers. σ σ σ σ N N N N N N O N N O O N N O N N N N N + 41 + N N 42 N O- - O Radicals 41 and 42 showed a strong temperature dependence of their EPR envelopes in diluted (10-4 M in toluene) solutions (Figure 4.16B and 4.17B respectively), although the observed dominating nine lines pattern for 41 and thirteen lines pattern for 42, it suggested that still the strong exchange limit of |J/aN | >>1 holds in both cases. As observed in the other biradical systems previously discussed, also in these cases the EPR features were 82 Chapter 4- EPR Analysis ___________________________________________________________________________ independent of the solvent used. The alternating line-width effect can be rationalized upon assuming the occurrence of different rotamers in solution (due to the σ bonds) [30], with a much easier interconversion from more planar to twisted radical-core conformations. No J modulation is involved in such process since the peak-positions and total spectral widths (apart from large broadening upon lowering the temperature) did not change in the temperature range analyzed (see Figure 4.16 B). 2 Observed Simulated (B) 1 332 333 334 335 336 337 293 K 0 282 K -1 272 K (A) -2 260 K 3320 3340 3360 B / Gauss 246 K Double Integration EPR signal 0.7 Curie-Fit 236 K 211 K 226 K 226 K 236 K 0.6 246 K 211 K 260 K 332 333 334 335 336 337 0.5 272 K (C) Magnetic field (mT) 282 K293 K 0.0036 0.0040 0.0044 0.0048 1/ T (K) Figure 4.16: The alternating line-width effect in solution (toluene, B) observed upon cooling for the radical nitronyl nitroxide biradical 41. Figure (C) shows the theoretical line (dashed line) for the Curie-like behaviour while the symbol (●) represents the double integrated EPR envelope at the selected temperature. The spectra in (A) (solid line) report the observed biradical feature recorded at 293 K and its related computer simulation (dashed line). The spectra in (A) and (B) were base-line and frequency corrected. Experimental and spin- Hamiltonian parameters: 9.400086 GHz for (A) and (B), then 0.03 mT modulation amplitude, 2.0 mW power. giso =2.0065(1) and the apparent aN/2 = 0.374(2) mT. The double integrated signal intensities increased either for 41 and 42 upon decreasing the temperature, with 2.0 ± 0.25 spins for 41 and 2.0 ± 0.15 for 42 in the full temperature range analyzed, showing however larger deviations from the theoretical linear Curie-plot especially for 41 (Figure 4.16C). The observed EPR spectrum of 41 recorded at 83 II DI (a.u) dχ / dB (a.u.) Chapter 4- EPR Analysis ___________________________________________________________________________ room temperature could be simulated only upon assuming a large percentage (~ 35%) of uncoupled biradical (with ׀J10-4 × 7 >> ׀ cm-1) superimposed on the biradical component (~65%) under strong exchange limit (with ׀J/a 40 ≤ ׀, i.e. > 2J > 0.056 cm-1 with aN = 20.94 MHz) (Figure 4.16A, dashed line). In the similar way, the simulation of the solution EPR spectrum of 42 (Figure 4.17A, black dashed-line) gave ~ 25% of uncoupled biradical (with ׀J׀ << 7 × 10-4 cm-1) superimposed on the biradical component (~75%) under strong exchange limit (with 2J ≥ 0.055 cm-1 with aN1 = 25.20 MHz and aN2 = 12.32 MHz). 8000 Observed Simulated 4000 0 -4000 (A) -8000 3320 3340 3360 3380 B / Gauss Double integrated EPR signal intensity Curie-Fit 4.0 Temperature (K) (C) 8000 293 283 3.6 273 4000 263 3.2 253 0 240 220 2.8 205 -4000 (B) 2.4 3320 3340 3360 3380 B / Gauss 2.0 0.0032 0.0036 0.0040 0.0044 0.0048 1 / T (K-1) Figure 4.17. The alternating line-width effect in the solution EPR spectra observed upon cooling for the imino nitroxide biradical 42 (toluene, 10-4 M). The figure (C) shows the theoretical line for the Curie-like behaviour (dashed line) while the circles (●) correspond to the measured double integrated EPR signals. In (A) is reported the thirteen line pattern of 42 recorded at 293 K together with its computer simulation (dashed line). The spectra in (A) and (B) were base-line and frequency corrected. Experimental and spin-Hamiltonian parameters: 9.402404 GHz, 0.03 mT modulation amplitude, 2.0 mW power. giso =2.0060(1), with the apparent aN1/2 = 0.450(2) mT, aN2/2 = 0.220(2) mT. 84 d IIχ / dB (a.u.) d IIχ / dB (a.u.) DI (a.u.) Chapter 4- EPR Analysis ___________________________________________________________________________ 4.5. The EPR of mono and biradical systems in frozen solutions. In frozen solution or in a powder the relevant spin-Hamiltonian describing the EPR transitions for monoradical and biradical systems are reported in equations (25) and (26) respectively: Ĥ = βe g ŜiBo + S.A.I (25) Ĥ = βe g Ŝa,bB0 – 2JŜaŜb+ S.D.S + S.A.I (26) Where g and A are now tensors with different space components (anisotropy) such as g = (gxx +gyy +gzz /3), and A = (Axx +Ayy +Azz /3) [22b,c]. The origin of the anisotropy in A comes from the interaction between the magnetic moment of the electron and the nuclei, therefore it depends on r-3 (µ 2eµn [1-3 cos θ] /r3), with r being their distance, Ф the angle between the µ vectors, and µeµn electron and nuclear magnetic moments. In the general case for the dipole expression, when two µA,B vectors are not aligned mutually parallel to each other, it assumes the forms (µAµB [cos φ -3 cosθ1cosθ 32] /r ) with r being the distance between the centres O1 and O2. This is shown in the small figure on the right. Remarkably enough it is independent on the sign of r. In S ≥ 1 systems assuming that the D- and g-tensors have the same principal axis, the equation (26) reported above can be rewritten as (27): Ĥ = gβeŜa,bB0 – 2JŜaŜb+ D{S2z – S (S+1)/3} + E(S2x -S2y) + S.A.I (27) D and E are related to the principal values of the D-tensor through D = 3Dzz/2 and E = |Dxx- Dyy|/2 and they represent the fine structure parameters (axial and rhombic) also called zero- field–splitting parameters. They originate from the electron-spin-electron-spin dipole interaction [20], and this causes the three fold degeneracy of the triplet state to be removed even in zero magnetic fields. The D, J and the Zeeman contributions expressed in the spin-Hamiltonian (27) are depicted in the Figure 4.18. Their in-depth theoretical descriptions are treated in several monographs [20, 22]. 85 Chapter 4- EPR Analysis ___________________________________________________________________________ Figure 4.18. Energy diagram for the eigenstates of the spin-Hamiltonian (27) for two S = ½ interacting spins. The term J is assumed as positive (ferromagnetic interaction); when it is negative (antiferromagnetic) the singlet lies below the triplet in energy. The zfs parameter D is also assumed as positive; when it is negative the transition │1, 0> lies higher in energy. The first-order perturbation theory is applied (e.g the Zeeman term is larger than the zfs terms). The powder spectrum (absorption) and its first derivative have been simulated with non zero line-width for a pure axial system (E=0). The terms Bh and Bl represent the high and low field components of the two allowed absorption lines. The forbidden half-field transition, ∆Ms = 2, is also shown. 86 Chapter 4- EPR Analysis ___________________________________________________________________________ 4.6.1 The observed EPR spectra of the monoradical systems in frozen solutions. The monoradical spectra for the nitronyl- (NN) 8 (gav = 2.0066) and imino nitroxide (IN) 9 (gav = 2.0061) recorded at low power in diluted frozen solutions [10-4 M] are reported in Figure 4.19A and 4.19B respectively. The other two monoradical systems based on the pyridylpyrazolyl moieties (37 and 38) show the same EPR envelopes with respect to 8 and 9, thus are not reported in the figure. However they feature slightly different g values with gav = 2.0065 for 37 and gav = 2.0060 for 38. Note that the rhombic pattern (i.e. gxx ≠ gyy ≠ gzz) of the monoradical systems cannot be resolved at X-band frequency [31] where they appear axial (with gxx = gyy ≠ gzz). All the monoradicals synthesized, followed nicely the Curie-Law, with linear increase of the double integrated signal intensities upon decreasing the temperature, and therefore they can be used as spin-standards for the biradical systems at concentration lower than 10-2 M. Figure 4.19. The EPR powder spectra of the monoradical systems 8 (A, nitronyl nitroxide) and 9 (B, imino nitroxide) in dilute (8 × 10-5 M) toluene solutions recorded at low microwave powers. In (C) for 8 and (D) for 9 are shown their correspondent signal- saturation at 110 K in the range of powers from 0.05 mW up to 32 mW, in which are visible the evolutions of additional lines in the low-field region. Those absorptions are indicated with thin-arrows and asterisks (*) in the figures (C) and (D). The monoradicals 37 and 38 follow the very similar trend. The spectra were both base-line and centre-field corrected. Background signals were subtracted. 87 Chapter 4- EPR Analysis ___________________________________________________________________________ 4.6.2 . The observed EPR spectra of the biradical systems in frozen solutions. The EPR spectra in frozen solution (toluene) for the biradicals 12 (NN), 13(IN), 26 (NN), 28 (IN), 33 (NN), and 34 (IN) are collected in Figure 4.20, to allow quick comparison, while the spectra for 41 (NN) and 42 (IN) are shown again separately in Figure 4.21. It is clear from the spectra, although part of the anisotropic patterns are retained, that most of the hyperfine couplings are not resolved. This effect arises from dipolar interactions characteristic of randomly oriented triplet species. Increasing the intramolecular radical distances led to a decrease in the magnitude of the zfs component, since the │D│ value in axial systems is related with the averaged radical distances according to equation (28)(the point-dipole approximation) │D│ (MHz) = 77924 (gobs / ge) /r3 (28) With r (in Angstrom, Å) considered as their averaged through-space separation. In Table 4.3 are collected for comparison the spin-Hamiltonian parameters for 12 (NN), 13(IN), 26 (NN), 28 (IN), 33 (NN), and 34 (IN), together with their estimated r. In the biradical (NN) 33 the spectrum appeared quite complex (Figure 4.20 C) in which eleven resolved lines are observed (∆Bpp = 0.75 mT for the outermost pairs) with an unusual resolution of nitrogen hyperfine coupling. All the biradical systems even at ~ 120 K showed the forbidden (∆ms ± 2) half-field transition at g ~ 4.01 - 4.02, therefore supporting their biradical nature. For two biradical systems, 26 (NN), and 28 (IN), and perhaps also in 35 (IN) (vide infra), even after repeated purification steps through column chromatography or preparative TLC, a small amount of monoradical impurity (~ 3% in 26 and ≤ 6% in 28) can be observed in the ∆ms ± 1 transition in the recorded powder spectra. The monoradical-signal contributions are indicated with asterisks in Figure 4.20. Those impurities can be easily discriminated in 26 and 28, because their saturation trends are much different from the major biradical absorption lines. This is shown in the Figure 4.21 for 28, as one example in which the monoradical signal contribution is highlighted by enclosing it into a box. A double-quantum transition has to be ruled out because, it occurs, it requires far higher power [22d, 22e]. In 34 indeed, the overall EPR signal (including those marked with the asterisk in Figure 4.19 D) broadens in the same way when increasing microwave power has been applied. Upon subtraction of a small percentage (~5%) of monoradical contribution from the major biradical absorption line, the overall EPR envelope shows almost a full isotropic line, making very hard the estimation of │2D│. Although one might suggest that those extra signals are originating from the presence of different conformers in the powder spectra, with strongly twisted radical sites with respect to the coupling unit core, and not from monoradical impurities, this appears clearly in contradiction 88 Chapter 4- EPR Analysis ___________________________________________________________________________ with the fact that for 26 and 28 the solution spectra did not show strong line-width dependencies, upon decreasing the temperature. However, such hypothesis cannot be excluded for 34 in which such effect was observed. Figure 4.20.The EPR powder spectra of the biradical systems (A) 12, (B) 13, (C) 33, (D) 34, (E) 26 and (F) 28. The relevant experimental parameters together with the estimated zfs are reported in Table 5.3. Note that the spectra were recorded at different powers, since the biradicals featured different saturation trends. The asterisks ( ) indicate a monoradical impurity and the arrows the estimated 2׀D׀ range. 89 Chapter 4- EPR Analysis ___________________________________________________________________________ Table 4.3: The estimated zfs parameters for the biradical systems. ∆ms = 1 Microwave T (K) Frequency gobs │2D│ │D/hc│ r transition Power [C] GHz Gauss 10-3 Å mWatt Solvent cm-1 110 12 (NN) 0.8 4 ×10-4 M 9.3970 2.0066 74 3.46 9.1 toluene 110 13 (IN) 0.8 6 ×10-4 M 9.4007 2.0061 79 3.69 8.9 toluene 120 33 (NN) 0.6 1.0 ×10-3 9.3989 2.0066 81 3.79 8.8 M toluene 120 34 (IN) 0.6 1.1 ×10-4 9.4027 2.0061 ~ 80 3.73 ~ 8.8 M toluene 120 26 (NN) 0.2 1.0 ×10-4 9.4168 2.0065 ~ 96 4.50 ~ 8.3 M toluene 120 28 (IN) 0.2 4 ×10-4 M 9.4144 2.0060 102 4.78 8.2 toluene Figure 4.21. The EPR powder spectra at 122 K for the biradical system 26 recorded at different microwave power. The red box encloses the signal coming from the monoradical impurity. For the biradical 41 (NN) and 42 (IN), conformational interconversion towards strongly twisted radical sites with respect to the coupling unit core are more easily accessible. The zfs envelope for the pure biradical character in the ∆ms ± 1 transition for 41 was obtained by subtracting a weighted percentage of uncoupled biradical (Figure 4.22 C). This was made possible by comparison with the saturation trend of the monoradical system 37. The estimated zfs accounts for │2D│ ~ 62 Gauss (i.e. │D/hc│~ 2.90 × 10-3 cm-1) corresponding to the 90 Chapter 4- EPR Analysis ___________________________________________________________________________ averaged distances between the radical fragments of 9.6 Å. This is the largest distance in the NN biradical series. Figure 4.22: EPR spectra of the ∆Ms = 1 transition for the nitronyl nitroxide radical 41 (A) and the imino nitroxide biradical 42 (C) with anelling procedures. The asterisks in (A) and (C) are fingerprints for more strongly twisted conformations vs coupling unit core for the radical moieties. Experimental parameters for (A): 9.41591 GHz, 100 kHz mod. frequency, 0.6 Gauss mod. amplitude, 2.0 × 104 gain, 84 sec sweep time, 41 msec t. constant, 122 K, 0.2 mWatt power, [concentration] = 4.5 × 10-4 M in toluene. Experimental parameters for (C): 9.41699 GHz, 100 kHz mod. frequency, 0.6 Gauss mod. amplitude, 4.0 × 104 gain, 84 sec sweep time, 82 msec t. constant, 0.16 mWatt power, 122 K, [C] = 4.9 × 10-4 M in toluene. Note that the filling factors were different for (A) and (C). The spectra were base-line corrected. Figure (B) shows in the lower traces (subtracted) the biradical EPR envelope for the planar conformer. (D) and (E) show the signal-saturation trends for 41 (D) and 42 (E) at 122 K. The insets in (A) and (C) represent the observed EPR half-field transition for 41 and 42 respectively. Note on anelling: fast cooling in N2 bath at 77 K→ raise temperature at 170 K in the EPR cavity (keep for 20 min) → lower the temperature slowly to 122 K. 91 Chapter 4- EPR Analysis ___________________________________________________________________________ However, in 42 since the overall absorption line broaden inhomogeneously at all available power (Figure 4.21 E), such a comparison was not possible. Therefore a very crude estimation of │2D│ ~ 68 Gauss (│D/hc│~ 3.18 × 10-3 cm-1, r = 9.4 Å) has been obtained. A further comment might be useful at this point; one might suggest an indirect correlation between size of J and the zero-field-splitting │D│ by comparing equation (28) and (13). As much as r is decreasing, J is increasing, hence J and │D│ should feature the same trend. From the values reported in Table 4.3, it seems that the nitronyl nitroxide biradical systems possess smaller or comparable │D│ once related with the imino nitroxide biradicals. Therefore J should be smaller or comparable (within the same coupling unit series). However, even taking under consideration errors in the estimation of the zfs parameters for the imino and nitronyl nitroxide biradicals, we shall see that larger │D│ does not imply larger J, and therefore correlation between │D│ and J based on the through-space interaction model would not be appropriate for discussing the strength of through-bond interaction between radical centers. As pointed out earlier by Ullman [33], the cis conformation of imino biradicals (e.g. in meta- phenylene derivatives) although seems to represent not the dominating conformers in frozen solutions, show larger zfs compared to the nitronyl nitroxides, even though the J value are smaller. All the biradical systems presented in this thesis, exhibit Curie-like behaviour also in their powder spectra, and the spin-concentration accounted for 2.0 ± 0.2 uncorrelated spins. These results indicate that thermally activated spin states still survive even at around 100 K, and therefore for all the biradical system the ∆EST < 100 K (i.e. < 69.5 cm-1). 4.7. The EPR saturation behaviour of mono and biradical systems in frozen solutions. Clear differences between mono and biradical systems are observed upon following the saturation trends of their ∆ms = 1 transitions in frozen solution. These results are collected together, for most of the spin systems synthesised, in Figure 4.23. The parameters obtained from fitting the variation of the double integration of the signal intensities (DI) versus applied microwave power (P) are reported in Table 4.4. The data of the EPR signal saturation were fitted using the empirical expression provided by Portis [36] and Castner [37], written in the following form: DI = k × √P / [(1+P/P b/21/2)] (a) where b represents the relaxation factor (b = 1 for inhomogeneous line broadening and b = 3 for homogeneous line broadening) for a first derivative spectrum as it is usually acquired within common EPR experiment, P1/2 the power at which half of the signal is being saturated 92 Chapter 4- EPR Analysis ___________________________________________________________________________ and k a normalization factor associated with the instrument. The logarithmic form of equation (a) is more often used to present the experimental data: Log10 (DI / √ P) = Log10 k – (b/2) Log10 [1 + (P / P1/2)] (b) If [Log10 (DI / √P)] is plotted versus Log10 P, two linear regions are obtained that intersect at P = P1/2. The value of b depends on which mechanism is dominating within the relaxation process of the quanta being adsorbed, i.e. if the spin-lattice relaxation (TL) represents the dominant factor, the power-dependent line broadening is inhomogeneous (Gaussian line) and b assumes the minimum value of 1. When Ts is dominant, the line broadening is homogeneous (Lorentzian line) and b assumes the maximum value of 3. The relaxation factor b is often allowed to fluctuate in the fitting in order to account for intermediate cases (1≤ b ≤ 3), when the line-shape observed is a mixture of Lorentzian and Gaussian line. This effect particularly holds for frozen solutions. In order to apply equation (a), the following experimental conditions must be satisfied: the samples should be in the region of the cavity with the maximum microwave field, H1, thus the filling factor has to be optimized, while the sample temperature, the Zeeman modulation amplitude, the frequency and possibly the gain must also be constant. Equation (a), on the other hand, is not strictly applicable when dipolar couplings are present in the system, because: P1/2 = [α / (TS × TL)] (c) With α = 1/2 × (V/Q γ2) where V represents the cavity volume, Q the cavity quality factor (Q = H 21 V/ 2P), P the power dissipated in the cavity and γ the gyromagnetic ratio. Equation (c) assumes that all spins at resonance saturate equivalently and hence they show the same product for (TL × TS). When dipolar interactions are present, the product (TL × TS) is no longer constant, and b is found sometimes smaller than 1, nevertheless a much more crude estimation of P1/2 can be still obtained. Therefore, from the trend in the saturation data and the corresponding theoretical fitting the following comments are drawn: 1) The monoradical systems start to saturate at very low power (<< 1 mWatt), already at 122 K, with the clear tendency for the imino nitroxide radical to saturate at higher power with respect to the nitronyl nitroxide. The line-shapes are mostly described by a Gaussian-line (b ~ 1) although severe distortions occur at high power suggesting that the gzz component saturates faster than the gyy ~ gxx components. 93 Chapter 4- EPR Analysis ___________________________________________________________________________ 2) The same trend is observed in the biradical systems with the exception of 28 (if we consider the estimated P1/2 value) despite the additional contribution coming from dipolar interactions. The shape factor (b) is generally smaller in the case of the nitronyl- (NN) with respect to the imino nitroxide (IN) radicals with again the exception of 28 in which such trend is reversed. This should be related with the “magnitude” of the dipolar interaction (smaller b → larger D) in disagreement with the zfs parameters of the radical systems discussed previously (D seems usually smaller in NN with respect to IN). 3) No cross correlation among magnitude of P1/2 or the shape factor (b) in similar biradical systems (e.g. all the nitronyl- and all the imino nitroxides) and corresponding strength of J can be made. For example in Figure 4.23 B the difference in saturation between radical 26 (NN) and 41 (NN) is negligible, and it is very small as compared with 28 (IN). However, 26 shows clear triplet ground state, while 41 and 28 feature almost singlet-triplet degeneracy (vide infra). 4) The use of the double integration (DI) of the ∆Ms = 1 transition in these biradical systems down to cryogenic temperature (in order to estimate J) would be precluded, since already at less than 20 dB at 122 K ~10-15% signal saturation is observed. 5) The empirical correlation used to evaluate │D│ is based on the intensity ratios between the ∆Ms=2/∆Ms=1 transition (1:[D/B 20] ). This would be hardly applicable in these systems in whatever temperature range. Over 10 mWatt of microwave power, very much of the ∆Ms=1 signal at 122 K is saturated, for the majority of these biradicals, while at such power hardly the ∆Ms=2 can be observed; therefore no direct comparison can be made and no estimation of │D│ can be obtained. Table 4.4: Saturation data at 122 K for the dilute (10-4 M) radical systems with simulation parameters according to Log10 (DI / √ P) = Log10 k – (b/2) Log10 [1 + (P / P1/2)]. The k term represents the instrument constant. Upon normalization its value is ~1.000 and thus has not been included in the Table. Note that the estimated P1/2 values in the biradicals are certainly approximated. Radical Type Shape factor P 2 1/2 R (b) mWatt > 41 NN biradical 0.467 ± 0.009 0.82 ± 0.06 0.999 42 IN biradical 1.078 ± 0.029 1.12 ± 0.11 0.999 8 NN monoradical 1.048 ± 0.013 0.71 ± 0.04 0.999 9 IN monoradical 1.017 ± 0.026 1.06 ± 0.10 0.999 33 NN biradical 0.802 ± 0.014 1.17 ± 0.08 0.999 34 IN biradical 0.918 ± 0.016 1.25 ± 0.08 0.999 12 NN biradical 0.287 ± 0.019 0.61 ± 0.19 0.991 13 IN biradical 1.217 ± 0.031 1.41 ± 0.13 0.999 26 NN biradical 0.524 ± 0.009 1.19 ± 0.07 0.999 28 IN biradical 0.450 ± 0.001 1.09 ± 0.10 0.996 94 Chapter 4- EPR Analysis ___________________________________________________________________________ Figure 4.23: (A) and (B) Power saturation for the mono and biradical systems performed at 122 K. The radical concentrations were 10-4 M. The different symbols represent the double integration (DI) of the signal intensities and the solid lines the theoretical fit. Before performing the double integration of the signals, each spectrum was base-line corrected, and adjusted for frequency shift (diode current 200 µA, lock offset to zero). 95 Chapter 4- EPR Analysis ___________________________________________________________________________ 4.8. Determination of the electronic ground state in the magnetically diluted biradical systems. 4.8.1. The terpyridine based biradicals. In order to analyze the trends in the magnetic properties of the isolated molecules, the temperature dependence of the ∆Ms = 2 transitions have been followed down to cryogenic temperature. The microwave powers applied in the measurements were kept in such a way that the signals were proportional to the root of power, in order to avoid saturation effects. In the case of the terpyridine based biradicals 12 (NN) and 13 (IN) the double integration of the ∆Ms=2 signal increases upon decreasing the temperature for both (Curie Plot) as shown in Figure 4.24A and 4.24B. Moreover, the fact that also the product DI·T increases with decreasing temperature indicates that the ground states are certainly the triplets (S=1), and the singlets (S=0) have to be associated with thermally accessible excited states. Fitting the DI·T data according to the Bleaney-Bowers model [34a] for two interacting spins S=½ system where DI is expressed as DI = χ 2 2EPR ∝ {(2Ng β /3kbT)×[1/3+exp(-2J/kbT)]} (29) it gave a separation for 12 of |∆E | = 2J/k of 15.4 ± 2.0 cm-1ST b between the magnetic ground (S=1) and the excited states (S=0) (Figure 4.24C), while a smaller singlet-triplet gap has been observed for 13 with |∆EST| = 2J/kb of 8.7 ± 1.0 cm-1 (Figure 4.24D). In order to predict theoretically the ∆EST and compare it with the experimental findings, the ground spin-state for 12 and 13 has been obtained towards quantum chemical calculations, carried out by P.D. Dr. Martin Baumgarten and Dr. Anela Ivanova (University of Sofia, Bulgaria). The geometry of the biradicals was optimised both with UHF/AM1 and ROHF/AM1, the two methods yielding virtually identical lowest energy structures. Furthermore, no essential differences were witnessed between the geometry of the unrestricted singlet and triplet molecules. Therefore, the ROHF/AM1 structures of 12 and 13 were used further on for calculation of the spin-state energies. The imino nitroxide and the nitronyl nitroxide rings are essentially planar (intraring torsion angles ≤ 10o), with the N–O bonds about 15o out of the ring plane. The calculated bond lengths fall within the range of those measured for radicals of this class. The N–O bonds are slightly shorter than the most commonly encountered experimental value of 1.28 Å, namely RN–O = 1.209 and 1.204 for 12 and 13, respectively. This result is similar to previous AM1 calculations. An interesting structural feature of the molecules are the torsional angles Θ 1 and Θ 2 and in the case of 12 those results might be compared with its X- ray structure (see Figure 4.25 for both spin densities distributions and optimized structure). Since angles Θ 1 and Θ 2 torsional angles reflect the possibility for free rotation around the two 96 Chapter 4- EPR Analysis ___________________________________________________________________________ single bonds, they can be used to measure the effective π-conjugation through the spacer. Therefore, a conformational search, i.e. systematic variation of Θ 1 and Θ 2 with calculation of the corresponding energy, was performed at the UHF/AM1 level. 92% of the conformations of 12 and 100 % of those of 13 differ in energy by less than 5 kcal/mol. This result indicates relatively high flexibility of the two molecules, the rotation barrier being lower in 13 than in 12. The calculated singlet-triplet splitting and heats of formation of 12 and 13 are collected in Table 4.5. The simulations predict triplet ground states for both biradicals, as experimentally found, evidenced by the positive values of ∆EST. The triplet state is more stable with respect to the singlet in 12 than in 13, which is probably due to more effective π-conjugation resulting from the less flexible structure of 12. This is also in agreement with the experimental findings. It is apparent that the exchange coupling is extremely sensitive to rotation around the single bonds. Increase of Θ1 to ~90o results in practically degenerate singlet and triplet states. This is an indication for hampered intramolecular coupling between the radical sites due to prevented spin transfer into the spacer. The calculated spin densities are presented in Figure 4.23. Both biradicals feature alternating signs of the spin densities at neighbouring atoms throughout the whole molecule. Thus, the main prerequisite for effective ferromagnetic exchange coupling is fulfilled. Increase of Θ 1 to ~80o (conformation a in Table 4.5) leads to zero value of the spin density at the carbon sites connecting the central pyridine ring to the two outer ones. A direct consequence is the inability for spin transfer and hence a breakdown in spin polarization. Although conformation a has slightly lower heat of formation in the gas phase, the small rotation barrier allows stabilization of more planarized structures like b. However, the optimised structure appears quite different from that observed in the crystal for 12 with an over-estimation of the singlet-triplet gap of ~ 87 cm-1 (i.e. ~ 126 K), whereas in the case of 13 the estimated ∆EST gap of ~ 9.4 cm-1 (~ 13.5 K) is nearly identical to that one experimentally found. The intramolecular radical distances calculated using the approximation introduced by Mukai and co-workers [34b,c] defined as: D = ¾ g2β2∑ [r2 2ij – 3m ij]ρ 5iρj / r ij (30) Where rij is the distance between the atoms i and j, mij is the distance vector along the axis, which give rise to the largest dipole-dipole interaction, and ρi and ρj are the spin-densities on atom i and j. The relation (30) provides a better estimation of the averaged radical separation distances of r = 0.83 nm in case of 13 and r = 0.88 nm in case of 12. Hence, larger D for 13 should be observed as compared with 12. This is also in agreement with the experimental findings. 97 Chapter 4- EPR Analysis ___________________________________________________________________________ Figure 4.24: Curie-Plot for the terpyridine biradical systems: (A) nitronyl nitroxide 12 and (B) imino nitroxide 13 recorded in diluted (10-3 M) toluene solutions. DI represents the double integrated signal intensity of the half-field transition. The small inset in (A) and (B) show the observed half-field transition recorded 4.0 and 7.3 K respectively. (C) and (D) show the DI.T vs T plot; the best fitting according to the Bleaney-Bowers equation is shown as bold line with parameters given in the text. IN NN 13 0.000N0.005 12 -0.001 -0.005 0.002 N 0.001 0.007 0.012 N 0.405 -0.007 -0.039 O -0.014 -0.001 -0.005Θ1 -0.056 O0.294 0.022 0.058 0.382 N 0.003 0.011 N 0.288 - 0.014 Θ2 -0.013 -0.060 -0.126 N 0.286N 0.141 O 0.292 Figure 4.25. ROHF/AM1/CIS(20,20) calculated spin densities of conformation b of biradicals 12 (right) and 13 (left). 98 Chapter 4- EPR Analysis ___________________________________________________________________________ Table 4.5: ROHF/AM1/CAS [8,8] calculated singlet-triplet gap (∆EST) and heat of formation (Hf) of biradicals 12 and 13 for different torsions Θ1 and Θ2. Radical o oConformation Θ1, Θ2, ∆EST, kcal/mol Hf, kcal/mol a 86 0 0.0006 192.898 12 b 49 27 0.2499 195.846 13 a 84 7 0.0000 188.968 b 47 20 0.0269 192.376 4.8.2. The bispyrazolylpyridine based biradicals. In the case of 26 (NN) the double integration (DI) of the ∆Ms=2 signal increases upon decreasing the temperature (Figure 4.27A, ●, and 4.27C, ○) but also the quantity DI·T (Figure 4.27C, ●). Such findings indicate that in 26 the ground state is the triplet (S=1), and the singlet (S=0) represents a thermally accessible excited state. Figure 4.27: Curie-Plot for the bispyrazolypyridine biradical systems: (A) nitronyl nitroxide 26 and (B) imino nitroxide 28 recorded in dilute (10-3 M) toluene solutions. DI represents the double integrated signal intensity of the half-field transition. The small inset in (A) and (B) show the observed half-field transition recorded at 4.1 and 4.2 K respectively. (C) and (D) show the DI.T vs T plot; the best fitting according to the Bleaney-Bowers equation are described in (C) and (D) as bold line (for S=1) and dashed line (for S = 0) with parameters given in the text. In (D) constraints in the fitting have been used according to the equations (32) and (33). 99 Chapter 4- EPR Analysis ___________________________________________________________________________ The fitting of the data according to the Bleany-Bowers model for two interacting spins gaves a separation ∆EST = 2J/kb of 11.8 ± 4.8 cm-1 between the magnetic ground (S=1) and the excited states (S=0). However, in the imino nitroxide biradical system, 28, while the ∆Ms=2 signal clearly increased upon decreasing the temperature (Figure 4.27B,▲, and 4.27D, ○), the product DI·T appeared nearly constant (Figure 4.27D, ●), within the experimental error. Therefore, no thermal population or depopulation of the spin state of the molecules within the examined temperature range could be envisaged. In such case, either the triplet state (S=1) is the lowest-energy state separated from the singlet (S=0) by a substantial gap relative to the thermal energy (|2J| >> KbT), or the energy difference between triplet (S=1) and singlet (S=0) is extremely small leading to near degeneracy of those levels. Any change in temperature does not shift the thermal equilibrium between the two states, and they remain statistically populated according to the Boltzmann distribution (exp–[∆E/KbT], e.g. 75% of the molecules occupies the triplet state and 25% the singlet). Since from the solution and frozen state EPR studies discussed for 28, it was clear that the energy gap between singlet and triplet cannot be large, such separation has to be very small, ranging far below the temperature available with a conventional experimental setting (4 K). Fitting the DI.T data according to the Bleany-Bowers model for two interacting spin systems, gave an upper limit for the singlet-triplet separation ∆EST = 2J/kb of 0.7 cm-1 (~ 1 K) assuming triplet ground state (J>0). It is worth to emphasize that in this case in the fitting process, some assumptions have been made by using the limiting expressions (32) and (33) if J > 0, then {[lim T → 0 K (D.I0)] / D.IT}→ 0.333 (32) if J < 0, then {[lim T → 0 K (D.I0) ] / D.IT} → 0 (33) Where DI0 represents the double integrated ∆Ms = 2 signal intensity extrapolated at T = 0 K and DIT the double integration at the temperature in which the equilibrium between singlet and triplet state still satisfies the Boltzmann distribution (see equations 23 and 24). If the singlet ground state (S=0, J<0) is considered, then the separation ∆EST is further reduced by a factor of ten (2J/kb of ~ -0.07 cm-1), in agreement with the lower limiting value obtained from the simulation of the solution EPR line. Because the complete spin reversal of the ground state multiplicity (from nitronyl- to imino nitroxide) seems improbable, although cannot be fully excluded, it is possible that also in 28 the ground state might be the triplet. 100 Chapter 4- EPR Analysis ___________________________________________________________________________ 4.8.3. The pyrazolylbipyridine based biradicals. Similarly as found in the previous cases for the terpyridine and bispyrazolylpyridine biradical systems, in the nitronyl nitroxide biradical 33 the ground state is the triplet with ∆EST = 2J/kb of 13.3 ± 4.9 cm-1 (19.0 K ± 7.0) while in the imino nitroxide 34, the smaller range of - 0.07 cm-1 ≤ ∆EST ≤0.7 cm-1 has been obtained. These results are shown in Figure 4.28 with the related Curie and Bleaney-Bowers fitting. We might conclude that the nitronyl nitroxide biradicals in the π-conjugated coupling unit series posses stronger exchange interaction with respect to the imino nitroxide systems. Figure 4.27: Curie-Plot for the bispyridylpyrazolyl biradical systems: (A) nitronyl nitroxide 34 and (B) imino nitroxide 35 recorded in diluted (10-3 M) toluene solutions. DI represents the double integrated signal intensity of the half-field transition. The small insets in (A) and (B) illustrate the observed half-field transition recorded at 4.2 and 4.6 K respectively. (C) and (D) show the DI.T vs T plot; the best fitting according to the Bleaney-Bowers equation are depicted in (C) as bold line (S=1), and in (D) bold line for S=1 and dashed line for S=0. The constraint equations used for the fitting in (D) correspond to the equations (32) and (33) with parameters given in the text. 101 Chapter 4- EPR Analysis ___________________________________________________________________________ 4.8.4. The σ - conjugated biradicals. The introduction of two σ bonds between the coupling unit and radical moieties in the NN biradical 41 (dipyrazolyldimethylpyridine) severely destabilizes the triplet state, with an upper limit for the singlet-triplet gap ∆EST = 2J/kb of 2.8 ± 0.8 cm-1 (assuming triplet ground state). This is shown in the fitting of the DI.T data in Figure 4.29. Again, if the singlet ground state is considered, such gap has the size estimated from the solution EPR studies of ~ – 0.06 cm-1. Judging from the data trend, still the triplet seems the ground state in 41. However, in the imino nitroxide biradical 42, most likely the singlet or its close degeneracy with the triplet is present. Figure 4.29: (A) Curie and (B) Bleaney-Bowers Plot for the bispyrazolyldimethyl-pyridine (NN) biradical 41 (10-3 M in toluene). DI represents the double integrated signal intensity of the half- field transition. The small inset in (A) shows the half-field transition recorded at 4.1 K and the solid lines the best fitting according to the Bleaney-Bowers equation with parameters given in the text. 4.9. Conclusion. The detailed studies of nitronyl and imino nitroxide biradicals are scarce, fragmented and often contradictory to each other, as compared with the much more studied carbene, nitrene and nitroxide systems. Furthermore, in the plethora of the biradical derivatives found in literature, the unambiguous characterisation of the magnetic ground-states for the magnetically isolated molecule are, surprisingly, rarely treated, although such knowledge should constitute the essential prerequisite prior their use in extended magnetic structures. In this perspective, the literature available in the field makes very hard to compare the results obtained by different groups, since, very often, the experimental details and the procedures used for assessing the magnetic properties are not exhaustively reported. In this chapter much effort was thus devoted in presenting a comprehensive and detailed EPR analyses for 102 Chapter 4- EPR Analysis ___________________________________________________________________________ both the nitronyl and imino nitroxide mono- and biradical entities. We provided the necessary step-by-step methodologies for assessing the range of conditions in which the properties of the molecules were unambiguously analysed (radical purities, concentration range, presence or absence of solvent effects, radical stabilities, saturation trends), and then we clearly defined the ground spin state and the relative size of the exchange interactions for the biradical cases. In order to put this work into perspective, and to compare the results obtained with those available in literature, the following points need to be recollected. 1) In general, as discussed in Chapter 1, two unpaired electrons linked by m-phenylene bridge should afford triplet ground state (S=1). 2) However, heteroatomic substitution in the coupling unit core, like in pyridines, the presence of non alternant and competitive pathways for the spin polarization, like in the pyrazoles, the molecular conformations and the presence of substituents on coupling unit core, would affect strongly the ground spin state and the spin densities distribution in the molecule. This in turn would influence not only the size of the exchange term J but also its sign. Consequently the preference for S=1 state is not guaranteed. The effect of the heteroatomic substitution in the coupling unit core is documented in literature by doubtful results. This is shown, as one example, by comparing the very simple biradical a (benzene core) versus b (pyridine core) and c, all reported in Figure 4.30. - -O O 2J / k =+ b 40 K Ref. 38a, b O N + O N N N2J / kb > 200 K Ref. 39 N N a N N+ N + 2J / kb = 0 K Ref. 40O - O b O -O 2J / kb = 36.0 +- 10 K Ref. 41 2J / kb = 18.8 K Ref. 42a, b 2J / kb = 16 K Ref. 43 N O O-+ N N N+ c N -O O 2J / kb = ? Singlet Ref. 43 Figure 4.30: The estimated singlet-triplet energy gap (∆EST) in a serie of NN biradical systems based on benzene (a), 2,6 pyridine (b), and 3,5- pyridine (c). The reader can recognize that even for a, in which the isophthaldehyde as key-radical precursor is commercially available, and therefore does not require much synthetic effort to access to large quantities for the radical system, at least four very different estimations of the 103 Chapter 4- EPR Analysis ___________________________________________________________________________ intramolecular exchange interaction J have been provided over the years (ranging from 0 up to > 200K) [38 - 41]. They were usually obtained by susceptibility measurements on powder samples. In only one reference [41] the EPR technique was applied in parallel with the magnetic susceptibility in the bulk, in order to evaluate and compare the strength of the through-bond interaction, J, with these two different techniques. The result reported in Figure 4.30 (2J / kb = 36 ± 10 K) was obtained from the low temperature EPR studies, and judged as the best estimation for the intramolecular interaction between the radical moieties. Nevertheless several doubts about the purity of the sample examined remain, especially if we consider that no EPR spectra in frozen solution were provided, together with the knowledge of the sample concentrations. Those are both crucial points that need to be documented. The intramolecular exchange interaction can only be probed in dilute solution (≤ 8 * 10-3 M), and when no specific interactions with the solvent molecules are present. In concentrated phases, dipolar through space interactions among paramagnetic molecules are active, and therefore one would still analyse the cooperative effects coming from both inter- and intramolecular interactions. Also the sample purities needs to be specified, since as we emphasized in this chapter, monoradical impurities on biradical samples can be easily assessed by using the EPR technique. From the data reported in Figure 4.30, and supposing that a and b might adopt similar molecular conformation, one would conclude that the pyridine ring hamper the propagation of the spin-polarization induced by the two radical systems through the coupler, although J still is kept ferromagnetic in b. In contrast with this assessment, we showed indeed that non “zero-spin-densities” are delocalised from the radical moiety (NN) into either the pyridine ring (see EPR analyses of monoradical NN 8) or even in the pyrazole (see EPR analyses of monoradical NN 37). This is also in agreement with similar findings envisaged from other referenced data (Figure 4.4). More surprising are the results obtained for biradical c in which the ground spin-state is stated to be reversed [43], but no information about the magnitude of J was ever provided. In the section 4.2.2, we clearly showed that the spin concentration, once a suitable spin-standard is employed, combined with the EPR simulation in solution for the biradical envelope, can undoubtedly provide at least the upper and lower limit for the singlet-triplet energy gap. Very few examples of biradical systems attached on pyridine units are known, and are therein provided. These structures are collected in Figure 4.31. In all these biradicals systems (see structures from d to i) no significant intramolecular interactions were found by the authors. No ∆Ms = 1 transition envelopes were ever reported, no estimation of the zero-field splitting terms, D and E, were ever made, nor any ∆Ms = 2 transition were observed. From the literature results it appeared that a tremendous attenuation of the spin polarization effect is observed when a second or a third pyridine ring is present. However, it should be noted that for all of these NN pyridine derivatives, due to their topological design, the low spin ground states are expected. All the values reported in Figure 104 Chapter 4- EPR Analysis ___________________________________________________________________________ 4.31 correspond to the Weiss (Φ) constant, accounting for only weak intermolecular antiferromagnetic interaction, although in the case of g the authors estimated a very crude through-bond interaction ~ -10 K. However, it is not understandable why no similar estimation was provided in e, which clearly should features in principle larger interaction, at least due to the shorter through-space radical distances. O- O- -O N N +N + N N O N N+ N N O d O Φ = -24 K- O N Ref. 42a Φ = -1.9 K Ref. 42a,c N e N+ O - - - O O O N N N + N +N N O N N + N N N Φ = -3 K O - O Nf O N 2J / kb ~ -10 K + g Φ = -1.3 K Ref. 42a,c N Ref. 42c O - O N N + O - N h N N N+ O N N NO- i + + Φ = -3.5 K Ref. 42a,c O N N O O- N N O- Φ = -2.2 K Ref. 35 Figure 4.31: Some of the known π-conjugated NN biradical derivatives based on pyridine units, with their calculated intermolecular through-space interactions (Φ). The molecule i represented the only other known example of NN biradical in the terpyridine system. The intramolecular interactions between the radical moieties was excluded, but it exhibited only weak intermolecular and antiferromagnetic contributions in the bulk with Φ = -2.2 K [45]. These analyses of the electronic and magnetic properties, did not allow us to clearly compare our findings with such reference data. In all the new biradical systems reported in this work, we provided either the zero-field splitting parameters and always, although weak, the forbidden half-field transitions were observed. Our terpyridine NN system 12 showed in contrast with i a fairly large ferromagnetic through-bond interaction (2J / kb ~ 22 K), that was decreasing in the IN system 13 (2J / kb ~ 12.5 K). For that reason, not only the through-bond interactions between the radical units were present, could be observed and clearly defined, but they were really large compared with those, for example, estimated in a and b (see Figure 4.30), where there is only one aromatic ring connecting the radical units. Our results accounted for a very efficient propagation of the spin polarization through the coupler, besides theoretical predictions, and therefore those would constitute unprecedented 105 Chapter 4- EPR Analysis ___________________________________________________________________________ findings. Our clear attribution of the electronic ground-state made it possible to visualize the variation of the ∆EST in the π-conjugated systems 12, 13, 26, 28, 33 and 34 (see Figure 4.32). This illustrates experimentally the weight of the five-member (pyrazole) ring in slightly hampering but unexpectedly not quenching the through-bond exchange interaction, due to the presence of two different pathways for the - spin-polarization (Figure 4.32). + * + *- + + Under the assumption that in the imino N N -* *- -* nitroxide biradicals the coupling unit core N N -* + would be as planar as those observed in path (a) path (b) the X-ray structures of 12, 26 and 34 (all with = NN or IN nitronyl nitroxides, NN, see Chapter 5), no dominant geometrical factors seemed Figure 4.32 responsible for modulating the observed different singlet-triplet stabilities. Noteworthy, the averaged intramolecular radical distances in the terpyridine systems 12 and 13 are larger than in 26, 28, 33 and 34 (see Table 4.3) being accompanied with small zfs. However, their exchange couplings are the strongest one. Even though in 41 a decrease in the magnitude of J is expected, overall the trends in J experimentally obtained would be hard to acquire quantitatively by other means (e.g. with the aid of theoretical predictions). Therefore the following efficiency in the coupling unit series could be drawn, according to the radical type and size of J estimated. 12 (NN) > 33 (NN) > 26 (NN) > 13 (IN) > 41 (NN) >34 (IN) ~ 28 (IN) > 42 (IN) One note of criticism we might furthermore add, upon comparison with the other class of relatively stable and much more studied nitroxide radical systems [45]: the very common assumption [44] that the nytronyl nitroxide should exhibit much weaker coupling with N O respect to the nitroxide moiety (Figure 4.33), need to be reconsidered in view of these Reference 45 J / kb = 5.3 K recent findings. N N O O J / kb = 5.3 K ~ 3.68 cm - 1 Figure 4.33 106 Chapter 4- EPR Analysis ___________________________________________________________________________ Figure 4.32: Comparison among the observed zfs (D, upper figure) with the singlet-triplet energy gap (∆EST = 2J/kb, lower figure,) in the synthesized biradical systems. 107 Chapter 4- EPR Analysis ___________________________________________________________________________ References [1] (a) Gerson F., Huber W., Electron Spin Resonance Spectroscopy of Organic Radicals, 2004, Wiley-VCH. (b) Barth U., Hedin L.,"A Local Exchange –correlation Potential for Spin Polarized case: I". J. Phys. C: Solid State Phys., 5, 1972, 1629. (c) Gerson, F. and Huber, W., Electron Spin Resonance spectroscopy of Organic Radicals, Wiley-VCH, 2003.(d) Bales, B.L.; Peric, M.; Dragutan, I., J. Phys. Chem. A, 107, 2003, 9086. [2] (a) Catala L., Feher R., Amabilino D.B., Wurst K., Veciana J., Polyhedron, 20, 2001, 1563. (b) Vasilevsky S.F., Tretyakov E.V., Usov O.M., Molin Y. N., Fokin S.V., Shwedenkov Y.G., Ikorskii V.N., Romanenko V.G., Sagdeev R.Z., Ovcharenko V.I., Mendeleev Commun., 1998, 216. (c) Cirujeda J., Hernández-Gasiò E., Rovira C., Stanger J-L., Turek P., Veciana J., J. Mat. Chem., 5, 1995, 243. (d) Nagashima H., Inoue H., Yoshioka N., Polyhedron, 22, 2003, 1823. (e) Takui T., Miura Y., Inui K., Teki Y., Makoto I., ItoH K., Mol. Cryst. Liq. Cryst., 271, 1995, 55. (f) Davis M.S., Morokuma K., Kreilick R.W., J.Am.Chem. Soc. 94, 1972, 5588. See also ˝Electron Carbon Couplings of Aryl Nitronyl Nitroxide Radicals˝ Neely J.W., Hatch G.F., Kreilick R.W., J.Am.Chem. Soc., 96, 1974, 652. [3] Heisenberg W., Z. Phys., 49, 1928, 168. [4] Dirac P.A.M., The principles of Quantum Mechanics, 3rd Ed., Clarendon, Oxford. [5] Anderson P.W., Solid State Phys., 14, 1963, 99. [6] (a) Löwdin P.O., Phys. Rev. 97, 1995, 1509. (b) Rev. Mod. Phys. 34, 1962, 80. [7] (a) Nesbet R. K., Ann. Phys. (Leipzig) 3, 1958, 397. (b) Nesbet R. K., Ann. Phys. (Leipzig), 4, 1958, 87. (c) Nesbet R. K., Phys. Rev. 122, 1961, 1497. (d) Nesbet R. K., Phys. Rev. 135, 1964, A460. [8] Hay P.J., Thibeault J.C., Hoffmann R. J. Am. Chem. Soc., 97, 1975, 4884. [9] (a) Kahn O., Briat B., J. Chem. Soc. Faraday Trans., 72, 1976, 268. (b) Kahn O. Molecular Magnetism, Wiley-VCH, New York, 1993, Chapter 8. [10] Noodleman L., J. Chem. Phys., 74, 1981, 5737. [11] Noodleman L., Davidson E.R., Chem. Phys., 109, 1986, 131. [12] Schleyer P.V., Allinger N.L., Clark T., Gasteiger J. Kollman P.A., Schaefer H.F. III, Schreiner P.R. (Eds.) Encyclopedia of Computational Chemistry, 1998, Wiley, Chichester. [13] Bauschlicher C.W. Jr., Langhhoff S.R., Taylor P.R., Adv. Chem. Phys. 77, 1990, 103. [14] Roos B.O. (Ed.), Lecture Notes in Quantum Chemistry; Lecture Notes in Chemistry, Vols 58 and 59, 1992, Springer, Berlin-Heidelberg, New York. [15] Parr R.J., Yang W., Density Functional Theory of atoms and Molecules, 1989, Oxford University Press, New York. [16] Dreizler R.M., Gross E.K.U., Density Functional Theory: An approach to the Quantum Many Body Problem, 1990, Springer, Berlin Heidelberg New York. 108 Chapter 4- EPR Analysis ___________________________________________________________________________ [17] (a) Seminario J.M., Politzer P. (Eds), Modern Density Functional Theory: a tool for chemistry, Vol.2, 1995, Elsevier, Amsterdam. (b) Illas F. and Martin R., J. Chem. Phys., 7, 2001, 2887. [18] Illas F., Moreira I. de P.R., Graaf de C., Barone V., Magnetic Coupling in Biradicals, Binuclear Complexes and wide-gap insulator: a survey of ab initio wave function and density functional theory approaches, Theor. Chem. Acc., 104, 2000, 265. [19] White R.M. Quantum Theory of Magnetism. Springer Series in Solid-State Sciences, Vol 32, 1983, Springer, Berlin Heidelberg New York. [20] O. Kahn, Magnetism: A Supramolecular Function, Eds., Kluwer, Dordrecht, 1996. [21]Casper W.J. Spin Systems, 1989, World Scientific, Singapore. [22] (a) Yoshida K, Theory of Magnetism, 1998, Springer, Berlin Heidelberg New York. (b) J. E. Wertz, J. R. Bolton, Electron Spin Resonance, Elementary Theory and Practical Applications, 1986, Chapman & Hall. (c) Mabbs F.E., Collins D., Electron Paramagnetic resonance of Transition Metal Compounds, Elsevier, 1992. (d) De Groot, M.S. and van dr Waals J. H., Physica, 29, 1963, 1128. (e) Grivet, J. –Ph. and Mispelter J., Mol. Phys. 27, 1974, 15. [23] McConnell H. M., J. Chem. Phys. 33, 1960, pp.115-121. [24] McLachlan A. D. in “Self-consistent field theory of the electron spin distribution in π- electron radicals” Mol.Phys. 3, 1960, pp.233-252. [25] (a) Salem L., The molecular orbital theory of conjugated systems, Benjamin, New York, NY, 1965. (b) J. E. Wertz, J. R. Bolton, Electron Spin Resonance, Elementary Theory and Practical Applications, 1986, Chapman & Hall, p.254. [26] (a) Fang S., M-S. Lee, Hrovat D.A., Borden W.T., J. Am. Chem. Soc., 117, 1995, 6727. (b) Chiarelli R., Gambarelli S., Rassat A. “Exchange interactions in nitroxide biradicals”, Mol.Cryst. Liq. Cryst., 305, 1997, 455. [27] Kobori Y., Takeda K., Tsuji K., Kawai A., Obi K., J. Phys. Chem. A, 102, 1998, 5160 (in particular page 5166) and references cited therein. [28] (a) Anderson P.W., Phys. Rev., 115, 1959, pp.2-13. (b) J. E. Wertz, J. R. Bolton, Electron Spin Resonance, Elementary Theory and Practical Applications, 1986, Chapman & Hall, pp153-154. [29] Bales L.B., Peric M., Dragutan I., J. Phys. Chem. A, 107, 2003, 9086. [30] Kanaya T., Shiomi D., Sato K., Takui T., Polyhedron, 20, 2001, 1397. [31]Tretyakov E.V., Samoilova R.I., Ivanov Y.V., Plyusnin V.F., Pashchenko S.V., Vasilevsky S.F., Mendeleev Commun., 1999, 92. [32] (a) M. Tinkham, M. W. P. Strandberg, Phys. Rev. 1966, 97, 937.(b) K. Mukai, T. Tamaki, Bull. Chem. Soc. Jpn. 1977, 50, 1239. [33] E. F. Ullman, J. H. Osiecki, D.G.B.Boocock, J. Am. Chem. Soc. 1972, 94, 7049. 109 Chapter 4- EPR Analysis ___________________________________________________________________________ [34] (a) B. Bleanay, K. Bowers, Proc. R. Soc. London 1952, A214, 451. (b) Mukai K., Tamaki T., Bull Chem. Soc. Jpn. ,50, 1977, 1239. (c) Mukai K., Sakamoto J., J. Chem. Phys., 68, 1978, 1432. [35] C. Stroh, R. Ziessel, Tetrahedron Lett. 40, 1999, 4543. [36] Portis A.M. Phys. Rev. 91, 1953, 1071-1078. [37] Castner T.J .Jr. Phys. Rev. 115, 1959, 1506-1515. [38] (a) Hase S., Shiomi D., Sato K., Takui T., J Mat. Chem. 11, 2001, 756. (b) Izuoka A., Fukada M., Sugawara T., Mol. Cryst. Liq. Cryst., 232, 1993, 103. [39] Shiomi D., Tamura M., Sawa H., Kato R., Kinishita M., , J. Phys. Soc. Jpn., 62, 1993, 289. [40] Caneschi A., David L., Gatteschi D., Sessoli R., Inorg. Chem., 32, 1993, 1445. [41] Catala L., LeMoigne J., Kyritsakas N., Rey P., Novoa J. J., Turek P., Chem. Eur. J., 7, 2001, 2466. [42] (a) Ziessel R., Ulrich G., lawson R.C., Echegoyen L., J. Mat. Chem., 9, 1999, 1435. (b) Ulrich G., Ziessel R., Tetrahedron Lett., 35, 1994, 1215. (c) Romero F. M., Ziessel R., Tetrahedron Lett., 40, 1999, 1895. see also the short review R. Ziessel, Mol. Cryst. Liq. Cryst., 273, 1995, 101. [43] Shiomi D., Ito K., Nishizawa M., sato K., takui T., Itoh K., Synthetic Metals, 103, 1999, 2271. [44] (a) Joe A. Crayston, John N. Devine and John C. Walton, Conceptual and Synthetic Strategies for the Preparation of Organic Magnets, Tetrahedron, 56, 40, 2000, 7829. (b) Nakatsuji S., Kiroyuki A., J. Mat. Chem. 7, 1997, 2161. [45] Kanno F., Inoue K., Koga N., Iwamura H., J. Phys. Chem., 97, 1993, 13267. For other examples of nitroxides biradicals see also: Matsumoto T., Ishida T., Koga N., Iwamura H., J. Am. Chem. Soc., 114, 1992, 9952. 110 ____________________________________________________________________________ Chapter 5 – Crystal Stuctures of the Radicals Experimentally the mechanism of through-bond spin coupling can be elucidated only by understanding the phenomenological relationships among spin delocalisation, spin polarisation, molecular conformations, radical separation distances, and hetero-ring substitutions. The spin delocalization/polarization effects and the studies of molecular conformations in solution for all the radicals synthesized have been discussed in details in Chapter 4. In this chapter the central point is devoted to the analyses of the structural factors that characterize three nitronyl nitroxide biradicals (namely 12, 26 and 33). It is also reported the structure of the nitronyl nitroxide monoradical 8 since it disclosed in the EPR solution studies (Chapter 4) evidence of spin polarization from the radical moiety to the protons of the neighbour pyridine ring. Regardless of several attempts, unfortunately, no suitable crystals for 41 and all the correspondent imino nitroxides were obtained. The three systems 12, 26 and 33 shared nearly planar arrangements of the coupling-core. Rather small torsions between the imidazolyl and the pyridyl/pyrazolyl rings were observed. Therefore the geometrical prerequisites to enable excellent conjugation between the radical moieties through the coupler were ensured. Such structural findings further enhance the potential advantages that similar hetero-ring cores present with respect to those based on phenylene moieties, since the larger torsions between adjacent benzene rings (> 30°) in turn would induce far less efficient radical conjugation. 5.1. The crystal structure of 12 and the supramolecular π-stacking chain formation. Crystals with the P21/c (No. 14) symmetry for the nitronyl nitroxide biradical based on terpyridine 12 were grown in a solution of acetone upon slow diffusion of hexane at 4°C within few days. The structure is shown in Figure 5.1. Details of the structural refinements are given in Table 1 at the end of this chapter. The torsional angles φ1, φ2, φ3, and φ4 are meaningful parameters for the transmission of effective π-conjugation between the two spin carriers through the spacer, because very strong torsions would hamper or even cancel the conjugation effect through the coupler. The φ1 angle (N5-C18-C5-C4) is ~ -0.7° and the φ2 angle (N2-C6-C1-C2) is -4.0° thus the overall terpyridine backbone is fairly planar. The torsional angle φ3 between the imidazolyl and one of the pyridyl-moiety (N6-C23-C21-C20) is ~ -30.1°; this large torsion arises from the interaction between the imidazolyl oxygen O3 involved in hydrogen bonding with one of the two water molecules (O11) found in the crystals. Similarly, the φ4 angle (N3-C11-C9-C8) follows the same trend, with 31.9°, since again the O1 oxygen interacts via hydrogen bonding with the second water molecule O12. 111 Chapter 5- The Radical`s X-ray Stuctures ___________________________________________________________________________________ Figure 5.1. Crystal structure of 12 with ORTEP drawn at the 50% of probability level. The hydrogen atoms have been omitted for clarity. The intraring-torsional angles of the imidazolyl rings are ~ -19.3° (C23-N7-C25-C24), -19.1° (C23-N6-C24-C25), -13.7° (C11-N4-C13-C12), and -14.8° (C11-N3-C12-C13). The intramolecular (O3-O1) distance between the oxygens is 11.24 Å. The intramolecular distances between the two ONĈNO groups (C23-C11) is d = 12.11 Å, while that between the pyridyl carbons (C9-C21) is d = 9.52 Å. Two of the N-O bond distances, are slightly longer than the others, and those in fact are associated with hydrogen bonding between the radical oxygen and the water molecules (N3-O1, d = 1.29 Å and N6-O3, d = 1.29 Å), while the other two have d = 1.26 Å (N7-O4) and d = 1.28 Å (N4-O2). The crystal packing of 12 is shown in Figure 5.2. The stacking among the terpyridine moieties is gated by these two bridging water molecules, allowing the formation of an infinite chain with 180° Figure 5.2. ORTEP sketch for the crystal rotated units in top to each other. The Figure packing of 12 (nitrogen ●, oxygen ●, carbon 5.3 shows a section of the chain (trimer A-B-C) ●).The unit cell is depicted as thin yellow solidline. where the nitrogen of the pyridine A (N5) is connected with the water molecules O12 (dN5-O12 = 2.92 Å) and hydrogen bonded with the radical oxygen (O1) of the molecule B (dO12-O2 = 2.92 Å). Then, the nitrogen of the pyridine B (N2) is connected with the water molecules O11 (dN2-O11 = 2.93 Å) and hydrogen bonded with 112 Chapter 5- The Radical`s X-ray Stuctures ___________________________________________________________________________________ the radical oxygen (O3) of the molecule C (dO11-O3 = 2.82 Å). This allows a stacking motif with distances C6(A)-C18(B) of 3.28 Å and C18(B)-C6(C) of 3.29 Å as repeating unit for the chain. Figure 5.3. Capped stick sketch of the molecular packing in 12 with a perspective of the chain section. Such appealing supramolecular organisation found in 12 gave an unprecedented model for probing the bulk magnetic properties of the material. The static magnetic susceptibility was therefore measured down to cryogenic temperature with a Faraday balance. This experiment was performed by Dr. K. Falk (Prof. W. Haase Group, TU Darmstadt). As shown in the Figure 5.4, the effective magnetic moment is 2.37µB hence slightly smaller than the expected 2.43µB for two uncorrelated spins. This value appears almost constant down to 50 K. Then below 20 K the magnetic moment decreases. Nevertheless a small increase is observed in the range 40-20 K, which is consistent with the intramolecular singlet-triplet gap of 15.4 ± 2.0 cm-1 obtained from the EPR studies. Presumably, the sharp decrease of µeff below 20 K arises from dominating inter-chain antiferromagnetic interactions; however the full evaluation of these data is still under examinations. This is due to the large numbers of intermolecular interactions (Jinter, see Figure 5.5), and a suitable model for their parameterization has not been developed yet. This observation underlines once more the better suitability of the EPR technique for evaluating at least the through-bond exchange interaction, because by working with magnetically dilute samples, it allows minimizing intermolecular contributions. Although 12 did not show bulk ferromagnetic properties, still it represents a unique example in the terpyridine family, where only rather strong antiferromagnetic interactions in the bulk are found, with negligible through-bond interaction, as observed for the other one example reported so far in literature (see Figure 5.4 B). 113 Chapter 5- The Radical`s X-ray Stuctures ___________________________________________________________________________________ (A) Figure 5.4. (A) The static magnetic susceptibility χ (red circles, ●) and the effective magnetic moment µeff (blue circles, ●) for the π-stacking of the polycrystalline biradical 12 showing the dominant contribution arising from the weak intermolecular antiferromagnetic interactions below T = 20 K. (B) The product χ.T (open circles, ○) in the other known NN biradical based on the terpyridine system, measured on polycrystalline powder (taken from Stroh C., R. Ziessel, Tetrahedron Lett., 40, 1999, 4543.) Figure 5.5. Perspective of the short (< 3.5 Å) magnetic contacts (red dotted lines) and the H- bonding (light blue dotted lines) found in the crystals of 12. 5.2. The crystal structure of 26 and the zig-zag chain formation. Diffusion of hexane into a CHCl3 solution of radical 26 within 2 days allowed the formation of a blue crystalline material that was structurally characterized by X-ray diffraction (Figure 5.6). Since the molecules are located on a twofold rotation axis (C2/c, N° 15), the torsional angles φ1 and φ2 are identical for each half-molecule. The φ1 (C6-N2-C1-N1) angle is 4.6° and the torsional angle (C6-N2-N3-C4) is ~ 0.3° thus the two pyrazolyl-rings and the pyridyl central-core are nearly coplanar. The torsional angle φ2 between the imidazolidyl and 114 Chapter 5- The Radical`s X-ray Stuctures ___________________________________________________________________________________ pyrazolyl-moiety (N5-C7-C5-C6) falls in very similar range with 4.2° hence, the overall planar structure of 26 (more planar than 12), satisfies in the crystals the geometrical prerequisite for effective π-conjugation between the two radical fragments. The intraring torsion of the imidazolyl rings are respectively - 18.9° (C8-C9-N5-C7) and -22.6° (C9-C8-N4-C7). The intramolecular (O2-O*2) and the shortest intermolecular through-space distances between the oxygens are 5.64 Å and 4.74 Å. The intramolecular distances between the two ONĈNO groups (C7-C*7), Figure 5.6. Crystal structure of 26 with ORTEP drawn at the 50% of probability level. The hydrogen where most of the radical-spin atoms were omitted for clarity. densities is about are located is about 9.13 Å, while that one between the pyrazolyl-carbons (C5-C*5) is 7.71 Å. The N-O bond distances are both 1.28 Å for O2-N5 and for O1-N4. The Figure 5.7 shows the zig-zag chain motif found in the crystals, in which the units are rotated by 180°, running along the b-axis. Such motif arises by short electrostatic contacts, (shorter than the sum of the van der Waalls radii minus 0.1 Å) between the imidazolyl-oxygens and the methyl groups of the neighbouring radicals. Figure 5.7. The representation of the molecular packing in 26 drawn with ORTEP32. 115 Chapter 5- The Radical`s X-ray Stuctures ___________________________________________________________________________________ 5.3. The structure of 33 and the dimers formation. After diffusion of ether into a CHCl3 solution of the radical 33, a blue crystalline material was obtained in the P-1 (No.2) symmetry. Two molecules for the biradical plus two solvent molecules (CHCl3) were found per unit cell. In Figure 5.8 is reported the structure of the isolated 33. Figure 5.8. Crystal structure of 33 with ORTEP drawn at the 50% of probability level. The hydrogen atoms were omitted for clarity. The φ2 angle (C10-C9-C11-N3) is ~ -3.2° and the φ3 angle (C20-N5-C5-N1) falls in the similar range (– 4.3°), thus the bispyridyl-pyrazolyl rings, are nearly coplanar. The large torsional angle φ1 between the imidazolyl and pyridyl-moiety (N3-C11-C9-C10) accounts for - 27.8°. Similarly, the φ4 angle between imidazolyl and pyrazolyl residue (N7-C21-C19-C18) is - 14°. The overall structure of 33 is more planar with respect to 12 but less than 26 (see Figure 5.1 and 5.6). The intraring torsional angles of the imidazolyl rings are -8.4° (N7-C21-N8-C23), -21.9° (C21-N7-C22-C23), ~ -8.3° (N4-C11-N3-C12), and 16.7° (C11-N4-C13-C12). The presence of two solvent molecules allows the formation of dimeric structures (Figure 5.7) with units rotated 180° in top of each other. This feature is similar to those observed in the radicals 12 and 26. The intramolecular (O4-O2) and the shortest intermolecular through-space distances between the imidazolyl-oxygens (short magnetic contacts) are respectively 8.5 Å and 4.3 Å. The later is associated with the distance between oxygen in the dimer. The intramolecular distances between the two ONĈNO groups, where most of the radical-spin densities are located, is d(C21-C11) = 10.64 Å, and that between the pyrazolyl-pyridyl carbons (C9-C19) is d = 8.64 Å. The intradimer distance between two ONĈNO groups (C21-C11, d = 4.668 Å) is indeed far more close (Figure 5.10). 116 Chapter 5- The Radical`s X-ray Stuctures ___________________________________________________________________________________ The N-O bond distances are 1.27 Å for O2-N4 and 1.29 Å for O4-N8. As previously found in 12, also for 33 there are short interdimer contacts. Those arise from the interaction between the imidazolyl-oxygens and the methyl groups of the neighbouring radicals as schematically depicted in the Figure 5.10 (B, red lines). Figure 5.9. The view of the dimeric form of (33) found in the crystals. The two CHCl3 molecules are drawn in green. Figure 5.10. The molecular packing of 33. 5.4. The structure of 8 and the dimers formation. Suitable crystals for the bipyridine based monoradical 8 were grown in a solution of acetone upon slow diffusion of ether within one day, in the P-21/n (No.14) symmetry. Structurally, as shown in Figure 5.11, the two pyridine-rings are nearly coplanar with a torsional angle of – 2.80° (N2-C6-C5-C4). However, the torsion between the imidazolyl and the pyridyl ring (N4-C11-C9-C8) is large, Figure 5.11. Crystal structure of 8 with ORTEP accounting for -35.4°. In addition, also the drawn at the 50% of probability level. The hydrogen atoms were omitted for clarity. 117 Chapter 5- The Radical`s X-ray Stuctures ___________________________________________________________________________________ intraring torsion for the imidazolyl moiety is large, with ~21.7° (N4-C13-C12-N3) and ~ 5.0° (C12-N3-C11-N4). The shortest intermolecular through-space distances between the radical- oxygens (short magnetic contact) is 4.25 Å. The N-O bond distances for N3-O1 and N4-O2 are both 1.28 Å, and no solvent molecules were found to fill the unit cell. In the cell unit (Figure 5.12 C), the radical 8 (C) forms π stacking dimers (see Figure 5.12 B, and A) with short contacts (O2-C14, d = 3.32 Å) between imidazolyl-oxygens and the methyl groups of the neighbouring radical dimer. This is schematically depicted in the Figure 5.12 A, with black arrows. Such dimers organization is unusual as compared with the other available bipyridine structures deposited in the Cambridge Crystallographic Data Centre. (A) (B) Figure 5.12. (A), (B) and (C) show the molecular packing for the monoradical 8. 5.5. Conclusion. In the present chapter the crystal structures of three nitronyl nitroxide biradical systems were presented, together with one example of monoradical. The structural analyses showed that the small torsions between radical moieties and coupling unit core (bispyrazolyl- pyridine, terpyridine and pyrazolyl-bispyridine) favour significant through-bond exchange 118 Chapter 5- The Radical`s X-ray Stuctures ___________________________________________________________________________________ interaction, despite the large through-space radical distances. Such arrangement of the molecular units underlined the advantage on using nitrogen-containing heterocycles with respect to the more often used benzene cores. Other than the greater proclivity to adopt planar conformations, these radical systems offer in addition various coordination sites for metal chelation, a further advantage in order to extend the magnetic structures. Table 5.1. X-ray data: experimental details, structure solutions and refinements. Compound 12 26 33 8 CCDC number 234179 217301 229003 229004 Formula C29H37N7O6 C25H31N9O4 C28H33Cl3N8O4 C17H18BrN4O2 Formula Weight M 579.66 521.57 651.98 390.25 Crystal System Monoclinic Monoclinic Triclinic Monoclinic Space group P21/c (N° 14) C2/c (N°15) P1 (No 2) P-2 /n (No1 14) a (Å) 6.5690 18.5194(7) 11.5796(4) 6.6600(4) b (Å) 21.5610 9.6934(5) 12.3160(5) 10.9300(5) c (Å) 20.6620 14.4537(6) 12.7410(5) 23.1570(7) α (°) 90 90 69.9597(13) 90 β (°) 90.8 103.4960(10) 65.7760(145) 96.3620(13) γ (°) 90 90 88.9354(12) 90 V (Å3) 2926.16 2522.92(19) 1540.85(11) 1675.31(14) Z 4 4 2 4 ρ -3calc (g×cm ) 1.316 1.373 1.405 1.547 µ (MoKα) (mm-1) 0.094 0.085 0.346 2.471 F(000) 1232 548 680 796 Crystal Size (mm) 0.09×0.14×0.42 0.43×0.26×0.18 0.09×0.21×0.38 0.08×0.15×0.41 Colour Blue Blue Blue Blue Shape Prism Needles Prism Needles Temperature (K) 120 120 120 120 Radiation, λ(Å) MoKα,0.71073 MoKα,0.71073 MoKα,0.71073 MoKα,0.71073 θ Min-Max (°) 4.1-27.4 4.1-27.5 4.0-29.0 4.0-29.5 Total data 5870 3057 31506 19655 Unique data 1924 2889 8016 4621 Rint 0.060 0.060 0.060 0.000 1911 1895 3793 2764 Observed data Nref 1911 1895 3793 2764 Npar 379 175 388 217 S (GooF) 1.07 1.05 1.07 1.08 aR1 0.0645 0.0385 0.0512 0.0405 bwR2 0.0666 0.0453 0.0571 0.0448 a R1 = ∑||F0| - |Fc||/∑|F0| b wR2 = {∑w(|F0| - |Fc|)2/∑w|F 2 ½ 0| } The CCDC code corresponds to the deposition number provided by the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK. 119 ____________________________________________________________________________ Chapter 6 - Summary and Outlook Eight novel potential high spin ligands based on terpyridine, bispyrazolylpyridine and pyrazolybipyridine cores decorated with nitronyl nitroxide (NN) and imino nitroxide (IN) radicals were synthesized, together with four monoradical molecules. The preparation of the radicals involved several types of reactions (bromination, iodination, N- and carbaldehyde protecting groups, Stille coupling, Grignard reaction, etc.) in order to assemble the mono and the biscarbaldehyde hetero-derivatives as key precursors for the radical systems. Such intermediates were subjected to condensation reaction with 2,3-dimethyl-2,3- bis(hydroxylamino)-butane (generally in dioxane under argon for ~ 7 days), followed by oxidation of the bis-hydroxylimidazolyn precursor under phase transfer conditions (NaIO4/H2O). The use of other oxidants (e.g. PbO2) was found to be far less effective in the oxidation step for all the presented systems. The structures of these radicals are given in Figure 6.1. Figure 6.1: The novel mono and biradical systems synthesised based on nitronyl (NN, drawn in blue) and imino nitroxide (IN, drawn in red) moieties. 120 Chapter 6. Summary and Outlook __________________________________________________________________________ The nitronyl nitroxide radicals were always obtained in lower yields with respect to the imino nitroxides. This was explained by the difficulties in controlling the over oxidation/dehydratation of the hydroxylimidazolyn-derivatives, even upon working under stoichiometric conditions of oxidizing agent, and with argon saturated solutions. In particular, those based on pyridine were much more sensitive towards the dehydratation processes than the others attached to the pyrazole. UV/Vis solution of the radicals witnessed no specific interaction between solvent and radical systems, in a broad range of solvent polarities. However the radical stabilities strongly decreased in protic solvents, especially in THF and MeOH, where they could not be stored even for few days under argon. In toluene solutions they were all very stable (up to a year), and those based on pyrazole could be even heated up to 60°C for few hours without any decomposition, providing that the medium was maintained free of oxygen. Nitronyl nitroxides gave blue colour solutions, with two characteristic absorption bands: one defined as n→π* transition related to the aminoxyl-oxide residue that occurs around 600 nm with several vibronic compononents, and a second one with a much higher intensity in the UV region (π→π*, ~ 390 nm). These bands were strongly dependent for both intensity and position on the type of hetero-ring to which the radical units were connected. The radicals based on pyrazole exhibited enhanced optical properties in the visible (ε ≥ 1000 M-1× cm-1) and a relatively strong blue-shift of the π→π* absorption (< 390 nm), while those grafted to pyridine rings showed hampered optical properties on the visible (ε ≤ 500 M-1 × cm-1) and a red shift for the π→π* absorption (≥ 390 nm). These findings were consistent with the enhanced aromatic character of the systems. The imino nitroxide radicals provided orange-red solutions, with a broad absorption band around 470 nm (n→π*), while the second transition (π→π*) of the amino-oxyl moiety appeared embedded into the organic backbone absorptions, and could not be discriminated. The trends observed previously in the NN radicals were followed also in the IN radicals, in which the radical molar extinction in the visible decreased in the pyridine (ε << 1000 M-1 × cm-1) and increased when connected to pyrazoles (ε ≥ 1000 M-1 × cm-1). The UV/Vis transitions for three biradical systems (namely 12, 13, and 26) were simulated, after geometry optimization, using ROHF/AM1/CIS (20,20) in order to assign the type of the experimentally observed absorptions. Although the calculated transitions corresponded fairly well with the observed ones, their relative intensities did not reproduce the trends experimentally found for ε (pyridine hampered and pyrazole enhanced). The FTIR absorption spectra of the bis-hydroxylimidazolidyn precursors always showed a characteristic broad absorption (νOH, 3100 - 3400 cm-1) that was lost upon oxidation. The NN radicals exhibited a strong band around 1350 cm-1 (νN-OH), while for the IN radicals 121 Chapter 6. Summary and Outlook __________________________________________________________________________ such transition were both shifted to longer wavenumbers and much less intense (~1370 cm-1). The room temperature X-band EPR studies in solution for the mono and biradical NN systems gave respectively five and nine lines pattern, while the correspondent IN mono- and biradicals showed seven and thirteen lines. All the radicals did not witness line-width dependency upon changing solvent polarities (from hexane to MeOH), in agreement with the UV/Vis findings. However, they were all instable in protic solvents. This was corroborated by monitoring the decrease of their double integrated signal intensities versus time. On the other hand, 2- propanol constituted an exception; here the radicals were fairly high stables, but only those based on pyrazoles showed a slight increase in the observed nitrogen hyperfine interaction, (where aN, raises from ~ 0.75 mT to ~ 0.77 mT) without any substantial variation in the UV/Vis envelopes. The observed giso for the NN radicals directly attached on pyridine moieties were centred at 2.0066(1) and 2.0061(1) for the NN. The slight shift at giso = 2.0065(1) (NN) and 2.0060(1) (IN) were found for the radicals attached on pyrazoles. The biradical systems featured half of the spacing between the lines, as compared with the monoradicals (e.g. aN/2 ~ 0.374 mT for the nitronyl nitroxide biradicals where aN = 0.748 mT for the monoradicals, and aN1/2 ~ 0.430 mT, aN2/2 ~ 0.225 mT for the imino nitroxide biradicals with aN1 = 0.885 mT, aN2 = 0.430 mT for the related monoradicals, all recorded in toluene). These findings revealed that the radical moieties were strongly exchange coupled (J) within the EPR limit. Estimation of the lower limit for J were obtained by fitting the EPR isotropic spectra, yielding in all cases ׀J/a 40 ≤ ׀, i.e. 2J= ∆EST > 0.056 cm-1. For the monoradical 8 (NN), in addition to the major five lines, the high resolution spectra showed a more complex pattern with at least 13 visible additional splittings overlapped on each N hfc, originating from the presence of additional couplings of the single unpaired electron (S = ½) with twelve hydrogen nuclei (I=½,) of the four methyl groups (aH = 0.022 mT), the 6’ (aH = 0.041 mT) and 4’ hydrogens (aH = 0.044 mT) of the pyridyl moiety. Such finding experimentally demonstrates that non "zero spin density" resides also in the pyridine ring. The very similar resolved pattern occurred also in the case of monoradical 37, where the spin carrier (nitronyl nitroxide) is connected to the pyrazole ring. The solution EPR studies did not show any appreciable broadening in line-width upon cooling for all the monoradical systems; in addition they followed nicely the Curie-law in dilute system with [C] < 7 X 10-3 M, making them suitable standards for the biradicals spin concentration. Some of the coupled systems (e.g. 12, 13, 26, 28) did not feature a strong temperature dependence of the EPR line-width upon cooling. However, in the cases of 33, 34, 41, and 42 the EPR spectra in solution witnessed strong temperature dependency. For 33 and 34, such alternating line-width were explained by assuming the out-of-phase rotation of the radicals attached to the pyridine unit versus that one attached to the pyrazole moiety, in the EPR time 122 Chapter 6. Summary and Outlook __________________________________________________________________________ scale. In the biradicals 41 and 42, where two CH2 σ bonding are present, conformational interconversions in solution (rotamers) more easily allowed strong torsions between the radical moieties and the coupling unit core, leading to a fraction of uncoupled biradicals (~ 35%, with J< 200°C (760 mbar). 1H NMR (CDCl3, 250 MHz, 298 K, 16 scan) 8.43 δ (s, 1H, H-6), 7.62 δ (dd, 1H, 3J = 2.2, 7.9 Hz, H-3), 7.49 δ (d, 1H, 3J = 8.2 Hz, H-4), 5.80 δ (s, 1H, -CH), 4.32 δ (m, 4H, -CH2). 13C NMR (CDCl3, 63 MHz, 298 K, 9000 scan) δ (ppm): 147.7, 141.8, 135.7, 129 Chapter 7-Experimental Session ___________________________________________________________________________________________ 132.1, 126.8, 100.2, 64.4. MS-FD (70 eV, CHCl3) 230.1 (100%, M+), MW calculated (C8H8BrNO2) 230.06. 7.2.3. Synthesis of 2-tributylstannyl-5-[1,3]dioxolan-2-yl-pyridine (3) O O N Sn 2-Bromo-5-[1,3]dioxolan-2-yl-pyridine 2 (1.086 g, 4.7 mmol) was charged into a round flask together with freshly distilled diethylether (40 mL), and kept under argon. The solution was cooled to –78°C using dry ice/acetone bath and n-BuLi (1.6 M in hexane, 3.5 ml, 5.7 mmol) was added slowly within 10 min. The mixture was kept for 90 min at this temperature under stirring and rigorous argon atmosphere. Initially, after the complete addition of n-BuLi the solution appeared deep-green and after 90 min turned black. Tributyltin-chloride (Bu3SnCl, 97%) (1.66 mL, 6.2 mmol) was added within 5 min. The solution slowly turned into deep-red. The low temperature (-78°C) was maintained for 120 min and after that the mixture was allowed to warm slowly to room temperature overnight. The oily solution was filtered from the inorganic salts and the solvent evaporated under reduced pressure. Finally 3.8 mL of bright orange oil (ρ = 1.05 g/mL) of 3 were obtained and used for the Stille coupling reaction without purification. (Note that a part from the known toxicity of the stannyl-compounds, the product decomposes either on silica or alumina columns. 7.2.4. Synthesis of 2-tributylstannyl-6-bromopyridine (4) Br N Sn 2,6-Dibromopyridine (1 g, 4.22 mmol) was charged into a flask, evacuated and put under argon. Dry diethylether (70 mL) was added from a syringe under stirring and the solution was cooled to –50 °C in dry ice/ acetone bath. The white solution was very dense due to the low solubility of 2,6-dibromopyridine in diethylether. Then excess of n-BuLi (1.6 M in hexane, 5.6 mL, 8.96 mmol) was added dropwise from a syringe within 3 min. As soon as the addition was completed the mixture turned deep green. The reaction mixture was stirred for 30 min under 130 Chapter 7-Experimental Session ___________________________________________________________________________________________ continuous cooling ( -50 °C) then the temperature was lowered till – 60 °C. Tributhylstannyl- chloride (Bu3SnCl, 97%) (2.72 mL, 10.1 mmol) was added from a syringe within 5 min. When half of the addition was completed the solution became pale green and very limpid. The temperature was maintained for 60 min, then lowered till – 78 °C and kept for an additional hour. Finally the mixture was allowed to warm slowly overnight till room temperature. The solution was filtered from the white inorganic salts and dry toluene (2 mL) was added. The organic solution was collected and the diethylether was evaporated under reduced pressure avoiding heating to afford 3.7 mL of 2-tributylstannyl-6-bromopyridine 4 as pale yellowish oil (ρ = 1.134 g/mL) which was used for the Stille coupling reaction without purification (the product decomposes either in silica or alumina columns.) 7.2.5. Synthesis of 6’-bromo-[2,2]’-dipyridinyl-5’-carbaldehyde (5) N Br H N O The crude oily 3 (3.7 mL, 4.22 mmol) was transferred into a two-necked round bottomed flask together with an excess of 6-bromo-3-pyridinecarbaldehyde (1.8 g, 9.7 mmol). The mixture was degassed and kept under argon. Then dry and degassed toluene (70 mL) was added with a syringe together with dichlorobis(triphenylphosphine)-palladium(II) (148 mg, 0.21 mmol, 5% with respect to 3 and triphenylphosphine (110 mg, 0.42 mmol) as catalyst. The solution was then heated to reflux, in argon under stirring, for 72 hours. During the reaction the initially light- yellow solution became very dark after 72 hours. The solvent was evaporated under reduced pressure and then dichloromethane (50 ml) together with a saturated solution ammonium- chloride (30 mL) and a solution of EDTA (5%, 10 ml) was added to the resulting black-slurry. The mixture was shaken vigorously in a separator funnel and the phases were separated. The aqueous layer was extracted with portions of dichloromethane (2 × 30 mL). The combined organic layers were collected and the solvent evaporated under reduced pressure till small volume. The crude oily mixture was subjected to column chromatography (silica gel, ethyl- acetate/dichloromethane/hexane, 1/3/4). The main fraction eluted was 6’-bromo-[2,2]’- bipyridinyl-5-carbaldehyde (Rf = 0.65) as pale yellow powder. The product 5 was further washed with small cold portions of light petroleum ether (2 × 5 mL) (b.p. 30 - 40°C) and it was collected as highly pure white crystalline powder (750 mg, 68%). The second clear fraction collected was found to consist of [2,2’]-bipyridinyl-5-5’-dicarboxaldehyde (Rf = 0.25, 90 mg) as homocoupled product of 3. Starting from 8.44 mmol (7.4 mL) of 2-tributylstannyl-6- 131 Chapter 7-Experimental Session ___________________________________________________________________________________________ bromopyridine and 3.6 g of 6-bromo-3-pyridinecarbaldehyde (19.4 mmol) was obtained 1.29 g of 6’-bromo-[2,2]’-bipyridinyl-5-carbaldehyde (yield 58%). [2,2’]-Bipyridinyl-5-5’-dicarboxaldehyde: 1H-NMR (DMSO-d6, 250 MHz, 298 K, 32 scan) δ (ppm): 10.26 δ (s, 2H), 9.31 δ (s, 2H), 8.75 δ (d, 3J = 8.2 Hz, 2H), 8.52 δ (dd, 3J = 2.0, 8.0 Hz, 2H). MS-FD (70eV, CH2Cl2) 212.30 (M-H, 100%), MW calculated (MW+H) 213.30. Elemental analyses, found C 68.11, H 3.65, N 13.30 %, C/N = 5.12. C12H8N2O2 required C 67.92, H 3.80, N 13.20 %, C/N = 5.14. Procedure for the homocoupling: 6-bromo-3-pyridinecarbaldehyde (300 mg, 1.6 mmol) dissolved in dry xylene (20 mL) was added tributyltin-chloride (Bu3SnCl) (0.13 mL, 0.16 mmol), dichlorobis(triphenylphosphine)-palladium(II) (24.4 mg) and triphenylphosphine (18.2 mg) and heated to reflux under argon for 60 hours. The work up followed the same procedure as reported above (204 mg, yield 60%). 6’-Bromo-[2,2]’-bipyridinyl-5-carbaldehyde:M.p. 176 – 177 °C. 1H NMR (CDCl3, 250 MHz, 298 K, 64 scan) δ (ppm): 10.15 (s, 1H, -CHO), 9.09 (s, 1H, H-6'), 8.57 (d, 3J = 8.5 Hz, 1H, H- 3), 8.45 (d, 3J = 7.7 Hz, 1H, H-3'), 8.28 (d, 3J = 8.3 Hz, 1H, H-4'), 7.70 (t, 3J = 7.8 Hz, 1H, H-4), 7.56 (d, 3J = 7.8 Hz, 1H, H-5). 13C NMR (CDCl3, 63 MHz, 298 K, 8000 scan) δ (ppm): 188.7, 157.2, 154.2, 149.8, 140.2, 137.7, 135.4, 129.7, 127.4, 119.9, 119.2. MS-FD (70eV, CH2Cl2) 264.1 (M-H, 100%), MW calculated (MW+H) 264.1. UV-Vis (CHCl3) λ/nm (ε, mol-1 × cm-1) 321 nm (21380), 310 nm (25850), 260 nm (12030). FT-IR (KBr pellet, ν cm-1) 3042 (w, νC-H), 2962 (w, νC-H), 2883 (w, νC-H), 1682 (s, νC=O), 1592 (s, pyr), 1546 (s, pyr), 1435(m, pyr), 1369 (m), 1263 (m, pyr), 1209 (m, pyr). Elemental analyses, found C 47.20, H 3.30, N 10.01%, C/N = 4.72. C11H7N2O × H2O required C 47.00, H 3.20, N 9.97 %, C/N = 4.71. 7.2.6. Synthesis of 2,3-dimethyl-2,3-bis(hydroxylamino)-butane (6) HO N N OH A suspension of 2,3-dimethyl-2,3-bis(hydroxylamino)-butane sulfate salt (85%, C6H16N2O2 × H2SO4 , MW 264.30, 5 g, 18.92 mmol) in THF (70 mL) was kept cold with an ice bath and a solution of NaOH (1.56 g, 39 mmol, dissolved in 10 mL H2O) was added drop by drop while stirring within 10 min. The suspension became limpid after the addition of the base was completed. Then the resulting mixture was left under stirring at 4°C for further 10 min. A white solid slowly precipitated that consisted of Na2SO4. The salt was filtered off and the filtrate was slowly evaporated under air stream in a crystallizing disk. A white powder very hygroscopic was left, and it was further washed with cold hexane and dried under vacuum. In case that this 132 Chapter 7-Experimental Session ___________________________________________________________________________________________ powder slowly became pale rose, it was indicative for decomposition of the product into oxyme. Thus, the powder was washed first with cold and degassed water, followed by hexane. A white solid consisting of 2,3-dimethyl-2,3-bis(hydroxylamino)-butane (free base) was collected and stored at – 10°C in a closed bottle (1.4 g, MW 148.20, yield 50%). M.p. 160-161°C. FT-IR (KBr) ν/cm-1: 3257 (vs and broad, νOH), 2987 (vs, νC-H), 1479-1374 (vs, several bands), 1261 (s), 1178 (vs), 1145 (vs), 1080 (s), 1035 (vs), 989 (m), 952 (vs), 904 (vs), 852 (m), 790 (m), 690 (m). 1H NMR (D2O, 250 MHz, 298 K, 16 scan) δ (ppm): 1.25 (s, - CH ). 133 C NMR (D2O, 63 MHz, 298 K, 256 scan) δ (ppm ): 21.5 (CH3), 63.3 (CH3 - C - CH3). 7.2.7. Synthesis of 6-bromo-5'[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'- bipyridine (7) HO N Br N N N OH 6’-Bromo-[2,2]’-dipyridinyl-5’-carbaldehyde 5 (280 mg, 1.06 mmol) and 2,3- bis(hydroxylamino)-2,3-dimethylbutane 6 (444 mg, 3.0 mmol) were dissolved in 30 mL of dioxane and CH2Cl2 (1/1) and stirred under argon for four days until a fine white precipitate was formed. The solvent was removed by slow evaporation at room temperature and the solid product was collected, washed with a small amount of cold methanol/water mixture (1/1), and air dried to give crude 7 (354 mg, crude yield 85 %) that was subjected to the oxidation step without further purification. It decomposes before melting (from white powder to bright red powder) at 166-167 °C (from methanol). FT-IR (KBr pellet, ν/cm-1) 3251 (s and very broad, -OH), 2983 (s, νC-H), 2920 (s, νC- H), 1570 (m), 1548 (m), 1431 (s), 1378 (s), 1255 (w), 1155 (s), 1124 (s), 875 (s) 790 (s). 1H NMR (DMSO-d6, 250 MHz) δ (ppm) 8.65 (s, 1H, -CH), 8.31 (d, 3J = 7.53 Hz, -CH), 8.18 (d, 3J = 8.17, 1H, -CH), 7.96-7.92 (dd, 3J = 8.16 Hz, 1.9 Hz,1H, -CH), 7.87 (s, 1H, -OH), 7.82 (t, 3J = 7.85 Hz, 1H, -CH), 7.62 (d, 3J = 7.54 Hz, 1H, -CH), 4.56 (s, 1H, -CHimid), 1.00 (d, 3J = 7.84 Hz, 12 H, 4-CH ). 133 C NMR (63 MHz, DMSO-d6) δ (ppm) 156.7, 152.7, 149.7, 149.4, 141.0, 140.6, 138.3, 137.2, 128.2, 119.7, 87.8, 66.3, 24.2, 17.2. Elemental analyses, found C 47.76, H 5.75, N, 12.81%. C17H21BrN4O2 × 2 H2O (429.31) required C 47.56, H 5.87, N 13.05. 133 Chapter 7-Experimental Session ___________________________________________________________________________________________ 7.2.8. Synthesis of 6-bromo-5'[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'- bipyridine (8) O- N Br N+ N N O 6-Bromo-5'[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'-bipyridine 7 (250 mg, 0.58 mmol) was oxidized under phase transfer conditions in water/chloroform mixture (30mL, 1/3) using NaIO4 (186 mg, 0.87 mmol) for 30 min; during this period of time the solution gradually became blue-greenish. The phases were separated and the aqueous layer was extracted with chloroform (3 × 15 mL). The organic phases were collected and evaporated under air. The green-blue solid was dissolved in the minimum amount of acetone and chromatographed on neutral alumina using as eluents a mixture of acetone/petroleum-ether (low boiling point, 1/3). The first blue fraction eluted was collected and the solvent was removed under reduced pressure to afford pure 6-bromo-5'[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'- bipyridine 8 (56 mg, yield 25 %) was obtained as deep blue powder. M.p. 165 – 166 °C (from acetone, upon melting changes from deep-blue to bright orange). FT- IR (KBr pellet, ν/cm-1) 3112 (w, νC-H), 2981 (s, νC-H), 2933 (s, νC-H), 2875 (s, νC-H), 1583 (m), 1548 (m), 1454 (s), 1430 (s), 1413 (s), 1352 (s, N-O), 1213 (m), 1124 (m) (pyridine and pyrazolyl- moieties). UV/Vis, λ/nm (toluene), (ε, M-1× cm-1): 316 (23270), 327 (24480), 342 (20235), 377 (5580), 394 (8960), 468 (84), 498 (71), 537 (115), 574 (187), 620 (259), 669 (221), 746 (65). Elemental analyses, found C 47.76, H 5.45, N 12.90%. C17H18BrN4O2 × 2H2O (425.08) required C 47.90, H 5.20, N 13.14%. EPR (298 K, 9.399870 GHz, 7 × 10-5 M in toluene): five lines, giso = 2.0066(1), aN = 0.748 mT, aH= 0.022 mT (12 H, -CH3), aH = 0.044 mT and 0.041 mT (2H, 6’ and 4’). 7.2.9. Synthesis of 6-bromo-5'[1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'- bipyridine (9) N Br N N N O 6-Bromo-5'[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'-bipyridine 7 (100 mg, 0.23 mmol) was oxidized under phase transfer conditions in water/chloroform mixture (20mL, 1/3) 134 Chapter 7-Experimental Session ___________________________________________________________________________________________ using excess of NaIO4 (242 mg, 1.63 mmol), previously dissolved in 15 mL of water, for 60 min and upon warming the reaction mixture up to 40°C. During this period of time the organic phase became bright red. The phases were separated and the aqueous layer was extracted with chloroform (3 x 20 mL). The organic phases were collected and evaporated under reduced pressure. The crude red solid was dissolved in acetone and purified by chromatographic separation on neutral alumina, using as eluents acetone /petroleum ether (low boiling point, 1/3) mixture. The red fraction eluted (Rf = 0.67) was collected and upon drying it gave pure 6-bromo-5'[1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl]-2,2'-bipyridine 9 (24 mg, yield 28 %) as bright red solid very hygroscopic. M.p.147 – 148°C (from acetone). FT-IR (KBr pellet, ν/cm-1) 3072 (w, νC-H), 2964 (s, νC-Hal), 2923 (s, νC-H), 2855 (s, νC-H), 1594 (m), 1544 (s), 1455 (m), 1428 (s), 1373 (s, N-O), 1261 (m), 1155 (s), 1124 (s) (pyridine and pyrazolyl- moieties). UV/Vis, λ/nm (toluene) (ε, M-1 × cm-1) 308 (19086), 343 (3222), 392 (282), 438 (278), 464 (321), 493 (287), 534 (157). Elemental analyses, found C 47.43, H 5.85, N, 13.00%. C17H18BrN4O × 3 H2O (428.30) required C 47.67, H 5.65, N 13.08. EPR (298 K, 9.399947 GHz, 10-4 M in toluene): seven lines, giso = 2.0061(1), aN1 = 0.885 mT, aN2= 0.430 mT. 7.2.10. Synthesis of 5, 5"-diformyl-2,2':6',2" terpyridine (10) N H N N H O O 2-Tributylstannyl-5-[1,3]dioxolan-2-yl-pyridine 3 (3.8 mL, 4.7 mmol) was placed in a two- necked round bottomed flask together with the 6’-bromo-[2,2]’-bipyridinyl-5-carbaldehyde 5 (900 mg, 3.4 mmol). The mixture was degassed and kept under rigorous argon atmosphere. Then dry and degassed toluene (50 ml) was added with a syringe together with dichlorobis(triphenylphosphine)-palladium(II) (165 mg, 0.235 mmol) and triphenylphosphine (123 mg, 0.47 mmol) as catalyst. The solution was heated to reflux in argon under stirring for 60 hours. The resulting dark solution was washed with a saturated solution of ammonium- chloride (20 ml). The mixture was shaken vigorously in a separator funnel. The phases were separated and the aqueous layer was extracted with toluene (2 × 15 mL). The combined dark yellow organic layers were collected and the solvent evaporated under reduced pressure. The oily solution was then treated with HCl (20 mL, 6N) and was heated to reflux for 6 hours under stirring. This ensures for the complete hydrolysis of the dioxolane. The mixture was treated 135 Chapter 7-Experimental Session ___________________________________________________________________________________________ with CH2Cl2 (10 mL) and the aqueous phase was collected. Basification with a saturated solution of K2CO3 shows the formation of a white flocculate at pH 7-8 that consists mainly of 5,5"-diformyl-2,2':6',2"-terpyridine. The crude product was extracted with dichloromethane (3 × 40 ml). The organic layers were collected and the solvent evaporated till small volume under reduced pressure. The mixture was separated by column chromatography (silica gel, ethyl- acetate/dichloromethane/hexane, 1/4/4). The fraction eluted (colorless) was 5,5"-diformyl- 2,2':6',2"-terpyridine (Rf = 0.2) as pale yellow powder. The product was further washed with small portions of light petroleum ether (2 × 10 mL) (b.p. 30 - 40°C) and it was collected as highly pure white crystalline powder (720 mg, 73 %). M.p. 246-247 °C. FT-IR (KBr) ν/cm-1: 3036 (w, νC-H), 2961 (w, νC-H), 2877 (νC-H), 1693 (s, νC=O), 1591(s, pyr), 1560 (s, pyr), 1482 (w, pyr), 1450 (w, pyr), 1371 (s), 1264 (m, pyr), 1205 (m, pyr). 1H NMR (DMSO d6, 250 MHz, 298 K, 64 scan) δ (ppm): 10.19 (s, 2H, -CHO), 9.22 (s, 2H, H-6, H-6''), 8.84 (d, J = 8.2 Hz, 2H, H-3, H-3''), 8.62 (d, 3J = 7.9 Hz, 2H, H-3', H-5'), 8.44 (dd, 3J = 2.2, 8,2 Hz, 2H, H-4, H-4''), 8.24 (t, 3J = 7.9 Hz, 1H, H-4'). 13C-NMR (DMSO-d6, 63 MHz, 298 K, 10000 scan), δ (ppm): 192.5, 159.3, 154.3, 151.9, 139.5, 137.7, 131.7, 123.2, 121.5. MS-FD (70eV, CH2Cl2) 290.4 (100%), (MW+H calc 290.4). UV/Vis (CHCl3) λ/nm (ε, mol-1 × cm-1) 321 nm (31800), 312 nm (30870), 256 nm (23020). Elemental analyses, found C 70.42, H 3.92, N 14.40%. C17H11N3O2 (289.29) required C 70.58, H 3.83, N 14.53 %. 7.2.11. Synthesis of 5,5"-bis(1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"- terpyridine (11) HO N OH N N N N N N OH HO 5,5"-Diformyl-2,2':6',2"-terpyridine 10 (590 mg, 2.04 mmol) was placed in a round bottomed flask together with 2,3-bis-hydroxylamino-2,3-dimethylbutane 6 (906 mg, 6.12 mmol). A mixture of 1,4-dioxane, trichloromethane (CHCl3) and methanol (20 mL /15 mL /15 mL) was used as reaction solvents due to the low solubility of the diformyl-derivative. The mixture was then stirred for 7 days under argon at room temperature. The white precipitate 5,5"-bis(1,3- dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"-terpyridine was filtered, washed with small portions of cold methanol/acetone (4/1, 3 × 5 mL) and dried in high-vacuum. The radical precursor was then recovered as fine white powder, and was used without further purification (960 mg, yield 86%). 136 Chapter 7-Experimental Session ___________________________________________________________________________________________ It decomposes before melting at 186 – 187 °C. 1H NMR (DMSO d6, 250 MHz, 298 K, 64 scan) δ (ppm): 8.76 (s, 2H, H-6, H-6''), 8.57 (d, 3J = 7.8 Hz, 2H, H-3, H-3''), 8.40 (d, 3J = 7.5 Hz, 2H, H-3', H-5'), 8.05 (m, 3J = 7.8, 8.2, 6.7 Hz, 3H, H-4, H-4', H-4''), 7.94 (s, 4H, OH), 4.66 (s, 2H, - CH), 1.09 (d, 3J = 6.6 Hz, 24H, -CH3). 13C NMR (DMSO- d6, 63 MHz, 298 K, 8000 scan) δ (ppm): 154.5, 153.9, 148.9, 137.2, 136.6, 129.1, 120.0, 119.4, 87.5, 65.8, 23.9, 16.8. MS-FAB (NBA matrix) m/z 549.4 (100%)(M+H)+, calculated (MW) 549.66. UV/Vis (DMSO) λ/nm (ε, mol-1× cm-1) 290 nm (11900). FT-IR (KBr) ν/cm-1 = 3252 (s, broad, νOH), 2988 (s,νC-H), 2930 (s, νC-H), 1596 (m, pyr), 1560 (m, pyr), 1374 (s), 1262 (s, pyr). Elemental analyses, found C 59.40, H 7.55, N 16.63%, C/N = 3.57. C29H39N7O4 × 2H2O required C 59.50, H 7.35, N 16.75%, C/N = 3.55. 7.2.12. Synthesis of 5,5"-bis(1-oxyl-3-oxo-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"- terpyridine (12) O- -+ N O N N N +N N N O O Compound 11 (217 mg, 0.39 mmol) was charged into a small flask together with 20 mL of CHCl3 and CH2Cl2 (4/1). Then it was degassed and kept under argon while stirring at room temperature. Separately, a solution of NaIO4 (211 mg, 0.987 mmol) dissolved in 10 mL of H2O was first degassed and saturated by argon exchange; then it was added to the solution of the radical precursor by using a syringe under argon. After 30 min a deep green solution was obtained. The organic layer was extracted by using portions of CHCl3 (3 × 10 mL) until the aqueous phase was almost colorless. The organic solution was evaporated by continuous argon bubbling until a small volume of solvent was left (~2 mL). The mixture was separated by column chromatography (Aluminium oxide, acetone/light petroleum ether, 3/7). The 5,5"-bis(1- oxyl-3-oxo-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"-terpyridine 12 (Rf = 0.23) was obtained after solvent evaporation (by argon bubbling) as deep green powder (35 mg, yield 16 %). FT-IR (KBr) ν/cm-1 = 3066 (w, νC-H), 2990 (m, νC-H), 2924 (m, νC-H), 2853 (m, νC-H), 1591 (m, pyr), 1482 (m, pyr), 1452 (m, pyr), 1387 (m), 1351 (s, νN-O), 1216 (m, pyr). MS-FAB (NBA matrix) m/z 544.6 (100%) [M+H]+, calculated (MW+H, 544.66). UV/Vis (CHCl3) λ/nm (ε, mol-1× cm-1) 718 (145), 650 (425), 605 nm (480), 562 (350), 387 nm (13100), 342 nm (28300), 328 nm (33400), 282 nm (31550). Elemental analyses, found C 59.96, H 6.50, N 16.82%. 137 Chapter 7-Experimental Session ___________________________________________________________________________________________ C29H33N7O4 (543.62) × 2 H2O required C 60.09, H 6.43, N 16.91%. EPR (260 K, 9.400220 GHz, toluene 10-4 M): 9 lines, giso =2.0066(1), aN = 0.374 mT. 7.2.13. Synthesis 5,5"-bis(1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"- terpyridine (13) N N N N N N N O O Compound 11 (240 mg, 0.44 mmol) was charged into a small flask together with 20 mL of a mixture of CHCl3 and CH2Cl2 (4/1), under stirring at room temperature. A solution of NaIO4 (377 mg, 1.76 mmol, 10 mL H2O) was slowly added to the solution of the radical presursor and then the mixture was slightly warmed at 40 °C. After 30 min an orange-red solution was obtained. The organic layer was extracted by using portions of CHCl3 (3 × 10 mL). The organic solution was evaporated under reduced pressure up to small volume (1 mL) . The mixture was separated by column chromatography (Aluminium oxide, acetone/light petroleum- ether low boiling point, 3/7). The 5,5"-bis(1-oxyl-4,4,5,5-tetramethylimidazolidin-2-yl)2,2':6',2"- terpyridine 13 (Rf = 0.25) was obtained after solvent evaporation as orange powder (63 mg, yield 28 %). M.p. 139-140 decompose by changing color from bright red to pale yellow and melt to dark mass at 239-240°C. FT-IR (KBR) ν/cm-1 = 3070 (w, νC-H), 2957 (s, νC-H), 2924 (s, νC-H), 2870 (m, νC-H), 1596 (m, pyr), 1554 (s, C=N), 1469 (s, pyr), 1448 (s, pyr) 1425 (s), 1378 (s, νN-O), 1262 (s, pyr), 1216 (s, pyr). MS-FAB (NBA matrix) m/z 512.2 (100%) [M+H]+, calculated MW 511.6. UV/Vis (CHCl3) λ/nm (ε, mol-1×cm-1) 528 (170), 490 (315), 459 nm (380), 300 nm (16200), 257 nm (24100). Elemental analyses, found C 63.36, H 6.91, N 17.74%. C29H33N7O2 (511.62) × 2 H2O required C 63.60, H 6.81, N 17.90%. EPR (260 K, 9.402404 GHz, toluene 10-4 M): 13 lines, giso =2.0061(1), aN1 = 0.430 mT, aN2 = 0.225 mT. 138 Chapter 7-Experimental Session ___________________________________________________________________________________________ 7.2.14. Synthesis of triformylmethane (14) CHO H CHO CHO To a cooled solution of DMF (C3H7NO, 40mL, 0.518 mol, at ~ 0°C), POCl3 (13.8 mL, 0.147 mol) was added drop wise over a period of 1 hour keeping the temperature below 5°C. The solution appeared initially greenish and turned at the end of the addition into pale orange. Then the ice-bath was removed and the dense mixture was left under stirring at room temperature for one hour. The bromoacetic acid was added in small portions (7.15 g, 0.051 mol) and the mixture was heated for 24 hours at 70 °C. The brownish mixture was decomposed with ice/water (200 mL) and solid Na2CO3 was carefully added in excess (till pH ~ 8). To this dense solution, absolute ethanol was added (2 L) and the inorganic salts were filtered off. The organic filtrate was evaporated slowly and the pale yellowish residue left was neutralized with H2SO4 (50%, 10 mL), extracted with CHCl3 (3 × 200 mL) and dried over MgSO4. After solvent removal, the triformylmethane was obtained as yellowish crystals that were further purified by sublimation (2.3 g, yield 45% calculated based on bromoacetic acid). M.p. 102-103°C (lit. M.p, 101°C). FT-IR (KBr) ν/cm-1: 2828 (w, νC-H), 2731 (m, νC-H), 1699 (s, C=O), 1599(s), 1566 (s), 1561 (m), 1207 (s), 1166 (m), 826 (s), 723 (m). 1H NMR (CDCl3, 250 MHz, 298 K, 32 scan) δ (ppm): 9.92 (s, 1H, -CHO), 8.13 (s, 2H). Elemental analyses, found C 48.25, H 4.11%; C4H4O3 (MW 100.07) required C 48.01, H 4.03%. 7.2.15. Synthesis of 4-formyl-1(H)-pyrazole (Method A) (15) H N O N H To a mixture of hydrazine-monohydrate (0.33 g, 6.6 mmol) dissolved in methanol (20 mL) a solution of HCl was added (6N, 3 mL). Separately a solution of triformylmethane 14 (0.6 g, 6 mmol) dissolved in methanol (20 mL) was prepared, and then the acidified methanolic solution of hydrazine was added drop wise very slowly over 3 hours. The solution was stirred at room temperature for other 20 hours and then was air dried. The yellowish residue dissolved in water (20 mL) was neutralised with solid NaHCO3. Then the crude product was extracted with ethylacetate (3 × 20 mL). The organic phase was collected and the solvent extracted under reduced pressure. The crude mixture was chromatographed over silica (hexane/ethylacetate 139 Chapter 7-Experimental Session ___________________________________________________________________________________________ mixture, 2/1). The 4-Formyl-1(H)-pyrazole (15) (MW 96.09) was collected (Rf = 0.2) and upon drying, it gave 0.58 g of pure compounds in the form of yellowish powder very hygroscopic (yield 52%). M.p. 83-84°C (from ethylacetate). 1H NMR (250 MHz, CDCl3, 298 K, 16 scan ) δ (ppm): 9.92 (s, 1H, 1-CHO), 8.13 (s, 2H, -CH). 1H NMR (250 MHz, DMSO-d6, 298 K, 16 scan): δ (ppm): 13.56 (broad, s, -NH), 9.83 (s, 1H, 1-CHO), 8.47 (s, broad, 1H, -CH), 7.99 (s, broad, 1H, -CH). 13C NMR (DMSO-d6, 63 MHz, 298 K, 256 scan) δ (ppm), 185.4, 139.6, 134.1, 123.8. 13C NMR (CDCl3, 63 MHz, 298 K, 2000 scan) δ (ppm), 182.8, 134.6, 122.21. FT-IR (KBr) ν/cm-1: 3176 - 2790 (strong and broad band with several components), 1689 (vs, C=O), 1650 (s), 1558(s), 1509 (s), 1413 (s), 1386 (s), 1213 - 609 (several strong absorptions). Elemental analyses, found C 49.96, H 4.33, N 29.00. C4H4N2O (96.09) required C 50.00, H 4.20, N 29.15%. 7.2.16. Synthesis of 4-formyl-1(H)-pyrazole (Method B) The following methodology represents an alternative route to the synthesis of 4-formyl-1(H)- pyrazole. It relies on three synthetic steps: iodination of pyrazole, N-H protection, Grignard reaction followed by N-H deprotection. Synthesis of 4-iodo-pyrazole (16): H N N I Pyrazole (3.4 g, 50 mmol) was dissolved in acetic acid (30 mL) and left under stirring. Separately, a solution containing HIO3 (1.8 g, 10 mmol), I2 (5.1 g, 20 mmol), H2SO4 (2 mL, 30%) and acetic acid (15 mL) was prepared, leading to a deep red-violet mixture. The pyrazole solution was heated up to 60°C under argon, and then the iodine mixture was slowly added, taking care that before any further addition, all the iodine was consumed. When half of the addition was completed, the rest of the iodine was dropped fully, and the pyrazole solution was left to react for 60 min. Then, a saturated solution of NaHCO3 (15 mL) was added in order to smoothly start the quenching of the acetic acid. A saturated solution of Na2CO3 was finally added slowly until no evolution of CO2 was observed and a fine white flocculate was formed. The flocculate was extracted with CHCl3 (3 × 30 mL), the organic layer dried over MgSO4, filtered and left to crystallise under air. Finally 8.6 g as white needles of 4-iodo-pyrazole have been obtained (yield 88%). M.p. 107-108°C (from CHCl3) (literature M.p. 108.5°C). 1H NMR (250 MHz, CDCl3, 298 K, 16 scan), δ (ppm): 9.51 (s, broad, 1H, -NH), 7.62 (s, asymmetric, 2H, 2-CH). 1H NMR (250 MHz, 140 Chapter 7-Experimental Session ___________________________________________________________________________________________ DMSO-d6, 298 K, 16 scan), δ (ppm): 13.20 (s, broad, 1H, -NH), 7.90 (s, broad, 1H, -CH), 7.62 (s, broad, 1H, -CH). 13C NMR (DMSO-d6, 63 MHz, 256 scan), δ (ppm): 56.9, 133.6, 143.8. FT- IR (KBr): ν/cm-1 = 3114 – 2788 (m, several broad vibronic band), 1624 (m), 1537 (m), 1475 (m), 1452 (m), 1365 (s), 1328 (m), 1267 (m), 1178 (m), 1141 (m), 1033 (s), 954 (s), 937 (vs), 871 (s), 810 (vs), 609 (vs). FD-MS (8 kV, CHCl3) m/z: found 194.1 (100%), requires for C9H7BrN3O (MW+H+) 193.97. Elemental analyses, found C 18.40, H 1.62, N 14.32. C3H3N2I (193.97) required C 18.58, H 1.56, N 14.44%. 7.2.17. Synthesis of 1-(1-ethoxyethyl)-4-Iodo-pyrazole (17) (Method B): N N I O Into a 4-iodo-pyrazole 16 (3.0 g, 15.47 mmol, 1 eq.) solution, previously dissolved in 20 mL of benzene, were added 3 drops of HCl (33%) and ethylvinylether (ethoxyethene, 2.0 mL, d = 0.754 g/mL, 21 mmol, 1.36 eq.) and left under stirring for 6 hours at 45-50°C. A bright yellow- orange solution was slowly formed. The solution was cooled to room temperature and neutralised with a saturated solution of NaHCO3 in water. Then the organic phase was separated and the water phase extracted with portions of benzene (3 × 10 mL). The collected organic portions were dried over MgSO4 and reduced to small volume. The crude mixture was chromatographed on Al2O3 column using a mixture of hexane/CHCl3/etylacetate (2/4/1). The first fraction eluted (Rf = 0.75) was concentrated under reduced pressure (60°C, 20 mbar) and 1-(1-ethoxyethyl)-4-iodo-pyrazole (17) was collected highly pure (3.9 g, 4.2 mL, 3.49 mmol/mL, yield 94.7 %) as very pale yellowish oil. 1H NMR (250 MHz, CDCl3, 298 K, 16 scan), δ (ppm): 7.61 (s, 1H, -CH), 7.47 (s, 1H, -CH), 5.49-5.42 (q, 3J = 6 Hz, 2H, -CH), 3.48-3.23 (m, 2H, -CH ), 1.60 (d, 32 J = 6 Hz, 3H, -CH3), 1.11 (t, 3J = 7 Hz, 3H, -CH ). 133 C NMR (CDCl3, 63 MHz, 298 K, 256 scan), δ (ppm): 14.7, 22.1, 57.3, 64.2, 87.9, 130.6, 143.7. FD-MS (8 kV, CH2Cl2) m/z: found 266.1 (100%), requires for C7H11IN2O (MW+H+) 266.08. 7.2.18. Synthesis of 4-formyl-1(H)-pyrazole (18) (Method B) H N O N H 141 Chapter 7-Experimental Session ___________________________________________________________________________________________ In a small round bottomed flask (20 mL) closed at the top with a rubber cup, it was placed 1- (1-ethoxyethyl)-4-iodo-pyrazole (1.5 mL, 5.2 mmol) and THF (6 mL), then the solution was cooled in ice bath (0-4°C), carefully evacuated and kept under argon. A cold solution of EtMgBr (Grignard reagent, CH3CH2MgBr, 1.9 mL, 5.7 mmol) was slowly added drop by drop with a syringe through the rubber septum into the cold solution of iodo-pyrazole within 5 min. A milky solution that became a solid paste was slowly formed due to the low solubility of the pyrazolyl-magnesiate derivative. The solution was then aged under stirring for 60 min. Dry DMF (0.5 mL, ρ = 0.946 g/mL, 6.5 mmol) was added drop by drop while keeping the temperature below 4°C, and left under stirring for 60 min. Then the ice bath was removed and left at room temperature for 30 min. A pale yellowish solution very dense was formed. A saturated solution of NH4Cl (5 mL) was added, left under stirring for 20 min, and the organic layer collected, while the water phase was extracted with portions of CHCl3 (3 × 10 mL). The organic phases were collected and allowed to dry under stream of air affording a pale yellowish oil. This oil, that contained the aldehyde still protected with the ethoxyethyl-group (TLC SiO2, ethylacetate/CHCl3/ hexane, 1/4/2, Rf = 0.5-0.7), was placed in a flask together with dioxane (10 mL) and HCl ( 10 mL, 20% in water) and left under stirring at 60 °C overnight. The solution was neutralised with a saturated solution of K2CO3 in water, and the water phase extracted with ethylacetate (3 × 10 mL). The organic phases were collected, reduced to a small volume and chromatographed over a small SiO2 column (ethylacetate/ hexane, 1/2), and pure 4-formyl-1(H)-pyrazole 18 (Rf = 0.2) as been collected as very sticky yellowish powder (0.4 g, yield 80%). 7.2.19. Synthesis of 2,6-bis(4'-formylpyrazol-1'-yl)-pyridine (19) H N N N H O N N O To a solution of 4-formylpyrazole 18 (1 g, 10 mmol) dissolved in dry diglyme (30 mL, b.p. 162 °C), potassium metal was added (0.4 g, 10 mmol) then the mixture was heated under argon at 70 °C under stirring until all of the potassium was reacted to form the pyrazolate-derivative. Then, 2,6-dibromopyridine (1.18 g, 5 mmol) was added in portions, and the reaction mixture was heated at 110 °C for further 72 h under argon. The solid product formed upon cooling to room temperature the reaction mixture was filtered off, washed first with cold water, and then with small portions of ethanol. Finally the residue was air dried. Recrystallisation from toluene 142 Chapter 7-Experimental Session ___________________________________________________________________________________________ afforded 2,6-bis(4'-formylpyrazol-1'-yl)-pyridine 19 (0.6 g, yield 45%) as pale yellowish crystals. M.p. 283-284°C (from toluene). FT-IR (KBr): ν/cm-1 = 3129 (w, νC-H, aromatic), 3097 (w, νC-H, aromatic), 2857 (w, νC-H), 1682 (s, νC=O), 1610 (m), 1583 (m), 1552 (s), 1473 (s), 1409 (m), 1336 (w), 1290 (w), 1217 (m) (pyridinyl- and pyrazolyl- moieties). 1H NMR (250 MHz, DMSO- d6, 298 K, 32 scan): δ = 9.97 (s, 2H, 2-CHO), 9.73 (s, 2H, 2-CH), 8.34 (s, 2H, 2-CH), 8.25 (t, 1H, 3J = 7.2 Hz, -CH), 7.95 (d, 2H, 3J = 7.9 Hz, 2-CH). 13C NMR (63 MHz, DMSO-d6, 298 K, 8000 scan): δ = 185.3, 148.9, 143.0, 142.4, 133.1, 125.8, 111.3. FD-MS (8 kV, CH2Cl2): m/z (%) = found 266.90 (100%), calculated for C13H9N5O2 (267.24). Elemental analyses, found C 58.28, H 3.56, N, 26.10%. C13H9N5O2 (267.24) required C 58.43, H 3.39, N 26.21. Bispyrazolylpyridine derivatives 7.2.20. Synthesis of 2,6-bis-pyrazol-1-yl-pyridine (20) N N N N N Pyrazole (5.5 g, 0.081 mol) dissolved in dry diethylene glycol dimethyl ether (diglyme, 100 mL) was stirred under argon with potassium metal (3 g, 0.0765 mol) and heated up to 50°C until all the potassium was consumed. Then, 2,6-dibromopyridine (5.9 g, 0.0248 mol) was added with a TLC cannula and the reaction mixture heated at 140 °C for four days, keeping 1 rigorous argon atmosphere. The mixture was cooled to room temperature 2 and the solvent evaporated under air stream. The pale yellowish residue was pored into cold water (4°C) and stirred for 10 min and filtered. The 3 residue left consisted of the product, the monocoupled 2-bromo-6-pyrazol- 1-yl-pyridine and unreacted 2,6-di-bromopyridine. The crude mixture was purified by column chromatography using a mixture of CHCl3/hexane/etehylacetate (6/2/1). The 2,6-di-bromopyridine (1) was eluted first, then 2- bromo-6-pyrazol-1-yl-pyridine (2) (Rf = 0.8), and finally 2,6-bis-pyrazol-1-yl-pyridine (3) (Rf range 0.1 - 0.5) was collected (see TLC, SiO2). After solvent removal and recrystallisation from methanol/ethylacetate (1/1), 2.6 g of the product were obtained as white crystalline plates (yield 49.6%, M.p. 137-138°C). The monocoupled 2-bromo-6-pyrazol-1-yl-pyridine (2.2 g, yield 39%) was recrystallised from CHCl3 as white crystalline needles. 143 Chapter 7-Experimental Session ___________________________________________________________________________________________ 2-Bromo-6-pyrazol-1-yl-pyridine : 1H-NMR (250 MHz, CDCl3, 298 K, 16 scan),δ (ppm): 8.52- 8.50 (dd, 3J = 3 Hz, 0.7 Hz, 1H, -CH), 7.92-7.88 (dd, 3J = 8.5 Hz, 1 Hz, 1H, -CH), 7.71 (d, 3J = 1 Hz, 1 H, -CH), 7.63 (t, 3J = 7.9 Hz, 1H, -CH), 7.35-7.31 (dd, 3J = 6.9 Hz, 0.7 Hz, 1H, -CH), 6.45-6.43 (m, 1H, -CH). FT-IR (KBr): ν/cm-1 = 3160 (w), 3082 (w), 3048 (w), 2923 (w), 2852 (w), 1588 (s), 1565 (s), 1465 (s), 1398 (s), 1208 (s), 1116 (vs), 1035 (s), 985 (s), 940 (vs), 795 (vs), 759 (vs), 651 (s), 621 (s). FD-MS (70 eV, CHCl3): m/z = 224.1 (100%), calculated MW 224.06. Elemental analyses, found C 43.00, H 2.85, N 18.60%. C8H6N3Br (224.06) required C 42.88, H 2.70, N 18.75%. 2,6-Bis-pyrazol-1-yl-pyridine: 1H-NMR (250 MHz, CDCl3, 298 K, 64 scan) δ (ppm): 8.55 (s, 2H, -CH), 7.95 – 7.81 (m, 3H, -CH), 7.74 (s, 2H, -CH), 6.49 – 6.47 (dd, 3J = 1.90 Hz, 0.95 Hz, 2H, -CH). 13C-NMR (CDCl3, 63 MHz, 298 K, 256 scan), δ (ppm): 148.7, 141.1, 140.1, 125.7, 108.1, 106.7. FT-IR (KBr): ν/cm-1 = 3162 (w), 3103 (w), 1607 (s), 1582 (s), 1526 (s), 1480 (vs), 1459 (s), 1391 (vs), 1334 (m), 1216 (m), 1153 (m), 1127 (m), 1127 (m), 1072 (m), 1035 (vs), 950 (s), 935 (vs), 807 (vs), 758 (vs), 607 (m). MS-FD (70 eV, CH2Cl2): m/z = 211.3 (100%), calculated MW 211.22. UV/Vis, λmax (CH2Cl2)/ nm (ε, mol-1 × cm-1): 312 (7830), 272 (6600). Elemental analyses, found C 62.25, H 4.40, N 32.92%. C11H9N5 (211.22) required C 62.55, H 4.29, N 33.16%. 7.2.21. Synthesis of 2,6-bis-(4-Iodo-pyrazol-1-yl)-pyridine (21) I N N N I N N 2,6-Bis-pyrazol-1-yl-pyridine 20 (0.3 g, 1.42 mmol) was charged in a round bottomed flask with acetic acid (4 mL), H2SO4 (30% in water, 0.5 mL) and heated up to 60°C under argon. Separately, a deep CH3COOH solution (10 mL) containing HIO3 (0.1 g, 0.57 mmol), I2 (0.29 g, 1.14 mmol) and two drops of concentrated H2SO4 was slowly added into the 2,6-bis-pyrazol-1- yl-pyridine solution in such way that the iodine was consumed before adding the next drop. Only when half of this solution (5 mL) was added, the remaining 5 mL were dropped into the reaction mixture that appeared heterogeneous due to the presence of a white flocculate. Then, the solution was left at 60°C for further 3h under TLC argon. After cooling to room temperature, Na2S2O3 was added just in 2 enough amounts to quench the pale rose solution. A NaHCO3/Na2CO3 (1/1) 3 1 water solution was added until the pH reached neutrality (pH 7-8) and the product 2,6-bis-(4-iodo-pyrazol-1-yl)-pyridine 21 was extracted with portions of CHCl3, air dried and recrystallised from benzene to afford white crystalline 144 Chapter 7-Experimental Session ___________________________________________________________________________________________ needles (0.64 g, yield 97%). The product obtained was controlled by TLC using a mixture of CHCl3/hexane/ethylacetate (4/2/1) as shown in the figure, where 1 represents the starting material (Rf = 0.6), 2 the bis-iodo derivative, and 3 a very small amount of, presumably, mono- iodo derivative since it has Rf exactly in-between with respect to 1 and 2. When the TLC was run in presence of 5% of Ethyl3N, 3 was not found, therefore part of the bis-iodo-derivative seemed to decompose on silica. No starting material was left in the collected crystals. The reaction can be scaled up, upon working in more diluted solution using the following proportions: 2,6-Bis-pyrazol-1-yl-pyridine (1.2 g, 5.68 mmol) dissolved in acetic acid/H2O/H2SO4 (16 mL / 2 mL / 2 mL 30%), then HIO3 / I2 (0.408 g, 1.16 g) dissolved in acetic acid/H2SO4 (60 mL / 2 drops) and heated at 70°C for 1 h. The yield is however lower (2.2 g, 82%). M.p. 188 – 189°C. MS-FD (CHCl3, 70 eV) m/z = 463.1 (100%), calculated MW 463.02. FT-IR (KBr): ν/cm-1 = 3149 (w), 3096 (w), 2874 (w), 1611 (m), 1586 (s), 1514 (m), 1466 (vs), 1422 (m), 1372 (s), 1314 (m), 1198 (m), 1145 (m), 963 (m), 950 (s), 801 (s), 601 (s). 1H-NMR (250 MHz, CDCl3, 298 K, 16 scan) δ (ppm): 8.57 (s, 2H, -CH), 7.96 – 7.90 (dd, 3J = 6.95 Hz, 1.90 Hz, 1H, -CH), 7.81 – 7.78 (dasymmetric, 3J = 7.27 Hz, 2H, -CH), 7.72 (s, 2H, -CH). 13C-NMR (CDCl3, 63 MHz, 298 K, 256 scan), δ (ppm): 150.0, 148.1, 142.7, 132.4, 110.4, 61.2. UV/Vis, λmax (CH2Cl2)/ nm (ε, mol-1 × cm-1): 262 (17020), 279 (sh, 13250), 286 (14590), 314 (23650). Elemental analyses, found C 28.74, H 1.65, N 14.98%. C11H7I2N5 (463.02) required C 28.53, H 1.52, N 15.13%. 7.2.22. Synthesis of 2,6-bis-(4-bromo-pyrazol-1-yl)-pyridine (22) Br N N N Br N N 2,6-Bis-pyrazol-1-yl-pyridine 20 (1.0 g, 4.73 mmol) was charged in a round bottomed flask with acetic acid (15 mL), H2SO4 (10% in water, 2.0 mL) and heated up to 60°C. Separately, a solution of Br2 (d = 3.11 g/mL, MW 159.82, 0.364 mL, 7.102 mmol) dissolved in acetic acid (10 mL) was slowly added into the 2,6-bis-pyrazol-1-yl-pyridine TLC solution drop by drop. When half of this solution (5 mL) was added, a dense orange-yellowish solution was formed, showing the presence of a white flocculate. Then, the Br2 solution was added faster, and the temperature rose up to 85 °C. The mixture left under stirring for 1 h. After cooling to room temperature the mixture was quenched with NaHCO3 and then Na2CO3 water 145 Chapter 7-Experimental Session ___________________________________________________________________________________________ solutions until pH ~ 9 was reached. Na2S2O3 was added just in enough amounts to destroy the Bromine left. The white precipitate formed was extracted with CHCl3 and the solvent evaporated under reduced pressure. The 2,6-bis-(4-bromo-pyrazol-1-yl)-pyridine 22 was purified by column chromatography (SiO2) using a mixture of ethylacetate/CHCl3/hexane (1/1/6) (Rf = 0.6). Recrystallisation from ethylacetate/hexane mixture (1/1) afforded 1.0 g of pure product (yield 57%). FT-IR (KBr): ν/cm-1 = 3157 (w), 3097 (w), 2954 (w), 2923 (ws, νCH aliphatic), 2852 (w), 1735 (m), 1612 (vs), 1587 (vs), 1525 (m), 1471 (vs), 1428 (s), 1378 (vs), 1325 (s), 1273 (m), 1198 (ms), 1145 (s), 1032 (ms), 958 (vs), 848 (ms), 796 (vs), 775 (ms), 646 (m), 598 (s). 1H-NMR (250 MHz, CDCl3, 298 K, 16 scan) δ (ppm): 8.55 (s, 2H, -CH), 7.98 – 7.91 (dd, 3J = 6.95 Hz, 2.21 Hz, 1H, -CH), 7.84 – 7.80 (d 3 13asymmetric, J = 8.53 Hz, 2H, -CH), 7.68 (s, 2H, -CH). C-NMR (CDCl3, 63 MHz, 298 K, 256 scan), δ (ppm): 147.6, 141.3, 140.1, 125.5, 107.8, 95.2. MS-FD (CHCl3, 70 eV) m/z = 369.2 (100%), calculated MW 369.01. Elemental analyses, found C 35.64, H 2.02, N 18.74%. C11H7Br2N5 (369.01) required C 35.80, H 1.91, N 18.98%. 7.2.23. Synthesis of 2,6-bis-(4-trimethylsilamylethynyl-pyrazol-1-yl)-pyridine (23) Si N N N Si N N 2,6-Bis-(4-iodo-pyrazol-1-yl)-pyridine 21 (0.36 g, 0.78 mmol) was charged in a round bottomed flask with anhydrous triethylamine (Et3N, 10 mL) and dioxane (2 mL), together with dichlorobis(triphenylphosphine)-palladium(II) (55 mg, MW 701.89, 0.078 mmol), triphenylphosphine (41 mg, MW 262.28, 0.156 mmol) and CuI (20 TLC mg, 0.1 mmol) as catalyst. The solution was further evacuated and left under argon. A solution of trimethylsilylacetylene (0.71 g/mL, MW 98.22, 0.33 mL, 2.35 mmol) was added with a syringe under argon, and the resulting mixture was heated up to 80°C for 30 min until it became dark green/black. Then, the mixture was cooled at room temperature and left further under stirring for an additional hour. The solution was partially 5% Et3N neutralized with HCl (20% in water) (the pH > 7) and the product extracted with portions of CH2Cl2 (3 × 20 mL). The orange-yellow organic phase was collected, and it was washed further with a saturated solution of NH4Cl. The organic phase was collected for the second time, dried over MgSO4, filtered from the inorganic salt and the CH2Cl2 evaporated under air stream affording a pale yellowish powder. The product 2,6-bis-(4- trimethylsilamylethynyl-pyrazol-1-yl)-pyridine was purified on silica column (SiO2) using a 146 Chapter 7-Experimental Session ___________________________________________________________________________________________ mixture of CHCl3/hexane (5/3) in presence of Et3N (5%) (Rf = 0.55) since showed some hint of decomposition due to the acidity of the silica gel. After collection of the fractions containing the product (see TLC) and recrystallization from Et2O, the 2,6-bis-(4-trimethylsilamylethynyl- pyrazol-1-yl)-pyridine 23 was obtained as fine yellowish powder (285 mg, yield 91 %). The product can be stored at room temperature for long time without decomposition. M.p. 132 – 133°C (decompose before melting from yellow to dark brown powder). FT-IR (KBr): ν/cm-1 = 3139 (w), 3056 (w), 2958 (s, νCH aliphatic), 2896 (w), 2164 (vs, C≡C−Si[CH3]3), 1608 (s), 1582 (s), 1554 (m), 1471 (vs), 1435 (s), 1402 (s), 1348 (m), 1249 (s), 1180 (m), 1010 (vs), 966 (m), 952 (m), 858 (vs), 800 (s), 759 (m), 657 (m). 1H-NMR (250 MHz, DMSO- d , 298 K, 16 scan) δ (ppm): 9.47 (s, 2H, -CH), 8.16 (t, 36 J = 8.2 Hz, 1H, -CH), 8.04 (s, 2H, - CH), 7.83 – 7.80 (d, 3J = 7.9 Hz, 2H, -CH), 0.23 (s, 18 H, -CH3). 13C-NMR (DMSO-d6, 298 K, 1024 scan), δ (ppm): 148.7, 144.7, 143.0, 131.4, 128.7, 109.7, 105.1, 96.4, -0.1. Elemental analyses found C 62.16, H 6.42, N 17.10%. C21H25Si2N5 (403.63) required C 62.49, H 6.24, N 17.35%. 7.2.24. Synthesis of 2,6-bis-(4-ethynyl-pyrazol-1-yl)-pyridine (24) N N N N N 2,6-Bis-(4-trimethylsilamylethynyl-pyrazol-1-yl)-pyridine 23 (150 mg, 0.372 mmol) was initially dissolved in MeOH (99 % dry)/THF (5mL/5mL) and kept under argon. Then solid K2CO3 (52 mg, 0.376 mmol) was added, and the mixture was left under stirring in argon for 3 hours at room temperature. Initially the solution appeared bright yellow that became very dark at the end of the reaction. The organic solution was evaporated under air stream, and the brown- yellowish residue was dissolved in CH2Cl2 (10 mL). A saturated solution of NaHCO3 (5 mL) was added and the organic phase extracted using portions of CH2Cl2 (2 × 10 mL). Collection of the organic phase followed by drying over MgSO4 and solvent evaporation, afforded a pale yellowish residue containing the product. The residue was dissolved in the minimum amount of CH2Cl2 and chromatographed on alumina column (CH2Cl2/hexane/ethylacetate, 3/2/0.5). The first fraction eluted contained the 2,6-bis-(4-trimethylsilamylethynyl-pyrazol-1-yl)-pyridine 24 as pale yellowish powder that was further recrystallised from ethylacetate (92 mg, yield 95%). M.p. > 390°C. FT-IR (KBr): ν/cm-1 = 3280 (s, ν C≡C−H), 3157 (w), 3097 (w), 2962 (w, νCH aliphatic), 2921 (w), 1604 (s), 1581 (m), 1552 (m), 1479 (vs), 1434 (m), 1400 (m), 1346 (w), 147 Chapter 7-Experimental Session ___________________________________________________________________________________________ 1261 (m), 1207 (mw), 1095 (m), 1005 (s), 952 (s), 871 (m), 798 (vs), 671 (m). 1H-NMR (250 MHz, DMSO-d6, 298 K, 16 scan) δ (ppm): 9.41 (s, 2H, -CH), 8.16 (t, 3J = 8.2 Hz, 1H, -CH), 8.04 (s, 2H, -CH), 7.84 – 7.81 (d, 3J = 8.2 Hz, 2H, -CH), 4.23 (s, 2H, ≡−CH). 13C-NMR (DMSO-d6, 298 K, 1024 scan), δ (ppm): 149.1, 145.2, 143.4, 132.1, 110.0, 104.8, 83.3, 75.1. UV/Vis, λmax (CH2Cl2)/ nm (ε, mol-1 × cm-1): 263 (9810), 279 (sh, 9080), 286 (9570), 319 (12480), 351 (sh, 4100), 379 (broad, 850). Elemental analyses found C 69.62, H 3.66, N 26.88%. C15H9N5 (259.27) required C 69.49, H 3.50, N 27.01%. Synthesis of 2,6-bis(4'-formylpyrazol-1'-yl)-pyridine by Grignard reaction (see also 19.) 2,6-Bis-(4-iodo-pyrazol-1-yl)-pyridine (300 mg, MW 463.02, 0.648 mmol) was placed in a round bottomed Schlenk with dry THF (15 mL) evacuated and kept under argon while stirring. The solution was cooled using an ice bath (0 - 4°C). Through the rubber septum, a cold solution of Grignard reagent (EtMgBr in diethylether, 1.43 mmol, 0.47 mL) was slowly added within 10 min through with a syringe under argon. Upon addition of the Grignard reagent, the reaction mixture became slowly milky and pale rose after aging for 90 min (temperature should not rise above 4°C). Then, dry DMF (0.12 mL, 1.6 mmol) was added, and left for 60 min at 4°C. The ice bath was removed and the solution was left to warm up to room temperature. A creamy solution was obtained showing the presence of a fine precipitate. The solvent (THF) was removed and the residue washed with a solution of EDTA in water (10 %) and filtered. The pale yellowish residue left that was further washed, first with cold portions of acetone (2 × 2 mL) and then CH2Cl2 (2 × 2 mL). A fine yellowish powder of 2,6-bis(4'- formylpyrazol-1'-yl)-pyridine was finally obtained (125 mg, yield 72 %). 7.2.25. Synthesis of 2,6-bis[4'-(1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2- yl)pyrazol-1'-yl]-pyridine (25) OH OH N N N N N N N N N OH OH A mixture of 2,6-bis(4'-formylpyrazol-1'-yl)-pyridine 19 (300 mg, 1.12 mmol) and 2,3- bis(hydroxylamino)-2,3-dimethylbutane 6 (600 mg, 4 mmol) in dioxan (40 mL) was stirred under argon for 10 days. The solvent was removed under reduced pressure and the solid product was collected, washed with water, ethanol and air dried to afford 265 mg of crude 2,6- bis[4'-(1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl)pyrazol-1'-yl]-pyridine 26 that was subjected to the oxidation step without further purification (yield of the crude 42%). 148 Chapter 7-Experimental Session ___________________________________________________________________________________________ M.p. 253-254°C (from ethanol). FT-IR (KBr): ν/cm-1 = 3215 (s, broad, νOH), 3121 (s, νC-H, aromatic), 2977 (s, νC-H), 2932 (s, νC-H), 1620 (s), 1472 (s), 1405 (s), 1320 (m), 1207 (m) (pyridinyl- and pyrazolyl- moieties). 1H NMR (250 MHz, DMSO-d6, 298 K, 32 scan): δ = 8.85 (s, 2H, 2-CH), 8.17 (t, 3J = 2.8 Hz, 1H, -CH), 8.0 (s, 4H, 4-OH ), 7.86 (s, 2H, 2-CH), 7.82 (d, 3J = 8.2 Hz, 2H, 2-CH), 4.75 (s, 2H, 2-CH imidazolidyn), 1.15 (d, 3J = 8.4 Hz, 24H, 8-CH 133). C NMR (63 MHz, DMSO-d6, 298 K, 6000 scan) δ = 149.8, 142.7, 141.5, 126.7, 126.0, 108.5, 83.1, 66.2, 24.0, 17.5. Elemental analyses, found C 52.96, H 7.44, N 22.10%. C25H37N9O4 × 2 H2O (563.65) required C 53.27, H 7.33, N 22.37. 7.2.26. Synthesis of 2,6-bis[4'-(3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazol- 1'-yl]-pyridine (26) O- O N+ N N N N N N N +N O - O The compound 26 (200 mg, 0.38 mmol) dissolved in chloroform (30 mL) was oxidized under phase transfer conditions with NaIO4 (400 mg, 1.85 mmol), previously dissolved in water (20 mL), during 30 min to afford finally a deep blue organic solution. The organic phase was collected and evaporated under air. The bluish residue was dissolved in acetone (3 mL) and chromatographed on SiO2 column using mixture of acetone/petroleum ether (2/8) as eluents to afford the biradical product 27 as blue solid (Rf = 0.34) that was further recrystallised from CHCl3 (54 mg, yield 27 %). M.p. 220-221 °C (from CHCl3). FT-IR (KBr): ν/cm-1 = 3164 (w, νC-H, aromatic), 2984 (w, νC-H), 2936 (w, νC-H), 1596 (s), 1470 (s), 1429 (s), 1403 (s), 1359 (s, νN-O), 1312 (m), 1190 (m) (pyridinyl- and pyrazolyl- moieties). UV/Vis (toluene): λmax nm (ε, M-1 × cm-1) 668 (1460), 610 (1696) 563 (930), 518 (328), 375 (16960), 328 (24850), 323 (31220). FAB-MS (NBA matrix): m/z (%) = found 521.30 (100%) [M+H]+, calculated for C25H31N9O4 (521.57). Elemental analysis, found C 57.34, H 5.72, N, 23.94%. C25H31N9O4 (521.57) required C 57.57, H 5.99, N 24.17%. EPR (298 K, 9.403341GHz, 10-4 M toluene): 9 lines, giso = 2.0065(1), aN/2 = 0.380 mT. 7.2.27. Synthesis of 2,6-bis[4-(1-hydroxy-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]- pyridine (27) 149 Chapter 7-Experimental Session ___________________________________________________________________________________________ N N N N N N N N N OH OH A mixture of 2,6-bis-(4’-formylpyrazol-1’-yl)pyridine 19 (260 mg, 1 mmol) and 2,3- bis(hydroxylamino)-2,3-dimethylbutane 6 (300 mg, 2.02 mmol) in dioxan (40 mL) was heated at 60 оC under argon for 7 days. The solvent was removed under reduced pressure, and the solid product was collected, washed with water, ethanol and dried under nitrogen stream to give the highly air sensitive compound 2,6-bis[4-(1-hydroxy-4,4,5,5-tetramethylimidazolin-2- yl)pyrazolyl]-pyridine 28 (225 mg, crude yield 46 %) that was subjected for the next step without further purification. FT-IR (KBr): ν/cm-1 = 3225 (broad, -OH), 3121 (s, νC-H, aromatic), 2977 (s, νC-H), 2932 (s, νC-H), 1620 (s), 1472 (s), 1405 (s), 1320 (s), 1207 (m) (pyridinyl- and pyrazolyl- moieties). Elemental analyses, found C 60.78, H 6.52, N 25.34%. C25H33N9O2 (491.59) required C 61.08, H 6.77, N 25.64%. 7.2.28. Synthesis of 2,6-bis[4-(1-oxyl-3-4,4,5,5-tetramethylimidazolin-2- yl)pyrazolyl]pyridine (28) N N N N N N N N N O O Compound 28 (200 mg, 0.41 mmol) was oxidized under phase transfer condition (H2O / CHCl3, 20/40 mL) mixture using NaIO4 (400 mg, 1.87 mmol) for 30 min; within this period of time, the color of the organic layer gradually turned to deep orange. The organic phase was collected and evaporated under reduced pressure then the residue was dissolved in 5 mL of acetone and chromatographed over SiO2 column using acetone and petroleum ether as eluents (2/8, low boiling point, 30-40°C, Rf = 0.46). After solvent removal, pure 2,6-bis[4-(1- oxyl-3-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]pyridine 29 was collected as orange hygroscopic solid (105 mg, yield 51%). The solid was then recrystallised from CH2Cl2. M.p. 254-255 °C (from CH2Cl2). FT-IR (KBr): ν/cm-1 = 3142 (w, νC-H, aromatic), 2976 (m, νC-H), 2926 (w, νC-H), 1599 (s, νC=N), 1585 (m), 1470 (s), 1439 (m), 1405 (m), 1289 (w), 1257 (w), 1198 (w) (pyridinyl- and pyrazolyl- moieties). UV/Vis (toluene): λ -1 -1max nm (ε, M × cm ) = 552 (90), 500 (850), 468 (1400), 442 (1285), 415 (910), 315 (23850). Elemental analyses, found C 58.96, H 6.66, N 24.68%. C25H31N9O2 × H2O (507.59) required C 59.16, H 6.55, N 24.84%. 150 Chapter 7-Experimental Session ___________________________________________________________________________________________ EPR (298 K, 9.402690 GHz, 10-4 M toluene): 13 lines, giso = 2.0060(1), aN1/2 = 0.450 m, aN2/2= 0.210 mT. 7.2.29. Synthesis of 2-bromo-6-hydrazinopyridine (29) Br N NH NH2 2,6-Dibromopyridine (1.44 g, 6.1 mmol) was charged in a round-bottomed flask together with an excess of hydrazine-monohydrate (1.61 g, 32.2 mmol) and butanol (100 mL) and the mixture was stirred under reflux for 5 hours. Then, the solvent was evaporated slowly under air and yellowish crystals were obtained. The crystals were initially washed with small portions of cold water (3 × 10 mL) then the solid was dissolved in CHCl3. The solution was chromatographed on SiO2 column (CHCl3/Hexane, 1/2). The unreacted 2,6-dibromopyridine was eluted first (Rf = 0.41) then ethylacetate (EtOAC) was added, and pure 2-bromo-6- hydrazinopyridine was collected as pale yellowish needles after solvent evaporation (0.72 g, yield 63%). Starting from 3 g of 2,6-dibromopyridine (12.66 mmol) and 3.35 g of hydrazine monohydrate (66.92 mmol) it was obtained 1.36 g of 2-bromo-6-hydrazinopyridine 30 (7.23 mmol, yield 57%). M.p.117-118°C (from ethylacetate). FT-IR (KBr pellet, ν/cm-1): 3307 (m), 3104 (w), 3029 (m), 1564 (s), 1546 (s), 1513 (s), 1385 (s), 1167 (s), 1134 (s), 1093 (s), 980 (s), 787 (s), 742 (s), 650 (s). MS-FD (8 kV, CH2Cl2) m/z: found 188.1 (100%), C5H6BrN3 requires MW 188.03. 1H NMR (250 MHz, CDCl3, 298 K, 32 scan) δ (ppm): 7.20 (m, -CH), 6.74 (d, 3J = 7.58 Hz, -CH), 6.61 (d, 3J= 8.2 Hz, -CH), 6.42 (s, br, -NH-), 3.47 (s, br, -NH ). 132 C NMR (DMSO-d6, 63 MHz, 298 K, 1024 scan) δ (ppm): 162.3, 139.6, 139.2, 114.4, 104.5. UV/vis, λmax (CH2Cl2)/ nm (ε, mol-1 × cm-1): 278 (4740), 284 (sh, 3715). Elemental analyses, found C 31.77, H 3.32, N, 22.21%. C5H6BrN3 required C 31.94, H 3.22, N 22.35%. 7.2.30. Synthesis of 6-bromo-2-[4'-formylpyrazol-1'-yl]-pyridine (30) Br N N H N O 151 Chapter 7-Experimental Session ___________________________________________________________________________________________ To a mixture of 2-bromo-6-hydrazinopyridine 30 (2.0 g, 10.6 mmol) and triformyl methane 14 (1.06 g, 10.6 mmol) in methanol (60 ml), a solution of HCl (3.0 mL, 6 N) was added, and then stirred for 1 day at room temperature. The solvent was removed under reduced pressure and the residue was neutralized with aqueous Na2CO3 solution. The yellowish solid product was filtered off and dissolved in CHCl3. The mixture was chromatographed on silica gel using mixture of CHCl3/Hexane/Ethylacetate (1/3/1). The yellowish fraction eluted (Rf = 0.67) was collected and upon air-drying it gave pure 6-bromo-2-[4'-formylpyrazol-1'-yl]-pyridine 31 (1.87 g, yield 70%) as pale yellowish powder that strongly retains water. M.p.141-142°C. FT-IR (KBr pellet, ν/cm-1) 3110 (w, νC-H), 2865 (m, νC-Hal), 1668 (s, C=O), 1583(s), 1566 (s), 1544 (s), 1454 (s), 1363 (s), 1211 (s) (pyridine- and pyrazolyl- moieties). 1H NMR (DMSO-d6, 250 MHz, 298 K, 16 scan) δ (ppm), 9.96 (s, 1H, -CHO), 9.26 (s, 1H, -CH), 8.31 δ (s, 1H, -CH), 7.98 δ (m, 2H, 2-CH), 7.73 δ (m, 1H, -CH). 13C NMR (DMSO-d6, 63 MHz, 298 K, 1024 scan) δ (ppm), 185.4, 149.9, 142.8, 141.5, 139.5, 133.0, 127.1, 125.6, 112.0. FD- MS (8 kV, CH2Cl2) m/z: found 253.1 (100%), required for C9H7BrN3O (MW+H+) 253.08. UV/Vis, λmax (CH2Cl2)/ nm (ε, mol-1 × cm-1): 261 (sh, 11340), 265 (11730), 299 (16350), 338 (broad, 685). Elemental analyses, found C 39.82, H 3.30, N 15.20. C9H6BrN3O × H2O required C 40.02, H 2.99, N 15.50%. 7.2.31. Synthesis of 6'-(4-formyl-pyrazol-1-yl)-[2,2']-bipyridinyl-5-carbaldeyde (31) N N H H N N O O 2-Tributylstannyl-5-[1,3]dioxolan-2-yl-pyridine 3 (3.8 ml, 4.7 mmol) was placed in a two- necked round bottomed flask together with the 6-bromo-2-[4'-formylpyrazol-1'-yl]pyridine 30 (720 mg, 2.8 mmol). The mixture was degassed and kept under rigorous argon atmosphere. Then dry and degassed toluene (30 mL) was added with a syringe together with dichlorobis(triphenylphosphine)-palladium(II) (296 mg, 0.42 mmol), triphenylphosphine (220 mg, 0.84 mmol) and catalytically amount of CuI (20 mg, 0.1 mmol). The solution was then heated to reflux in argon atmosphere under stirring for 60 hours. A dark solution was formed and it was filtered from the inorganic salts, collected and washed with a saturated solution of ammonium-chloride (20 mL). The organic mixture was shaken vigorously in a separator funnel. The aqueous layer was extracted with toluene (2 × 15 mL) and the combined dark 152 Chapter 7-Experimental Session ___________________________________________________________________________________________ organic layers were collected and the solvent evaporated under reduced pressure. The oily solution was treated with HCl (20 mL, 6N) and heated to reflux for 8 hours under stirring; this procedure ensured for the complete hydrolysis of the dioxolane. Basification with saturated solution of potassium carbonate K2CO3 afforded the formation of a yellowish flocculate at pH 8 that consists of the product 31. The product was extracted with dichloromethane (3 × 20 mL). The organic layers were collected and the solvent evaporated to small volume under reduced pressure to afford a yellowish powder. The product was further washed with small portions of light petroleum ether (2 x 10 ml) (b.p. 30 - 40°C) and 4'',5'-diformyl-6-(pyrazol-1''-yl)-2,2'- bipyridine 31 was collected analytically pure as yellowish powder that strongly retains water (360 mg, yield 46 %). M.p. 221-222°C. FT-IR (KBr pellet, ν/cm-1) 3128 (w, νC-H), 2946 (w, νC-H), 2915 (w, νC-H), 2848 (w, νC-H), 1689 (s and asymmetric, C=O), 1592(s), 1546 (s), 1463 (s), 1454 (s), 1363 (s), 1211 (s) (pyridine- and pyrazolyl- moieties).1H NMR (DMSO-d6, 250 MHz, 298 K, 64 scan) δ (ppm), 10.20 (s, 1H, -CHO), 10.01 (s, 1H, -CHO), 9.70 (s, 1H,-CH), 9.22 (s, 1H, CH), 8.88 (d, 3J = 8.16 Hz, 1H, -CH), 8.47 (m. 2H, 2-CH), 8.36 (s, 1H, -CH), 8.26 (t, 3J = 7.85 Hz, 1H, -CH), 8.11 (d, 3J = 7.84 Hz, 1H, -CH). 13C NMR (DMSO-d6, 63 MHz, 298 K, 8000scan) δ (ppm): 192.1, 185.3, 158.0, 153.1, 151.4, 149.7, 141.6, 141.4, 137.40, 132.6, 131.5, 125.5, 121.4, 120.6, 114.0. MS-FD (8 kV, CH2Cl2) m/z: found 278.30 (100%), C15H10N4O2 required (MW. 278.27). Elemental analyses, found C 60.50; H, 4.23; N, 18.67%. C15H10N4O2 × H2O required C 60.81, H 4.08, N 18.91%. UV/Vis (DMSO) λ (ε, mol-1max × cm-1) 320 nm (17230). 7.2.32. Synthesis of 4'',5'-bis[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-6- (pyrazol-1''-yl)-2,2'-bipyridine (32) OH HO N N N N N N N N OH OH 4'',5'-Diformyl-6-(pyrazol-1''-yl)-2,2'-bipyridine (31, 180 mg, 0.65 mmol) was placed in a round bottomed flask together with 2,3-bis-hydroxylamino-2,3-dimethylbutane 6 (337 mg, 2.27 mmol). A mixture of 1,4-dioxane and trichloromethane (10 mL/10 mL) was used as reaction solvents due to the low solubility of 31. The mixture was then stirred for 7 days under argon at room temperature. The solution appeared yellowish-orange with no hint of precipitate formation. Afterwards the mixture was dried under air and a pale brownish powder was obtained. This powder was washed on filter paper with small portions of cold water (2 × 5 mL) 153 Chapter 7-Experimental Session ___________________________________________________________________________________________ and cold light petroleum ether (2 × 5 mL). Finally 255 mg of light brown powder consisting on crude 4'',5'-bis[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-6-(pyrazol-1''-yl)-2,2'- bipyridine 32 was obtained (yield of the crude 73%). A complex 1H-NMR (250 MHz, DMSO-d6) indicated that the residue was a mixture of the desired product 32, plus the excess of 2,3-bis- hydroxylamino-2,3-dimethylbutane and its decomposition products (oxime), nevertheless, it was judged adequate for the next oxidation step. M.p. 213-214°C. FT-IR (KBr pellet, ν/cm-1) 3257 (s, broad, -OH), 2977 (s, νC-Hal), 2925 (s, νC- Hal), 2859 (s, νC-Hal), 1594 (s), 1563 (m), 1465 (s) (pyridyl and pyrazolyl-moieties), 1386 (s), 1141 (s and wide). 7.2.33. Synthesis of 4'',5'-bis[3-oxide-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl]-6- (pyrazol-1''-yl)-2,2'-bipyridine (33) -O O- N N +N N+ N N N N O O The crude 4'',5'-bis[1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl]-6-(pyrazol-1''-yl)-2,2'- bipyridine 32 (130 mg, 0.24 mmol) was charged into a small flask together with CHCl3 (15 mL), degassed and kept under argon while stirring at room temperature. Separately, a solution of NaIO4 (128 mg, 0.6 mmol) dissolved in water (15 mL) was degassed and saturated by argon exchange; then it was added to the solution of the radical precursor by using a syringe under argon. After 20 min a deep blue-violet solution was obtained. The organic layer was extracted by using portions of CHCl3 (3 × 10 mL). The organic solution was evaporated by continuous argon bubbling to a small volume. The mixture was separated by column chromatography (Al2O3, acetone/light petroleum ether, 1/3). The 4'',5'-bis[3-oxide-1-oxyl- 4,4,5,5-tetramethylimidazolin-2-yl]-6-(pyrazol-1''-yl)-2,2'-bipyridine 33 (Rf = 0.37) was obtained after solvent evaporation as deep blue powder in poor yield (12 mg, yield 9 %). M.p. 209 – 210°C (from acetone). FT-IR (KBr pellet, ν/cm-1): 3174 (w, νC-H), 3097 (w, νC-H), 2983 (m, νC-H), 2937 (w, νC-H), 2869 (w, νC-H), 1592 (s), 1548 (m), 1462 (s), 1423 (m), 1387 (m), 1352 (s, N-O), 1217 (m). UV/Vis, λ -1 -1max (toluene)/ nm (ε, mol × cm ): 321 (33624), 336 (35017), 350 (31867), 373 (17510), 392 (10685), 470 (163), 518 (260), 564 (634), 610 (1114), 666 (1004), 746 (70). Elemental analyses, found C 60.64, H 6.01, N 20.84 (%).C27H32N8O4 154 Chapter 7-Experimental Session ___________________________________________________________________________________________ (532.59 MW) required C 60.89, H 6.06, N 21.04%. EPR (298 K, 9.400200 GHz, 10-4 M toluene): 9 lines, giso = 2.0066(1), aN = 0.37 mT. 7.2.34. Synthesis of 4'',5'-bis[-1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl]-6-(pyrazol-1''-yl)- 2,2'-bipyridine (34) N N N N N N N N O O The compound 33 (125 mg, 0.23 mmol) was charged into a small flask and dissolved in CHCl3 (15 mL). Then a solution of NaIO4 (197 mg, 0.92 mmol) dissolved in water (15 mL) was added. After 40 min a deep orange- red solution was obtained. The organic layer was extracted by using portions of CHCl3 (3 × 10 mL). The organic solution was evaporated under reduced pressure till small volume (~ 1 mL). The mixture was separated by column chromatography (Al2O3, acetone/hexane, 2/3) and 4'',5'-bis[-1-oxyl-4,4,5,5- tetramethylimidazolin-2-yl]-6-(pyrazol-1''-yl)-2,2'-bipyridine 34 (Rf = 0.42) was obtained after solvent evaporation as fine orange-red powder, that was further recrystallised from CHCl3 (21 mg, yield 18%). M.p. 205 – 206°C (from CHCl3). FT-IR (KBr pellet, ν/cm-1): 3160 (w, νC-H), 3104 (w, νC-H), 2977 (m, νC-H), 2929 (m, νC-H), 2865 (m, νC-H), 1595 (s), 1458 (vs), 1409 (m), 1371 (s, N-O), 1267 (m) 1157 (m) (pyridine and pyrazolyl moieties). UV/Vis, λmax (toluene)/nm (ε, mol-1 × cm-1): 321 (26100), 417 (791), 442 (997), 467 (1082), 496 (773), 536 (284). Elemental analyses, found C 61.74, H 6.10, N 20.97%. C27H32N8O4 × ⅓ CHCl3 required C 62.07, H 6.23, N 21.19. EPR (298 K, 9.400210 GHz, 10-4 M toluene): 13 lines, giso = 2.0061(1), aN1 = 0.45 mT, aN2 = 0.21 mT. 7.2.35. Synthesis of 2-(4-formylpyrazolyl)pyridine (35) N N H N O To a mixture of 2-hydrazinopyridine (0.55 g, 5 mmol) and triformyl methane 14 (0.5 g, 5 mmol) in methanol (50 mL), it was added HCl (1 mL, 6 N) and then stirred for 22 hours at room 155 Chapter 7-Experimental Session ___________________________________________________________________________________________ temperature. The solvent was removed under reduced pressure and the residue was neutralised with aqueous NaHCO3 solution. The solid product 35 was filtered off and recrystallised from ethanol to afford pale yellow crystals (0.66 g, yield 76%). M.p 96-97 °C (from ethanol). FT-IR( KBr) ν/cm-1 = 3105 (w), 3062 (w), 2855 (w), 1672 (s,νC=O) 1596 (s), 1549 (s), 1476 (s), 1455 (s), 1396 (m), 1207 (s), 1056 (m), 950 (s), 885 (m), 776 (s), 749 (s), 719 (m), 655 (m), 611 (m). 1H NMR (CDCl3, 250 MHz) δ (ppm): 7.2-7.26 (m, 1H, CH), 7.82-7.85 (m, 1H,-CH), 7.9-7.95 (d, 1H, J= 8.2 Hz, -CH), 8.1 (s, 1H, CH), 8.39-8.41(d, 1H, J=4.7 Hz, -CH), 9.0 (s, 1H,-CH), 9.93 (s, 1H, -CHO). 13C NMR (DMSO-d6, 63 MHz) δ (ppm): 113.3, 123.1, 125.8, 131.3, 139.4, 141.8, 148.7, 150.8, 184.5 Elemental analyses, found C 56.23, H 4.86, N 21.77%. C9H7N3O × H2O required C 56.54, H 4.74, N 21.98%. 7.2.36. Synthesis of 2[4-(1-hydroxy-3-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]- pyridine (36) O- + N N N N N OH A mixture of 35 (0.5 g, 3 mmol) and 2,3-bis(hydroxylamino)- 2,3-dimethylbutane 6 (0.45 g, 3 mmol) was dissolved in methanol/chloroform mixture (3/1, 40 mL) and heated at 60 °C under argon for 48 hours. The precipitated product was filtered off, washed several times with methanol and air dried to afford 270 mg (yield 30 %) of the yellowish product which was identified as the highly air sensitive derivative 36. M.p. 195-196 °C (from methanol, before melting the colour changes to red-brown). FT-IR (KBr), ν/cm-1 = 3160 (broad, νOH), 3091(m), 3056 (m), 2991 (m), 2935 (m), 2632 (broad), 2505 (m), 1618 (m), 1595 (s), 1575 (s), 1527 (s), 1475 (s), 1457 (s), 1407 (m), 1384 (m), 1292 (s), 1230 (s), 1207 (s), 1132 (s), 1054 (m), 1020 (s), 998 (s), 954 (s), 879 (m), 846 (w), 782 (s). 1H NMR (DMSO-d6, 250 MHz) δ (ppm): 1.25- 1.3 (d, 12H, 4-CH3), 7.4-7.5(m, 1H, -CH), 8.0-8.15 (m, 2H, 2-CH), 8.6-8.63 (d, 1H, -CH), 9.45 (s, 1H, -OH), 9.65 (s, 1H, -CH). Elemental analyses, found C 59.40, H 6.55, N 22.93%. C15H19N5O2 (301.34) required C 59.79, H 6.36, N 23.24%. 156 Chapter 7-Experimental Session ___________________________________________________________________________________________ 7.2.37. Synthesis of 2[4-(1-oxide-3-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]- pyridine (37) O N N N N +N -O Compound 36 (200 mg, 0.66 mmol) was oxidized under phase transfer condition with NaIO4 (150 mg, 1 mmol) using water/chloroform mixture (1/1) during 30 min. The organic phase became deep blue. After separation and evaporation of the organic phase, the product was chromatographed on SiO2 using a mixture of acetone /petroleum ether (30-40 °C, 3/7, Rf = 0.6) as eluents to finally afford the radical product 37 as blue solid (120 mg, yield 60%). M.p. 132-133 °C (from acetone). FT-IR (KBr) ν/cm-1 = 3170 (w), 3128 (w), 3064 (w), 2991 (m), 2929 (w), 1593 (s), 1577 (νC=N, s) 1462 (s), 1431 (m), 1415 (m), 1392 (w), 1357 (νN=O, s); 1325 (m), 1267 (w), 1219 (w), 1182 (s), 1136 (m), 1093 (w), 1051 (w), 1007 (m), 955 (m), 870 (w), 777 (m), 667 (m). UV/Vis (toluene) λmax (ε, M-1 × cm-1) 295 (18970), 312 (sh, 13470), 351 (6580), 368 (11100), 513 (152), 558 (460), 607 (870), 664 (785). Elemental analyses, found C 53.62, H 6.60, N 20.78%. C15H18N5O2 (300.34) × 2 H2O required C 53.56, H 6.59, N 20.82%. EPR (298 K, 9.399628 GHz, toluene, 10-4 M): 5 lines, giso = 2.0065(1), aN = 0.750 mT, L/G = 1/3. 7.2.38. Synthesis of 2[4-(1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolyl]-pyridine (38) N N N N N O Despite the sensitivity of compound 36 (the powder gets bluish on air exposition) it appeared surprisingly hard to over-oxidize the radical 37 into the imino nitroxide monoradical 38 (with NaIO4). Therefore as oxidizing agent it was used a solution of NaNO2 (20 mg, 0.29 mmol) dissolved in acidified water solution (10 mL H2O, 3 drops of HCl 33%). The compound 37 (70 mg, 0.23 mmol) was initially dissolved in chloroform (10 mL) and the oxidising reagent was 157 Chapter 7-Experimental Session ___________________________________________________________________________________________ added under stirring. The organic phase became reddish very fast (~ 1 min). After separation and evaporation of the organic phase, the product was chromatographed on SiO2 using a mixture of acetone /petroleum ether (30-40 °C, 1/2, Rf = 0.8) as eluents, to finally afford the radical product 38 as orange solid (33 mg, yield 50%). FT-IR (KBr) ν/cm-1 = 3074 (w), 2968 (ms), 2923 (m), 2860 (m), 1594 (m), 1546 (s), 1431 (vs), 1371 (νN-O, s) 1294 (s), 1126 (s), 1024 (w), 858 (w), 802 (s), 765 (m). UV/Vis (toluene) λmax (ε, M-1 × cm-1) 321 (sh, 2550), 415 (487), 442 (705), 469 (768), 500 (467), 552 (50). Elemental analyses, found C 63.01, H 6.11, N 24.32%. C15H18N5O (284.34) required C 63.36, H 6.38, N 24.63%. EPR (298 K, 9.400113 GHz, toluene, 10-4 M): 7 lines, giso = 2.0060(1), aN1 = 0.440 mT and aN2 = 0.885 mT, Lorentzian line. 7.2.39. Synthesis of 2,6-bis(4-formylpyrazolylmethyl)pyridine (39) N N N N N H H O O To a solution of 4-formylpyrazole 15 (1 g, 10.4 mmol) in 30 mL dry THF, NaH was added (0.24 g, 10 mmol), then the mixture was heated under Argon at 50 °C while stirring for 20 min. Then 1.32 g (5 mmol) of 2,6-dibromomethylpyridine were added and the reaction mixture was heated up to 60 °C for 3 h. The reaction mixture was cooled to room temperature and the solvent was removed under reduced pressure. The pale yellowish residue was poured into water/ice mixture. All the inorganic salts were solubilized in the water phase and a white solid was collected. Recrystallization from ethanol/water solution (1/1) afforded 0.8 g of 2,6-bis(4- formylpyrazolylmethyl)pyridine 39 (yield 54 %). M.p. 106 –107 °C (from water). FT-IR (KBr) ν/cm-1 = 3108 (m), 3061 (w), 2969 (w), 2845 (w), 2789 (w), 1683 (νC=O, s), 1595 (m), 1573 (m), 1544 (s), 1456 (m), 1420 (m), 1324 (m), 1197 (s), 1162 (s), 1091 (m), 997 (s), 985 (w), 967 (w), 894 (m), 872 (m), 761 (s), 631 (s).1H NMR (CDCl3, 250 MHz, 298 K, 16 scan) δ (ppm): 5.4 (s, 4H, 2-CH2), 7.11-7.14 (d, 2H, 3J = 8.0 Hz, 2-CH), 7.66-7.73 (t, 1H, 3J = 8.0, CH), 7.96 (s, 2H, 2-CH), 8.05 (s, 2H, 2-CH), 9.83(s, 2H, 2- CHO). 13C NMR (CDCl3, 63 MHz, 298 K, 1024 scan) δ (ppm): 57.6, 121.6, 124.7, 133.6, 138.4, 141.0, 154.8, 183.9. FD-MS (70 eV, CHCl3): m/z = 295.1 (100%). Elemental analyses, 158 Chapter 7-Experimental Session ___________________________________________________________________________________________ found C 57.30, H 4.94, N 22.24%. C15H13N5O2 (295.30) × H2O required C 57.50, H 4.83, N 22.35%. 7.2.40. Synthesis of 2,6-bis[4-(1,3-dihydroxy-4,4,5,5-tetramethylimidazolidin-2-yl)- pyrazolyl- methyl]-pyridine (40) N N N N N HO OH N N N N OH HO 2,6-Bis(4-formylpyrazolylmethyl)pyridine 39 (590 mg, 2 mmol) and 2,3-bis(hydroxylamino)-2,3- dimethylbutane 6 (600 mg, 4 mmol) were charged into a small flask with methanol (40 mL), and stirred under argon for 48 h. The white solid product was filtered off, washed with water, ethanol and air dried to give 690 mg of 2,6-bis[4-(1,3-dihydroxy-4,4,5,5- tetramethylimidazolidin-2-yl)-pyrazolyl- methyl]-pyridine 40 (yield 62 %) that was judged adequately pure for the oxidation step without further purification. M.p. 148-149 °C (from water, it decomposes before melting by changing colour to deep red). FT-IR (KBr) ν/cm-1 = 3238 (broad, νOH) 2976 (s), 2929 (s), 1599 (m), 1575 (s), 1459 (s), 1421 (s), 1363 (s), 1311 (w), 1209 (w), 1174 (s), 1145 (s), 1093 (w), 1022 (m), 999 (s), 964 (w), 935 (w), 912 (m), 843 (m), 802 (m), 762 (s), 695 (w), 665 (w). 1H NMR (DMSO-d6, 250 MHz, 298 K, 64 scan) δ (ppm): 0.65- 0.74 (dd, 24H, 4-CH3), 4.34 (s, 2H, 2-CH), 5.07 (s, 4H, 2-CH2), 6.50-6.53 (d, 2H, J=8.0 Hz, 2-CH), 7.15 (s, 2H, 2-CH), 7.30-7.34 (t, 1H, J=7.0 Hz, -CH), 7.37 (s, 2H, -CH), 7.53 (s, 4H, 4-OH), 13C NMR (DMSO-d6, 63 MHz, 298 K, 1024 scan) δ (ppm): 16.8, 23.4, 56.0, 65.4, 78.7, 82.9, 119.9, 122.6, 129.6, 138.6, 156.3. Elemental analyses, found C 58.04, H 7.66, N 22.36%. C27H41N9O4 (555.67) required C 58.36, H 7.44, N 22.69%. 7.2.41. Synthesis of 2,6-bis[4-(1-oxyl-3-oxide-4,4,5,5-tetramethylimidazolin-2- yl)pyrazolylmethyl] pyridine (41) N N N N N O O- N ++ N N - NO O 159 Chapter 7-Experimental Session ___________________________________________________________________________________________ The compound 40 (555 mg, 1 mmol) was oxidised with NaIO4 (540 mg, 2.5 mmol) by working under phase transfer condition in water/chloroform mixture (1/1), during 60 min. The organic phase became deep blue. After collection and evaporation of the organic phase, the product was chromatographed on SiO2 using a mixture of acetone/petroleum ether (low boiling point, 4/6) as eluents (Rf = 0.28) to give 355 mg of the radical product 41 as blue solid (yield 64%). M.p. 169-170°C (from acetone). FT-IR (KBr) ν/cm-1 = 3145 (w), 2977 (s), 2987 (m), 2931 (m), 1672 (w), 1598 (m), 1573 (w), 1541 (w), 1460 (s), 1427 (m), 1403 (m), 1365 (s, νN-O),1317 (m), 1219 (w), 1196 (m), 1174 (m), 1138 (m), 1016 (m), 993 (d), 868 (m), 813 (w), 773 (m) 659 (m). UV/Vis (toluene) λ/nm (ε, M-1 × cm-1) 651 (2340), 596 (2155), 550 (1000), 352 (44060), 336 (18660). FAB-MS (NBA matrix) 550.3 [M+H]+. Elemental analyses, found C 58.73, H 7.50, N 22.78%. C27H35N9O4 (549.63) required C 59.00, H 6.42, N 22.94%. EPR (293 K, 9.400086 GHz, toluene 10-4 M) 9 lines, giso =2.0065(1), aN = 0.374 mT. 7.2.42. Synthesis of 2,6-bis[4-(1-oxyl-4,4,5,5-tetramethylimidazolin-2-yl)pyrazolylmethyl] pyridine (42) N N N N N O N N N N O The compound 41 (45 mg, 0.09 mmol) dissolved in CHCl3 (15 mL) was oxidised with NaNO2 (24.8 mg, 0.36 mmol) dissolved in HCl/H2O mixture (5%, 5 mL) by working under phase transfer condition during ~ 2 min until the organic phase became deep red. After collection and evaporation of the solvent from the organic phase, the product was chromatographed on SiO2 using a mixture of acetone/petroleum ether (low boiling point, 4/6) as eluents (Rf = 0.42) to give 35 mg of the radical product 42 as orange-red solid (yield 75 %). M.p. 190 – 191°C (from acetone). FT-IR (KBr) ν/cm-1 = 3149 (m), 2977 (s), 2929 (s), 1614 (s), 1575 (s), 1548 (w), 1527 (w), 1458 (s), 1425 (m), 1387 (s), 1375 (s),1244 (s), 1213 (m), 1142 (s), 1116 (m), 1016 (m), 997 (m), 953 (m), 875 (m), 823 (m), 762 (m) 692 (s). UV/Vis (toluene), λ/nm (ε, M-1 × cm-1) 552 (156), 504 (1177), 470 (1990), 444 (1830), 414 (sh, 1245), 353 (80). Elemental analyses, found C 62.30, H 7.00, N 24.02%. C27H35N9O2 (517.63) required C 62.65, H 6.82, N 24.35%. EPR (293 K, 9.402404 GHz, toluene 10-4 M) 13 lines with strong temperature dependencies, giso =2.0060(1), aN1/2 = 0.450 mT, aN2/2 = 0.220 mT. 160 ___________________________________________________________________________________________ Acknowledgements I would like to thank the following persons for the constant support within these years at MPIP: The supervisors of the present work: P.D. Dr. Martin Baumgarten for giving the opportunity to pursue a PhD in the exciting field of Molecular Magnetism. I am indebted to him for his continuous support, the outstanding guidance, and the in-depth discussions. Without his driving passion on sharing his knowledge and curiosity with a beginner as I was, this work would never be completed. I will keep his teaching not only in my mind but also in my heart. I extend my sincere thanks to Prof. Dr. Klaus Müllen, for accepting me as a part of his prestigious group. I am grateful for his countless advices, suggestions and for carefully monitoring the progresses of this work. His tireless enthusiasm for Science is undoubtedly a source of inspiration for all of us. I had the opportunity to meet wonderful and unique people here at MPIP, both from the scientific and personal point. Although some of them left the Institute, they will remain forever in my thoughts. Because “verba volant scripta manent “ I would like to fully acknowledge them here. Dr. Anela Ivanova [University of Sofia] for her Quantum Mechanical Computational work for the high spin molecules, for being so patient in teaching the theoretical basis of QMC, but above all to be the best friend I ever had. Prof. Dr. A. Geies [University of Assiut, Egypt] for teaching me the beauty “hidden” in the heterocyclic chemistry, for his sense of humor combined with a great heart. Another special thank I wish to address to three extraordinary friends: Dr. Chris Clark, for all the precious discussions we had about scientific and non scientific issues, and for valuable advices on writing part of this work, Erhan Ergen, and Krasimir Vasilev (Prof. Knoll Group), for the joy, the optimism and sense of humor that are able to inspire. Thank to have been there when I needed most. I acknowledge all the people in the Prof. K. Müllen Group and particularly: the Group Leaders, Dr. Andrew Grimsdale, Dr. Manfred Wagner, Dr. Andreas Herrmann, Dr. Markus Klapper, Dr. Hans Joachim Räder, personnel and most close colleagues, Lileta Gherghel, Dr. Gueorgui Mihov, Moustafa Abdulla, Al-Hussaini Aymann, Dr. W. Wu [Prof. Knoll Group], Roland Bauer, Zeljco Tomovic, Luke Oldridge, R. S. Prabakaran, N.R. Kundu, Dr. Josemon Jacob, A. K. Mishra, Dr. P. Sonar and Dr. J. Sakamoto (ETHZ, Switzerland), Dr. Ingo Lieberwirth, Dr. Volker Enkelmann for solving the crystal structures of the biradicals presented in this work and Dr. R. Kita (Prof. Wegner Group), Yutta Schenee for all the help in the laboratory, and last but not least the wonderful people of the ChemLager Marcus Thull and Willi Lutz. 161 ___________________________________________________________________________ C.V. Giorgio Zoppellaro Personal informations: Marital status: single. Italian citizenship. Born on 22.12.1967. e-mail: zoppella@mpip-mainz.mpg.de, giorgio_zoppellaro@yahoo.com, giorgio.zoppellaro@int.fzk.de Academic Records: Laurea in Industrial Chemistry from The University of the Studies of Milan, Italy, Faculty of Industrial Chemistry on date 31/10/1997 (96 months). Research work on enzyme biomimics. Thesis title (work in Italian): “Uno Studio sui Complessi Binucleari e Trinucleari derivati da un Legante Octadentato Misto contenente donatori Azotati Aminici e Benzimidazolici”, pp. 1 – 131. Final result expressed in 110th base was 101. Research assistant at the CNR (National Research Council), Department of Analytical Chemistry, University of the Studies of Milan, Italy, till February 1998 (3 months). Research work: “Spectroscopic characterization of Ascorbate Oxidase”. Master in Pharmaceutical Sciences from The Graduate School of Natural Sciences and Technology, Kanazawa University, Japan on 22/03/2000 (24 months). Thesis title (work in English): “The Tetraelectronic Reduction of Dioxygen into Water by Laccase”, pp. 1 - 82. Final Result expressed on 30th base was 30. This work is available through the Library Collection of the Max Planck Society; Location MPIP Polymer Research, Mainz, Call number Dr-00, 6084- 10. Research assistant at The Royal Veterinary and Agricultural University (KVL), Department of Mathematics and Physics, Copenhagen, Denmark till December 2000. Research work: “Perturbed Angular Correlation Spectroscopy (PAC) applied on mononuclear copper proteins: electron transfer processes mediated by Plastocyanin and Mavicyanin” (3 months). Research assistant at the University of Oslo, Department of Biochemistry, till 15th November 2001 in the laboratory of K. K. Andersson (8 months). Ph.D. (Phys-Org. Chemistry) at the Max Planck Institute for Polymer Research, Mainz (Germany) obtained on19th November 2004, with a work in the field of Organic-based Molecular Magnets. (36 months). Thesis title (work in English): “A Study of Nitronyl and Imino Nitroxide Radicals Attached to Heterocyclic Cores. High Spin Building Blocks Towards Organic Magnets” pp. 1- 161. Final 162 ___________________________________________________________________________ result : Magna Cum Laude. This work is available through the Library Collection of the Max Planck Society; Location MPIP Polymer Research, Mainz, Call number Dr-20/4,8068. Awards: Year 1995- Honorary Diploma from the Ministry of the Italian Defence (Military Service). Year 1998 - Monbusho prize from the Ministry of Education, Sport and Culture of Japan. Year 2000- Monbusho prize from the Ministry of Education, Sport and Culture of Japan (3 months research grant). Year 2003 - Young Scientist Award, from the European Material Research Society. Others: 1991-1993 Teaching postion for the courses of Inorganic and Physical Chemistry at the Higher School Institute IPSIA, 21013 Gallarate (Varese) and for Mathemathics and Physics at the Institute Cavallotti, 21013 Gallarate (Varese) Italy. 1994-1995 Military Service in the Italian Army at 7° RGT “Cuneo”, Spaccamela Street N.128, Udine, Italy. 1999 Italian language’s Teacher at the Berlitz School Centre. Address: Berlitz Language Centre, 2- 35, Takaoka-machi, Kanazawa City, 920-0864 Japan. Teaching-assistant in the laboratory of Analytical Chemistry for undergraduate students, Kanazawa University, Division of Life Sciences, Kanazawa, Japan. 2003 Teaching-assistant (two semesters) in the laboratory of Organic and Analytical Chemistry for Biologists (72 students) at the University of Mainz. General Skills EPR (Models Varian, Bruker, Jeol) X and Q-band (9-33GHz) from cryogenic to room temperature with Oxford System; SQUID (Quantum Design, MPMS-5); CD-MCD (Jasco J 500C); UV-Vis/stopped-flow (Otsuka-Denshi RA, HP, Jasco) under aerobic/anaerobic conditions; FT-IR; Atomic Absorption. NMR (Bruker AMX, 250 MHz, 300 MHz); FD-MS, MALDI-TOF (the later applied especially for the analysis of biomolecules); Electrochemical apparatus; Languages English, Italian. 163 ___________________________________________________________________________ International Publications and Contributions 1. One Low Temperature EPR and Mössbauer Spectroscopic analysis of two cytochromes with His-Met axial coordination exhibiting HALS signals. Zoppellaro G.,Teschner T.,Benda R., Harbitz E., Karlsen S., Schünemann V., Arciero D. M., Ciurli S., Trautwein A.X., Hooper A. B., Andersson K.K. Submitted 2005. 2. One step synthesis of symmetrically substituted 4,4''- 2,6-bis(pyrazol-1-yl)pyridine systems. Zoppellaro G. and Baumgarten M., Eur. J. Org. Chem. 14 (2005), 2888. 3. 2,6-Bis(pyrazolylpyridine) functionalised with two nitronylnitroxide and iminonitroxide radicals. Zoppellaro G., Geies A., Enkelmann V., Baumgarten M., Eur. J. Org. Chem. 11 (2004), 2367. 4. Models for biological trinuclear copper clusters. Characterization and enatioselective catalytic oxidation of catechols by the copper (II) complexes of a chiral ligand derived from (S)- (-) -1,1’-binaphthyl-2,2’diamine. Mimmi M.C., Gullotti M., Santagostini L., Battaini G., Monzani E., Pagliarin R., Zoppellaro G., Casella L., Dalton Trans. (2004), 2192. 5. A multifunctional high-spin biradical pyrazolylbipyridine-bisnitronylnitroxide. Zoppellaro G., Volker Enkelmann V., Geies A. and Baumgarten M., Org. Letters, 6 (2004), 4929. 6. The reversible change in the redox state of type I Cu in Myrothecium verrucaria bilirubin oxidase depending on pH Zoppellaro G, Sakurai N, Kataoka K, Takeshi Sakurai BIOSCIENCE BIOTECHNOLOGY AND BIOCHEMISTRY 68 (2004)1998. 7. Synthesis, magnetic properties and theoretical calculations of novel nitronylnitroxide and iminonitroxide diradicals grafted on terpyridine moiety. Zoppellaro G., Ivanova A., Enkelmann V., Geies A. and Baumgarten M., Polyhedron, 22 (2003), 2099. 8. Spectroscopical studies of cytochrome cs, some exhibiting HALS EPR signals. Harbitz E., Zoppellaro G., Teschner T, Fauchald S, Katterle B, Schunemann V, Trautwein AX, Arciero D, Hooper A, Ciurli S, Andersson KK. J. Inorg. Biochem., 86 (2001), 249. 9. Kinetic studies on the reaction of the fully reduced laccase with dioxygen. Zoppellaro G., Huang HW., Sakurai T., Inorg. Reac. Mech., 2 (2000), 79. 10. Spectroscopic and kinetic studies on the oxygen-centered radical formed during the four-electron reduction process of dioxygen by Rhus-vernicifera laccase. Huang HW., Zoppellaro G., Sakurai T., J. Biol. Chem. 274 (1999), 32718. 11. Four electron reduction of dioxygen by Rhus vernicifera laccase. Huang HW., Zoppellaro G., Sakurai T. J. Inorg. Biochem. 74 (1999), 169. 164 ___________________________________________________________________________ 12. A novel mixed valence form of dioxygen by Rhus-vernicifera laccase and its reaction with dioxygen to give a peroxide intermediate bound to the trinuclear center. Zoppellaro G., Sakurai T., Huang HW., J. Biochem., 129 (2001), 949. 13. Synthetic models for biological trinuclear copper clusters. Trinuclear and binuclear complexes derived from an octadentate tetraamine-tetrabenzimidazole ligand. Monzani E., Casella L., Zoppellaro G. et al. Inorg. Chim. Acta, 282 (1998), 180. 14. The 48th Symposium on Coordination Chemistry of Japan, Kouchi 1998, Japan (Oral contribution) Magnetic Properties in the Intermediate Detected in the Four-electron Reduction Process of Dioxygen in Laccase. 15. ICBIC IX Poster Section, Minneapolis 1999, USA (Poster section). The Four-electron Reduction Process of Dioxygen in Multicopper Oxidases. 16. The 26th International Conference on Solution Chemistry, Section S6, Fukuoka 1999, Japan (Oral contribution). The Four-electron Reduction Process of Dioxygen in Laccase. 17. The 49th Symposium on Coordination Chemistry of Japan, Sapporo 1999, Japan (Oral contribution). Dioxygen Reduction by Multicopper-Oxidases and their Derivatives and Metastabilization of Activated Oxygen. 18. The Annual Meeting of the Iron-Oxygen protein Network, Les Arcs, 2001 France. Copper proteins and copper enzymes (Oral contribution). 19. The International Conference of Biological Inorganic Chemistry, ICBIC X 2001 Florence, Italy (Poster session). 20. The Fп5 International Conference, Functional π Electron System, Ülm, 2002, Germany (Poster session). 21. The EUROBIC VI, 29 July-03 August, Copenhagen, Denmark (Poster session). 22. The International Conference on Molecule Based Magnets, ICMM 2002, Valencia, Spain (Poster session and oral contribution). 23. E-MRS Spring Meeting, Strasbourg, France, 2003 (Poster Session). 24. The International Conference on Molecule Based Magnets, ICMM 2004, Japan (Poster Session). 165