UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Synthesis of novel imine bridged macrocycles Jasat, Ayub 1995

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

Item Metadata


831-ubc_1995-0474.pdf [ 11.17MB ]
JSON: 831-1.0059524.json
JSON-LD: 831-1.0059524-ld.json
RDF/XML (Pretty): 831-1.0059524-rdf.xml
RDF/JSON: 831-1.0059524-rdf.json
Turtle: 831-1.0059524-turtle.txt
N-Triples: 831-1.0059524-rdf-ntriples.txt
Original Record: 831-1.0059524-source.json
Full Text

Full Text

SYNTHESIS OF NOVEL I MINE BRIDGED MACROCYCLES by AYUB JASAT B.Sc.(Hons.), The University of Salford, England, 1992 A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Science in The Faculty of Graduate Studies Department of Chemistry We Accept This Thesis As Conforming to the Required Standard THE UNIVERSITY OF BRITISH COLUMBIA August 1995 © Ayub Jasat, 1995 In p r e s e n t i n g t h i s t h e s i s i n . p a r t i a l f u l f i l l m e n t o f the r e q u i r e m e n t s f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h C o l u m b i a , I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d by the head of my department o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g or p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f C K £ /iHST^-V The U n i v e r s i t y o f B r i t i s h Columbia V a n c o u v e r , Canada Abstract Expanded porphyrins have recently been the focus of much synthetic attention by various groups on account of their interesting photophysical and complexation properties. A great deal of the research in this area was driven by the potential biomedical applications of these large, conjugated aromatic macrocycles, particularly in the fields of magnetic resonance imaging and photodynamic therapy. In this regard, we set out to expand the basic porphyrin framework by introducing one or more imine-like bridges between the pyrrole subunits, whilst maintaining the overall aromaticity. We thus, envisaged the synthesis of macrocycles which would exhibit strong absorption bands in the far-red region of the electromagnetic spectrum and, contain a binding cavity suitable for complexing lanthanide (III) cations in a stable ligand:metal adduct. These efforts culminated with the recent discovery of the aromatic porphocyanine macrocycle (e.g. 282). In an effort to extend this work, we began a synthetic programme which would systematically substitute the dipyrromethane subunits of porphocyanine with bipyrroles and/or, difuryl- and dithienyl-methanes. The following account presents our attempts at 282 334 X= 0 340 335 X = S ii preparing these macrocycles. To date, we have synthesized three new analogues of 282, specifically, the difuryl compounds 334 and 340, and the dithienyl macrocycle 335. These bisimine macrocycles were prepared by acid catalysed condensation reactions of 5,5'-diformyl dipyrromethanes or bipyrroles with the 5,5'-bis(aminomethyl)- furans and thiophenes respectively. It is important to note, that, as formulated above, these macrocycle are both nonconjugated and nonaromatic, and therefore more closely resemble the porphyrinogens in terms of their spectroscopic and chemical properties. For instance, their electronic spectra contain no discernible bands beyond 400 nm, whilst their 1H NMR display no diamagnetic ring current effects associated with an extended, conjugated (4n + 2)7c-electron framework. However, unlike the porphyrinogens and the nonaromatic precursors to the tetrapyrrolic porphocyanine 282, these macrocycles do not readily oxidize to their aromatic counterparts. In fact, attempted oxidations with a range of chemical oxidants (e.g. DDQ, chloranil, l 2, etc.) have all failed to effect this critical step, which we believe is essential to ensure the long-term stability of these macrocycles. Interestingly, however, compounds 334 and 335 were found to undergo partial oxidation in air to the dipyrromethene-containing macrocycles 336 and 337. However, like their precursor macrocycles (334-335) they to do not completely oxidize to the fully aromatic species. Similarly, due to the inherent instability of these bisimine macrocycles, we were unable to isolate or characterize any non-labile metal complexes. Furthermore, due to their high basicity and propensity to hydrolyse, all attempts to purify these compound have met with little success. The formation of these macrocycles in preference to linear oligomers has been tentatively inferred from the proton NMR and mass spectral data. iii Nevertheless, many of the precursors required for the synthesis of these and other similar macrocycles have successfully been prepared as either their N-phthaloyl derivatives, or as the biscyano compounds. Thus, we have developed efficient strategies towards these key acyclic intermediates as presented in chapter 2, and therefore anticipate further developments in this area of porphyrin chemistry in the future. A successful methodology was developed for introducing cyano groups onto the oc-positions of the difuryl and dipyrrolic intermediates required in the reductive-macrocyclization route to porphocyanines. Attempts to elucidate the mechanistic aspects of this cyclization were carried out. 336 x= o 337 X= S The introduction immediately following this section presents a comprehensive review of the expanded porphyrin literature up until the end of April 1995. It should be appreciated that whilst there is some degree of flexibility in these assignments, every attempt has been made to group similar macrocycles within the same sections. iv Table of Contents ABSTRACT ii TABLE OF CONTENTS v LIST OF ABBREVIATIONS viii LIST OF FIGURES AND TABLES x LIST OF SCHEMES xv ACKNOWLEDGMENTS xx FORWARD xxi CHAPTER 1. LITERATURE REVIEW 1 1.0 INTRODUCTION 2 2.0 URANYL SUPERPHTHALOCYANINES 6 3.0 TEXAPHYRINS 9 3.1 SYNTHESIS AND PROPERTIES OF TEXAPHYRINS 11 3.2 METALLOCHEMISTRY 17 (a) Transition Metal Complexes 17 (b) Lanthanide Complexation 23 4.0 STRETCHED PORPHYRINS 26 4.1 ACETYLENE-CUMULENE PORPHYRINOIDS 26 4.2 VLNYLOGOUS PORPHYCENES 32 4.3 PENTAPLANAR EXPANDED PORPHYCENES 43 5.0 VlNYLOGOUS PORPHYRINS 51 5.1 BISVINYLOGOUS PORPHYRINS 51 5.2 TETRAVINYLOGOUS PORPHYRINS 63 6.0 PORPHOCYANINES 69 7.0 SAPPHYRINS AND HETEROSAPPHYRINS 74 7.1 SYNTHESIS AND SPECTROSCOPIC PROPERTIES 75 7.2 CHEMICAL PROPERTIES 81 7.3 COORDINATION CHEMISTRY 82 v (a) Metal Complexation 82 (b) Anion Binding 89 7.4 N-CONFUSED SAPPHYRINS 95 8.0 SMARAGDYRINS 97 9.0 ORANGARIN 99 10.0 PENTAPHYRINS 101 11.0 HEXAPYRROLIC EXPANDED PORPHYRINS 105 11.1 AMETHYRIN 105 11.2 ROSARIN 110 11.3 RUBYRIN 113 11.4 HEXAPHYRINS 116 12.0 TORAND EXPANDED PORPHYRINS 121 13.0 FUTURE OUTLOOK 124 CHAPTER 2. DISCUSSION 125 14.0 INTRODUCTION 126 15.0 PORPHOCYANINES 129 15.1 SYNTHESIS OF DIPYRROMETHANE INTERMEDIATES 129 15.2 SYNTHESIS OF AN OCTAALKYLPORPHOCYANINE DERIVATIVE 134 16.0 HETEROPORPHOCYANINES 140 16.1 SYNTHESIS OF ACYCLIC PRECURSORS 140 16.2 SYNTHESIS OFTETRAHYDRO-HETEROPORPHOCYANINES 148 16.3 SPECTROSCOPIC AND CHEMICAL PROPERTIES 151 16.4 ATTEMPTED PREPARATION OF 5-OXOPORPHOCYANINE ANALOGUES 159 16.5 SOME ALTERNATIVE APPROACHES TO HETEROPORPHOCYANINES 167 17.0 PORPHOCYANINE ANALOGUES OF PORPHYCENES 171 18.0 TEXAPHYRIN ANALOGUES 185 CHAPTER 3 CONCLUSION AND RECOMMENDATIONS 191 19.0 CONCLUSION AND RECOMMENDATIONS 192 20.0 FUTURE OUTLOOK 199 CHAPTER 4 EXPERIMENTAL 200 21.0 EXPERIMENTAL 201 vi 21.1 GENERAL AND INSTRUMENTAL 201 CHAPTER 5: BIBLIOGRAPHY 271 22.0 REFERENCES 272 vii List of Abbreviations 9-BBN 9-Borabicyclo[3.3.1]nonane b.p. Boiling point bs Broad singlet (1H NMR) BzIM Benzimidazole CSI Chlorosulphonyl isocyanate d Doublet (1H NMR) dd Doublet of doublet (1H NMR) DDQ 2,3-Dichloro-5,6-dicyano-1,4-benzoquinone DMAP 4-Dimethylaminopyridine DEA N,N-Diethylamine El Electron impact EPR Electron paramagnetic resonance FAB Fast atom bombardment HpD Hematoporphyrin derivative HRMS High resolution mass spectroscopy LRMS Low resolution mass spectroscopy m.p. Melting point MRI Magnetic resonance imaging OEP Octaethylporphyrin PCC Pyridinium chlorochromate PDT Photodynamic therapy PPE Ethyl polyphosphate PPTS Pyridinium tosylate py pyridine viii pyrr. pyrrole q Quartet(1H NMR) SPc Superphthalocyanine t Triplet (1H NMR) TFA Trifluoroacetic acid thio Thiophene TMEDA N,N,N',N'-Tetramethylethylenediamine TPP Tetraphenylporphyrin vis Visible ix List of Figures and Tables Page 2 6 6 10 11 14 Figure 1 Porphine Figure 2 Schematic representation of the two coordination geometries supported by the uranyl dication. Figure 3 Perspective view of S P c U 0 2 perpendicular to the 0=U=0 axis Figure 4 Possible routes to the formation of [1 +1] and [2 + 2] Schiff base macrocycles in metal template syntheses. Figure 5 Schematic representation of the binucleating "accordian" tetrapyrrole macrocycles Figure 6 Structures of some reduced texaphyrins Figure 7 Structures of Cd-texaphyrins (63-65) and water soluble lanthanide (III) complexes 67. 20 Figure 8 View of [53.BzlM]+ showing the six-coordinate Cd centre. 22 Figure 9 View of [53.(py)2]+ showing the overall seven coordinate Cd centre. 22 Figure 10 View of the La(lll) complex of 56 showing the ten coordinate La centre displaced from the N 5 plane by 0.914 A. Figure 11 View of the [Gd.67]+ nine-coordinate complex, with the Gd at 0.595 A above the N 5 plane. Figure 12 View of the [Tb.67.(N03)2] complex showing the 9-coordinate Tb 3 + , with the metal centre in the pentaaza ligand plane. Figure 13 View of the 8-coordinate Lu(lll) complex of 58, in which the Lu 3 + is 0.269 A above the mean-square pentaaza plane. 25 24 24 25 x Figure 14 Structural relationships between the (4n + 2)n annulenes (68-69) with porphine (1) and its structural isomer porphycene (70). 27 Figure 15 Molecular structure of 77. 31 Figure 16 21,23-Dithiaporphycene (83). 34 Figure 17 The annulenoid (89A) and porphyrinoid (89B) molecular descriptions of the 207t-electron antiaromatic diprotonated forms of 87 and 88. 35 Figure 18 X-ray crystal structure of annulenoquinone 88. 36 Figure 19 Structure of the all cis geometric isomer (90) of [22]porphyrin( 36 Figure 20 Molecular structure of the centrosymmetric stretched porphycene 91 38 Figure 21 Crystal structure of 99. 40 Figure 22 Crystal structure of ozaphyrin 106. 46 Figure 23 Schematic representation of 117 showing the interconversion between the two enantiomeric forms. 50 Figure 24 X-ray structure of 117 showing the twisted "figure eight" conformation. 51 Figure 25 Crystal structure of the bis-rhodium complex of 129; showing the saddle-shape and the c/s-coordination of the metal cations. 55 Figure 26 Molecular structure of the centrosymmetric dication 142. 60 Figure 27 Crystal structure of the all-trans conformation of 146.2TFA. 63 Figure 28 Helical conformation of the tetrapyrrole precursor for [26]porphyrinogens. 65 Figure 29 Porphobilinogen (165) 65 Table 1 Maximum 1 H NMR chemical shift differences, A8, between inner and outer protons of [4n + 2]vinylogous porphyrins, resulting from diatropic ring current effects. 69 xi Figure 30 View of 174.ZnCI2 showing the tetrahedral coordination around the Zn and hydrogen-bonding interactions between the pyrrolic NH and CI ligands. 73 Table 2 Optical Properties of typical Sapphyrins, Heterosapphyrins, and Sapphyrin Salts. 81 Figure 31 Proposed structures of the isomeric tetraligated Zn-sapphyrin complexes 206. 83 Figure 32 Structure of the iridium complex.211a. 85 Figure 33 Structure of the bimetallic complex 212a. 86 Figure 34 Crystal structure of the uranyl complex 218. 87. Figure 35 X-ray structure of [191.HF.HPF6], showing the centrally bound F~ anion, with equivalent N-F distances of 2.7 A. 90 Figure 36 Single crystal X-ray structure of the bis(phenylphosphate) salt of 191. The anions are held at ca 1.4 A above and below the sapphyrin plane 90 Figure 37 Structure of sapphyrin 219 and the X-ray structure of the monobasic phosphoric acid complex of 219. 92 Figure 38 Crystal structure of 191 .HCI. The Cl~ is displaced by 1.72 A above the sapphyrin plane. 92 Figure 39 Crystal structure of 191.HN3. The N3~ lies at 1.13 A above the sapphyrin plane. 92 Figure 40 Schematic representation of the cytosine-sapphyrin conjugate 220. 93 Figure 41 Schematic representation of the highly unstable pentapyrrolic smaragdyrin 226. 98 Figure 42 X-ray structure of 232b. 101 Figure 43 View of the uranyl complex of 236, parallel and perpendicular to the 0=U=0 axis, showing the pentagonal bipyramidal geometry of the U xii atom and saddle-shape of the pentaphyrin framework respectively. 103 Figure 44 Generalized structures of the two other possible (aromatic) oxidation Figure 45 Figure 46 Figure 47 Figure 48 Figure 49 Figure 50 Figure 51 Figure 52 Figure 53 Figure 54 Figure 55 Figure 56 Figure 57 Figure 58 Figure 59 states of amethyrin. 106 X-ray structure of 242.2HCI. 107 Molecular structure of the bis-Zn complex of 242. 108 View of the cobalt(ll) amethyrin complex. 109 X-ray structure of 248.3HCI. 111 Crystal structure of the bishydrochloride salt of rubyrin 254. 115 General structure of a torand macrocycle. 122 Generalized structure of porphocyanine 126 Generalized structures of the targeted expanded porphyrin systems 127 Computer generated, energy minimized structures of 340 and the aromatic species 341. 155 Molecular models of macrocycles 334 and 336, depicting the expected energy minimum conformation 157 Computer simulated structure of the fully conjugated macrocycle 338, showing the expected planar conformation, and some of the calculated key inter atomic distances 158 2,5:8,11:14,17-Triepoxy[17]annulenone (348) 162 Generalized structures of the three possible electronic states accessible by the porphycene analogue of porphocyanines. 171 Predicted conformation of bisimine macrocycle 373. 176 Computer generated structure of a bipyrrolic derivative (374) of porphocyanine, in its most stable conformation 177 xiii Figure 60 Schematic representation of the porphocyanine derivative (286), and its bisimine counter-part (375), of Vogel's tetraoxaporphycene 178 Figure 61 The p-hydroxy enamine intermediate formed between the thiazolium salt catalyst and pyrrole 380. 184 Figure 62 Generalized structure of a tetra-p-substituted porphocyanine analogue of porphycene. 194 xiv List of Schemes Page Scheme 1 Nickel induced ring expansion of an N-substituted TPP derivative (2) in the synthesis of homoporphyrins. 4 Scheme 2 Synthesis of uranyl superphthalocyanines. 7 Scheme 3 Ring contraction of the S P c U 0 2 macrocycle upon demetallation and transmetallation. 8 Scheme 4 Synthetic route to diformyltripyrranes 23 and 24. 12 Scheme 5 General synthetic approach towards the reduced-form Texaphyrins 34-44. 13 Scheme 6 Oxidation of macrocycle 34 to its aromatic, free-base texaphyrin congener 52. 16 Scheme 7 Texaphyrin metallochemistry: Metal insertion with concomitant aromatization of the non-aromatic methylene-bridged tripyrrane macrocycles 34-44. 19 Scheme 8 Synthesis of dipyrrolylacetylenes 72 and 75. 28 Scheme 9 Vogel's route to the acetylene-cumulene porphyrinoids 77 and 78. 29 Scheme 10 Synthesis of tetrathia[22]annulene[] 82. 32 Scheme 11 Synthesis of porphyrin-like [22]annulenoquinones. 34 Scheme 12 Two single-step approaches to [22]porphyrin( 91 37 Scheme 13 McMurry coupling of 2,5-diformylthiophene (93) to give macrocycles 94 and 95. 39 Scheme 14 Synthetic pathway to the paratropic macrocycle 99 and its dicationic aromatic 22TT, congener 100. 39 Scheme 15 Synthesis of a vertically expanded porphycene analogue. 42 xv Scheme 16. General syntheses of "pentaplanar" expanded porphyrins (106-109) 44 Scheme 17 Cava's syntheses of 28 annulenes (112-113) and the aromatic 267t-electron dication 114. 47 Scheme 18 Final coupling reaction in the synthesis of turcasarin 117. 49 Scheme 19 Synthetic route to [1,3,1,3]platyrin 125. 52 Scheme 20 Corriu's synthetic pathway to the conjugated nonaromatic platyrin-type macrocycles 129 and 130. 54 Scheme 21 LeGoff's route towards the [1,5,1,5]platyrin homologue (132) of macrocycle 125. 55 Scheme 22 General synthesis of [22]porphyrin(1,3,1,3) derivatives (137-138) via the MacDonald-type "2 + 2" coupling sequence. 57 Scheme 23 Franck's "biladiene-type pathway" in the synthesis of [22]octaethyl-porphyrin 142 and its electrophilic deuteration. 58 Scheme 24 Synthesis of octaethyl[26]porphyrin 146, and its 14-ethyl 147, 14-phenyl 148, and 14-aza 149 derivatives, from the tetravinylogous biladiene 145. 61 Scheme 25 Biomimetic synthesis of porphyrinogens 155-159, and their subsequent oxidation to the corresponding [26]porphyrins 160-163. 64 Scheme 26 Biomimetic synthesis of the aromatic [34]porphyrin( 169. 67 Scheme 27 Two single-pot strategies to the tetrapyrrolic porphocyanine macrocycles 174 and 175. 71 Scheme 28 The reductive-macrocyclization sequence in porphocyanine syntheses. 72 Scheme 29 The general [3+2] synthetic route to sapphyrins 191-200. 76 Scheme 30 An alternative route to the dioxasapphyrin 196, employing a sulphur extrusion reaction. 78 xvi Scheme 31 Synthesis of mono- and bimetallosapphyrins 210-214. 84 Scheme 32 Nucleophilic attack by methoxide anion on the sapphyrin periphery induced by uranyl insertion. 88 Scheme 33 Synthesis of a novel "crowned" sapphyrin (223). 94 Scheme 34 Condensation of excess pyrrole with benzaldehyde to furnish the "N-confused" tetraphenylsapphyrin 225. 96 Scheme 35 General synthetic strategy towards dioxasmaragdyrins. 98 Scheme 36 Synthesis of orangarin 232. 100 Scheme 37 Synthesis of pentaphyrins 236 and 237 via an oxidative "2 + 3" MacDonald-type coupling. 102 Scheme 38 Franck's biomimetic synthesis of an inverted nonconjugated pentaphyrin analogue (239). 104 Scheme 39 Synthesis of amethyrins 242 and 243. 105 Scheme 40 General synthetic route to protonated rosarins 248-251. 110 Scheme 41 Synthesis of rubyrins 254 and 255. 113 Scheme 42 Gossauer's route to hexaphyrins 263-266. 117 Scheme 43 The organometallic chemistry of hexaphyrins 263-265. 118 Scheme 44 Bis-palladium complexes of macrocycles 263 and 264, showing the different ligating modes in operation compared to the bis-Zn complexes 267-269. 120 Scheme 45 Synthesis of the only known torand expanded porphyrin 281 to date. 123 Scheme 46 Synthesis of the dipyrromethane intermediates 295 and 297. 130 Scheme 47 Introduction of a cyano moiety onto the 5-position of pyrrole 300. 131 Scheme 48 Attempted iodination of dipyrromethane 294. 132 Scheme 49 Direct introduction of cyano groups onto free a-carbons of dipyrro-methane 303. 133 xvii Scheme 50 Synthesis of porphocyanine 305. 134 Scheme 51 Preparation of the dipyrrolylimine 308. 137 Scheme 52. Preparation of the 2-phthalimidomethylpyrrole 310 and attempted elaboration to its dipyrromethane (311) and 2-chloromethyl (312) derivatives. 138 Scheme 53 Acid catalysed condensation reaction of furan and furfuryl alcohol 313. 140 Scheme 54 Synthesis of difurylmethane intermediates 318-320. 141 Scheme 55 Attempted strategies towards the synthesis of the 5,5'-bisaminomethyl-2,2'-difuryl-methane (321). 142 Scheme 56 Synthetic approach to the 5,5'-dicyano difuryl- and dithienyl-methanes 322 and 325. 143 Scheme 57 Synthesis of the bisphthalimidomethyl difuryl 330 and dithienyl 331 intermediates, and the subsequent deprotection of the bisamino groups. 147 Scheme 58 Synthesis of bisimine macrocycles 334 and 335, and the partially oxidized macrocycles 336 and 337. 150 Scheme 59 Synthesis of macrocycle 340. 151 Scheme 60 Some typical oxidants utilized: 156 Scheme 61 Schematic representation of 5-oxo-bisimine macrocycle 342, its cross-conjugated congener 343, and the meso-hydroxy tautomer 344. 160 Scheme 62 Retrosynthesis of 5-oxoporphocyanine derivatives. 161 Scheme 63 Attempted synthesis of the key diamine fragment 353. 163 Scheme 64 Reduction of the bridging keto group of 350b. 166 Scheme 65 Schematic representation of an alternative "2 + 2" coupling route to porphocyanines. 167 xviii Scheme 66 Synthesis of the precursor imine 358, and the attempted coupling of this intermediate. 168 Scheme 67 Synthesis of 2-chloromethyl and 2-acetoxymethyl furans 362 and 364. 169 Scheme 68 Ullmann coupling of iodopyrrole 300. 172 Scheme 69 Attempted preparation of the 1 8TT macrocycle 374. 174 Scheme 70 Synthetic routes to 5,5'-diformyl-2,2'-bifuran (185). 179 Scheme 71 Attempted preparation of diamine 379. 180 Scheme 72 Synthesis of dipyrrolylfuran 382. 182 Scheme 73 Synthesis of terfuran 386. 185 Scheme 74 General synthetic strategy towards tripyrrane bisimine macrocycles 294-295. 186 Scheme 75 Attempted preparation of 2,5-bisaminomethylfuran (388). 187 Scheme 76 Preparation of the 2,5-bisphthalimidomethyl furan (392) and thiophene (395). 188 Scheme 77 Synthesis of diformyltripyrrane 399. 189 Scheme 78 Proposed route to the tetraimino macrocycle 403. 195 Scheme 79 Proposed synthesis of macrocycles 409 and 359. 197 Scheme 80 Proposed derivatization of bifuran 185, in the synthesis of the diamine intermediate 379. 198 Acknowledgements Sincere thanks are due to Dr. David Dolphin for his advice during the course of this project. I would also like to thank all of the member of the group for their continued support, encouragement and advice, and particularly to Dr. Ross Boyle for kindly proof reading this document and the helpful advice. Many thanks to the staff in chemical stores, the NMR and mass spectrometry laboratories, Mr. P. Borda for the elemental analyses, and the Orvig group for the use of their IR spectrometer. Finally, and most importantly, I would like to extend my deepest thanks to my parents for their generous support and encouragement over the past six years. Financial support in the form of a teaching assistantship and research stipend from the University of British Columbia (September 1992-August 1994), and a research assistantship from Dr. David Dolphin (September 1994-January 1995) is greatly appreciated. xx Forward To my parents for their continued support, patience and understanding. xxi Chapter 1 Literature Review 1 1.0 Introduction The porphyrins (e.g. porphine 1) and related polypyrrolic macrocycles are probably the most widely studied of all known macrocycles. This can be undoubtedly attributed to (i) their ubiquitous presence in nature; (ii) the critical biological roles played by many porphyrins; (iii) their interesting physical and chemical properties; and (iv) by the fact that the basic porphyrin skeleton found in the natural pigments -heme, chlorophyll and bacteriochlorophyll- is probably one of the oldest known organic molecules on Earth. As ligands, their versatility is perhaps best exhibited by the large number of complexes isolated with almost every metal and semimetal in the periodic table.1 2 1 \ V-NH /13 5 \ V_12 11 7 A / — N Hh 8 I 9 ^ » -1 Figure 1. Porphine. More recently, attention in this area of chemistry has been focused on specific biomedical applications of porphyrins and their derivatives as a result of their physical properties. One such field is photodynamic therapy (PDT) for treatment of cancerous tissue, wherein a combination of red light and a photosensitising drug are used to bring about a therapeutic response.2 The efficacy of such photosensetisers depends largely (amongst other factors) on absorption of light at wavelengths greater than 630 nm, since at these wavelengths the natural chromophores in human tissue and blood become relatively transparent, thereby allowing a greater depth of light penetration. Certain porphyrins (e.g. hematoporphyrin derivative, HpD) 2 were found to concentrate in malignant tissue to a greater extent than in normal tissue. However, such porphyrins have weak absorptions in the desired red region of the visible spectrum. Fortunately, modification of the porphyrin periphery, for example, as in bacteriochlorin and chlorin-like systems, or by increasing the degree of conjugation, results in a bathochromic shift in wavelength of the desired Q-band farther into the red region of the electromagnetic spectrum. Consequently, this has stimulated research towards the synthesis of what is now generically termed the "expanded porphyrins". The latter compounds have similar physical properties to their porphyrin congeners, but contain an increasing number of 7i-electrons, additional coordinating heteroatoms, and/or a larger central binding core. In addition to the above, the introduction of other coordinating atoms and hetero-cycles (i.e. other than pyrrole) yield a new class of molecules containing novel chelating properties. For example, they may exhibit (i) an affinity for binding larger metal cations such as the lanthanides and actinides; and (ii) a capability to stabilize a range of unusual oxidation states and/or coordination geometries. This is of particular consequence for ligands which can form stable 1:1 adducts with the highly paramagnetic gadolinium (III) cation in aqueous media. Such complexes may be of potential significance as contrast agents in magnetic resonance imaging (MRI), another rapidly developing biomedical diagnostic technique.3 Here, the contrast agent, which ideally should show a greater affinity for the tissue under investigation, due to its paramagnetism will decrease the relaxation times of nearby nuclei thereby enhancing the difference between normal and abnormal tissue respectively. Although the metal complexation was the major focus in the study of expanded porphyrins, their unusual ability to chelate anions, under certain conditions, is noteworthy as a feat unparalleled in the chemistry of the parent porphyrins. 3 Finally, from a purely scientific point of view, macromolecules containing an extended, fully conjugated rc-electron network may also provide insight into questions on aromaticity as defined by Huckel's rules. Synthetically, such systems still remain a challenge in order to produce significant quantities for further investigations. This is further complicated by the fact that increments in ring size are accompanied by the occurrence of configurational isomers. Thus, introducing both some degree of conformational flexibility and skeletal strain. Evidently, only porphyrin homologues containing an odd number of methine (CH) units between the heterocyclic rings exhibit the essentially strain-free porphyrin-like geometry with D 2 h symmetry. This is best exemplified by the homoporphyrins (e.g. compound 3), which are macrocycles derived from porphyrins, but contain an extra atom between a meso and an oc-pyrrolic carbon (scheme 1). 4 Insertion of Ni(ll) into the mono N-Ph 5 Scheme 1. Nickel induced ring expansion of an N-substituted T P P derivative (2) in the synthesis of homoporphyrins. 4 substituted TPP 2 yields the expanded homoporphyrin 3 with the endo epimer (i.e. the configuration in which the ester group is pointing towards the nickel atom) as the major product. Sequential demetallation, followed by remetallation in nucleophilic solvent furnished macrocycle 4, which upon treatment with acid provided the stable cationic complex 5. 5 Although the latter is formally an aromatic 187t electron system, the spectral data were found to be more similar to that of the unconjugated macrocycle 3, than the corresponding data of porphyrins. This nonaromatic nature of compound 5 was further confirmed by the ease at which it can be electro-chemically reduced.6 The nonaromaticity of 5 most likely results from conformational deformity from a planar, conjugated system due to the Ni coordination requirements. Expanded porphyrins have been the subject of recent reviews.7"9 The first review of Sessler and Burrel dealt with both aromatic an nonaromatic macrocycles bearing at least one five membered heterocyclic ring system and containing at least 17 atoms in the internal ring pathway of the macrocycle.7 The most recent account by Sessler et al. dealt mainly with the texaphyrins and their applications in MRI,8 whilst that of Vogel examined the relationship between porphyrins and annulenes with respect to aromaticity.9 Therefore, the scope of the present review will be to briefly present the most significant of the earlier work, and expand on the more recent developments in this area of macrocyclic chemistry. For sake of brevity, only macrocycles which fulfil the following criteria will be included. Firstly, they must contain at least three heterocycles; and secondly, contain more than porphine's 16 atoms (see fig. 1) in their internal ring pathway. In addition, with the exception of a few interesting examples, mainly aromatic compounds will be included. Thus eliminating the porphyrin macrocycles such as the isomeric porphycenes,10 dicationic porphyrins 5 (and isophlorins), 1 0 9 , 1 1 isocorroles,12 the expanded benziporphyrin,13 Mertes's tetrapyrrolic macrocycles,1 4 and the recently reported corrphycene.15 2.0 Uranyl Superphthalocyanines The use of metal templates in the synthesis of macrocyclic ligands clearly has a profound effect on the outcome of the reaction, and in most cases may favour the formation of otherwise inaccessible cyclic products in good yield. However, the intrinsic limitations in coordination geometries and ionic radius of d-block transition metals severely restricts their potential to provide access to large cyclic ligands, thus diverting attention to the larger f-block elements. For instance, the general tendency of the uranyl ion to adopt a pentagonal bipyramidal, 6, or hexagonal bipyramidal, 7 (fig. 2) coordination geometry together with long U-N bonds (ca 2.5-2.6 A) has been exploited with great success by the Marks group.16 Figure 2. Schematic representation Figure 3. Perspective view of SPcU0 2 of the two coordination geometries perpendicular to the 0=U=0 axis. 1 7 supported by the uranyl dication. They devised a simple route towards the superphthalocyanines, which basically involved heating anhydrous uranyl dichloride and an o-dicyanobenzene derivative (e.g. 8a-c) in DMF (scheme 2), and upon cooling the product crystallizes out. X-ray diffraction analysis of this product confirmed the presence of five isoindoline 6 subunits with a pentagonal bipyramidal coordination geometry about the uranium atom (figure 3).17 It is apparent form the diagram, that the remainder of the molecule is severely and irregularly buckled. This wave-like nature is thought to arise from steric strain inherent in the "inner ring" of 20 atoms surrounding the uranyl moiety. R R R R 9 Scheme 2. Synthesis of uranyl superphthalocyanines. The severe strain within the macrocycle is clearly reflected in its chemical and physical properties. Traditional methodology employed in the demetallation of porphyrin and phthalocyanine complexes did not yield the expected metal-free ligand (SPcH 2), but instead the ring contracted phthalocyanine 10 was isolated (scheme 3).16ab Reducing the reaction temperature simply slowed this rate of contraction, thereby indicating that this does not simply result from a thermal rearrangement. Similarly, transmetallation reactions with anhydrous CuCI 2, CoCI 2, NiCI2, FeCI 3 gave the corresponding metallophthalocyanines 11. Larger metals such as P b 2 + and S n 2 + also induced contraction. This tendency for the superphthalo-cyanine ligand to undergo contraction to the four-subunit phthalocyanine suggests that the uranyl ion plays a significant role in stabilizing the macrocycle. Other than this facile ring contraction solely resulting from destabilizing ring strain; the radius of 7 the central core in the superphthalocyanine at 2.55 A,1 7 which is ideally suited for coordinating U 0 2 + , is generally not favourable for forming stable, planar metal complexes, in which there is good M-N overlap with transition metals. Even the larger P b 2 + and S n 2 + ions do not have sufficiently large ionic radii to efficiently overlap with the nitrogen atoms of the superphthalocyanines. The exact mechanism of the contraction has not yet been determined, however two possible schemes have been proposed.1613 R R M = Cu, Co, Ni, Zn, Pb, Sn, Er Scheme 3. Ring contraction of the SPcU0 2 macrocycle upon demetallation and transmetallation. The 1 H NMR indicates in solution the macrocycle is highly distorted from planarity as is the case in the solid state. The decreased shielding observed for the benzo protons in the superphthalocyanines 9 compared to those of an analogous phthalo-8 cyanine reflect the apparent impairment of the it system due to the severe ring buckling. Furthermore, variable temperature 1 H NMR experiments on the deca-methyl derivative 9b (R = Me) strongly indicate that the superphthalocyanines may also be conformationally dynamic. Conclusive evidence for this has not yet been obtained due to increased line broadening of the benzo resonance signals at low temperatures. The electronic spectra of the superphthalocyanines generally consist of an intense band at 914 nm (e = 66,700 M"1 cm - 1) with a shoulder at 810 nm and a second intense band at 420 nm (e = 54,100 M"1 cm"1) which are analogous to the Q and Soret bands observed in the electronic spectra of metalloporphyrins. The splitting of the Q type band has been shown to arise from a lifting of the degeneracy of the LUMO due to the non planarity of the ligand lowering the molecular symmetry from D 5 h . 1 6 The instability of the free ligand and other metallo complexes, has undoubtedly overshadowed the simplicity of the synthetic pathway to the superphthalocyanine. Consequently, further work in this area will be of little significance. 3.0 Texaphyrins The condensation reaction between primary 1,n-diamines and heterocyclic dicarbonyl compounds has been extensively utilized in synthetic routes towards new multidentate Schiff base type cyclic l igands. 7 , 1 4 , 1 8" 2 0 These reactions are often carried out in the presence of a suitable metal ion which can direct the steric course of the reaction preferentially towards cyclic products, the kinetic template effect, and/or stabilize the macrocycle once formed, the thermodynamic template effect.18 The ease and (in some cases) the mild conditions at which such reactions take 9 place has also been a contributing factor in favour of this approach (shown schematically in the figure below). Clearly, examination of figure 4 shows that there are two possible products i.e. the [1 + 1], 12, and the larger [2 + 2], 13, macrocycle. This has interesting implications as one can, theoretically, now design macrocycles to chelate virtually any size metal cations by choosing the appropriately configured metallic template and diamino chain. Mechanisms for these cyclization reactions have been proposed, but will not be discussed here. 2 0 Figure 4 . Possible routes to the formation of [1 +1] and [2 + 2] Schiff base macrocycles in metal template syntheses. Much of this earlier work was carried out with 2,6-dicarbonyl derivatives of pyridine as the heterocyclic dicarbonyl of choice. However, it was not until the mid-eighties II II 13 [2 + 2] 10 with appearance of Mertes's tetrapyrrolic "accordian" macrocycles 14 that the first "truly" expanded Schiff base porphyrins were reported. Unfortunately, these ligands could not be converted to a fully conjugated species due to the nature of the bridging tetraimino chains. 1 4 14 n =2,3, M =Zn, Pb, Cu, n = 3, M = H Figure 5. Schematic representation of the binucleating "accordian" tetrapyrrole macrocycles. 3.1 Synthesis and Properties of Texaphyrins Perhaps the most exciting development in this area came some two years later, when Sessler's group at the University of Texas at Austin, reported their tripyrrane based Schiff base macrocycle.21 Their approach hinged on the efficient synthesis of the key symmetric tripyrrane precursors 23-24, as shown in scheme 4 below. Thus, condensation of 3,4-diethylpyrrole (17)22 with two equivalents of the (acetoxy-methyl)pyrrole 15 2 3 under acidic conditions gave the tripyrrane 18 in good yield.2 1 Hydrogenolysis of the benzyl esters, followed by Clezy formylation24 of the intermediate diacid tripyrrane 21 furnished the requisite dialdehyde 23 in 68% yield. Also outlined in the scheme below is a related sequence, using an (acetoxy-methyl)pyrrole 16 with a (methoxycarbonyl)ethyl substituent at the pyrrole 3-postion as one of the precursors instead provided the tripyrrane 19. 2 5 Further manipulation 11 of these methyl esters via a diborane reduction, allowed access to the bis(hydroxy-propyl)-substituted diformyltripyrrane 24. R R R 23 R = H MeOH (for 22 21 R = H 24 R = CH2OH only) 22 R = CH2OH Scheme 4. Synthetic route to diformyltripyrranes 23 and 24. With the dialdehydes 23-24 in hand, the remaining sequence in the syntheses involved acid-catalysed condensations of the latter with an appropriately derivatised o-phenylenediamine, such as 25-33 (scheme 5). 2 1 , 2 5 " 2 8 The latter reaction gave the Schiff base expanded porphyrinogen 34-44 in good yield. In marked contrast to the dipyrromethane dialdehyde case, as reported by Mertes et a / . , 1 4 b no macrocyclic products were obtained when basic metal salt such as Ba(C0 3 ) 2 were used. Marked increases in yields were obtained when stoichiometric quantities of large metal cations (i.e. U0 2 CI 2 and Pb(SCN)2) were employed, but only in the presence of an acid catalyst.21 Nonetheless, the generality of the acid catalysed sequence (scheme 12 5) effectively provided access to a number of other related expanded porphyrinogens (45-51, fig. 6 ) 2 1 ' 2 6 ' 2 7 ' 2 9 23 = H 24 R, = CH2OH H2N. H2N R. 3 PhMe R i 25 R2 = R3 = H 26 R2 = R3 = Me 34 R, = R2 = R3 = H 27 R2 = R3 = OMe 35 R, = H, R2 = R3 = Me 28 R2 = R3 = 0(CH2CH20)4- 36 R, — H, R2 — R 3 —-OMe 29 R2 = R3 = 0(CH2)30H 37 Rx — H, R2 — R3 — •0(CH2CH20)4-30 R2 = H, R3 = OMe 38 R, = H, R2 = R3 = 0(CH2)3OH 31 R2 = H, R3 = CI 39 R, =CH2OH, R2 = R3 = 0(CH2)3OH 32 R2 = H, Rg — CO2H 40 R, =CH2OH, R2 = R 3=H 33 R2 = H, R3 = N02 41 R, — R2 — H, R 3 — OMe 42 R, — R2 — H, R3 = CI 43 R, = R2 ~ H, R3 — C02H 44 R1 — R2 — H, R 3 — N02 Scheme 5. General synthetic approach towards the reduced-form Texaphyrins 34-44. These reduced, methylene-bridged forms of the macrocycles (i.e. compounds 34-50) are nonaromatic, and essentially resemble porphyrinogens.11a"c,3° For instance, when pure, they are colourless, showing absorbances only in the UV region of the electronic spectrum. This resemblance is also reflected in the NMR spectra of these macrocycles. In the 1 H NMR the signals for the bridging methylenes appear in the same region as those of the N,N',N",N"'-tetramethylporphyrinogens of Franck, whilst 13 49 R = -(CH2)4- 51 50 R = -CH2CH2(OCH2CH2)3-Figure 6 Structures of some reduced texaphyrins. 14 in the 1 3 C NMR, these bridging carbons' signals were found to be identical. 1 1 a , c ' 2 1 One point of interest arises here, as the bridging methylenes of macrocycle 34 appear as a doublet in the 1 H NMR indicating a set of diastereotopic protons. This suggests, that in solution 34 adopts a conformation in which the ring deviates from planarity. X-ray crystallography of the HSCN adduct confirmed this non-planarity,21 which is thought to result from both internal hydrogen bonding to the SCN~ counteranion, and the presence of saturated methylene bridges which prevent conjugation between the central pyrrole and the remainder of the macrocycle. Furthermore, the X-ray data revealed that the HSCN was actually "coordinated" within the macrocyclic cavity. Similar observations were noted for anthraphyrin 48, with which this property was mirrored in solution-phase experiments.293 In fact, the latter nonaromatic system showed some promise as an effective anion carrier and as an anion-specific receptor in solution. Such anion coordination is unheard of in the related porphyrins, and appears to be a factor commonly associated with these expanded pyrrolic based systems.3 1 This, undoubtedly, is due to the fact that the core diameter (ca 4 A) in protonated porphyrins, is too small to accommodate ions within the cavity. Incidentally, macrocycle 51 is a very effective receptor for complexing neutral alcohol-type substrates in organic solvents.29 Although these expanded porphyrinogens, like the analogous porphyrinogens, are thermodynamically unstable in air, they do not readily oxidize to their aromatic congeners under these conditions. Moreover, chemical oxidation with wide variety of oxidants leads to decomposition of the macrocycle. 2 1 , 2 6 ' 3 2 Nevertheless, it was subsequently discovered that stirring the reduced macrocycle 34 in air-saturated chloroform-methanol with N,N,N',N'-tetramethyl-1,8-naphthalenediamine (Proton 15 Sponge™, a non-nucleophilic Bronsted base) provided the free-base aromatic texaphyrin 52 in ca 12% yield (scheme 6). 3 2 34 52 Scheme 6. Oxidation of macrocycle 34 to its aromatic, free-base texaphyrin congener 52. Texaphyrin 52 may be envisaged as a 22-7C benzannulene with both 18 and 22-71-electron derealization pathways. Further evidence of the aromatic nature of 52 was provided by the enhanced stability observed for the latter compared to the precursor 34, and from spectroscopic data. For example, a 10 ppm upfield shift in the internal pyrrole NH signal (5 = 0.9) is observed in the 1 H NMR spectrum of 52, as compared to the respective signals in the 1 H NMR of 34. 3 2 , 3 3 Incidentally this upfield shift in the NH signals parallels that seen when, for example, octaethylporphyrinogen (8NH = 6.9)3 4 is oxidized to the corresponding porphyrin (8NH = -3.74), 3 5 thereby indicating that the texaphyrins have a similar diamagnetic ring current to that of the porphyrins. The electronic spectrum is dominated by a Soret-like band at 422.5 nm (e = 60,500 M"1 cm - 1), flanked by N- and Q- like bands at higher and lower energies, with the lowest energy Q band at 752 nm (e = 36,400 M"1 cm" 1). 3 3 16 3.2 Metallochemistry (a) Transition Metal Complexes Although the discovery of the above conditions were a major breakthrough in the chemistry towards aromatic texaphyrins, it was severely limited synthetically in terms of reproducibility, generality and the reaction yield. Additionally, these aromatic texaphyrins showed no propensity to chelate metals. However, during preliminary metal complexation with the reduced macrocycle 34, these researchers encountered some interesting results. Specifically, treatment of 34 with either ZnCI2 or [Rh(C02)2CI]2 in benzene afforded a pink solid (34.ZnCI2) and a green micro crystalline solid (34.Rh(C02)2CI) respectively. 2 1 2 6 In both cases, the 1 H NMR spectra of the complexes contained signals corresponding to the internal pyrrole NH protons, of which, the two imine substituted pyrroles had the greatest upfield shift (cf. 8N H = 12.57 and 11.12 ppm respectively for 34 and 8N H = 10.3 and 9.99 ppm respectively (in CDCI3) for 34.Rh(C02)2CI). The fact that these signals were detected in the 1 H NMR infers that in these metal complexes, the metal centre was actually bound in an n 2 fashion with the two imine groups only (i.e. the tripyrrane subunit does not participate in metal binding). Evidence that this was the case, came from mass spectrometry of 34.ZnCI2 which was consistent with this 1:1 model; and more importantly, upon addition of pyridine to the 1 H NMR sample of this complex the spectrum reverts back to that of the free ligand, thereby suggesting the zinc is no longer chelated. In marked contrast, when CdCI 2 was employed as the coordinating metal, this time under aerobic conditions, a strongly absorbing green material was isolated. 2 6 , 3 2 The electronic spectrum of this material closely resembled that of the free-base 17 texaphyrin 52, with a prominent Soret-like band at 424 nm (e = 72,700 M"1 cm"1) and exceptionally strong N- and Q-like bands at higher and lower energies respectively.32 In addition, this optical data is consistent with that of the cadmium porphyrins, although the intensity of the Soret-like band is somewhat lower for the texaphyrins, than those of similar metalloporphyrins (cf. Cd.OEP.py: ? i m a x = 421 nm, e = 288,000 M _ 1 cm - 1 ) . 3 6 More striking, is the dramatic red-shift (by ca 200 nm) of the lowest energy Q-band (?im a x = 767.5 nm, e = 41,200 M"1 cm - 1) and increased intensity of this absorption (by a factor of ca 3) as compared to that of analogous cadmium porphyrins (e.g. Cd.OEP: Xmax = 571 nm, e = 15,400 M"1 cm" 1). 3 3 ' 3 5 Such behaviour typically reflects the larger delocalized aromatic system of the texa-phyrins, and is commonplace in the penta- and hexa-pyrrolic macrocycles and other ring expanded porphyrins. Once again, as with the free-base 52, the proton NMR exhibits the characteristic features of such aromatic systems containing strong diamagnetic ring currents, i.e. the alkyl, imine, and aromatic peaks are all shifted to lower field. The bridging methylene signals of 34 were also replaced by a sharp singlet (8 = 11.3 ppm) ascribable to the "meso" protons of 53. 3 2 Thus, from this spectroscopic data, it was evident that under these conditions, both metal insertion and oxidation of the ligand occurs concurrently. 18 34 R, = R 2 = H 35 R 1 = R 2 = Me 36 R 1 = R 2 = OMe 37 R 1 = R 2 = 0(CH2CH20)4 38 R, = R 2 = 0(CH2)3OH 41 R 1 = H, R 2 = OMe 42 RT = H, R 2 = CI 43 = H, R 2 = C02H 44 R 1 = H, R 2 = N02 53 M = Cd, R 1 = R 2 = H, n = 1 54 R 1 = R 2 = H, M = Zn, Mn, Hg, n = 1, M = Nd, Sm, Eu, n = 2 55 R 1 = R 2 = Me, M = Zn, Cd, n = 1, M = Sm, Eu, Gd, n = 2 56 R T = R 2 = OMe, n = 2 M = La, Ce, Pr, Nd, Sm " Lu 57 R, = R 2 = -0(CH2CH20)4-58 R 1 = R 2 = 0(CH2)3OH 59 R, = H, R 2 = OMe ' 60 R 1 = H, R 2 = CI n = 1 61 R 1 = H, R 2 = COzH k M = Cd 62 R, = H, R 2 = N02 , Scheme 7. Texaphyrin metallochemistry: Metal insertion with concomitant aromatization of the non-aromatic methylene-bridged tripyrrane macrocycles 34-44. This basic strategy has since been successfully employed to obtain numerous other cadmium texaphyrin complexes, and texaphyrin complexes with a wide range of other large cations, compounds 54-67 (scheme 7, fig. 7), provided an excess of a non-nucleophilic base was present. 7 ' 8 , 2 8 , 2 7 ' 3 2 Furthermore, once formed these metal complexes are extremely stable, except under aqueous acidic conditions which leads to hydrolysis of the macrocycle 3 2 Figure 7. Structures of Cd-texaphyrins (63-65) and water soluble lanthanide (III) complexes 67. Further investigation of the spectroscopic and redox properties of various functionalized cadmium texaphyrins (53, [55.Cd]+, 59-65) have established that, like the porphyrins, the Q-type band energies correspond to the relevant 7t-electron HOMO-LUMO energy gaps. 2 7 Since, in texaphyrins, the phenyl moiety is an intimate part of the macrocycle 7u-system, any variation of substitution to this phenyl ring will perturb the ^-conjugation and allow resonant interactions between the substituent and the main macrocyclic frame. Thus, by varying the nature of the substituents on the phenyl ring or extending the conjugation (compounds 53, [55.Cd]+, 59-65 respectively) the researchers were able to vary the Q-type band absorption maxima from 692 (for 63) to 864 nm (for 64) and similarly the Soret-type band from 417-489 20 nm. Here, for the o-phenylenediamine-derived complexes 53, [55.Cd]+, 59-62, (excluding solvent effects) it was found that electron-donating substituents (such as methoxy groups) on the phenyl ring shifted the Q-type band maximum to the higher energy side of the spectrum, whilst electron-withdrawing substituents caused a shift to the lower end. This effect is also clearly reflected in the reduction potentials of these compounds, where it was observed that electron-donating substituents (on the phenyl rings) shift the reduction potentials to slightly more negative values, relative to 53, and vice versa for electron-withdrawing substituents. One anomaly which is noteworthy is the phenanthrene derived texaphyrin 65, which although having a larger overall 7i-system (due to the three fused benzene rings), has its Q-type absorption peak situated at an even shorter wavelength (X m a x = 732 nm, e = 15,100 M"1 cm - 1) than that of 64. This unequivocally arises from restriction of the n-conjugation as a result of strain in the macrocycle which forces the phenanthryl ring to adopt a conformation that is not coplanar with the rest of the macrocycle ring. On a more practical level, this variability of the Q-type band, and the ease at which one can fine-tune the latter, make the texaphyrins viable PDT photosensitisers. In fact, some preliminary evaluations of these texaphyrins in PDT have appeared. 7 , 3 9 Examination of the structural details of the cadmium texaphyrins revealed some rather unexpected results. Firstly, when cadmium nitrate was used as the C d 2 + source, a mixture of crystalline and non-crystalline solids resulted. A single crystal X-ray diffraction analysis of the crystalline material revealed a six-coordinate pentagonal-pyramidal Cd(ll) complex wherein only one of the axial ligation sites was occupied by a benzimidazole, and the nitrate counteranion not being coordinated to the central Cd atom, Figure 8. 3 3 The source of the benzimidazole is believed to derive from degradation of the ligand during the metal insertion and oxidation reactions. More specifically, via an electrophilic aromatic deacylation of a tripyrrane 21 cx-carbon followed by condensation with o-phenylenediamine. The Cd atom in this complex lies 0.338(4) A above the N 5 donor plane, towards the benzimidazole. Treatment of the remaining inhomogeneous material with pyridine gave the bispyridine adduct of 53. The X-ray crystal analysis of this compound differed from the above complex, in that the bispyridine adduct [53.(py)2]+ consisted of a seven coordinate pentagonal-bipyramidal Cd centre, with the metal atom situated in the plane of the N 5 binding core (figure 9). 3 2 From the X-ray data it was established that in both complexes, the five nitrogen donor atoms are essentially coplanar and define a near-circular binding cavity with a centre-to-nitrogen radius of ca 2.4 A, which is approximately 20% larger than that of typical porphyrins (cf. r = 2.0 A).40 The unique ability of this ligand to support two rare coordination geometries about the same cadmium cation prompted Sessler's group to extend their structural studies to both solution phase NMR (i.e. 1 H and 1 1 3 Cd) and solid state cross-polarization MAS spectroscopy. 3 3 , 4 1 Titrametric experiments in the solution phase NMR revealed that 53 was also capable of supporting a pentagonal geometry (i.e. with no axial ligands) around the Cd centre. However, in the presence of the Figure 8. View of [53.BzlM]+ showing the six-coordinate Cd centre.8 Figure 9. View of [53.(py)2]+ showing the overall seven-coordinate Cd centre.8 22 strongly n basic imidazole ligand the monoligated complex is favoured (fig 8), whereas with the pyridine (a weaker n base) the bisligated complex is preferentially formed (fig. 9). This can generally be rationalized as being due to the steric bulk of the former ligand. (b) Lanthanide Complexation More recently, on the basis of the similar ionic radii of the trivalent lanthanides to that of the divalent Cd atom (CN = 6), metallotexaphyrin complexes ([M.54]2+, [M.55]2+, 56, 57, 58, 67) of the entire lanthanide series have been synthesized, with the exception of the radioactive Pm(ll l). 2 8 , 3 7 ' 3 8 X-ray diffraction analyses of the [La.56.MeOH.(N03)2], [Gd.67.(MeOH)2.N03]+, [Tb.67.(N03)2], [Lu.58.MeOH.N03]+ complexes, which may be considered representative of the Ln(lll) series, have been reported.8 , 3 8 The Ln(lll) cation is coordinated to all five nitrogen atoms of the macrocycle in a bona fide 1:1 adduct (figs.10-13). This is a clear reflection of the increased core size of these ligands in comparison with the porphyrin congeners where similar complexes were deduced as being 2:1 or 3:2 sandwich complexes, or 1:1 complexes in which the Ln(lll) cation is significantly displaced above the N 4 plane (e.g. by as much as 1.9 A).4 2 More importantly, these texaphyrin complexes allow evaluation of the intrinsic lanthanide contraction in relation to the basic pentagonal planar coordination environment. Thus, as this Ln series is traversed, the size of the cation successively decreases as indicated in the diagrams below. With this contraction a parallel reduction in both the total coordination around the metal centre and degree of ligand distortion is observed. Another noticeable feature is the migration of the metal centre further into 23 Figure 10. View of the La(lll) complex of 56, showing the ten coordinate La centre (1.27 A) 4 3 displaced from the N 5 plane by 0.914 A . 3 8 Figure 11. View of the [Gd.67]+ nine-coordinate complex, with the Gd (1.11 A) 4 3 at 0.595 A above the N 5 plane.8 the ligand's mean-square pentaaza plane. However, it should be appreciated that this migration may not entirely result from the decreasing metal centre, but also due, in part, to dissymmetry in the apical ligation around the metal. Evidence for this is best provided by the Tb(lll) and Lu(lll) complexes (figs. 12 and 13), where in the case of Tb-complex the larger Tb 3 +(r 9 c = 1.10 A)43 cation, with a completely symmetrical apical ligation sphere, is accommodated within the N 5 plane. On the other hand, in the [58.Lu.MeOH.N03]+ complex the L u 3 + (r8c = 0.98 A)43, with one bidentate nitrate counterion and one methanol ligand above and below the texaphyrin plane respectively, is held at 0.27 A above this plane. Similar observations were made with the cadmium texaphyrins- see figs 8 and 9. Further characterization of these Ln-texaphyrins has recently appeared in the literature.44 Through a combination of detailed 1 H NMR experiments, line width and isotropic shift analyses, the solution phase structures of the dinitrate complexes 56 (M = Ce, Pr, Nd, Sm, Eu, Tb, Dy, Ho, Er, Tm, and Yb) were evaluated. Generally, for these Ln.56.(N0 3) 2 complexes, the observed isotropic shifts were found to be in 24 0 Figure 12. View of the [Tb.67.(N03)2] complex Figure 13. View of the 8-coordmate Lu(lll) showing the 9-coordinate Tb 3 + (1.10 A)4 3, with the complex of 58, in which the Lu 3 + (0.98 A) 4 J metal centre in the pentaaza ligand plane.8 is 0.269 A above the mean-square pentaaza plane. good agreement with theoretical models, based on dipolar (through-space) contributions being the dominant factor for the isotropic shifts. Provided, the imino protons, wherein the contact (through-bond) contributions are considerable, are excluded in the calculations. Furthermore, these NMR studies enabled investigation of the ligand-based reactivity of these complexes in the presence of nitrate and phosphate-type counterions. Addition of the former anionic species to methanolic solutions of Ln.56.(N0 3) 2 had little effect on the paramagnetic proton NMR spectra. However, addition of phosphate-type ligands (e.g. diphenyl phosphate monoanion) to solutions of these complexes dramatically alters the resulting spectrum, particularly for complexes with large dipolar shifts, wherein, completely new spectral patterns result. Titration NMR experiments with the Eu(lll) complex revealed that as the concentration of sodium diphenyl phosphate was increased, the imino proton signal shifts from 8 = -20 ppm (for the original nitrate complex), through a time averaged signal in the -20 to -16.5 ppm range to 8 = -14 ppm corresponding to the bis(diphenyl phosphate) adduct. Addition of L iN0 3 to this new complex, reproduces the original spectra of the nitrate complex Eu.56.(N0 3) 2, thereby indicating that the monoanionic phosphate ligand replaces the axial nitrate counteranion in the lanthanide coordination sphere, and not just the more-loosely-bound methanol 25 molecules. The authors, in this instance, related the above spectral changes (i.e. changes observed in going from nitrate to phosphate-type axial ligands in the Ln(lll) coordination sphere) to changes in magnetic anisotropy invoked by changes in the relevant crystal field, (whilst maintaining the overall macrocyclic geometry and equatorial binding modes) as a consequence of differences in strengths of anion-to-lanthanide cation interactions. Conclusive evidence for this ligand exchange was confirmed by an X-ray crystallographic analysis of the Dy.56.[(PhO)2P(0)0]2 which, incidentally, closely resembles that of the terbium complex of 67 (fig12), with the Dy(lll) ion displaced by 0.073 A from the mean pentaaza plane. NMR studies of this dysprosium complex further indicate that in solution partial dissociation of one of the • axial phosphate ligand occurs. Nonetheless, as with the Cd complexes, these Ln-texaphyrins show both high stability and hydrolytic nonlability, a feature which potentially make the lanthanide texaphyrins ideal candidates for MRI studies. 8 , 3 8 4.0 "Stretched Porphyrins" 4.1 Acetylene-Cumulene Porphyrinoids More recently, other isomers of the porphyrin C 2 0 H 1 4 N 4 system have been attracting some interest. 1 0 , 1 5 , 4 5 The idea that such structural variants of 1 should exist came about as a result of Vogel's proposal to relate porphyrin 1 to the [18]annulene model (fig 14). 7 , 1 0 a , g The first such isomer to be reported in the literature was porphycene 70, which is formally derived from 1 by merely reshuffling of the pyrrole and methine moieties.10 Since the aromatic 1871 conjugation is still maintained 70 exhibits similar properties to the parent porphyrin, particularly in terms of spectroscopic features. 26 69 70 Figure 14. Structural relationships between the (4n + 2)n annulenes (68-69) with porphine (1) and its structural isomer porphycene (70). Although (4n + 2)n annulenes played a pivotal role in testing the validity of Huckel's (4n + 2) Rule and diamagnetic ring current effects as fundamental criteria in defining aromaticity, they become ineffective models, with increments in the value of n, for fulfilling these requirements.46'47 This obviously results from the increased conforma-tional mobility and steric requirements associated with these larger rings where, the resonance energy is henceforth no longer sufficient to maintain a stable planar conformation. The structural relationship between the porphyrins and annulenes has therefore led to the discovery of newer expanded porphyrins, as further proof of the above criteria for aromaticity. Vogel's group used the acetylene-cumulene [4n + 2]dehydroannulenes47b'48 as templates for expanding the basic porphycene framework, whilst maintaining a 27 planar structure with a (4n + 2)7i electronic conjugation pathway. An intriguing feature of these dehydroannulenes is the presence of a pair of linear CSp2(CSpCSp)nC s p2 units, and are thus best envisaged as equivalent Kekule resonance structures containing both acetylene and cumulene bonds. The inclusion of this structural feature into the porphycene ring relied heavily on McMurry-type reductive coupling of carbonyl compounds with low valent titanium,49 a strategy used with much success in porphycene synthesis.10 Moreover, the ease at which large-membered rings are formed with this type of coupling reactions50 and the observation that the parent N,N'-dihydroporphycene spontaneously oxidize to the aromatic product made this route very attractive. Therefore, the success of this synthetic strategy hinged upon an efficient approach to the bis(5-formyl-2-pyrrl)acetylene 72 and 1,4-bis(5-formyl-2-pyrrl)butadiyne 75 (scheme 8) . 5 1 , 5 2 73 74 75 Scheme 8. Syntheses of dipyrrolylacetylenes 72 and 75. 28 77b Scheme 9. Vogel's route to the acetylene-cumulene porphyrinoids 77 and 78. Here, palladium chemistry proved the most effective methodology for transforming the pyrrole iodoaldehyde 71 directly to the dialdehyde 72, the immediate precursor 29 to the 22TT, macrocycle 77. 5 1 The final step in the synthesis involved a critical reductive coupling of two moles of the dialdehyde 72 with TiCI4/Zn/Cul in THF to yield the expected N,N'-dihydrodiacetylenicporphycene intermediate 76, which readily oxidized in air to the aromatic acetylene-cumulene porphycene 77 (scheme 9). The spontaneous dehydrogenation of 76 is unequivocally a good reflection of the stabilization incurred upon aromatization from the resonance energy, as the formation of acetylene-cumulene bonds in such a fashion is unprecedented. Homologation of 77 was readily accomplished under the same conditions, but with(trimethylsilyl)acetylene as the ethynylating agent to furnish 73 (scheme 8). Basic hydrolysis of the silyl group followed by oxidative coupling of the terminal acetylene 74 with a Pd°/Cu(l) catalyst gave the desired diacetylenic bisaldehyde 75. 5 2 Exposing the latter to the same coupling reagents as described above, gave the novel 26TT porphyrinoid 78 (scheme 9). The 1 H NMR of both the 227t- and the 267t-macrocycles (77 and 78 respectively) were found to be virtually identical and confirmed the presence of a diamagnetic ring current (and hence the aromaticity of these macrocycles). For instance, the bridge protons of 77 and 78 appear as singlets at 5 = 9.99 and 10.08 ppm respectively. The internal NH protons, however, are located at a relatively low field (8 = 2.28 ppm for 77, and 8 = 2.04 ppm for 78) which is indicative of strong NH--H hydrogen bonding.53 The absence of a typical NH stretching band (3360-3300 cm"1) in the IR spectra of these compounds further supports the latter. In accord herewith, the 1 3 C NMR spectra of 77 and 78 exhibit the same resonance signals for the porphycene halves, but 77 shows a single resonance for the acetylene-cumulene type carbons (at 8 105.75) whilst 78 contains two (8 95.94 and 94.82). This would suggest that these compounds are planar, with an effective D 2 h symmetry, and should therefore 30 be formulated as resonance hybrids (77a-77b). This is further supported by the temperature independence of the proton NMR spectra. Single-crystal X-ray analysis (figure 15) corroborated the above conclusions, since, in both cases the molecules were elucidated as being centrosymmetric with a planar ring framework and equivalent C s p 2(C s pC Sp) nC S p2 structural units on opposite sides. The geometric parameters of the latter unit are not only virtually identical in these two macrocycles, but also in good agreement with those of the corresponding dehydroannulenes. The porphyrin-like nature of these macrocycles is clearly manifested in the UV-visible spectra which bears more resemblance to that of porphycenes than porphyrins. The noteworthy features are the double Soret-like band at X, m a x = 405 (e = 188,700 M-1 cm 1) / 439nm (76 500) and three Q-bands at Xmax = 677 (15,300), 724 (73,300) and 766 nm (79,400) for 77, and similarly for 78 Xmax = 448/495 (158,000 and 103,400), and 792 (42,900), 804 (45,400), 839 (98,300) and 889 nm (119,600). As expected a pronounced bathochromic shift towards longer wave-lengths is observed as the main conjugation pathway increases from the 187c Figure 15. Molecular structure of 77. Left: top view; right: side view. 51 31 system of 68 (cf. ? i m a x = 358/ 370 (139,200 / 106,900), 558 (34,200), 596 (30,400), 630 (51,900))10a to the 267t system of 78. Although both compounds 77 and 78 posses absorption bands at wavelengths greater than 630 nm and are highly photostable, further evaluation of their photophysical/photochemical properties revealed them to be devoid of any sensitiser activity.52 4.2 Vinylogous Porphycenes Introducing an additional atom between the pyrrole moieties of the bipyrrolic unit of porphycene is perhaps the most obvious way of expanding the basic ring, whilst maintaining the (4n + 2)n electronic conjugation pathway. This strategy has since only very recently been successfully implemented by one group, at the university of Alabama. 5 4 5 5 Their effort yielded tetrathia[22]annulene[] 82- the first neutral aromatic porphyrinoid derived solely from thiophene and methine units. 1. CH2(OMe)2, AcOH, H 2S0 4 2. BuLi, THF/-70°C, DMF OHC CHO 79 80 TiCyZn, Py., THF, reflux. 1. DDQ 2. H2NNH2 82 81 Scheme 10. Synthesis of tetrathia[22]annulene[] 82. 32 As with other porphycene syntheses, the reductive McMurry coupling of a dialdehyde played a central role in this sequence. Cava's route towards the 5,5'-diformyl-2,2'-dithienylmethane represents a much improved synthesis of such dithienyl compounds than previously reported.56 Firstly, the presence of the bromo group at an a position of 79, prevents the formation of oligomers during the initial condensation reaction, and secondly, these groups allow for a greater regio-selective lithiation at the a-positions of the thiophene rings of the 5,5'-dibromo-thienylmethane intermediate. A competing lithiation at the central methylene group is also known to occur.5 7 Thus, in two steps the requisite dialdehyde was obtained (scheme 10), which when subjected to the McMurry coupling afforded 81. Subsequent dehydrogenation with DDQ followed by reduction with hydrazine gave the fully aromatic 22TI macrocycle 82 in 82% yield. The aromaticity of 82 is evident from the 1 H NMR where the deshielding effect of the ring current is manifested in the low field resonance signals of the methine and vinylic protons (8 12.34 and 11.36 respectively), and the thiophene p protons at 10.86 and 10.84 ppm. Whilst in the dihydro precursor 81 no resonance signals are observed below 6.8 ppm. The UV-visible spectrum is dominated by a Soret band at 417 nm (E = 151,356 M"1 cm"1), and several weaker bands at 503 (5011), 540 (10,232), 579 (51,286) and 771 nm (4168). The latter spectrum bears marked resemblance to that of the 21,23-dithioporphycene 83, 1 0 j but with the anticipated bathochromic shift associated with increased 7t-conjugation. Given the above spectral data and that X-ray structure analysis of 83 revealed this systems planarity, the adoption of a similar planar conformation with a fully conjugated rc-electron periphery is predictable for the present less crowded macrocycle 82. 33 83 Figure 16. 21,23-Dithiaporphycene (83). In the year prior to this work, Markl and Striebl reported two structurally related macrocycles, the [22]annulenoquinones 87-88 (scheme 11).5 8 These compounds were obtained via a double Wittig cyclization of the dialdehydes 85, 86 with the bis-ylide derived from 84 (respectively), under high dilution conditions. The electronic spectra of compounds 87 and 88 typify their overall quinoid structure, with broad absorption bands in the ? i m a x = 297-452 nm region of the electromagnetic spectrum. In acidic media, these bands dramatically increase in intensity and exhibit a pronounced bathochromic shift. For example, in its dicationic form, the unsym-metrical macrocycle 88 now displays two bands at ? i m a x = 328 and 438 nm. 85X=0 87X=0 86X=S 88X=S Scheme 11. Synthesis of porphyrin-like [22]annulenoquinones. 34 It is in this protonated form that 87 and 88 may be described by either an annulenoid 89A or a porphyrinoid 89B antiaromatic 207t-electron system as depicted in the figure below. However, the 1 H NMR of these dications give no indication of any paratropicity, instead, the protons exhibit a low field shift compared to those of the neutral species. Thus, for the dication of the tetrafuryl compound, the furyl (3 protons resonate at 8 = 7.60 and 7.40 ppm and the vinylic protons appear at 6.40 ppm, whereas in 87, these protons' signals are observed at 8 = 7.30, 6.50 and 6.40 respectively. The corresponding dicationic species of 88, exhibits a similar low field shift in the 1 H NMR (A8 = 0.98-1.16 ppm). This deshielding has been interpreted as being more of a consequence of the double positive charge on the macrocycles outweighing the effects of a paratropic ring current. Others have rationalized similar increased deshielding effects observed in the 1 H NMR in the outer ring protons of the [4n + 2]annulene dications on a lowering of the 7t-charge by the positive charges.5 9 Moreover, the 1 H NMR of 87 and 88 serve to indicate that in solution the structure is fairly dynamic. However, the interconversion between the conformers is extremely rapid compared to the NMR time scale that the spectra for an average OH OH A 89X=0, S B Figure 17. The annulenoid (89A) and porphyrinoid (89B) molecular descriptions of the 20n-electron antiaromatic diprotonated forms of 87 and 88. symmetric, planar conformation is recorded. In the solid state, as revealed by X-ray 35 crystallography, the molecule adopts a puckered conformation in which one of the furan rings is twisted out of the general macrocyclic plane (fig. 18). Vogel's earlier foray into the acetylene-cumulene porphyrinoids51 inevitably led to the discovery of another expanded porphycene. Here, they realized that partial catalytic hydrogenation of the acetylene-cumulene bonds of 77 would yield the next higher homologue of [18]porphyrin 1- namely the [22]porphyrin( 90 (fig. 19).6 0 They anticipated the latter reaction should yield the all c/'s-compound 90, as hydrogenation over Lindlar-catalyst proceeds via c/'s-addition of hydrogen across the multiple bond. However, the c/s,frans,c/'s,frans-[22]porphyrin( 91 was the sole expanded porphyrin product isolated from such a hydrogenation reaction 90 Figure 19. Structure of the all cis geometric isomer (90) of [22]porphyrin( (scheme 12). That this isomer forms preferentially is not entirely surprising, as both molecular models and force field calculations indicate that 90 with an all cis-configuration is too highly strained to adopt a planar configuration. This, on the other 36 hand, is not the case for the c/s,frans,c/'s,frans-isomer 91, where a planar structure with minimal skeletal strain is not only more favourable, but would be additionally stabilized by the presence of strong N H - H hydrogen bonds. Compelling evidence for the cis,trans,cis, frans-configuration and the aromaticity of 91 is provided by the 1 H NMR, where the internal protons of the trans bridge resonate at extremely high-field (6 = -7.5 ppm), whilst those on the exterior together with those of the cis bridges occur at low-field (6=11.7, and 9.83 and 9.88 ppm respectively). The presence of N H - H hydrogen bonds is convincingly demonstrated by the down-field shift of the NH protons (8=1.34) compared to that of porphyrins (6 = -3.74) 3 5 b which is consistent with other porphycenes (8 = 3.15), 1 0 a and by the lack of NH stretching vibrations in the IR spectrum. The proton NMR does however show temperature dependence, thus leading to the conclusion that at temperatures in excess of 70°C 90 undergoes an isodynamic transformation involving rotation about the frans-CH CH bonds - tantamount to a transition from C 2 h to effective D 2 h molecular symmetry. The UV-visible spectrum of 90 is almost superimposable on that of the isoelectronic 22n system 77, and is typical of porphycene-like macrocycles. A slight bathochromic shift to longer wavelengths of the Soret and lowest energy Q bands is, however, observed with compound 90. The alternating cis,frans-configuration of the ethylene bridges and, in particular, the postulated 77 91 Scheme 12. Two single-step approaches to [22]porphyrin( 91. 92 37 planarity of the ring framework were both confirmed by X-ray crystallography (fig 20). Interestingly, this vinylic porphycene shows much improved photophysical properties than it's isoelectronic precursor 77, and has been found to be of comparable phototherapeutic activity to that of porphycene.52 Figure 20. Molecular structure of the centrosymmetric stretched porphycene 91, left: top view; right: side view.60 An alternative approach to [22]porphyrin-(, involving a one-pot Ti° mediated reductive coupling of 3,4-diethyl-2,5-diformylpyrrole 92 (scheme 12), gave this same isomer as the only low-molecular-weight product.60 The formation of this isomer is rather remarkable, considering that the analogous reductive coupling reaction with 2,5-diformylthiophene (93) favours the formation of the cyclic [18]annulene trisulphide 94 and the [24]annulene tetrasulphide 95 with no evidence of the thiophene derived analogue of [22]porphyrin( 91 (scheme 13).5 5 Hereby further supporting the above conclusions that 91 is the most energetically favoured isomer of [22]porphyrin( Another noteworthy feature of these thiophene based porphyrinoids is that the 18n system 94 and the related expanded 30K hexa-thia[30]annulene[], (derived from reductive coupling of 2,2'-bithiophene-5,5'-dicarboxaldehyde)55 are both atropic, since they cannot assume a planar configuration. In these examples, as with the tetrathiaporphycene, the severe 38 repulsive forces between the lone pairs of electrons of the sulphur atoms force the heterocyclic rings to assume a conformation in which these atoms on adjacent thiophene rings are directed away form each other. s 93 TiCI4/Zn, py, C H 0 THF, reflux 94 38% 95 4.7% Scheme 13. McMurry coupling of 2,5-diformylthiophene (93) to give macrocycles 94 and 95. Recently the tetraoxa analogue of [22]porphyrin( 91 has been described (scheme 14). 6 1 A Wittig reaction of the monoacetal 96 of 2,5-diformylfuran and the phosphonium ylid derived from 97 gave the frans-dialdehyde 98, which was easily separated from the isomeric E/Z mixture by recrystallization. Dimerization of this aldehyde under the standard conditions employed in porphycene synthesis provided the paratropic [24]annulene 99. rv (MeO)2HC- ~cf CHO Ph,PH,c' V "CH(OMo). "*OHC 96 l o l I l n V . 3 O 2 Br CHO 97 98 'TiCI4,Zn/ Cu.THF B^, CH2CI2 o°c 100 99 Scheme 14. Synthetic pathway to the paratropic macrocycle 99 and its dicationic aromatic 22n congener 100. 39 The presence of a paramagnetic ring current in 99, as expected for cyclic systems with 4n 71-electrons, was deduced from the low temperature (-78°C) 1 H NMR in which the inner and outer protons of the trans double bond appear at 8 = 12.38 and 5.30 ppm respectively as an AX system with 3 J = 16.17 Hz. With increasing temperature, rotation of the single bonds around the latter double bonds occurs and these protons begin to equilibrate, until at 65°C where only an averaged signal is observed at 8 = 8.52 ppm. From the coalescence temperature, A G * was calculated as approximately 46 kJ mol"1. A single-crystal X-ray structure analysis of this compound (fig. 21) confirmed the alternating cis,trans,cis,trans configuration, and showed the molecule to be planar with an inversion centre and the nonequivalent furan rings twisted in opposite directions by 10°. Figure 21. Crystal structure of 99. Top: view from the top; bottom: side view.D' Exposing compound 99 to an equimolar amount of bromine at 0°C results in oxidation to the aromatic tetraoxa[22]porphyrin( dication 100, isolated as the bisperchlorate salt. Retention of the alternating cis,trans,cis,trans configuration, with C2-symmetric conformation, and aromaticity in 100, is apparent in the 1 H NMR where the protons inside the aromatic (diamagnetic) ring current are shifted 40 drastically upfield to 8 = -8.08 ppm (3J = 14.47 Hz), whilst the protons of this trans double bond on the outside are located further downfield at 8 = 13.66. This complete reversal of the resonance positions of the inner perimeter protons (AS = 20.46 ppm) and the outer protons (A8 = 8.36 ppm) in going from an antiaromatic (4n)7i to an aromatic (4n + 2)n system typifies the opposite effects of paramagnetic and diamagnetic ring currents on the nuclear shielding of protons in a magnetic field. Incidentally, these NMR observation are analogous to those observed in similar experiments with the 4n and (4n + 2) annulenes.4713 The signals of the (3 proton of the furan rings, and those of the cis bridge (8=11.5 and 11.42 ppm, 3 J = 13.8 Hz) also exhibit a downfield shift associated with the ring current compared to those of the antiaromatic compound 99 (cf. for the cis bridges 8 = 4.72 and 4.66, 3 J = 13.8 Hz). The porphyrin-like nature of this macrocycle is revealed in the electronic spectra where an intense split Soret band at 438 (295,600) and 460 nm (229,000), and weaker Q-type transition between 643 and 678 nm with two slightly more intense absorptions at 712 (46,000) and 723 nm (30,000) are observed. The antiaromatic system 99, for sake of interest, exhibits two bands with high molar extinctions at short wavelength ( ? i m a x = 346 (128,000) and 358 nm (175,000)), and one of lower molar extinction at 471 nm (2,300). In marked contrast to the highly dynamic 24n electron system 99 and the isodynamic pyrrole analogue 91, the dication 100 is conformationally rigid and shows absolutely no dynamic effects in the 1 H NMR at temperatures ranging up to 180°C! A common feature of all the tetrapyrrolic based porphycenes and its expanded variants is the presence of strong N H - H hydrogen bonds. Any question about the extent to which this bonding plays in maintaining the planar ring framework and the 41 stability of these molecules are perhaps best answered by the tetrapropyl[26]-porphyrin( 102- the only example of a vertically expanded porphycene reported to date.9 The synthetic route employed follows the general pattern of porphycene synthesis, and in striking parallel the primary N,N'-dihydro product of the coupling reaction readily dehydrogenates to the aromatic species 102. 101 102 Scheme 15. Synthesis of a vertically expanded porphycene analogue. As expected for larger macrocycles, 102 was found to be extremely conforma-tionally mobile and the proton NMR could not be resolved, even at low temperatures (down to -50°C). However, protonation of 102, to yield the dicationic species, efficiently arrests the molecule in a planar conformation (as with the tetra-oxaporphycene macrocycle 100) thus enabling spectroscopic examination. The extreme chemical shifts observed for the inner protons (Hj, 5 = -10.19 ppm) and the outer ring protons (H 0, 8 = 13.04 ppm) clearly revealed the aromaticity of the macrocycle. It should be appreciated that this large difference in chemical shifts between the outer and inner protons (A8 = 23.23 ppm) arises from the combined effects of the diamagnetic ring current and the positive charges, 5 9 since addition of 42 trace quantities of methanol to the NMR sample- which also slows down the conformational processes presumably via coordination- the resultant spectra now contains a smaller chemical shift difference (A8 = 15.95 ppm). These observations undoubtedly underline the role of hydrogen bonding, particularly in regards to the remarkable stability of this series of expanded porphyrins. Unfortunately, as ligands these compounds are not expected to have any impact. This primarily arises from the unfavourable rectangular coordination centre, however complexes in which bridging multiple bonds participate cannot be outrightly dismissed. Nonetheless, the porphycenes still remain of considerable interest in terms of their photophysical properties. 4.3 "Pentaplanar" Expanded Porphycenes Fairly recently a novel way of expanding the basic porphycene ring framework has appeared in the literature. This approach is based on the inclusion of additional five-membered heterocycles in between either one, or both bipyrrolic units of porphycene. 5 5 , 6 2 ' 6 5 As with all other porphycene synthesis the McMurry coupling of the appropriate bisformyl units was the key step towards these new cyclic systems. Thus, a simple route to the precursor dialdehydes 103 and 104 was required. Ibers et al. approached this by expanding on chemistry unravelled by Merrill and LeGoff , 6 6 where in an electron-deficient pyrrole aldehyde and divinyl sulphone were coupled under Stetter conditions67 to provide a 1,4-dipyrrolylbutane-1,4-dione. Subsequent cyclization with either an acid catalyst64 or, with Lawesson's reagent,62 conveniently provided the additional furyl or thienyl rings in 103 and 104 (respectively) between the two terminal pyrroles. Coupling of 103 with the bipyrrole 105 in the presence of 43 low valent titanium in refluxing THF, as outlined in scheme 16 below, provided macrocycle 106,64 to which the researchers assigned the trivial name ozaphyrin, interestingly, after the Emerald City of O z 6 8 in view of its emerald green chloroform solutions! In a similar fashion the thiophene derivative, thioozaphyrin 107, was obtained by the "2 + 3" coupling of bipyrrole 105 and dialdehyde 104.65 OHC CHO 103 X= O 104 X=S 106 X=0 108X=Y = S 107 X=S 109 X=S, Y = 0 Scheme 16. General syntheses of "pentaplanar" expanded porphycenes (106-109). The 1 H NMR spectra of both 106 and 107 are consistent with presence of a diatropic species, whereby the internal protons are profoundly shielded whilst those on the outer periphery are deshielded. Thus, the internal pyrrole NH resonates at 8 = -2.16 ppm, and the methine, pyrrolic and furyl (3 protons of 106 resonate at 10.31, 10.47, and 10.5 ppm respectively,64 and vis a vis for 107.65 In marked contrast to 44 their porphycene counter parts, as inferred from the NH chemical shifts in the proton NMR, N H - H hydrogen bonding is not an intrinsic feature of these macrocycles. The UV-visible spectrum of 106 exhibits a split Soret band at 414 (120,000) and 430 nm (99,000), consistent with a C 2 v symmetry, and three intense Q transitions at 640 (33,880), 677(21,380) and 735 nm (57.540).64 This data is reminiscent of the porphycenes, but with the expected bathochromic shifts to longer wavelengths associated with increased ^-conjugation. The UV-visible spectra of thioozaphyrin 107, 6 5 whilst containing the same essential features as described above, is even more red-shifted (cf. ? i m a x = 425 (100,000), 447 (74,130), 660 (30,200), 697 (28,800), 713 (28,800) and 755 nm (50,120)) with respect to porphycene (and to a smaller extent to that of 106). This additional red-shifting is not too surprising, since it has been previously shown by others that successive replacement of one or more pyrrole rings in porphine with thiophene produces a concomitant bathochromic shift of the absorption bands. 1 0 j ' 6 9 In addition to providing conclusive evidence for the planarity and C 2 v symmetry of ozaphyrin, the X-ray analysis (fig 22) revealed that in the solid state the latter compound consists of layers of staggered macrocycles (4 molecules per unit cell) with the solvate chloroform molecules occupying the interstitial voids. The average intralayer distance is 3.4 A, and the ozaphyrins are arranged with slight overlapping of adjacent propyl side arms, such that no macrocycle is directly above the other. The central cavity, whilst being of comparable size (= 5.0 A) to that of the texaphyrins and superphthalocyanines, is highly distorted so as to accommodate the overall planarity of the molecule. Despite this lack of symmetry, it is highly plausible that this geometry may be suitable for binding lanthanide cations or even two first-row transition metals. 45 Figure 22. Crystal structure of ozaphyrin 106. Alternatively, self-coupling of the bisformyldipyrrolylthiophene 104 with low valent titanium and subsequent air oxidation of the initial product yields the 26n bronzaphyrin 108 (scheme 16). 6 2 , 6 5 In addition, this reaction sequence does not only allow access to the desired products, but also to all other statistically possible cyclic products such as the hetero-coupled product 109 (from coupling of 103 and 104) and the tetrapropylporhycene derived from self-condensation of bipyrrole 106. All of which were found to be highly stable and separable by extensive chromatography. The only exception being the homo-coupled product 108 (X = Y = O) which, although it does actually form via self-condensation of 103, is apparently highly unstable. Although, at first glance, the separation techniques may appear rather tedious, the versatility of this reductive coupling reaction far out weighs this. For instance, Michael Cava's group has extended this work to isolate the 28K annulenes 112-113 (scheme 17) in good yield (>60%).55,63 Preliminary electrochemical studies of these annulenes revealed that only 112 could be expected to undergo a two-electron oxidation to form the corresponding 26TI aromatic dication 114. 6 3 The N-alkylated 46 molecules 113, on the other hand, only show a single quasi-reversible peak in the cyclic voltammogram (E = 0.332 V for 113a and E = 0.415 V for 113b) corresponding to a one-electron transfer to form the radical cations. This is clearly a consequence of the steric requirements imposed by the internal alkyl substituents, which henceforth prevents the assumption of a planar conformation. In fact, the nonplanarity of all these [24]annulenes (i.e. 112-113) is fairly evident in their spectroscopic properties which are more in accordance with partially conjugated 4n7i -systems that are devoid of any paramagnetic ring currents. Nevertheless, as indicated in the scheme below the oxidation of 113 was achieved by treatment with 98% sulphuric acid, which acts as both oxidant and a counterion source. 5 5 114 Scheme 17. Cava's syntheses of 28n annulenes (112-113) and the aromatic 267i-electron dication 114. 47 The electronic spectra of this series of analogous 26n macrocycles 108, 109, and 114 display features reminiscent of the porphycenes. The UV-visible spectra of bronzaphyrin 108, for example, contains a widely split Soret band at 460 (200,000) and 501 nm (95,000) consistent with the D 2 h symmetry, and four Q transitions at 745 (69,200), 780 (60,255), 790 (63,000), and 858 nm (83,200).62 These transitions are considerably red-shifted compared to that of its 22n congener 107. Rather interestingly, oxabronzaphyrin 109 exhibits an almost identical absorption pattern to 108, but has one less Q transition (cf. Xmax = 455, 489, 745, 783, and 855 nm),6 5 whilst the dication 114 shows absorption bands which are even farther shifted towards longer wavelengths, with bands at 521 (63,100), 858 (18,600) and 886 nm (19,500) respectively.55 Perhaps, the most prominent feature of the UV-visible spectra of the porphycenes is that the Q bands are approximately 30-50% as intense as the Soret transition. With the parent porphyrins and metalloporphyrins, the Q transitions are typically only about 1-15% as intense. The 1 H NMR of bronzaphyrin 108 and oxabronzaphyrin 109 show peaks in the range of 9.96 to 11.65 ppm corresponding to the deshielded external furan, pyrrole, thiophene and methine protons, and a sharp peak corresponding to the diamagnetically shielded internal pyrrole protons at -2.2 ppm for 108 and 0.41 ppm for 109. Not surprisingly, the four different types of protons in dication 114 are even further deshielded, ranging between 11.28 and 12.10 ppm, indicating the cumulative effects of the positive charges and a more pronounced diamagnetic ring current.55 Finally, research in this area has culminated with the recent synthesis of turcasarin 117, which incidentally is the largest expanded porphyrin reported to date. 7 0 Here, the acid-catalysed condensation of 5,5'-diformyl-2,2'-bipyrrole 116 with terpyrrole 115 (scheme 18) furnished the target macrocycle in 20% yield. Being a 407i-electron system 117, is not expected to be aromatic but rather antiaromatic, and therefore 48 should exhibit the effects of a paramagnetic ring current in the NMR. This was not the case, as the pyrrolic inner proton resonances in the 1 H NMR spectrum show no ring current effects. Though, like other polypyrrolic aromatic macrocycles, it displays a rather intense absorption in the visible region of the electronic spectrum that is considerably red-shifted in comparison to the porphyrins. Thus suggesting turcasarin is an atropic (nonaromatic) molecule with a delocalized ^-electronic system. 117 Scheme 18. Final coupling reaction in the synthesis of turcasarin 117. The lack of any ring current effects can be attributed to the unusual conformation adopted by this large macrocycle. Both 1 H and 1 3 C NMR were consistent with C 2 -symmetric structure, but the complex splitting patterns for the methylene protons of the alkyl side chains observed in the 1 H NMR led these researchers to believe that in solution 117 must exists in two limiting conformations. Consequently, they proposed a pair of enantiomeric twisted "figure-eight" conformations as depicted in figure 23. Such conformational chirality, though a novel concept within the context of expanded porphyrins, would be in accord with the observed complex splitting patterns by virtue of the diastereotopicity imposed on the alkyl side chains. 49 Moreover, significant exchange crosspeaks were observed in the NOESY spectrum indicating that the rate of interconversion between these two enantiomers at room temperature is relatively slow. This probably, reflects conformational stability imparted to the macrocycle as a result of strong hydrogen bonding between the pyrrolic NH's and the chloride anions. Figure 23. Schematic representation of 117 showing the interconversion between the two enantiomeric forms. These solution-phase inferences were ultimately confirmed by solid-state structural information derived from single-crystal X-ray crystal analysis (figure 24), wherein both of the twisted enantiomers were observed. As depicted in the figure below, the assumed C2-symmetric, twisted "figure-eight" conformation is clearly evident. Within each of the four smaller "pockets" defined in this arrangement are located the chloride counteranions, held in place by hydrogen bonding interactions with the N-H groups. This chloride anion ligation will undoubtedly pave the way for future investigations of other anion chelation and even possibly the use of turcasarin as a solution-phase transport agent. In addition to the above cavities, each loop serves to further define two distinct pentapyrrolic "hemi-macrocycles" which should display reasonable propensity for complexation of large metal cations. This has indeed been realized, with the successful synthesis of the bis-uranyl chelate by Sessler's 50 group. Interestingly, the twisted geometry is maintained in this complex as was inferred from the diastereotopicity observed in the 1 H NMR. Figure 24. X-ray structure of 117 (top view), showing the twisted "figure eight" conformation, with an inter-plane separation at the "crossing point" of 3.268 A, and the CI" anion binding.70 5.0 Vinylogous Porphyrins 5.1 Bisvinylogous Porphyrins In the late 70's Berger and LeGoff realized that by inserting odd numbers of carbon alternately between the pyrrolic rings of porphine would give rise to a series of coplanar porphyrin vinylogues, the platyrins.71 The 227r-electron [1,3,1,3]platyrin 125 was the first in this series to appear in the literature. The extra bridging atoms in this macrocycle were introduced in the early stages of the synthesis by an acid catalysed condensation of 3,4-dimethylpyrrole (118) with 1,3-cyclohehanone (119) followed by partial reduction of the intermediate salt 120. The trimethine 121, thus obtained, when coupled with its bisformyl product 122 in the presence of air 51 provided, presumably through macrocycle 123, the diprotonated platyrin 124 and upon deprotonation the corresponding free-base 125. 124 125 Scheme 19. Synthetic route to [1,3,1,3]platyrin 125. 52 The 1 H NMR of the latter free-base species has not been reported, but that of the dication 124 is consistent for a highly diatropic system. Typically, a single resonance for the methine protons is observed at 11.64 ppm, with the internal vinylic and pyrrole protons appearing at higher field positions (-8.97 and -5.6 ppm respec-tively). The methyl groups also show a 2.1 ppm downfield shift for the latter molecule when compared to those of the trimethine salt 120. A similar comparison of the respective signals of the outer peripheral protons of the cyclohexenyl group of this same pair of compounds reveals that these protons do not appear to be affected by the diamagnetic ring current at all. The UV-visible spectra of both 124 and 125 contain a dominant Soret-like absorption at 477 nm (398,000), with several other low energy absorptions. These bands are generally more red-shifted in the free-base platyrin 125 than in 124, with the lowest energy absorption band at 846 nm (1,850) for the former compound and at 788 nm (6,220) for the dication. A group of French researchers at Montpellier, whilst working with bis(trimethylsilyl)-amines as the nitrogen source in synthetic routes towards five-, six-, or seven- N-membered heterocycles inevitably devised a rather elegant sequence to two novel platyrins.72'73 In their approach (scheme 20), the precursor dipyrrolic units were obtained by an initial carbocupration of [bis(trimethylsilyl)amino]methylpropriolate 126 followed by treatment with an appropriate diacid chloride (e.g. isophthalic acid dichloride). Although this reaction proceeds with moderate yields, it does however, provide a facile, short access to substituted bis(pyrrolyl) derivatives with free ex-positions at the pyrrole rings. Cyclization of the dipyrrolic units 127 and 128 with 53 benzaldehyde in acetic acid, and subsequent oxidation with DDQ gave the fully conjugated nonaromatic macrocycles 129 and 130 respectively. Both of which can be formulated to fully aromatic 26n systems, but oxidation to the latter species can only be achieved at the expense of the 67c aromatic systems. Berlin and Breitmaier have, however, very recently shown that such oxidations to reveal the aromatic (4n + 2)71 porphyrinoid are unattainable.13 ,74 1. Me(Hexenyl)CuLi, Et02C C02Et 129 X= CH 130 X= N Scheme 20. Corriu's synthetic pathway to the conjugated nonaromatic platyrin-type macrocycles 129 and 130. The potential of these new macrocycles as ligands, was aptly demonstrated by the bimetallic complexes formed with Pd, Ni and Rh by 129. 7 2 , 7 3 An X-ray structure determination (fig. 25) of the bis-rhodium complex revealed that 129 accommodates both Rh atoms on the same side of the molecule by adopting a saddle shaped conformation which, in addition, probably relieves any resultant strain.7 3 Contrary to 54 the latter findings, similar bimetallic complexes of the porphyrins are reported to contain the metal atoms above and below the mean plane of the macrocycle.7 5 Figure 25. Crystal structure of the bis-rhodium complex of 129; side view, showing the saddle-shape and the c/s-coordination of the metal cations.73 Almost a decade later, LeGoff's group reported the bisvinylogous homologue of 125, the expanded [1,5,1,5]platyrin 132.76 The additional rc-bond was introduced between the pyrroles by using 2,7-dimethoxy-1,4,5,8-tetrahydronaphthalene 131 as the bridging group (scheme 21). The synthetic sequence followed their previous attempt as shown in scheme 19. Unfortunately, this platyrin (in both the free-base and diprotonated forms), unlike the 227x-platyrin 125, proved to be highly reactive in both 132 Scheme 21. LeGoff's route towards the [1,5,1,5]platyrin homologue (132) of macrocycle 125 55 solution and in the solid state, thereby indicating that there is very little resonance stabilization in this molecule. Such an observation is, nonetheless, consistent with the theoretical view that for the higher [4n + 2]annulenes, the resonance energy is inversely proportional to the number of 7t-electrons.77 Despite its instability, [1,5,1,5]platyrin 132 exhibits a significantly pronounced dia-magnetic ring current in the 1 H NMR, in which, the meso-like methine protons are located at 8 11.75 ppm, whilst the internal vinylic and pyrrole NH protons are shifted to extremely high field positions (8 = -14.26 and -10.58 ppm respectively). Thus, the 26TC macrocycle 132 has, as expected,7 7 an even greater ring current effect than that of the 22TC compound 125. The electronic spectra of this expanded system has two intense bands at 495 (123,000) and 536 nm (144,000) and several weaker bands ranging between 651 and 780 nm. The dicationic salt of 132, displays a virtually identical UV-visible spectrum to its neutral precursor 132, but with an additional red-shifted band at 830 nm (19,000). Shortly thereafter, Franck and coworkers reported their contributions to this area of expanded porphyrins,78 with the synthesis of what could formally be considered the parent form of LeGoff's [1,3,1,3]platyrin 125. For their initial approach, they looked towards the time-honoured MacDonald "2 + 2" methodology frequently employed in total syntheses of normal [18]porphyrins.79 Of the two possible pathways presented by this "2 + 2" strategy, they found the route employing dipyrromethanes as the precursors more successful than the alternative approach (which is analogous to scheme 19) utilizing the unstable dipyrrylpropenes (i.e. compounds analogous to 121 and 122).78Thus as shown in the scheme below, dipyrromethane 133 served as the common intermediate, which when treated with dimethylaminoacrolein and 56 POCI 3 at -20°C provided the second half of the macrocycle, the vinylogous dialdehyde 135. A subsequent acid catalysed coupling of these two dipyrro-methane units (134 and 135), followed by dehydrogenation with bromine and neutralization yielded the deep-green octaethyl[22]porphyrin(1,3,1,3)-octaacetate 137a. Me2N POCI 137a R 1 = R 2 = C02Et 137b = 137a.2 HBr 138a = H, R 2 = CH2C02Me 138b = 138a.2 HCI 135 R 1 = R 2 = C02Et 136 RT = H, R 2 = CH2C02Me Scheme 22. General synthesis of [22]porphyrin(1,3,1,3) derivatives (137-138) via the MacDonald-type "2 +2" coupling sequence. In a similar fashion, they later synthesized the [22]coproporphyrin II 138 (scheme 22). 8 0 The idea behind this synthesis being that by incorporating the most favourable properties of hematoporphyrin (i.e. the propionic ester moieties) for photomedicinal applications coupled with expected more intense long wavelength absorption of the resultant expanded porphyrin (compared to that of the blood pigment hemin and hematoporphyrin), would produce a vastly improved photo-57 sensitiser. As was anticipated, 138 did actually display a slightly improved 1 0 2 production in comparison to isohematoporphyrin (1.7 times greater). 1. CH20, MeOH HBr 2. DDQ, CH2CI2 143 142 Scheme 23. Franck's "biladiene-type pathway" in the synthesis of [22]octaethylporphyrin 142 and its electrophilic deuteration. In the same year, Franck et al. also described an alternative more efficient route, this time based on a "biladiene-type pathway", to furnish the extremely stable octaethyl analogue of 137a. 8 1 Here, the requisite bisvinylogous biladiene 141, was obtained in high yield by condensation of the vinylogous formyl pyrrole 140 with the bis-oc-free dipyrromethane 139 under acid conditions. The bisvinylogous octaethyl-porphyrin 142 was then obtained in a two step sequence involving ring closure with formaldehyde and an in situ oxidation with DDQ . 58 The UV-visible spectra of these macrocycles are all dominated by an intense Soret absorption at 463 nm (142 and 138a), 8 0 8 1 and 469 nm (137a), with, e.g. for 137a, additional bands at 498 (103,100), 597 (280,000), 723 (6600) and 812 (2700).78 Upon protonation, to give 137b, the Soret band becomes extremely narrow and significantly more intense (e =1,090,000 M"1 cm"1) due to an increased symmetry of the molecule giving rise to more similar electronic transitions. The 1 H NMR of 137a reveals a substantial diamagnetic ring current effect, with the triplet of the inner protons at the trimethine bridge located at 8 = -8.19 ppm, and the doublet of the outer protons at the same part of the molecule at 8 = 11.91 ppm, giving A8 a value of 20.1 ppm. The meso protons are also strongly shifted to a lowfield position of 10.59 ppm. The analogous [22]porphyrins 138 and 141 exhibit similar chemical shifts in the proton NMR. The most significant information that can be drawn from this NMR data and that of the platyrin 125 is an obvious enhancement of the diatropic ring current as the macrocycle is forced to adopt a stable planar conformation. That this results from an increased rc-electron derealization brought on by the incorporated pyrrole and cyclohexane units, one needs only compare the former NMR data with that of Sondheimer's [22]annulene 4 6 , 4 7 b for which A8 = 10.3 ppm. In the latter annulene, the planar conformations necessary for sustaining a diatropic ring current in an applied external magnetic field are only sufficiently stable at low temperatures. Single crystal X-ray analyses of the bistriflouroacetate of the dication of 142 (figure 26) and of the dihydrochloride 138b confirmed the planarity of this set of expanded porphyrins, with a mean deviation (excluding the (3-side chains) from the cyclic framework of 0.6 pm. 8 0 , 8 1 In both solid state structures the counteranions were each coordinated to a dipyrromethane unit above and below the macrocyclic plane via hydrogen bonding. Furthermore, the bond lengths within the 227c-electron perimeter 59 of these bisvinylogous porphyrins are virtually identical, and are within the range of the ideal C-C bond lengths of benzene (139 ppm). Thus, in conclusion these expanded porphyrins not only fulfil the ring current criterion for aromaticity, but also the second most important criterion pertaining to C-C bond lengths.4 7 3 Figure 26. Molecular structure of the centrosymmetric dication 142. Left: top view perpendicular to the arene plane. Right: Side view.81 A final noteworthy feature of the octaethyl analogue 142 is the differing chemical reactivity observed for the methine protons. Particularly, the outer peripheral protons are exclusively susceptible to electrophilic substitution, whereas the inner protons are inert. For instance, deuteration of 141 quantitatively yields the hexadeutero derivative 142 in which all the external methine protons are exchanged.8 1 Fairly recently, Franck's group synthesized the next higher homologue in this series, the vinylogous [26]octaethylporphyrins 146-149,82 in an analogous fashion to their earlier synthesis of the 227t compound 142. The preference for this "biladiene synthetic pathway" is clearly underlined by (i) the ease at which the appropriate tetrapyrrolic precursors can be synthesized, and (ii) it further allows for the simultaneous preparation of meso- and hetero-substituted derivatives. Thus, an HBr catalysed condensation of the bisvinylogous pyrrole aldehyde 144 with dipyrro-60 methane 139 gave the open chain tetrapyrrole 145 (scheme 24). Cyclization of the latter with either formaldehyde, propionaldehyde, or benzaldehyde and in situ dehydrogenation with DDQ, gave the [26]porphyrin 146, and its 14-ethyl and 14-phenyl derivatives 147 and 148 respectively. Alternatively, reaction with ammonia followed by oxidation of the hydroporphyrinoid intermediate transformed 145 to the novel 14-aza[26]porphyrin 149. 149 146 R = H 147 R = Et 148 R = Ph Scheme 24. Synthesis of octaethyl[26]porphyrin 146, and its 14-ethyl 147, 14-phenyl 148, and 14-aza 149 derivatives, from the tetravinylogous biladiene 145. As expected, 146 is highly diatropic, sustaining a pronounced aromatic ring current in the 1 H NMR. For instance, the internal pentamethine bridge and pyrrolic protons 61 resonate at 8 = -9.79 and -5.77 ppm respectively. Conversely, the monomethine protons appear as a singlet at 12.26 ppm and the exterior pentamethylene protons as a triplet at 14.35 ppm. A comparison of the diatropicity of this system (A8 = 24.1 ppm) with that of its smaller 22n; homologue 142 (A8 = 22.3 ppm)81 reveals an increased diamagnetic ring current effect, which is apparently similar to that of LeGoff's [26]platyrin(1,5,1,5) 132 (A8 = 25.3 ppm).76 In contrast, the diatropicity of the meso-substituted derivatives decreases from A8 = 21.6 (for 148), through 20.7 (for 147) to 18.9 ppm for the aza derivative 149, but are still within range of that of 142. In their bisprotonated forms, these [26]porphyrins all exhibit a strong Soret band between 503 and 525 nm, which are typically 40-60 nm more red-shifted than that of 142. Of these four compounds, the aza derivative 149 exhibits the longest wave low-energy absorption at 735 nm (e = 27,500). Single crystal X-ray analysis of the bistrifluoroacetate of 146 revealed some unusual results. For instance, the unit cell contained two different conformers of the molecule in a 2:1 ratio. These conformers simply differ with regards to the arrangement of the P-ethyl substituents of the dipyrrylmethane units. Thus, there are two pseudo-C2 symmetrical a\\-trans conformations in which the ethyl side chains are bent away from each other (fig. 27). The third conformation being that of the centrosymmetric cis-trans arrangement. Nevertheless, the ring framework in both conformers are largely planar, and more interestingly, the bond lengths in the conjugated perimeter are almost equal (137.1-139.6 ppm) and in good agreement with those of its 227c homologue 142. 62 Figure 27. Crystal structure of the a\\-trans conformation of 146.2TFA. Left: Side view, Right: Top view of the arene plane.82 In marked contrast to the homologous [22]porphyrin 142, electrophilic deuteration of the more stable 14-phenyl[26]porphyrin 148 shows no selectivity over the outer peripheral protons. In this instance, from NMR and mass spectroscopy of the product, all the inner bridge protons and some of the corresponding outer bridge proton are exchanged. This differing chemical reactivity, though consistent with the aromaticity of 148, may be rationalized in terms of the size of the porphyrinoid rings, and the inductive effects of the phenyl substituents. 5.2 Tetravinylogous Porphyrins Progressing from the above series, the next obvious expanded porphyrins encompass compounds in which all four of the single atom meso bridges are enlarged. In 1986, Franck and Gosmann synthesized the first member in this series of tetravinylogous porphyrins via a biomimetic sequence analogous to porphyrin biosynthesis. 8 3 , 8 4 Here, an acid catalysed cyclotetramerization of the N-methylpyrryl-propenol 150 followed by dehydrogenation of the resultant macrocycle 155 gave the bisquaternary octaethyl[26]porphyrin 160 in just two steps. This expanded porphyrin 63 R R 160 R = Et 161 R,R'= -(CH2)4-162 R= nBu 163 R = 'Bu Scheme 25. Biomimetic synthesis of porphyrinogens 155-159, and their subsequent oxidation to the corresponding [26]porphyrins 160-163. exhibits a remarkably strong diatropic ring-current effect in the 1 H NMR, such that the internal protons of the methine bridges and the N-methyl groups appear as a triplet at 8 = -11.64 ppm and a singlet at -9.09 ppm respectively. The outer protons 64 of the methine bridges are shifted to the low field position (8 = 13.67 ppm). The aromatic porphyrin-like nature of 160 is manifested in its absorption spectrum by the presence of an intense Soret band at 546 nm (840,00), and a remarkably intense long wave absorption at 783 nm (28,800) with a shoulder at 830 nm. 8 5 Photophysical studies of this 26TC system, coupled with its high thermal and photochemical stability, have served to identify 160 as a potential candidate for application in PDT. 8 5 R R R H O H * C R 164 Figure 28. Helical conformation of the tetrapyrrole Figure 29. Porphobilinogen (165) precursor for [26]porphyrinogens. The most striking feature of this synthetic strategy, is the selectivity observed for the formation of macrocycle 155 (e.g. 27%). Evidence that this results from a conformational helical effect came from the synthesis of the p-unsubstituted and (3-substituted derivatives 156-159 (scheme 25); where, an increase in yield (from 0.1% to 52%) of the latter macrocycles is observed as the bulkiness of the side-chains progressively increases.8 4 Clearly, the steric congestion of the pyrrolic (3-side chains force the tetrapyrrolic precursors to adopt the more favourable helical conformation 164 shown in figure 28 immediately prior to the final ring closure. Molecular 65 modelling of the acyclic tetrapyrroles further support this supposition. Incidentally, this parallels the results of analogous biomimetic syntheses of normal [18]porphyrins from porphobilinogen (165) wherein, the [18]porphyrin is almost exclusively formed. 1 1 0 In this instance, however, this helical effect is more obvious due to the reduced distance between the pyrrole moieties. Dehydrogenation of macrocycles 157-159, furnished the corresponding aromatic congeners 161-163, which exhibit similar spectroscopic properties to 160. 8 4 Moreover, all these [26]porphyrins exhibit an equivalent diamagnetic ring current effect in the NMR (cf. A8 = 25.0 to 25.3 ppm) to the isoelectronic [26]porp-hyrin( 146 (A5 = 24.1 ppm).82 Subsequent application of the above principles culminated in the synthesis of the next higher homologue of 160, the octavinylogous [34]porphyrin 169. 8 6 The analogous acid catalysed condensation of pyrrole 166 gave both the predicted cyclic tetramer 167 and, surprisingly, the larger cyclic pentamer 168 in low yield (scheme 26). Obviously, here the ethyl side-chains have no effect on the outcome of this reaction via the helical effect described previously, due (simply) to the much increased distances between the pyrrole rings upon condensation. Oxidation of the nonconjugated [34]porphyrinogen 167 with bromine afforded the highly stable aromatic [34]porphyrin( 169 as its dibromide salt. The increased ^-conjugation of this expanded porphyrin is manifested in its UV-visible spectrum which contains an intense absorptions at 663 nm (e = 370,000), which is considerably red-shifted from the corresponding Soret absorption of the [18]porphyrins by ca 260 nm. An additional band (A,max = 997 nm, e = 24,000) is also observed in the near IR region of the spectrum. Unlike its 26TC counterpart (160), 66 169 Scheme 26. Biomimetic synthesis of the aromatic [34]porphyrin( 169. 67 preliminary photophysical data revealed 169 does not generate 1 0 2 , thereby rendering it unsuitable for PDT application.85 The latter does, however, attest to the recently reported long wavelength photochemical limitation of about 800 nm. 2 a Characteristically, the 1 H NMR exhibits effects associated with an extraordinarily large diamagnetic ring current. Here, the internal methine proton signal is found at 8 = -14.27 ppm and the corresponding external protons resonate at 8 = 17.19 ppm, thereby giving a maximum shift difference, A8, of 31.5 ppm. The signal for the N-methyl protecting group also appear at higher field (8 = -11.44 ppm). This extension of the proton resonances over such a wide chemical shift range is highly atypical for a non-organometallic compound. Nevertheless, it does attest the diatropic character of this expanded porphyrin, and even more importantly to the assumption of a stable planar conformation. The efficiency with which the pyrrolic rings arrest this 347X ring system in stable conformation is particularly noteworthy in view of the fact that [22]annulene 68b, for example, is extremely mobile at ambient temperatures. However, it should be appreciated that part of the stabilization incurred in Frank's bisvinylogous (160-163) and octavinylogous (169) expanded porphyrins may be attributed to the positive charges as was observed with Vogel's tetrapropyl-[26]porphyrin( 102.9 This series of vinylogous porphyrins are perhaps the most interesting of the expanded porphyrins in terms of their size and spectroscopic properties, particularly with regard to their NMR. Table 1 summarizes the maximum differences between chemical shifts (1H NMR) of the vinylic inner and outer ring protons of a representative sample of [22]-, [26]-, and [34]porphyrins, and for comparison, Sondheimer's [22]annulene. The A8 values given below provide a reasonable qualitative assessment of the size of the aromatic ring current in these diatropic 68 systems. 4 7 , 5 9 This data conclusively confirms the validity of Huckel's (4n + 2) rule in regards to aromaticity in the higher cyclic annulene-type systems, provided sufficient stabilization of a planar conformation is possible. Furthermore, the extreme diatropicity of 169 and its 267U congeners (146 and 160) lays to rest previous theoretical predictions that such annulenes with 267i-electrons and beyond ([22]annulene being the maximum limit)87 would essentially be atropic and contain localized bonds. 4 7 b Also evident from the figures above is an increase in diamagnetic ring current with the number of re-electrons, which is consistent with Haddon's unified theory of aromatic character76- a mathematical approach that predicts ring currents to be directly proportional to the number of Tt-electrons in these (4n + 2)% annulenes. Table 1. Maximum 1H NMR chemical shift differences, A5, between inner and outer protons of [4n + 2]vinylogous porphyrins, resulting from diatropic ring current effects. Compound A8 /ppm Solvent (Temperature/ °C) [22]porphyrin( 142 8 1 20.1 CDCI 3 (20) [26]porphyrin( 146 8 2 24.1 DMSO-d 6 (20) [26]porphyrin( 1 6 0 8 3 8 4 25.3 CDCI 3 (20) [34]porphyrin( 169 8 6 31.5 C 6 D 6 (20) [22lannulene 68b 4 6 , 4 7 b 10.3 THF (-90) 6.0 Porphocyanines Dolphin et al. have recently reported a new series of aromatic tetrapyrrolic expanded porphyrins which incorporate the combined structural features of porphyrins and phthalocyanines.8 8 , 8 9 Hence, in view of the latter features they 69 christened these new macrocycles the porphocyanines e.g. structure 174. Additionally these macrocycles resemble structurally Franck's [22]porphyrin( 142 but, with the central methine groups of the expanded vinylogous bridge replaced by a pair of imine linkages. Consequently, as ligands, the porphocyanines do not suffer from the limitations imposed upon the bisvinylogous porphyrins by the E geometry of the bridging methines and the presence of H atoms within the central binding core. The octaethylporphocyanine 174 was obtained via a sequence involving hydride reduction of the biscyanodipyrromethane 170, followed by self-condensation of the resultant unstable bis(aminomethyl)dipyrromethane 171 (with concomitant loss of ammonia) in refluxing MeOH/THF, and a final oxidation of the intermediate porphyrinogen. An alternative approach, also reviewed in the scheme below, furnished porpho-cyanine 174 in much improved yield (24%).89 This synthesis involved heating the bisformyldipyrromethane 172 in ethanol saturated with ammonia in a sealed vessel, under rigorously anhydrous conditions. Analogously, the 4,4'-methoxypropylate-5,5'-bisformyldipyrromethane 173 undergoes cyclo-condensation, with concomitant trans-esterification under the basic conditions, to give the aromatic macrocycle 175, which due to the labilitly of the ester groups towards LiAIH4 cannot be prepared via the previous reduction sequence. This preferential formation of the cyclic compound over higher linear oligomers is particularly remarkable in that, this reaction occurs in the absence of any metallic (or anionic) templates. The exact mechanism of this process is as yet not clearly understood but, is currently under further investigation. 70 H H H H 170 171 1. THF/ MeOH, reflux, N2 2. 0 2 > CH2CI2 T 174 R 1 = R2 = Et 175 R., = Me, R2 = (CH2)2C02Et Scheme 27. Two single-pot strategies to the tetrapyrrolic porphocyanine macrocycles 174 and 175. These researchers subsequently discovered that the cyclization step in refluxing THF/MeOH was not actually a prerequisite in the self-condensation of the presumed 5,5'-bis(aminomethyl)-2,2'-dipyrromethane intermediate products (e.g. compound 171) of the hydride reduction reaction. Thus, addition of DDQ to the crude reaction mixture, after destroying the excess LiAIH4, directly provided the aromatic porpho-cyanines 180-183 in modest yields (scheme 28). 9 0 , 9 1 Additionally, this oxidation could equally be effected by simply bubbling a stream of air through the reaction vessel following the reduction reaction.90 Thus, like the porphyrinogens, the initial unconjugated macrocyclic products of this "2 + 2" cyclization are highly unstable towards oxidation to their aromatic congeners. Moreover, the versatility of this one-71 pot sequence now enables syntheses of asymmetric macrocycles (e.g. 184), by co-reducing the two appropriately substituted dipyrrolic subunits within the same vessel as exemplified in the scheme below. 184 Scheme 28. The reductive-macrocyclization sequence in porphocyanine syntheses. The conjugated aromatic nature of the 227i-electron porphocyanines is evident in their optical spectra, which contain a dominant Soret-like absorption at 457 nm (e = 240,000) for 174, and Q-type absorptions at Xmax = 592 (17,000), 633 (5,800), 728 (3,200) and 797 nm (27,000).88 The introduction of meso-phenylene substituents on the porphocyanine ring, as anticipated, increases the overall derealization pathway, 72 and thus brings about an additional ca 15 nm bathochromic shift of the major Q-type band for macrocycles 180-182.91 With the more sterically congested porphocyanine 183, in contrast, a hypsochromic shift (relative to 180) is observed. The 1 H NMR of 174 exhibit a diamagnetic ring current of similar magnitude to bisvinylogous porphyrin 142. Thus, the single meso-bridge protons and the four other peripheral imino protons resonate at 8 = 11.95 and 13.75 ppm (as singlets) respectively, whilst the inner pyrrolic NH signals appear at 8 = -5.75 ppm. 8 8 CIS C17 Figure 30. Left: top view of 174.ZnCI2 showing the tetrahedral coordination around the Zn and hydrogen-bonding interactions between the pyrrolic NH and CI ligands. Right: Side view of this complex, showing the planarity of the ring framework and the metal centre located within the porphocyanine plane.88 Insertion of zinc into porphocyanine 174, gave a compound which exhibited a markedly different electronic spectra, with a Soret band at 464 nm and lowest energy Q band at 736 nm. Addition of base effects a further red-shifting of the latter band to 762 nm, thereby indicating the presence of acid pyrrolic protons. This, incidentally, was confirmed by the presence of a high field singlet (8 = -5.50 ppm) in the 1 H NMR. A single crystal X-ray analysis established the 1:1 zinc: macrocycle complexation and the planarity of the macrocyclic framework (Figure 30). 8 8 The zinc 73 is totally encapsulated within the macrocyclic cOre coordinated to two of the pyrrolic nitrogens in one half of the macrocycle, and two chlorine atoms in a tetrahedral fashion. The remaining two pyrrolic nitrogens being protonated form hydrogen bonds with the chloride anions and presumably account for the stability of the complex. The imine nitrogens, however, are not implicated in complexation. 7.0 Sapphyrins and Heterosapphyrins The sapphyrins, so called due to their intense blue-green colour, were the first known examples of the expanded porphyrins. Initially discovered serendipitously by Woodward some 30 years ago 9 2 , 9 3 during the course of their earlier attempts towards the synthesis of vitamin B 1 2 , and later isolated in small quantities (as the 25,29-dioxosapphyrins) as by products in the synthesis of 21,24-dioxacorroles by Grigg's group in Nottingham.94"96 As originally formulated by Woodward, they constitute a series of 22 7t-electron pentapyrrolic macrocycles containing a single direct link and four methine bridges between the pyrrolic subunits. The extended conjugated 7t-electron system, and aromatic character of the sapphyrins are particularly noticeable in their spectroscopic properties. The electronic spectra bear much resemblance to that of the normal porphyrins, but the Soret band, for instance, being significantly shifted (by ca 30-60 nm) to longer wavelength than that of porphine 1 . Similarly, the proton NMR exhibits characteristic diatropic ring current effects. Although known for sometime, the full potential of this class of expanded porphyrins was not exploited till fairly recently, when the emergence of newer biomedical 74 applications regenerated interest in these compounds. These efforts, have additionally revealed more intriguing chemical properties of the sapphyrins and other related macrocycles, such as, for example, anionic chelation.31 Much of the earlier work in this area largely focused on the aromaticity of such systems, as a means for verifying Huckel's rules. Nonetheless, these studies did, however, lay the groundwork for the subsequent chemistry related to the sapphyrins. 7.1 Synthesis and Spectroscopic Properties The basic strategy in the synthesis of sapphyrins has remained substantially unchanged since Woodward's pioneering work. 9 2 , 9 3 , 9 8 As outlined in scheme 29 it involved a MacDonald-type "3 + 2" condensation between a diformyl bipyrrole such as 105, 116, 186 and a bis(pyrrolylmethyl)pyrrole diacid, e.g. 187, as the key step. The other approaches of Grigg and Johnson 9 4 , 9 6 , 9 7 that followed shortly, essentially mirrored this procedure, whilst the more recent efforts of Sessler 9 9" 1 0 2 differed mainly in the efficiency of producing the requisite bipyrrolic and tripyrrane precursors. Variations on the above theme were initially reported by Grigg and Johnson in the early 70's and more recently by Sessler, in which, the basic sapphyrin structure was modified by replacing pyrroles with thiophene and furan rings to furnish the 25,29-dioxasapphyrins 195 1 0 0 , 1 0 1 and 196 9 4 , 9 6 3 , 9 7 and the 27-thiasapphyrin 19496 ,97 and 197100 in accordance with the sequence shown in scheme 29. Recently, inspired by their successes at attaining the prerequisite symmetric tripyrranes 187 via a vastly improved convergent approach 2 1 , 3 2 Sessler's group detailed a more efficient versatile strategy towards these heterosapphyrins 195 and 197, and three 75 R2 R2 R i OHC CHO C02H C02H 105 R1 = nPr, R2 = H, X= NH 116 R 1 = Et, R2= Me, X= NH 185 R 1 = R2 = H, X= O 186 R^ Me, R2 = Me, Et, (CH2)2C02Me, X= NH 187 R3, R4, R5, R6 = Me, Et, (CH2)2C02Me, etc., Y = NH 188 R3 = Me, R4 = Et, R5 = R6 = H, Y = S 189 R3 = Et, R4 = Me, R5 = R6 = H, Y = S 190 R3 = Et, R4 = Me, R5 = R6 = H, Y = O H+/ EtOH O, 191 R, = R4 = R5 = R6 = Et, R2 = R3 = Me, X = Y = NH 192 R1 = R2 = R3 = R4 = R5 =R6 = Me, X= Y = NH 193 R,, R2, R3> R4, R5> R6 = Me, Et, (CH2)2C02Me, etc, X = Y = NH 194 R, = R3 = Me, R2 = R4 = Et, R5 = R6 = H, X = NH, Y = S 195 R, = R2 = H, R3 = Me, R4 = R5 = R6 = Et, X= O, Y = NH 196 R, = R2 = H, R3 = R5 = Me, R4 = R6 = Et, X= O, Y = NH 197 R, = R3 = Et, R2 = R4 = Me, R5 = R6 = H, X= NH, Y = S 198 R, = R3 = Et, R2 = R4 = Me, R5 = R6 = H, X= NH, Y = O 199 R, = R2 = R5 = R6 = H, R3 = Et, R4 - Me, X= S, Y = O 200 Ri = R2 = R5 = R6 = H, R3 = Et, R4 = Me, X = Y = O Scheme 29. The general [3 + 2] synthetic route to sapphyrins 191-200. 76 other novel furyl and thienyl analogues 198-200.101 The general sequence (scheme 29), however, still relied on this critical oxidative "3 + 2" condensation between the key dipyrrolylfurans 190 or the dipyrrolylthiophene 189 with either the diformyl bipyrrole 116 or the diformyl bifuran 185, to furnish the stable macrocycles 197-199 in ca 35% yield. The 25,27,29-trioxasapphyrin 200, unfortunately, proved too unstable to be isolated. However, evidence for its formation was derived from the optical spectrum of the reaction mixture where, the presence of Soret-like band at 450 nm was observed. 1 0 1 An alternative synthesis of the 25,29-dioxasapphyrin 196 was achieved by an HBr catalysed condensation of bis(formylfuryl) sulphide 202 with tripyrrane 201 followed by an in situ oxidation of the initially formed dicationic macrocycle 203 with iodine. 9 6 , 9 7 , 1 0 3 This reaction proceeds via extrusion of sulphur in a cheletropic process from the intermediate thiiran 205, which incidentally is itself, the product of a disrotatory cyclization (204->205, scheme 30) . 9 4 9 7 The mechanism of this process is believed to be analogous to that elucidated for the formation of corroles from rneso-thiophlorin.94,95 The 1 H NMR spectra of the sapphyrins and the heteroatom analogues are consistent with their formulation as aromatic structures sustaining a large induced diamagnetic ring current. Typically, the signals ascribable to the two sets of meso protons are well resolved, appearing at 8 = 11.51 and 11.70 ppm (for the bis(hydrofluoroacetate) salt of decamethylsapphyrin 192).98 The ten methyl groups gave signals in the ratio of 1:2:1:1 at 4.04, 4.08, 4.19, and 4.22 while the internal pyrrole NH exhibit broader resonances at -5.46, -5.0, and -4.84 in the ratio of 2:1:2. Similarly, with dioxasapphyrin H2195.2CI the meso protons resonate at the low field positions of 11.40 and 11.97 ppm, the furan (3-H's at 12.32 and 12.35 ppm, 77 196 Scheme 30. An alternative route to the dioxasapphyrin 196, employing a sulphur extrusion reaction. 78 and the internal pyrrolic protons at -6.63 and -5.21 ppm. 1 0 1 The same group of protons in 27-thiasapphyrin 197 (for the free-base) resonate at 10.77 and 10.99 ppm (meso protons), 10.77 for the thiophene-(3-protohs, and -3.34 ppm for the NH's. 1 0 1 The larger diamagnetic ring current effects observed in the spectra of 195 and 192, compared to 197, results primarily from the cumulative effects of an additional charge-induced deshielding associated with such cationic species and the ring current.9 ,59 Increasing the solvent polarity has a marked effect on the 1 H NMR observed for 191.104 The meso H and pyrrolic N-H signals in CD 3 CN and CD 3 OD are broadened and shifted to higher field positions in comparison to the corresponding signals in CDCI 3. Thus in CD 3 CN the former protons resonate at 10.65 and 10.36 ppm whilst the N-H's are shifted to 5 = -9.58, -9.93 and -10.16 ppm now in a 1:2:2 ratio. In d 4-methanol, the pyrrolic N-H signals are, however, lost due to rapid exchange, but the meso protons are centred at ca 11.1 ppm as a broad peak. These spectral changes arise from strong association of the dicationic sapphyrins to form dimers in polar media, but in less polar media the monomeric form predominates. Stacking of the macrocycle planes introduces strong n-n interactions between adjacent rings which perturb the ring currents thereby inducing significant upfield shifts in the NMR signals. Typically, as with all other porphyrinoid-type compounds, the electronic spectra of the sapphyrins and its furan/thiophene derived analogues, are dominated by an intense Soret band in the 435-470 nm region for the free-base and mono- and dicationic forms (vide infra). In addition, a series of less intense Q-bands are observed in 600-750 nm spectral region. 9 6" 1 0 1 The table below summarizes the UV-visible properties of a representative sample of the sapphyrin, where it can be seen 79 that thiasapphyrins (e.g. 194, 197) exhibit a more red-shifted Soret with respect to the all nitrogen compound, whilst the oxa analogues exhibit the reverse effects.101 Upon protonation, thereby increasing the symmetry, the Soret-like bands noticeably shift slightly, and increase in intensity by approximately one order of magnitude. The exact position of this absorption band however, is significantly influenced by the nature of the counter anion (see Table 2) . 9 9 " 1 0 1 , 1 0 5 The Q-bands, on the other hand, shift more towards the blue region (between 570 and 690 nm) of the visible spectrum, but also show an increase in intensity. Additionally, the spectra are strongly dependent on the solvent polarity,1 0 5"1 0 7 for example, the UV-visible spectrum of 191 (in ethanol) exhibits a 12 nm blue-shift in the Soret band as compared to that in toluene.1 0 7 This hypsochromic shift is indicative of the formation of dimers in more polar media. The Q-bands on the other hand do not show the expected red-shift associated with the increased 71-71 interactions in the dimers, but show a blue-shift instead, i.e., 674-714 nm in toluene vs. 668-704 nm in ethanol. This blue-shifting in the Q-bands (upon acidification of toluene solutions or increased solvent polarity) has been interpreted as resulting more so from the effects of the dielectric constant on the energy gap between the (n-7t*) and (71-71*) states. 1 0 7 These effects are obviously substantial, as they effectively screen out the bathochromic shifts induced in the Q-bands upon dimerization. 1 0 5 , 1 0 6 Conformation of the different physical states of these dications in media of high and low dielectric constant was provided from fast EPR-magnetophotoselection (MPS) spectroscopy experiments on the photoexcited triplet states of H21 91 2 + . 1 0 5 , 1 0 7 Furthermore, these experiments revealed that: (i) the blue-shifts observed around 700 nm may result from a possible admixture of the (n,7t*) and (71,71;*) states; and (ii) the dimers involve both specific solvent-mediated 80 hydrogen bonding and n-n interactions. Nevertheless, collectively this optical and 1 H NMR data suggest that sapphyrins have a low energy HOMO-LUMO gap. Table 2. Optical Properties of typical Sapphyrins, Heterosapphyrins, and Sapphyrin Salts. Sapphyrin/Salt Soret /nm Q-bands /nm 191 456 616, 668, 711 195 442 555, 590, 673, 684, 752 197 463 644, 691 198 455 582, 636, 682, 722, 749 199 438 587, 670, 744 H2191.2CI 456 576, 624, 675, 686 H 2191-2Br 458 578, 624, 678, 689(sh) H 2 191.20Ac 450 622,676 H 2 191.2N0 3 449 620, 674 H 2191.2CI04 447 619, 672 H 2191.2F 446 572,619,670,676 7.2 Chemical Properties In the free-base form, the sapphyrins contain three normal protonated pyrroles and two "imine-type" nitrogens which are effectively pyridine-like. These latter pyrrolic nitrogens are relatively basic, and are readily protonated by weak acids such as silica gel. The basicity of these nitrogens are such that in the mass spectra of these macrocycles, the M + 2 peaks (i.e. for the dicationic species, in which all five N-atoms are protonated) are unusually intense (19-68%).9 6 , 9 7 Thus, the diprotonated dicationic species are the most stable form of sapphyrins 191-193. The mono-81 cationic species, existing only in the 3.5 < pH < 10 range, 1 0 8 has, however, been observed in the UV-visible spectra by Bauer et. a/. 9 8 The pK a values of 191 have recently been determined and found to be pK a 1 = 3.5 (H21912 +/H191+) and pK a 2 = 3.5 (H1917191).1 0 8 Likewise, the heterosapphyrins 195, 196, 198 are all reportedly highly basic and are isolated as the more stable diprotonated cationic salts. 9 7" 1 0 1 In contrast, the 27-thiosapphyrins 194 and 197 are markedly less basic, and can be isolated and characterized in their free-base form. 9 7 ' 1 0 0 , 1 0 1 The inclusion of heterocycles other than pyrrole in the sapphyrin framework also imparts differing reactivity at the meso positions. In electrophilic deuteration reactions of the dioxasapphyrin 195 in d r T F A at 100° C two of the four possible meso protons are exchanged within 2.2 h. 9 7 Prolonging the reaction (up to 100 h) only effects a slight exchange at the remaining meso positions. Precise assignment as to which pair of meso protons preferentially exchange has not proved possible. On the other hand, with decamethyl sapphyrin 192 all four meso protons are readily exchanged on exposing the latter to d r T F A over-night at room temperature, thus indicating the ease of electrophilic attack on the sapphyrin nucleus in this case. 9 8 However, extending this reactivity to nitration and bromination failed to yield the corresponding products.98 7.3 Coordination Chemistry (a) Metal Complexation With a planar pentadentate ligation core and an N-centre radius of about 2.7 A, one can anticipate a rich coordination chemistry for the sapphyrins, a chemistry that should differ from that of the smaller tetrapyrrolic porphyrins. However, to date the 82 organometallic chemistry of the sapphyrins largely remains in its infancy. The first reported metal complexes were those of the first row transition metals, with the C o 2 + and the Z n 2 + compounds isolated as crystalline material,98 which were believed to be highly symmetric structures (such as 206a) in which only four of the five pyrrole nitrogens participate in the chelation of the metal centre. A more recent independent 1 H NMR evaluation of Zn.H191 not only confirmed this incomplete ligation, but also provided strong evidence for the presence of two isomers 206a and 206b (i.e. from the splitting of the meso H signals from two into four singlets).1 0 0 In view of these observations, Sessler's group focused their attention on complexes of second and third row transition metals to fully exploit the coordination properties of the sapphyrins. Figure 31. Proposed structures of the isomeric tetraligated Zn-sapphyrin complexes 206. Treating the free-base sapphyrins 191 and 207 with 0.5 equivalents of [RhCI(CO)2]2 or 1 equivalent of lrCI(CO)2(py) furnished the corresponding metallo dicarbonyl complexes 208 and 209, which were isolated in their more stable cationic forms 210a, 211a, and 210b respectively- scheme 31.100-109 The asymmetric nature of these cationic complexes were clearly evident in the 1 H NMR, where four meso proton signals were observed in the 8 = 11.44-10.98 ppm (for 210) regions, together 206a 206b 83 H2191 R1 = Et, R2 = Me 212a M = M' = Rh(CO)2 H2 07 R1 = n Pr , R2 = H 212b M = M' = Rh(CO)2 213a M = M' = lr(CO)2 214a M = Rh(CO)2, M' = lr(CO)2 Scheme 31. Synthesis of mono- and bimetallosapphyrins 210-214. (i) [RhCI(CO)2]2, Et3N; (ii) [lrCI(CO)2(py)], Et3N; (iii) silica gel; (iv) Et3N; (v) HCI. a, = Et, R2 = Me; b, R1 = n Pr , R2 = H. 84 with three distinct high field signals (6 = -2.60, -2.71, -3.22 ppm) corresponding to the pyrrolic NH's. 1 0 9 A single crystal X-ray analysis of the iridium complex 211a shown in figure 32 confirmed this structural arrangement, and further showed that the metal centre was held out of the general macrocyclic plane. Treatment of compounds 208 and 209 with an additional equivalent of the metal carbonyl salts, or more conveniently reacting the free-base sapphyrins with an excess of the transition metal salt, gave access to the bimetallic complexes 212 and 213. Additionally, this stepwise metallation provides a convenient approach to the hetero-bimetallic complexes of sapphyrins 214a as reviewed in the scheme above. Figure 32. Structure of the iridium complex 211a.IUVJ All of these sapphyrin complexes show remarkable stability and are much less reactive towards mild oxidizing agents (such as methyl iodide and acetic anhydride) than similar porphyrin complexes. 1 1 0 On the other hand, iodine effects partial demetallation of the bimetallic complexes to yield the corresponding monometallo compounds 210 and 211a respectively, whilst with HCI complete cleavage of the metal centres from both mono- and bimetallic complexes results. The bimetallic complexes are, however, reportedly more stable than their monometallo counter-85 parts. For instance, solutions of the rhodium dicarbonyl sapphyrin complex 210a disproportionates on standing in air to 212a and the sapphyrin dication H2191. Furthermore, the CO ligands at the rhodium centres exhibit unusually high thermal and photo stability, despite the relatively high v(CO) values in the infra-red spectrum. Figure 33. Structure of the bimetallic complex 212a. I U U The trans arrangement of the metal centres in the bimetallic complexes 212 and 213a were confirmed by single crystal X-ray determinations (fig. 33) . 1 0 0 , 1 0 9 Thus, unlike the bisrhodium complex of platyrin 129 (fig. 25), 7 2 , 7 3 these bis(metallodi-carbonyl)sapphyrins bear structural resemblance to [Rh(CO) 2] 2.OEP 7 5 and the N-methylcorrolebis[dicarbonylrhodium(l)]111 complexes in both the arrangement of the metal centres in relation to the macrocyclic plane, and in that each metal atom is bridged between an imine and an amine nitrogen atom. The metal centre assumes a general square planar geometry, with the metal planes inclined at approximately 47° to the overall sapphyrin plane. Moreover, to accommodate this geometrical arrangement the sapphyrin framework is considerably ruffled, with the pyrrole rings distorted in such a way that the chelating N-atoms are directed towards the metal atoms. Changes in the electronic spectra are also observed with successive 86 coordination of transition metals. Generally, shifts in the Soret bands from ca 450 (for the free-bases), through 480 (for complexes 208-209) to 500 nm for the bimetallic complexes, are observed. Figure 34. Crystal structure of the uranyl complex 218.1 Extension of this work by the Austin, TX, group to the actinide series revealed an unexpected result. Contrary to previous reports,98 sapphyrin 191 was found to react rapidly with U0 2 CI 2 in a mixture of methanol, pyridine and triethylamine.112 The resultant uranyl chelate, however, exhibited a rather complex 1 H and 1 3 C NMR, whilst the UV-visible spectrum simply consisted of two broad absorptions at 479 and 508 nm. Such spectra are clearly atypical of an aromatic metallosapphyrin structure. Precise formulation of this complex was determined from an X-ray analysis (fig. 34), which showed that the uranyl moiety had indeed coordinated to the macrocycle. But, the sapphyrin ring had additionally undergone attack by a methoxide anion at a meso position (scheme 32), thus disrupting the inherent two-fold symmetry and the aromaticity of the macrocycle. The overall result is a diasteriomeric mixture which accounts for the intricate NMR spectroscopic data. Moreover, in the solid state, the 87 macrocycle is severely distorted so as to accommodate the uranyl group within the nitrogen donor plane in a pentagonal arrangement. 218 Scheme 32. Nucleophilic attack by methoxide anion on the sapphyrin periphery induced by uranyl insertion. 88 A tentative mechanism for this unusual metallation reaction is shown schematically above. Thus, from further experimental work, it appeared that attack by the methoxide anion on the initial uranyl complex 216 plays a vital role in stabilizing this uranyl sapphyrin complex. The resultant complex 217 then rapidly oxidizes at the uranyl moiety yielding the nonaromatic complex 218. Such reactions involving nucleophilic attack at the porphyrin periphery are known, but in all these examples prior activation of the macrocycle by the introduction of electron-withdrawing substituents on the porphyrin ring was necessary. 1 1 3 Moreover, this uranyl insertion is sensitive to both the conditions employed and the substrates, with a high specificity for pentapyrrolic sapphyrins only. 1 0 1 , 1 1 2 To date, of the heterosapphyrins, only the 27-thiasapphyrin 197 is reported to form metallo complexes. The bisfrhodiumdicarbonyl] complex of 197 1 0 1 has been isolated and, from X-ray crystal analysis, was found to be structurally identical to the metallosapphyrin 212a. 1 0 0 Details of the structure have as of yet, however, not been published. (b) Anion Binding In their most stable form, sapphyrins comprise of a pentameric NH-containing core, in well defined planar cyclic array. Constrained within a ca 5.5A diameter, this imparts unusual anionic binding properties to the sapphyrins via hydrogen bonding.31 Initial evidence for this unique binding came serendipitously from a single crystal X-ray analysis of the mixed salt [191.HF.HPF 6]. 9 9 ' 1 0 0 Here an unexpected electron density was observed in the macrocyclic central core, which from a combination of independent synthesis and 1 9 F NMR was deduced as being a fluoride anion. 9 9 Thus, it appears that the sapphyrin core is ideally suited for totally 89 encapsulating the F anion in a stable hydrogen bonded complex in the solid state (figure 35). Figure 35. X-ray structure of [191.HF.HPF6], showing the centrally bound F~ anion, with equivalent N-F distances of 2.7 A . 1 0 5 This anionic binding was not restricted to the small fluoride anion, as subsequent findings in this area led to the discovery of several other dicationic sapphyrin-anion adducts. X-ray crystal data of the bischloro1 0 5 and the monobasic bis(phenyl-phosphate) 3 1 , 1 1 4 anion complexes of H 2 191 2 + have been reported. In the solid state, the latter two complexes are structurally similar in that, both counter anions form 2:1 adducts (i.e. anion:sapphyrin), and that the anions are now complexed in a near Figure 36. Single crystal X-ray structure of the bis(phenylphosphate) salt of 191. The anions are held at ca 1.4 A above and below the sapphyrin plane.31 90 symmetric fashion above and below the sapphyrin plane as shown in the figure below- fig. 36. In either of these two complexes the anions are chelated by 3 and 2 hydrogen bonds respectively. The solution phase structures of the halo complexes were elucidated as being identical to that of the solid state. 1 0 5 Analysis of the high field signals in the 1 H NMR of the dihydrofluoride, dihydrochloride, dihydrobromide and the mixed hydrofluoride-hexafluorophosphate salts of 191 were particularly revealing, as the nature of the counter anion had a profound effect in this region. Thus, for 191.2HCI and 191.2HBr the peaks ascribable to the pyrrolic NH's as three broadened peaks in the 5 = -4.2 to -5.1 ppm range integrating in a 2:1:2 ratio as expected for the three types of non-equivalent protons. On the other hand, 191.2HF and 191.HF.HPF 6 the corresponding resonances were shifted to an even higher field (e.g. for 191.2HF 5 = -4.6. -5.8, -6.0 ppm at 1 mM) integrating for 1:2:2 protons, with an obvious 1 H- 1 9 F coupling pattern. In addition, in both of these complexes these signals were highly concentration dependent (cf. 5 = -7.6, -8.6, -8.8 ppm at 30 mM for 191.2HF), indicative of aggregation and/or dimerization at higher concentrations. Whereas, in the dihydrochloride and dihydrobromide salts these signals show no concentration dependency. This has been assumed to reflect a tight external ion pairing in the latter complexes, whilst with the fluoro complexes, the aforementioned dependency together with the general upfield shifts observed in these proton signals are more consistent with a structure in which encapsulation of the fluoride anion is prominent. These inferences were further supported by UV-visible spectra and time resolved fluorescence studies. In contrast to the previous phosphate complex, the dication 219 forms a 1:1 complex with monobasic phosphoric acid, in which the bound oxygen atom is 91 displaced by 0.83 A from the sapphyrin plane (fig. 37). 1 1 5 These two phosphate complexes serve to indicate a flexibility in phosphate chelation, wherein anywhere from 2 to 5 NH-to-0 interactions can effect phosphate-to-protonated sapphyrin ligation in the solid state. 219 Figure 37. Left: structure of sapphyrin 219. Right: X-ray structure of the monobasic phosphoric acid complex of 219. 1 1 5 The monocationic forms of sapphyrin also demonstrate similar monoanionic substrate recognition and chelation capabilities. To date, two such complexes, the chloro31 and azido 1 0 0 of H.191+, have been fully characterized by X-ray crystallo-graphy, as shown in figures 38 and 39 respectively. Here, as with the dicationic species, the anion are held out of the mean sapphyrin plane at distances well within those typical for hydrogen bonding interactions. Figure 38. Crystal structure of 191.HCI. Figure 39. Crystal structure of 191 .HN 3 . The Cl~ is displaced by 1.72 A above the The N3" lies at 1.13 A above the sapph-sapphyrin plane.31 yrin plane.31 92 Collectively, these crystal structures serve to indicate that not only are different binding patterns operative in the solid state, but the sapphyrins may additionally show selectivity in terms of anion recognition. Initial evidence for this has come from extensive studies of the hydrohalo complexes of 196. 1 0 5 The latter dicationic species, for example, shows a high selectivity for F~ anions over CI7Br~ with K s values of ca 105 M" 1 and <102 M~1 respectively. This affinity for F~ can simply be attributed to the sapphyrin core possessing the correct dimensions for accommodating the latter anion when compared to that of the larger chloride and bromide anions. Furthermore, as in the solid state, H 2196 2 + also exhibits phosphate anion binding in solution. 1 0 8 , 1 1 6 The potential utility of this property has recently been demonstrated by Sessler et. a / . 1 1 6 Thus, by attaching a derivative of this sapphyrin to a silica gel support, they were able to separate mixtures of various monobasic phosphate, phosphonate and arsenate anionic species at neutral pH. Moreover, this system was found to be capable of separating both mixtures of simple phosphorylated nucleotides (e.g. AMP, ADP, and ATP), and even more complex oligonucleotides at this pH. 220 Figure 40. Schematic representation of the cytosine-sapphyrin conjugate 220. These observations now may have potentially useful implications, for example, this anion recognition ability of the sapphyrins has been exploited by Sessler's group 93 with much success. These workers have shown that 196 in either its mono- or dicationic forms can efficiently effect transport of: (i) nucleotide monophosphates (specifically GMP, AMP, and Ara-AMP) at pH < 3.5; 1 0 8 and (ii) cyclic-AMP 3 1 and, chloride and fluoride anions 1 1 7 at neutral pH's, in a standard three phase Aq I-CH 2CI 2-Aq II U-tube liquid membrane cell. Modification of the peripheral substituents engenders enhanced recognition properties as exemplified by the cytosine-sapphyrin conjugate 220. This nucleobase-sapphyrin conjugate reportedly shows highly selective GMP transport at neutral pH in this standard three phase membrane system 114 222 Scheme 33. Synthesis of a novel "crowned" sapphyrin (223). Another rather elegant modification of the peripheral substituents has been very recently described in the literature. Here, the novel "crowned" sapphyrin 223 was obtained by coupling of the sapphyrin diacid chloride 221 with the bisaminopropyl-94 diaza-18-crown-6 derivative 222 as shown in the scheme above. 1 0 2 This system is unique in that it enables simultaneous coordination of both cationic and anionic substrates. Preliminary work carried out by this group of researchers has shown that 223.HCI is indeed capable of supporting this type of binding. Thus far, an adduct in which an ammonium cation and a fluoride anion are complexed within the crown ether and sapphyrin subunits, respectively, has been characterized by 1 H NMR and UV-visible data. Further proof of this crowned sapphyrins ability to act as a ditopic receptor have come from mass spectrometric work, where upon addition of the appropriate salts, signals corresponding to [223.2H + Mg(S0 4) 2], [223.4NH3], and [223.4H + Fe(CN)6] have been observed. Clearly, this area represents an exciting new development in the ever-increasing chemistry, and applications of expanded porphyrins. Thus one can anticipate an extremely rich and varied chemistry towards this direction in the near future. 7.4 "N-Confused Sapphyrins" Recent developments in porphyrin chemistry has seen the emergence of what perhaps could be regarded as the most intriguing of all the known porphyrin isomers to date. These so-called "N-confused porphyrins" contain the same basic framework as porphine 1, but with one inverted pyrrole ring, were independently discovered by Polish 1 1 8 and Japanese 1 1 9 groups within the past year. Nevertheless, an analogous isomer pertaining to the sapphyrins has inevitably been isolated by the Polish team of Latos-Grazynzki 1 2 0 during their evaluation of the Rothemund synthesis 1 2 1 as a facile route to expanded porphyrins. They found that a BF 3 .E t 2 0 catalysed condensation of benzaldehyde with an excess of pyrrole (1:3 molar ratio) followed by oxidation with chloranil gave in addition to the expected tetraphenyl and "N-95 confused" porphyrins, the "N-confused" tetraphenylsapphyrin 224 as a minor by-product. Scheme 34. Condensation of excess pyrrole with benzaldehyde to furnish the "N-confused" tetra-phenylsapphyrin 225. Evidence for the presence of an inverted pyrrole moiety initially came from the unusual spectroscopic properties of the isolated macrocycle. The initial formulation of the macrocyclic product as a sapphyrin was derived from its mass spectral data. The electronic spectrum typically displays a split Soret-type absorption at 493 and 518 nm, and four Q-bands in the 640-790 nm spectral region as would be expected for a conjugated polypyrrolic aromatic macrocycle. However, the 1 H NMR which, although consistent with a diatropic system, differed markedly from that of a typical sapphyrin. Here, the exterior [3-pyrrolic protons and the exterior 27-NH proton appear at 8 = 9.0-10.2 and 12.2 ppm respectively, whilst the (3-pyrrolic protons located within the core appear at -1.5 ppm and those of the pyrrole NH's at -2.58 ppm. 2D 1 H NMR, COSY and NOESY experiments further served to confirm these assignments and the skeletal arrangement. The smaller ring current effect observed in this macrocycle, in comparison to other analogous 22 7t-electron expanded porphyrins, indicates that the molecular framework deviates somewhat from planarity. Molecular models predict a structure in which the inverted pyrrole is forced out of the general plane defined by the inner tetraaza core. Ph 224 225 96 Moreover, the sapphyrin 224 was found to exhibit an unusual degree of flexibility in acidic solution, wherein, the inverted pyrrole heterocycle readily undergoes a dramatic 180° flip to reveal the planar dicationic tetraphenylsapphyrin 225 as evidenced by the changes in the upfield NH resonance pattern in the 1 H NMR. Such conformational mobility is clearly a concept unique to meso substituted sapphyrins within the context of pentapyrrolic expanded porphyrins. The formation of the inverted sapphyrin structure in this synthetic pathway is believed to be a consequence of the acyclic pentapyrrole adopting a stable helical geometry, which is orientated in the precise fashion for an acid catalysed oc-a pyrrole coupling. 8.0 Smaragdyrins The smaragdyrins (nor-sapphyrins) constitute a series of pentapyrrolic macrocycles formally derived from sapphyrins by a mere replacement of one of the four methine bridges with a direct pyrrole linkage. Being 227t-electronic systems, they exhibit spectroscopic features akin to the sapphyrins. Unfortunately, a very limited knowledge of the chemistry of these intriguing molecules is available to date, largely due to their poor stability. For instance, evidence for the formation of the hexa-methyl smaragdyrin 226 has, so far, only come from the presence of a strong Soret absorption at ca. ? i m a x = 450 nm, and additional broad bands in the 700-725 nm region in the electronic spectra of the reaction solutions. 9 3 , 9 7 Attempts to isolate the corresponding metallo complexes were further hampered by its extreme sensitivity towards light and acid media. 97 226 Figure 41. Schematic representation of the highly unstable pentapyrrolic smaragdyrin 226. Nonetheless, Grigg and coworkers have had some success in this area, with the synthesis of the dioxasmaragdyrins 229 and 230.97 ,122 This was achieved via an HBr catalysed "2 + 3" MacDonald coupling of the pyrrolyldipyrromethanes 227 and 228 with 5,5'-diformyl-2,2'-bifuran (185), respectively. Chemically, the dioxasmaragdyrins are reported to differ from the corresponding dioxasapphyrins in that they are markedly less basic. Scheme 35. General synthetic strategy towards dioxasmaragdyrins. The electronic spectra of the free base forms of 229 and 230 closely resemble that of the dioxasapphyrins, with an intense Soret band at 442 nm (e = 201,200) and several additional weaker absorptions in the 547-730 nm region. 1 2 2 However, the cationic species (e.g. H230+) display a split Soret band at 448 and 459 nm. The 1 H NMR spectra of these compounds contain features typical of an aromatic structure 98 sustaining a significant diamagnetic ring current. The meso protons resonate at 8 = 10.52 and 10.06 ppm, whilst the internal pyrrolic NH typically appear at a high field position (8 = -4.85 ppm). The above represents the state of current research in this area, most of which was developed in the early 1970's. Perhaps, a reevaluation of this work, together with the present insight into the unusual properties of the related sapphyrins, may provide a plethora of new chemistry relating to this obscure class of expanded porphyrins. 9.0 Orangarin In recent months, the Austin group led by Sessler introduced a new general synthetic approach to (^-substituted terpyrroles based on a key oxidative coupling between an LDA-derived enolate and an a-propipnylpyrrole.123 This in turn, has inspired the synthesis of a novel pentapyrrolic expanded porphyrin derived solely from this tripyrrolic subunit and a bipyrrole linked by a pair of methine bridges. This novel macrocycle 232, christened orangarin due to its bright orange coloured organic solutions, is formally a nonaromatic, conjugated 20 7t-electron annulene. The synthesis of the target macrocycle 232 was based on a "1 +1" condensation of terpyrrole 231 with bipyrrole 116 under conditions identical to those employed in the turcasarin synthesis7 0 (cf. scheme 18). The most notable feature here is that in the latter synthesis, the "2 + 2" adduct was the major macrocyclic product with no evidence for the formation of corresponding orangarin. Whereas, in the present example, inclusion of (3-alkyl substituents on the terpyrrolic moiety dramatically 99 reduces the formation of the larger turcasarin system, favouring the "1 + 1" adduct (orangarin) instead. Thus, it is fairly apparent that these alkyl side chains enhance the formation of the orangarin molecule 232 through a controlling "helical effect". 1 1 0 , 8 3 This view is, in fact, supported by an X-ray crystal analysis of a precursor 5,5"-diesters to a diethyl derivative of terpyrrole 231. In this solid state structure the molecule was found to adopt a helical twist. Scheme 36. Synthesis of orangarin 232. The UV-visible spectrum of 232.2HCI contains three absorption bands in the visible region. The Soret-like absorption band at 462.8 nm, however, is rather broad and considerably weaker than that of its larger congener turcasarin. This has been equated to a disruption in the conjugated pathway resulting from distortions within the overall planar framework. The NMR spectra of the dihydrochloride salt 232a are consistent with its C 2 symmetric structure and are devoid of any aromatic ring current effects. Interestingly though, the internal NH signals are located at an unusually high field position (cf. 8 = -0.2 ppm). The meso-like protons' signals, on the other hand, are shifted further upfield than would be expected for typical vinylic protons. Given that orangarin is a 4n7t-electron system, the upfield shifting in the resonances of these external peripheral protons can be accounted for in terms of a 100 paratropic ring current effect. This is, however, inconsistent with the positions of the NH signals, and thus is believed to be more reflective of some kind of localized anisotropy effects possibly associated with the positive charges on the macrocycle. Figure 42. X-ray structure of 232b. 1 2 3 The single crystal X-ray structure analysis of the free base form served to confirm the above conclusions and in particular its nonaromatic 207c-electronic formulation. This is evidenced by (i) the localization of the three hydrogens to three of the five pyrroles; and (ii) the disparate bond lengths of the double bonds typifying more localized double bonds within the structure. Moreover, as depicted in the figure above, the molecule adopts a fairly planar conformation, with only a slight distortion around the bipyrrolic units resulting from unfavourable steric interactions between the p-methyl substituents as was inferred from the electronic spectrum. 10.0 Pentaphyrins The pentaphyrins, as implied by their trivial name, are a series of pentapyrrolic macrocycles in which the heterocyclic subunits are all interconnected by methine bridges. The first prototypical member 237 of this family was synthesized in the mid-101 80's by Gossauer and coworkers. 1 2 4" 1 2 7 The general synthetic strategy, as with other related expanded porphyrins (e.g. the sapphyrins), towards the pentaphyrins relied on an oxidative "2 + 3" MacDonald coupling of an cc-free dipyrromethane with a diformyl tripyrrane as shown in scheme 37 . 1 2 4 - 1 2 8 234 R, = n Bu 235 R 2 = -(CH2)2C02Me, R 3 = Me 237 R, = Me, R 2 = -(CH2)2C02Me, R 3 = Me Scheme 37. Synthesis of pentaphyrins 236 and 237 via an oxidative "2 + 3" MacDonald-type coupling. The aromatic nature of these compounds is clearly apparent in their spectroscopic properties. The electronic spectra of both the free base pentaphyrins and the triprotonated forms display the characteristic intense absorption features of porphyrinoid-type compounds. The free-base 237, for instance, exhibits an intense Soret-like absorption at 458 nm, with three weaker bands at 367, 642, and 695 nm. The 1 H NMR confirms the presence of a diatropic ring current within a delocalized 22 jt-electron perimeter system. Thus, the peripheral methine protons typically resonate in the 12.4-12.6 ppm region, whilst the inner ring protons (NH's) appear at high field (ca - 5 ppm). More recently, Sessler et al. have synthesized the uranyl complex of 236, by refluxing the trication H 3 2 3 6 3 + with U0 2 CI 2 in a mixture of pyridine/ j PrOH. 1 2 8 The resultant complex exhibits high stability towards competing ligands, hydrolysis and 102 transmetallation reactions, but readily undergoes demetallation in the presence of strong acids. The UV-visible spectrum of the uranyl complex displays a considerably red-shifted Soret at 500 nm (cf. Xmax = 467 and 462 nm for 236 and the tricationic species H 3 236 3 + respectively). The 1 H NMR spectra of this complex, on the other hand, is remarkably similar to that of the unmetallated pentaphyrin 236, showing features consistent with an overall C 2 symmetry. The single crystal X-ray structure (fig. 43), however, reveals a markedly distorted structure reminiscent of the uranyl superphthalocyanines (cf. Fig. 3).17 Thus, in order to accommodate the bonding requirements of the centrally-located uranyl cation, the pentaphyrin is forced to assume a saddle-shaped conformation. This results in an essentially symmetrical pentagonal bipyramidal coordination geometry around the uranium atom with U-N and U-0 distances of ca. 2.5 A and 1.75 A, respectively. Thus from the X-ray data and symmetry observed in the 1 H NMR, clearly, in solution U0 2.236 exists in a fluxional ensemble of distorted conformations which on average contain the uranyl cation within the macrocyclic plane. Figure 43. View of the uranyl complex of 236, parallel (left) and perpendicular (right) to the 0=U=0 axis, showing the pentagonal bipyramidal geometry of the U atom and saddle-shape of the pentaphyrin framework respectively. 2 8 103 An interesting feature of these findings is the markedly differing chelation properties engendered on this series of pentapyrrolic macrocycles over the sapphyrins, by the introduction of an additional methine bridge into the sapphyrin molecular frame. In the present macrocycles this structural transformation is sufficient to stabilize the resultant uranyl complex whilst maintaining the overall aromatic character. Whereas, with sapphyrins, as described previously, insertion of the uranyl cation appears to not only destabilize the corresponding aromatic metallocomplex, but concurrently activate one of the meso carbon atoms towards nucleophilic attack.1 1 2 Complexes with gadolinium(lll)128 and with zinc, cobalt and mercury7 are reported to form, however, no details are available for these. Bn Bn 239 240 Scheme 38. Franck's biomimetic synthesis of an inverted nonconjugated pentaphyrin analogue (239). A related compound which is worthy of mention is the inverted pentaphyrin 239, reported by Franck. 1 2 9 This macrocycle is unusual in that all five pyrrolic nitrogens are on the periphery of the macrocycle instead of on the interior. Synthetically, this macrocycle was exclusively obtained via a biomimetic condensation of the pyrrole 237 in oxygen-free glacial acetic acid with p-toluenesulphonic acid (scheme 38). Unfortunately, attempts to oxidize this macrocycle to its aromatic counterpart 240 met with no success. The inability to access the aromatic 22 rc-electron inverted 104 pentaphyrin was rationalized on account of a destabilizing accumulation of charge occurring in 240. 11.0 Hexapyrrolic Expanded Porphyrins 11.1 Amethyrin The recent advancements in the syntheses of linear terpyrrolic compounds has seen the emergence of the smallest members of the hexapyrrolic based macro-cycles. 1 2 3 These new macrocycles (e.g. 242 and 243), termed the amethyrins (from Greek amethus, in view of the dull purple colour of the dication in organic solution), are comprised of two terpyrrole fragments linked via a pair of methine bridges. Structurally, they resemble the recently reported porphocyanines 8 8 , 8 9 , 9 1 of Dolphin, 242 R = H 243 R = p-N0 2 C 6 H 4 X = TFA~, CI" Scheme 39. Synthesis of amethyrins 242 and 243. 105 and Corriu's platyrin 130 7 2 , 7 3 however, like the latter compound they are non-aromatic 24 re-electron systems. Here, the general macrocyclization step employed an acid catalysed "2 + 2" condensation between the hexaalkyl terpyrrole 241 and an appropriate aldehyde. Subsequent oxidation with chloranil or DDQ furnished the fully conjugated dicationic species in good yield. The free-bases are readily obtained upon exposure to aqueous NaOH, and display good stability. The above 24rc formulation represents one of three possible oxidations states accessible to the amethyrins. The other two correspond to the formally aromatic 22TC - (244) and 26Tc-electron (245) annulenes as depicted in the figure below. Interestingly, neither the oxidative conditions employed in this reaction sequence nor do attempts with other oxidants yield either of these aromatic systems. 244 245 Figure 44. Generalized structures of the two other possible (aromatic) oxidation states of amethyrin. This contention that the amethyrins are conjugated nonaromatic macrocycles is self-evident in their spectroscopic properties. The UV-visible spectrum of the dihydrochloride salt of 242 contains three bands, a weak N-type absorption at 384 nm, a strong Soret-like band at 492.8 nm, and a fairly intense Q-type absorption at 597 nm. The free base form, on the other hand, exhibits a single broad band at 467 nm. The 1 H NMR spectrum of this salt displays some anomalous features with 106 regards to the resonance positions of the peripheral protons. For instance, the internal NH signals are dramatically shifted downfield to 8 = 24 ppm, and the meso protons exhibit an upfield shift to 8 = 3.4 ppm. The origin of these induced shifts in the proton signals, as in the case of orangarin, is as of yet not fully understood. Notably, the 1 H NMR of meso-substituted amethyrin 243 presents some interesting results. Here, the incorporation of the p-nitrophenyl substituents at these positions induces diastereotopicity in the methylene protons of the ethyl groups. This effect has been attributed to a combination of two factors primarily resulting from steric interactions between the ethyl and phenylene groups. Namely, restriction in the free rotation of these ethyl groups, and a distortion of the macrocycle such that a number of low energy conformations may now prevail. The interconversion between these conformers is obviously slower than the NMR time scale, thus, as with turcasarin, the enantiomers are detected in the 1 H NMR. Moreover, due to the close proximity of these methylene protons to the aromatic 7c-cloud of the p-nitrophenyl group, a high field shifting is observed in the former protons resonances. These conclusions were corroborated by a single crystal X-ray analysis. Figure 45. X-ray structure of 242.2HCI. 123 107 X-ray crystallography further (fig. 45) confirmed the nonaromatic formulation of the amethyrins, and served to indicate that the molecular framework adopts a fairly planar geometry. However, steric interactions between the methyl substituents force the two central pyrroles into a conformation in which the NH groups are forced above and below the macrocycle plane. Additionally, the latter groups are coor-dinated to the chloride counteranions via defined hydrogen bond interactions. Figure 46. Molecular structure of the bis-Zn complex of 242. The amethyrins have so far demonstrated a varied coordination chemistry with several first row transition metals and the uranyl dication, in which different binding modes are operative. To date, the Zn, Co, Cu, Ni and the U 0 2 2 + complexes have been synthesized and tentatively characterized by mass spectra and/or X-ray crystallography. With Z n 2 + and Cu + , bimetallic complexes result, in the case of the latter, however, the exact details of this Cu-amethyrin complex are presently unresolved. A single crystal X-ray analysis of the bis-zinc complex of 242 has revealed that both metal centres are held within the macrocyclic plane, coordinated to the terminal pyrrolic nitrogens only. The central pyrroles, although not implicated in the metal coordination, are distorted from the general plane so as to form hydrogen bonding interactions with the two bridging axial metal ligands. An interesting feature of this complex is the lability of the bridging chloride anions. As 108 seen in the diagram above (fig. 46), one of these groups is readily exchanged with a hydroxide species, presumably during the course of purification. On the other hand, the cobalt(ll) complex of 242, is somewhat more reminiscent of the porphocyanine zinc complex reported previously by the Dolphin group (fig. 30), 8 8 both structurally and in its coordination of a single metal atom. Here, the amethyrin ligand is apparently only serving to fill the coordination sphere around the original cobalt(ll) centre. The cobalt as in the case with the bis-zinc complex, is held within the ligand plane, and the complex is more than likely further stabilized by the apparent strong hydrogen bonding interactions between the chloride anions and the central pyrrolic protons. The uranyl complex is also believed to be structurally similar to the cobalt amethyrin adduct, as inferred from high resolution mass spectroscopy. In terms of stability, all these metalloamethyrins are fairly acid sensitive, particularly that of the uranyl complex, which is so highly labile that demetallation occurs under weakly acidic condition. Nonetheless, these complexes clearly underline the versatility of amethyrins as ligands capable of stabilizing both mono- and bi-metallic complexes. More notably, the bis-zinc adduct is the only structurally characterized Figure 47. View of the cobalt(ll) amethyrin complex. 123 109 in-plane bis-metallated expanded porphyrin reported to date, and may be of significance as bioinorganic models. 11.2Rosarin During an attempt to develop a more general, facile synthetic route to expanded porphyrins, Sessler's group discovered another novel hexapyrrolic macrocycle 248, 1 3 0 which they coined rosarin (from the Latin rosa) in view of the bright-red-to-purple colour of its triprotonated derivative. In order to achieve their initial goal, these researchers adapted a time-proven strategy originally developed by Rothemund in 1936 1 2 1 a b and subsequently optimized by others, 1 2 1 c e which entails coupling of a di-cc-free pyrrole with an aryl aldehyde under acidic, oxidative conditions to yield tetra-meso-substituted porphyrins. R H H 246 + O 1.TFA, CH2CI2 247a R = H •R 247b R = p-N02 247c R = o-OMe 247d R = p-OMe 248 R = H 249 R = />N02 250 R = o-OMe 251 R = p-OMe Scheme 40. General synthetic route to protonated rosarins 248-251. 110 Thus, as shown in the scheme above, acid catalysed condensation of 5,5'-unsusbstituted bipyrrole 246 with benzaldehyde 247a followed by oxidation with DDQ provided the trifluoroacetate salt of H 3 248 3 + as a green metallic solid in surprisingly high yield (> 70%). The free-base form is readily obtained by treating this salt with 10% aq. NaOH. Analogously, the rosarins 249-251 were obtained via this procedure, by simply introducing the appropriate aryl aldehyde to the reaction mixture. With 24 7t-electrons, rosarins 248-251 do not obey the Hiickel 4n + 2 rule. Like amethyrin,123 the rosarins can exist at two other oxidation levels corresponding to the formally aromatic 22 n- and 26 rc-electron macrocycles (cf. fig. 44). That this macro-cyclization procedure only furnishes the nonaromatic structure, is self-evident in the spectral data of the isolated macrocycles. Thus, the UV-visible spectrum of 248 displays a considerably red shifted Soret-like absorption at X, m a x = 552.5 nm (e = 192,000), which apparently would be expected for a deformed, nonplanar structure.131 Moreover, the 1 H NMR is devoid of any substantial ring current effects. Conclusive proof of this assignment was derived from single-crystal X-ray diffraction analysis of the hydrochloride salt of rosarin 248, where the nonplanarity of the Figure 48. X-ray structure of 248.3HCI. 130 111 macrocycle is fairly evident. Two of the chloride counter anions are hydrogen bonded within the macrocyclic core whilst the third forms part of the general lattice structure and is not proximate to the macrocycle. Incidentally, this binding of chloride anions within the macrocyclic core is somewhat reminiscent of the anion chelation properties of the sapphyrins discussed previously, and increasingly appears to be a common property endowed upon the larger polypyrrolic macro-cycles. One question that remains to be addressed is the lack of propensity for the rosarins to attain either one of their corresponding fully aromatic species, despite the anticipated additional gain in resonance stabilization upon doing so. This is believed to be reflective of unfavourable ethyl-phenyl interactions about the meso-Wke positions, and/or destabilizing methyl-methyl interactions within the bipyrrolic subunits of the rosarins. This argument is further supported by the X-ray structure of a prototypical rosarin (fig. 48), where the molecule's deformed conformation is fairly obvious. Likewise, with the isoelectronic amethyrins wherein similar methyl-methyl interactions within terpyrrolic subunits do enforce a slight deviation from planarity around the central pyrrolic moiety (fig. 45). Thus, it would appear that these interactions within these directly linked pyrrolic units are sufficient to perturb the p-orbital overlap of the extended rc-framework, and henceforth prevent the assumption of a fully conjugated, planar (4n + 2)n electronic pathway required of an aromatic system. Interestingly, on the other hand, the congeneric rubyrins1 3 2 which also contain a pair of directly linked pyrrolic units, are isolated in their 26 7t-electron aromatic forms. However, the latter macrocycles additionally contain one and two more meso carbons than these rosarins and amethyrins respectively. 112 11.3 Rubyrin In the year prior to the publication of the above work towards the rosarins, this same group at the University of Texas had already successfully prepared a related hexapyrrolic macrocycle. This macrocycle, to which they assigned the trivial name rubyrin (from the Latin rubeus) in light of the intense bright red colour of its diprotonated derivatives, comprises a pair of bipyrrolic subunits interlinked to a pair of pyrroles via methine bridges.1 3 2 254 = Et, R2 = Me 116 R1 = Et, R2 = Me 255 = nPr, R2 = H 105 R1 = nPr, R2 = H Scheme 41. Synthesis of rubyrins 254 and 255. The synthesis of these macrocycles hinged upon two key acid catalysed condensations as outlined in scheme 41 above. The initial condensation between 113 bipyrrole 246 and acetoxymethylpyrrole 15 provided the tetrapyrrole 252 in ca 66% yield. Subsequent hydrogenolysis of the benzyl ester gave the diacid 253, which was immediately subjected to a "4 + 2" MacDonald-type oxidative coupling with a diformyl bipyrrole, e.g. 105 or 116, to yield the corresponding rubyrins as their dicationic salts. The rubyrins 254 and 255 display differing chemical properties. For example, 254 is unusually basic, and like the dioxasapphyrins, cannot be isolated in its free-base form, on the other hand, 255 is readily isolated as its free-base. This behaviour has been explained in terms of the ability of the latter system to adopt a more planar array as a consequence of some degree of rotational freedom about one of the bipyrrolic bonds. Nevertheless, both species are consistent with their aromatic formulations with a 26 7i-conjugated electronic pathway. Thus, as expected the absorption bands in their electronic spectra are considerably red-shifted in comparison to that of the equivalent porphyrin or sapphyrin dicationic species. For example, the UV-visible spectrum of 255 contains an intense Soret-like band at 513 nm, and three additional less intense Q absorptions at A,m a x = 711, 789, and 844 nm respectively. In the dicationic form (i.e. 255.2HCI) these bands all display a blue-shift, e.g. A,m a x = 501, 692, 771, and 817 nm, which is apparently indicative of anionic complexation occurring (cf. the effects of the counter anions on the observed UV-visible spectra of the sapphyrin). 1 0 0 ' 1 0 1 ' 1 0 5 Moreover, the presence of a strong diamagnetic ring current as evidenced in the 1 H NMR, where the meso protons resonate at 8 = 11.58 and 11.60 ppm, and the pyrrolic NH's at 8 = -4.97 and -5.3 (for 254.2HCI) respectively, provides decisive proof of the diatropicity of the rubyrins. 114 Figure 49. Crystal structure of the bishydrochloride salt of rubyrin 254. Left: View perpendicular to the ligand plane. Right: Side view.132 The figure above depicts the crystal structure of the bishydrochloride salt of rubyrin 254, thus confirming the overall planar geometry of the macrocycle. Moreover, figure 49 also confirms that the rubyrins, like the sapphyrins and rosarins, are capable of forming complexes with anionic substrates via hydrogen bonding. Like the sapphyrin bishydrochloride adduct 191.2HCI 3 1 , 1 0 5 in the solid state structure of 254.2HCI, the chloride counter anions are held above and below the macrocyclic plane by 1.6 A. Furthermore, these chloride anions are held closer to the plane in the rubyrins than in the case of sapphyrin (cf. 1.88 and 1.77 A). 1 0 5 This, could reflect a reduction in the intra-core N+H-to-N+H electrostatic repulsive interactions within the larger rubyrin cavity. Thus, considering the above observations, taken in context with the more detailed sapphyrin anionic substrate binding work, it is conceivable that the rubyrins, with a larger core may show similar recognition properties. This, has indeed been realized in preliminary studies with fluoride and phosphate anions which are reportedly bound in a strong and non-labile manner. 1 3 2 115 11.4 Hexaphyrins Chronologically, the hexaphyrins were the first members of this class of expanded porphyrins to be reported in the literature just over a decade ago. 1 2 5 Driven by their successful synthesis of pentaphyrin, Gossauer 1 2 5 and coworkers1 3 3 subjected the di-oc-free tripyrranes 256-259 and their bisformyl derivatives 235, 260-262 to a two step sequence encompassing an acid-catalysed condensation followed by oxidation with iodine/p-benzoquinone (scheme 42). Through this, they isolated the hexapyrrolic macrocycles 270-273 together with minor quantities of the corresponding pentaphyrins. On account of the substitution pattern on the precursor tripyrranes 256-265, 235, 260-261, the macrocycles 270-272 are obtained in two isomeric forms, namely structures A and B, in equal proportions. In either case, however, molecular models predict a planar geometry only for those structures containing a pair of methine bridges with an E configuration on opposite sides. The 1 H NMR data has served to unravel most of the structural details of the hexaphyrins. The isomeric macrocycles 263A and 263B exhibit three distinct low-field singlets ascribable to the protons at the Z-configurated methine bridges, which integrate in roughly a 2:1:1 ratio. Of these, the four homotopic protons of isomer 263A, which is of the D 2 h symmetry group, resonate at 8 = 12.42 ppm. Accordingly, the remaining two signals at 8 = 12.33 and 12.13 ppm were ascribed to the two pairs of homotopic methine protons of isomer 263B belonging to the C 2 h symmetry point group. The pairs of endo methine protons appear as two signals at 8 = -7.40 and -7.44 ppm corresponding to isomers 263A and 263B respectively. Further confirmation of these assignments was obtained from the 1 H NMR spectrum of the dodecamethylhexapyhrin 266, where the signals corresponding to the exo and endo methine protons appear as singlets at 8 = 12.5 and -7.3 ppm respectively. The UV-116 visible spectrum of 263 is characterized by three absorption bands at A,m a x = 572 (e = 76,000), 595 (47,000) and 789 (3,981) nm. Upon acidification, the resultant spectrum displays only a "Soret-like" absorption at 551 nm and a slightly red-shifted low energy absorption at 798 nm. 256 R = (CH2)2C02Me 235 R = -(CH2)2C02Me 257 R = CH2C02Me 260 R = CH2C02Me 258 R = CgH^ 261 R = CgH^ 259 R = Me 262 R = Me 1. HBr 2. I2 /p"benzoquinone 263 R = -(CH2)2C02Me 264 R = CH2C02Me 265 R = CgH^ 266 R = Me Scheme 42. Gossauer's route to hexaphyrins 263-266. 117 263 R = (CH2)2C02lvle 3 264 R = CH2C02Me , 265 R = CgH^ 268 R — CgH^ Scheme 43. The organometallic chemistry of hexaphyrins 263-265. Investigations of the organometallic chemistry of the hexaphyrins has provided an intriguing insight into their dynamic behaviour. The flexibility of the molecular framework has been elegantly displayed by the isolation of their bimetallic Zn and Pd chelates in which two substantially different geometrical arrangements predominate. Treating the isomeric mixtures of 263 or 265 with ZnCI2 furnished the symmetrical bimetallic zinc complexes 267 and 268 respectively of C 2 v point symmetry.133 Interestingly, the analogous methodology when applied to hexaphyrin 118 264, on the other hand, gave the less symmetric isomer 269 with C 2 h point symmetry as outlined in scheme 43. Structural analyses of these compounds has, however, relied heavily on a elemental analyses, whilst the geometrical conformations were gleaned from a combination of 1 H NMR spectroscopy and NOE difference experiments. Thus, the exact geometry about the metal centres cannot be inferred. These arrangements were assigned on the basis of molecular modelling. Nevertheless, the zinc complexes do display intense absorptions in the visible region of the electromagnetic spectrum. For example, 268, exhibits an intense absorption at A.m a x = 574 nm (e = 263,000), with additional bands at 450, 601, and 810 nm. In acidic media this "Soret-like" band is hypsochromically shifted to 556 nm, and is almost twice its original intensity (cf. e = 410,000 M"1 cm"1). Additionally, less intense Q-type absorptions are observed at 772, 795, 818 and 849 nm. The formation of single isomeric metallo complexes was originally thought to arise from an interconversion process mediated by Lewis acidic-type metal centres (e.g. Zn 2 +) during the insertion reaction.7 This has, however, recently been disproved by 1 H NMR studies of the crude reaction mixtures, which strongly indicate the presence of metallo-complexes corresponding to both isomeric forms of the precursor ligands. 1 3 3 Thus the isolation of single isomeric metallo complexes is merely a consequence of subsequent chromatographic separation of the reaction mixtures. With Pd, on the other hand, a totally different picture emerged. Treatment of basified crude reaction mixtures of hexaphyrins 263 or 264 with ammonium tetra-chloropalladate surprisingly furnished the bis-palladium complexes 270 and 271, respectively. Even more unexpected was their unusual geometry. In order to 119 263 R = (CH2)2C02Me 264 R = CH2C02Me 270 R = (CH2)2C02Me 272 R = (CH2)2C02Me 271 R = CH2C02Me Scheme 44. Bis-palladium complexes of macrocycles 263 and 264, showing the different ligating modes in operation compared to the bis-Zn complexes 267-269. accommodate the square planar geometry about the d 8 Pd centres, the two central pyrrolic rings coordinated to these metal atoms are rotated through 180° with concomitant E/Z isomerization of two formal C=C bonds as indicated in the scheme below. This geometry is clearly apparent in the 1 H NMR, where the highfield signal of the endo protons is now replaced by two singlets at 8 = -6.24 and -2.35 ppm with relative intensities of 12H and 2H respectively. These peaks were duly assigned to 120 the p-methyl substituents and the pyrrole NH protons located within the macrocyclic cavity. These assignments and overall geometry were corroborated by NOE difference experiments. Furthermore, exchange of the labile ammonia ligands at the Pd centre with pyridine, yielding the corresponding bis(pyridyl)dipalladium complex 272, confirmed (from its 1 H NMR and NOE data) both the assigned molecular conformations, and the absolute geometry about the square-planar Pd centres. Interestingly, the electronic spectrum of 270 contains several intense absorptions spanning a wavelength range from 276 to 840 nm, but with two dominant bands at 574 and 607 nm of roughly equal intensities (ca. e = 80,000 M"1 cm"1). Despite the lack of more conclusive X-ray structural analysis, the presence of a strong diamagnetic ring current in the 1 H NMR coupled with the observed strong absorptions in the visible region of the electromagnetic spectrum are consistent with Gossauer's assessments of these macrocycles as planar, conjugated aromatic structures. What remains to be seen, however, is whether bis-zinc adducts 267, 268 mirror their smaller amethyrin congeners in terms of in-plane metal chelation. 12.0 Torand Expanded Porphyrins Torands, e.g. structure 273, are planar macrocyclic ligands composed of a cyclic array of smaller fully fused rings, incorporating heterocycles bearing ligand atoms or hydrogen bond donors, interspersed at regular intervals.134 Generally, the majority of torands synthesized within the last decade, by Bell and his group, at the State University of New York, contained pyridine as the heterocyclic unit. More recently, however, they have described an expanded porphyrin-type torand 281 which contained both pyridine and pyrrole rings within the macrocyclic perimeter.134 121 273 Figure 50. General structure of a torand macrocycle. A brief outline of the major features of the synthesis of this hitherto expanded porphyrin is presented in the scheme below. The key steps in this sequence involve two thermal Piloty rearrangements of azines 275, and that derived from conden-sation of two moles of hydrazone 277 with the bisketone 276. This reaction, in combination with the deketalization, constitutes an effective homologation method for the formation of curved bands of alternating pyrrole and pyridine heterocycles mutually fused to a 6-membered carbocycle. Further elaboration of the ketal groups of the curved oligomer 278 provided the bis(semicarbazone) 279, which upon pyrolysis effects the final macrocyclization step, yielding torand 280 in 69% yield. Partial dehydrogenation of the ethano bridges, furnishing macrocycle 281, also occurs during the final thermal cyclization reaction. These researchers further report that the fully aromatic congener of torand 280 is readily accessible via dehydrogenation with DDQ. This oxidation would, however, have to occur at the expense of the pyridine aromatic system, which is contrary to more recent accounts by Breitmaier1 3 , 7 3 and previous work by Newkome. 1 3 5 Unfortunately, to date, details of the structural and physical properties of this torand and its oxidized derivatives remain sketchy. Nevertheless, it is conceivable that in 122 n B u n B u "Bu 274 275 281 Scheme 45. Synthesis of the only known torand expanded porphyrin 281 to date. 123 their anionic forms they may be of interest for metallo-complexation, whilst the polyprotonated forms may exhibit an affinity for large anionic substrates. 13.0 Future Outlook The preceding article represents the state of current research in the expanded porphyrins, and is by no means a comprehensive account of the chemistry in this field. However, it does provide an insight into the versatility of these macrocycles in terms of their diverse coordination properties which range from lanthanide(lll) cation coordination of the texaphyrins to the anionic recognition abilities endowed upon the sapphyrins and, its penta- and hexaphyrin congeners. Moreover, their unique photo-physical properties which can be altered by modifications of the conjugated pathway and/or peripheral substituents imparts optical properties unsurpassed by other organic chromophores. Clearly, the improvements in the chemistry towards the acyclic precursors over the last decade, has now made a vast array of larger, more intriguing macrocycles accessible. In view of the latter, coupled with their potential applications, one can only anticipate a plethora of exciting developments in this area of porphyrin research within the near future. 124 Chapter 2 Discussion 125 14.0 Introduction Our interests in this area of porphyrin chemistry were largely driven by their potential biomedical applications. These efforts resulted in the recent disclosure of a novel series of expanded porphyrins, coined the porphocyanines (e.g. 282). 8 8" 9 1 Structurally, these compounds are closely related to the bisvinylogous [22]porp-hyrins (e.g. 142) of Franck et a / . , 7 8 , 8 0 , 8 1 with the central methine groups of the latter systems replaced by a pair of imine bridges. Unlike the other congeneric macro-cycles which recently emerged from the Corr iu 7 2 , 7 3 and Sessler 1 2 3 groups, the porphocyanines retain the aromatic characteristics typical of porphyrins as evidenced by their spectroscopic properties. 282 Figure 51. Generalized structure of porphocyanine. With a hexaaza core in predefined planar "cyclic" array, one can envisage a rich and diverse metallochemistry with the transition metals and the larger lanthanide (III) cations. Of particular interest were the lanthanide complexes, more specifically with the paramagnetic Gd(lll) cation, as potential contrast agents in MRI diagnosis of neoplastic tissue given the propensity of certain HpD derivatives to aggregate in such tissue 2 Moreover, despite the growing number of fully aromatic expanded porphyrins available to date, the lanthanide complexation capabilities of these macrocycles is still by and large confined to obscurity. In fact, it has only been within 126 the last two years with Sessler's comprehensive investigations of the texa-phyrins. 8 ' 2 8- 3 7 ' 3 8 ' 4 4 that this goal was actually realized. These latter findings, at first, appeared very encouraging to the porphocyanines, however, only a zinc chelate has so far successfully been isolated and characterized by a single crystal X-ray analysis.8 8 283 X= 0 285X=Y = NH 288X=Y=NH, O 284 X=S 286X=Y = 0 289X=NH, Y = 0 287 X= NH, Y = 0 290X=Y = O 292 X=0 294 X=0 291X=NH, Y = 0 293 X=S 295 X=S Figure 52. Generalized structures of the targeted expanded porphyrin systems. Thus, in view of the above limitations, we sought to synthesize other hetero-substituted analogues of the tetrapyrrolic porphocyanines, namely the thienyl and furyl derivatives (e.g. 283 and 284), which consequently may show differing binding properties. Moreover, the documented changes in the electronic spectra of 127 porphyrins,69 porphycenes10' and sapphyr ins, 9 4 ' 9 7 ' 1 0 0 , 1 0 1 ' 1 0 3 upon substitution of one or more of the pyrrolic rings with other heterocycles rings, particularly with thio-phene, made these analogues more attractive as long-wavelength absorbing photosensitisers. Although much of the chemistry had already been unravelled, questions on the mechanisms of the unusual cyclization step still remained unanswered. Moreover, a more efficient strategy to the key biscyanodipyrro-methanes was required. Finally, in order to evaluate the generality of this reductive-cyclization sequence, we additionally targeted some other expanded porphyrins, as shown in the figure above, which also contain the imino linkages. Molecular modelling of all these aromatic expanded porphyrins predict essentially planar geometries.136 The macrocycles 285-289, are particularly intriguing in that they can potentially access a number of aromatic and nonaromatic oxidation states. The obvious question that arises here is which of these oxidation states would prevail. For instance, Sessler has shown with the congeneric hexapyrrolic amethyrins123 and rosarins,1 3 0 where similar possibilities exist in their oxidation states, that it was the 247t-electron nonaromatic system that was favoured. The following account details our attempts towards these newer macrocycles, and provides a tentative mechanism for the cyclization reaction. In summary, to date, only three of the target systems (i.e. 283, 284, and 292) have actually been isolated in their reduced forms. All attempts to effect the key 6-electron (and 4-electron) oxidative transformation (respectively) of these macrocycles have so far proved fruitless. Although, it should be noted, partial oxidation of the dipyrromethane units of the precursors to 283 and 284 has been serendipitously achieved. Moreover, efforts to coordinate various metals with these reduced-form macrocycles met with no success. This is probably due to the instability of the imine bridges. Of the other macrocycles, an initial attempt at synthesizing the 227t-porphocyanine analogue 285 128 similarly failed to materialize, whilst, for the remaining systems, the precursor acyclic systems have been obtained. 15.0 Porphocyanines 15.1 Synthesis of Dipyrromethane Intermediates In the preferred route to the porphocyanines (e.g. schemes 27 and 28), the 5,5'-bis-cyanodipyrromethanes play a pivotal role in the sequence. 8 8 , 9 1 Synthetically, these intermediates are in turn synthesized from the readily available 5,5'-bisformyl-dipyrromethanes via well established procedures in pyrrole chemistry. 1 3 7 , 1 3 8 The general sequence starts with a highly selective oxidation of the a-methyl group of benzyl 4-ethyl-3,5-dimethylpyrrole-2-carboxylate (292),139 with one equivalent of lead tetraacetate to furnish the acetoxymethylpyrrole 1523 in excellent yield (> 90%). This material was then subjected, without any further purification, to a self-condensation reaction in refluxing aqueous acetic acid (80%),1 4 0 thus giving the dipyrromethane 29323 in ca 50-70% from pyrrole 292. The final steps to the target bisformyldipyrromethane 295 entailed hydrogenolysis of the benzyl esters, a thermal decarboxylation of the resultant diacid 294 under an inert atmosphere, and finally a modified Vilsmeier-Haack formylation. 1 3 7 , 1 4 1 Transformation of the formyl groups of 295 to provide the requisite cyano functionalities in 295 was accomplished in a two-step sequence, which initially furnishes the requisite bisoximes 296 in an isomeric mixture. Subsequent dehydration of the latter species with acetic anhydride at 100° C, under anaerobic conditions, completed this synthesis, yielding the 5,5'-biscyano-2,2'-dipyrromethanes in 50% overall yield (from 295). 129 296 295 ^ / A C 2 0 297 298 299 Scheme 46 Synthesis of the dipyrromethane intermediates 295 and 297. Although the chemistry presented above provided the key dipyrrolic intermediates, for subsequent access to the porphocyanines, it at times lacked generality and proved inefficient. Moreover, the dehydration strategy utilized here was generally plagued by inherent side reactions. For example, during an initial procedure this approach gave, as the major products, the mono-N-acetylated dipyrromethane 298 and its tricyclic derivative 299 in roughly equal amounts, and the desired biscyano compound in only minuscule quantities. The formation of these by-products clearly result from acetic anhydride's efficient acylating properties to give initially the acetyl derivative 298 of the dicyano compound 297. Prolonged exposure of this species to 130 these harsh reaction conditions, not only favours the formation of 298, but additionally sets up the latter compound to undergo a condensation-dehydration sequence to yield its tricyclic derivative 299. Thus in order to avoid the formation of these by-products and the associated encumbersome separative procedures, an alternative dehydration methodology which would still be compatible with the dipyrromethanes was sought. However, a number of other dehydrating agents, e.g. 1,2-dibromotetrachloroethane/triphenyl-phosphine/triethylamine142 and trifluoroacetic anhydride/pyridine/1,4-dioxane143 sys-tems, only led to extensive decomposition. Alternatively, attempts at by-passing this dehydration process and directly access the unstable 5,5'-bisaminomethyl dipyrro-methanes, which were believed to be the final intermediates in the macrocyclization step, 8 8 from the bisoxime 296 through either LiAIH4 reduction1 4 4 or catalytic hydrogenation over P d 1 4 4 1 4 5 and Pt0 2 , 1 4 4 failed to effect the desired transformation. Scheme 47. Introduction of a cyano moiety onto the 5-position of pyrrole 300. The lack of success with these methods prompted an investigation into a different strategy. Here, encouraged by our success at effecting the nucleophilic displacement of the 5-iodo substituent of pyrrole 300 1 4 6 with CuCN in refluxing DMF (scheme 47),147 we focused our attention on applying this methodology to the corresponding dipyrrolic substrates. With this approach, however, unexpected difficulties were encountered in the key iodinative decarboxylation step. In this 300 301 131 instance, exposing basified methanolic148 or dichloromethane145 solutions of the diacid 294 to an aqueous solution of l2-KI at ca 60 °C failed to yield the requisite 5,5'-bisiododipyrromethane 302 despite precedent.149 The failure of this iodination reaction can, perhaps, be attributed to the iodine acting more as an oxidant than an electrophile, yielding the corresponding dipyrromethene salt, which is now much less susceptible to electrophilic attack than, for example, pyrroles. Evidently, such a sequence of events has been noted with Br2, where the reaction proceeded with simultaneous decarboxylation and bromination of a similar 5,5'-diacid, in addition to oxidation to the dipyrromethene.150 Moreover, despite the high yields realized in the above substitution reaction with pyrrole 300, this intermediate too proved unsuitable for further elaboration to the corresponding dipyrromethane in view of Clezy's work in this area. 1 4 1 In 1969, this Australian based group showed that self-condensation reactions of appropriately a-derivatized cyanopyrroles only leads to the formation of either intractable tars or monopyrrolic by-products. 294 302 Scheme 48. Attempted iodination of dipyrromethane 294. Undeniably, we required a more versatile strategy which would ideally be tolerant to a diverse range of substituents. After a considerable effort, this goal was eventually realized with the discovery of a direct approach for introducing nitrile groups onto active aromatic ring systems. 1 5 1 Through the one-pot sequence outlined in scheme 49 below, the key biscyanodipyrromethane was obtained in reasonable yields. Thus, diacid 294 was initially converted to the intermediate di-oc-free dipyrromethane 303, 132 which was subsequently diluted with an equal volume of CH 3 CN, cooled to -78 °C, and treated with an excess of chlorosulphonyl isocyanate (CSI). Upon slow warming, to -40 °C, the intermediate bis-N-chlorosulphonylamide undergoes solvolysis with DMF yielding the bis-nitrile 297 in 40% overall yield from the dibenzyl ester 293. This three-step sequence clearly was a vast improvement over our previous methods. In fact, the remarkable versatility and selectivity of this reagent-solvent combination is perhaps best demonstrated by the wide range of dipyrromethanes, and more importantly the p-unsubstituted dipyrromethanes, we were subsequently able to bis-cyanate at their free a-positions.1 5 2 CSI, CH3CN / DMF 78° - 20°C Scheme 49. Direct introduction of cyano groups onto free cc-carbons of dipyrromethane 303. 133 15.2 Synthesis of an Octaalkylporphocyanine Derivative and Some Mechanistic Considerations With the 5,5'-biscyanodipyrromethanes in hand, the remaining steps in the synthesis of porphocyanines now simply consisted of the self-condensation these biscyano compounds and subsequent oxidation of the intermediate macrocycle 304. As 307 Scheme 50. Synthesis of porphocyanine 305. 134 described previously, this was accomplished by a LiAIH4 reduction in THF at 0 °C, followed by dehydrogenation with DDQ at room temperature, furnishing the aromatic porphocyanine 305 in 18% yield. The ease with which macrocycle 304 undergoes this dehydrogenation reaction is particularly noteworthy. Like the congeneric porphyrinogens, this transformation occurs instantaneously upon addition of a dichloromethane solution of DDQ. This process is self-evident in the changes in the both the solution colour (from a straw-yellow to fluorescent green) and the UV-visible spectrum, where an intense Soret-type absorption rapidly develops at 454 nm. Three additional weaker bands were observed at 592, 632 and 798 nm. The 1 H NMR conclusively manifests the presence of a substantial diamagnetic ring current, and thereby confirming the aromatic nature. Here, for instance, the peripheral imine and meso protons appear at the lowfield positions of 12.7 and 10.4 ppm respectively. The proton signals of (3-alkyl side-chains also exhibit similar lowfield shifts. The internal pyrrolic NH signals, however, were not observed in the field range scanned (i.e. -6 to 15 ppm). Mechanistically, this intriguing cyclization is believed to result from the expected initial reduction product, the bisaminomethyldipyrromethane 306, ionizing to the di-carbocationic species 307 which then couples with a second molecule of 306 to provide the dihydro precursor of the bisimine macrocycle 304 as shown in the scheme above. This proposal is highly plausible on account of (i) Franck's biomimetic synthesis of N,N',N",N"'-tetramethylporphyrinogen form the correspond-ing N-methylnorporphobilinogen,11ac and (ii) Cookson's and Rimmington's pioneering work in this area which established that porphobilinogen 165 (fig. 29) in the presence of an acid catalyst gives uroporphyrinogen III,153 the biogenetic precursor of all natural porphyrins. The key point to note here, however, is that 135 these experiments were all conducted in the presence of an acid catalyst. In the present case, on the other hand, the proposed intermediate aminomethyl 306 was at no time exposed to any acidic media in this route. Moreover, several attempts to trap this intractable intermediate have met with no success. Nonetheless, this reductive-cyclization reaction has so far proved to be highly specific of the biscyanodipyrromethanes. What is certain, however, is that the cyclization occurs during the reductive process. In view of the above argument, this in turn has led us to propose an alternative reaction mechanism whereby the nitriles are only partially reduced, to the corresponding imines (by addition of 1 mole of hydrogen). It is conceivable, that the resultant imino groups are coordinated to the aluminium atoms which serve to hold two such units in close proximity, and thus enable the cyclization reaction at this oxidation level giving the intermediate bisimine macrocycle 304; or by further reduction to the unstable aminomethyl compounds, and follow the presumed "carbonium ion" route. Interestingly, independent work in our laboratories, by fellow colleagues, has recently shown that reducing pyrrole-2-carbonitrile leads exclusively to the formation of the dipyrroimino compound 308 in good yield- scheme 51 . 1 5 4 These findings are further corroborated by previous work of Anderson et a / . 1 5 1 d In this example, however, the resultant dipyrrolylimine, corresponding to 308, from hydride reduction of 2,4^dicyanopyrrole was not isolated, but hydrolysed to furnish 4-cyano-2-formylpyrrole instead. Such partial reductions of nitrile groups by metal hydride reagents, and in situ hydrolysis of the resulting imines, represents a well documented methodology for transforming nitriles to aldehydes. 1 5 5 These results do however, seemingly lend more weight to the second of the proposed mechanisms. Unfortunately, the inability to isolate any intermediates in these experiments coupled with the fact that both mechanisms 136 provide the same end results, makes the task of confirming which of the two pathway is actually followed rather difficult. H H H 308 Scheme 51. Preparation of the dipyrrolylimine 308. With the above view in mind, we pursued a strategy which may help to solve this mechanistic argument. The most obvious approach to tackle this problem was to directly access these bisaminomethyldipyrromethanes via an alternative pathway based on our earlier work with analogous furan and thiophene systems. In these instances, the synthesis starts off with the heterocycle already containing this key amino functionality in a protected form. With the pyrrolic series, this was readily accomplished by reacting the acetoxymethylpyrrole 15 with potassium phthalimide in DMSO at room temperature, giving the phthalimidomethylpyrrole 309 in 60% yield. 1 5 6 A simultaneous debenzylation and decarboxylation of the phthalimido-methylpyrrole 309 in acidic media in the presence of anisole, which, evidently acts as a benzyl carbonium ion scavenger, provided the key a-free pyrrole 310. Thus, with this compound in hand, the sequence now hinged upon the successful coupling of two of these pyrrolic units with an appropriate one carbon building block. Several attempts to effect this transformation with benzaldehyde or formaldehyde as the one carbon source, under various acid catalysed conditions failed to yield the target dipyrromethane 311. After considerable efforts, the 5,5'-bisphthalimidomethyl-2,2'-dipyrromethane 311 was eventually serendipitously isolated, albeit in 4% yield, after chromatographic 137 purification of the product mixture, from one such condensation with paraformalde-hyde and catalytic cone. HCI in refluxing ethanol; in an attempt to identify the major products of this condensation. However, subsequent efforts to repeat this reaction failed to provide any evidence for the formation of this key dipyrrolic intermediate, as judged by TLC. 309 10% H 2S0 4 / CF 3C0 2H, PhOMe, (68%) O 312 Scheme 52. Preparation of the 2-phthalimidomethylpyrrole 310 and attempted elaboration to its dipyrromethane (311) and 2-chloromethyl (312) derivatives. 138 Additionally, an alternative approach to access the target molecule 311, via the 2-chloromethylpyrrole 312 (and/or hence its acetoxymethyl derivative), was stopped short by the failure of pyrrole 310 to undergo this critical chloromethylation step. Here, TLC revealed the formation of complex mixtures, undoubtedly from degrada-tion of the starting material. The above observations suggest that the phthalimidomethyl groups, may not be sufficiently electron-withdrawing to stabilize the formation of the dipyrromethane systems, under the reaction conditions employed. Evidently, this finding contradicts two recent reports on the synthesis of 5,5'-unsubstituted dipyrromethanes, which are reportedly stable in the absence of air, light and acidic media. 1 5 7 However, adapting either of these methods into the current synthesis failed to provide the target dipyrrolic compounds. Thus far, we have been unable to access this key dipyrrolic intermediate in reasonable quantities for further elaboration to the critical aminomethyl adducts required to provide more concrete evidence for the above suppositions. Perhaps, in order to evaluate any template effects played by the aluminum ions, a corresponding hydride reduction of 2,5-dicyanopyrrole would be of particular relevance in regards to the above suppositions. Whether or not such a reduction would yield the corresponding tetraiminoporphocyanine, structurally analogous to Franck's tetravinylogous porphyrins,8 3 , 8 4 through a biomimetic type reaction path-way, however, remains to be seen. 139 16.0 Heteroporphocyanines 16.1 Synthesis of Acyclic Precursors Initially, we sought to access these heteroporphocyanines analogues through our original porphocyanine strategy (i.e. in a sequence analogous to scheme 46). Therefore, we now required efficient routes towards the 5,5'-bisformyl dithienyl- and difuryl-methanes. A survey of the literature revealed that the most direct method of obtaining the precursor a-free difurylmethane intermediate was by either an acid catalysed condensation of furan with aqueous formaldehyde, or a higher yielding reaction between furan and furfuryl alcohol in the presence of HCI. 1 5 8 In view of this, the latter strategy was adopted. Thus, treating a neat mixture of furan and furfuryl alcohol with cone. HCI at 12 °C, followed by successive neutralization, steam distillation, and finally a second distillation under reduced pressure yielded a colourless oil, which, upon spectroscopic analysis was found to comprise of a 1:1.5 mixture of the target difurylmethane 314 and unreacted furfuryl alcohol respectively. Consequently, this mixture proved difficult to handle, largely due to the presence of the unreacted alcohol, which readily polymerized. Several attempts to fractionally distill off the furfuryl alcohol or chemically remove this material by derivatization proved futile. (< 2%) 313 314 313 Scheme 53. Acid catalysed condensation reaction of furan and furfuryl alcohol 313. 140 The above strategy unquestionably suffered from the lack of a suitable protecting group at the furyl a-positions to prevent further condensations leading to poly-merization. Thus to remedy this situation, we looked into an alternative procedure which incorporates an easily manipulable blocking group, e.g. as an ester, at the 2-position. Here, once again, a procedure described in the literature in the 1930's proved the quickest route to the prerequisite difuryl molecules. 1 5 9 As reviewed in the scheme below, commercially available 2-furoic acid was initially elaborated to its ethyl ester in two steps, which was subsequently coupled with paraformaldehyde in cone, sulphuric acid to furnish the difurylmethane 318, in moderate yield. Subsequent base hydrolysis of the ester groups, yielded the intermediate diacid 319 in a more easily manipulable solid form. 315 R = OH^. R = CI-*-' R EtOH, reflux (74%) Or OEt (CHO)n, H 2S0 4 0°C, 0.5h (36%) 316 SOCI2, reflux 317 EtO OEt 318 20% aq. NaOH, reflux; (40%) 1. SOCI2, reflux 2. NH3, Et20, 0°C (72%) 320 319 Scheme 54 Synthesis of difurylmethane intermediates 318-320. 141 Although, others have reported the decarboxylation of this diacid, to give the a-free difurylmethane 314, 1 5 9 we specifically chose to avoid this procedure in view of the harsh conditions required, and the low yields in both this step and the anticipated subsequent formylation step to obtain the 5,5'-diformyl adduct;1 6 0 a key intermediate in this route to the porphocyanine. Attempted decarboxylations in high boiling solvents such as DMF (145 °C) or quinoline (240 °C) were found to reach insufficient temperatures to effect this process. Similarly, selective reductions of diacid 319 (Borane-dimethylsulphide/ P C C , 1 6 1 or 9-BBN/ teuLi162) or its immediate precursor diester 318 (LiAIH4/DEA/pentane)163 to directly yield the 5,5'-bisformyl-difurylmethane, were ineffective. Moreover, whilst the furyl esters are readily available, the difficulties associated in purifying the precious difurylmethane 318, from the intractable tars that inevitably result during the condensation, highly discouraged attempting the above inefficient decarboxylation-formylation sequence. NC CN 322 Scheme 55. Attempted strategies towards the synthesis of the 5,5'-bisaminomethyl-2,2'-difuryl-methane (321). Conditions: (i). (CF3CO)20, Py., Dioxane/DMF;143 (ii). SOCI2, reflux;f64 (iii). PPE, CHCI3, 80 ° C . 1 6 5 Nevertheless, diacid 319, when refluxed with thionyl chloride readily provided the intermediate diacid chloride, which was immediately elaborated into its bisamido 142 derivative 320, in 72% overall yield, upon saturating an ice-cold ether solution with gaseous ammonia. This bisamidodifuryl intermediate now paved the way for other strategies towards the target bisaminomethyl difurylmethane, as summarized in the scheme above. However, neither of these approaches succeeded in yielding this elusive target diamine. This, by and large, for the reduction and dehydration sequence (iii) was a consequence of the extremely poor solubility of the bisamido-difurylmethane 320, whilst the harsh dehydrating conditions described by methods (i) and (ii) led to extensive degradation. 319 314 323 CSI, DMFV CH3CN, -78 0 to -20 °C, (70%) T A y / \ / \ CSI, DMF/ CH3CN, \—S s—!l -78 °CtoRT 324 322 X=0 A 325 X=S CH20, ZnCI2, HCI, (29%) Scheme 56. Synthetic approach to the 5,5'-dicyano difuryl- and dithienyl-methanes 322 and 325. Despite these failures, the bisamide 320 is readily reconverted to its diacid 319 by refluxing in 10% aq. NaOH, and is therefore not wasteful of any difurylmethane 143 samples. Fortunately, our recent discovery of a less cumbersome methodology for directly introducing the cyano group onto 5,5'-unsubstituted dipyrromethanes with the low temperature CSI/DMF combination,152 not only appeared an attractive alternative, but additionally made the drastic decarboxylation step of the diacid 319 synthetically more feasible. Under the extremely brutal conditions of Dinnelli and Marini, 1 5 9 diacid 319 cleanly underwent decarboxylation to furnish the unsubstituted difurylmethane 314 in 22% yield, together with a minor quantity (ca 2%) of the monoacid 323, as shown in scheme 56. Exposing 314 to CSI under the conditions described previously, then provided the easily isolable biscyano adduct 322 in remarkably good yield (60%). The success of this reaction prompted an extension of this work to the thiophene analogues. Unfortunately, the corresponding dithienylmethane 324 5 6 failed to react with the CSI under these conditions. The lack of reactivity observed here is not too surprising, as thiophenes are known to be the least reactive towards electrophiles in this series of five-membered heterocycles. It should be noted, however, that subsequent to the submission of this work for publication, a recent paper by Vorbriiggen and Krolikiewicz was brought to our attention. In this account, the authors report that thiophene and its derivatives do indeed react with CSI, but only at room or elevated temperatures (e.g. in refluxing ether), to provide the N-chlorosulphonylamide in good yield. 1 6 6 Treatment with triethylamine, or solvolysis with DMF, then provides the nitriles. A preliminary investigation with 324 in benzene/THF with CSI at room temperature followed by reaction with triethylamine overnight, furnished a complex mixture (as inferred by TLC). However, this was not extensively evaluated any further, in view of the subsequent shortcomings encountered with the furan analogue 322 in this synthetic plan to heteroporphocyanines. Despite this, further examination of the literature for 144 related polythiophene adducts, increasingly indicates that 324 would yield the biscyano adduct given the appropriate reaction conditions. The successes achieved with the difurylmethane, however, proved to be short lived as indicated above. For instance, co-reducing the biscyanodifurylmethane 322 with the corresponding biscyanodipyrromethane 297, followed by oxidation with DDQ failed to furnish the heteroporphocyanine, but instead gave the previously isolated tetrapyrrolic macrocycle 305 in very low yield as inferred from their identical electronic spectra. Purification of this material, on deactivated alumina, however, gave a brown solid which failed to exhibit any features typical of porphocyanines in the 1 H NMR spectrum. More interestingly, hydride reduction of 322 did not yield the corresponding 5,5'-bisaminomethyldifurylmethane 321. The exact nature of the product(s) from this reduction, have yet to be determined. This latter finding is particular unusual, in that, unlike their pyrrolic congeners, the aminomethyl furans and thiophenes are typically more stable, and should be readily isolable. Therefore, these compounds are not expected to undergo the macrocyclization described previously for the analogous dipyrrolic compounds. Nonetheless, the above work undoubtedly demonstrated that a substantial amount of these strategies were largely rendered synthetically ineffective, by the inability to efficiently manipulate the functional groups of the intermediate compounds into those required for the final macrocyclization step. Therefore, during the course of this work, we increasingly placed stronger emphasis on introducing these critical substituents, in particular the intractable aminomethyl groups, during the preliminary stages of the synthesis in a protected form. Moreover, this will additionally impart more control on the number of subsequent steps required for manipulation of the acyclic precursors. 145 One of the most useful and facile route to emerge in organometallic chemistry for designing novel macrocyclic ligands is clearly that based on the condensation reaction between primary diamines and dicarbonyl compounds. 7 , 1 4 ' 1 8" 2 1 ' 2 5" 2 8 In context of the current synthesis, this approach appeared to be by far the most conducive strategy for introducing heterocycles other than pyrroles into the basic porphocyanine framework. It additionally benefited from the fact that, now we would be working with more stable and readily accessible pyrrolic intermediates, e.g. the 5,5'-bisformyldipyrromethane and, the two key precursor aminomethyl heterocycles-i.e. furfurylamine and 2-thiophenemethylamine- are readily available . This new approach to the heteroporphocyanines still relies on an initial coupling reaction akin to that of our previous work with ethyl 2-furoate. The critical feature of these reactions is their requirement of strongly acidic conditions to activate the free a-position of the furan and thiophene moieties for the ensuing electrophilic attack by the protonated formaldehyde species. With a few minor modifications of the original sequence, 1 5 9 this same methodology worked with great success for the 2-phthalimidomethylfuran 328, and subsequently for the analogous thiophene derivative 329, as shown in scheme 57. The synthesis started with furfurylamine 326, of which, the amino groups was protected as its phthalimide derivative in 87% yield, using a standard derivatization procedure.1 6 7 Subsequent condensation of the resultant phthalimide derivative 328 with paraformaldehyde in TFA at low temperature (< 0° C) furnished the difurylmethane 330 in moderate yield (32%). In a parallel sequence, 2-phthalimidomethylthiophene 330 1 6 8 provided the corresponding 5,5'-bisphthalimidomethyl-2,2'-dithienylmethane (331) in 27% yield. In this instance, however, the coupling was now effected with an aqueous solution of formaldehyde in the presence of catalytic ZnCI 2 . 5 6 The desired bisphthalimido products, 330 and 331, in both cases, were readily separable from unreacted starting material and the 146 inevitable resultant polymeric material, by initially passing through a short silica pad eluting with dichloromethane, followed by reerystallization in aqueous ethanol. 326 X = O 327 X= S (CHO)n, TFA < 0°C NAc2 HNAc 333 Scheme 57. Synthesis of the.bisphthalimidomethyl difuryl 330 and dithienyl 331 intermediates, and the subsequent deprotection of the bisamino groups. The formation of higher polymers in these coupling reactions is an unavoidable consequence of the presence of unsubstituted (3-positions in the difuryl and dithienyl products 330-331. No further evaluation of these compounds were attempted, since they were of no major interest to the present work. However, similar adducts 147 isolated from condensation reactions of furan esters have been extensively investigated by others, 1 6 9 , 1 7 0 and were found to comprise of polyether-type macro-molecules arising largely from electrophilic attack at the free p-positions of the initially formed difurylmethanes and difurylmethyl ethers. Therefore, it is predictable that analogous oligomers result in the above coupling sequences. Despite this undesirable polymerization, this strategy does represent the shortest possible route to two of the key precursors required in our synthesis of the targeted porphocyanine macrocycles. The final step in this sequence involved cleavage of the phthalimide groups of 330 and 331 with methanolic hydroxylamine156 at ambient temperature to provide the bisaminomethyl compounds 321 and 332, as oils, both in excellent yield (ca > 90%) and purity. More importantly, these critical precursors display good stability, and do not appear to degrade on standing in the open atmosphere at room temperature. The difuryl compound exhibited some interesting chemical properties. For instance, acetylation in refluxing A c 2 0 exclusively yields triacetate 333. However, the low yields attained in this reaction suggested that at elevated temperatures, degradation of 321 may readily occur. Nevertheless, the successful isolation and characterization of the above diamines, together with the subsequent derivatization of the difuryl analogue were extremely encouraging. 16.2 Synthesis of Tetrahydro-heteroporphocyanines On account of the low yields realized in the above acetylation reactions, model studies with the more readily available furfurylamine and an appropriate 5-formyl-pyrrole were carried out to determine the optimum conditions for effecting the condensation reactions. Through these, a titanium(IV) mediated reaction in toluene 148 at room temperature was initially chosen as the preferred strategy in light of the mild conditions and the purity of the crude product (as inferred from the 1 H NMR). Additionally, the use of titanium tetrachloride appeared advantageous for several other reasons. Firstly, TiCI4 is an extremely effective water scavenger, secondly, it can act "catalytically" as a Lewis acid to polarize the carbonyl bond, and, lastly, it can exert a template effect in the synthesis thereby promoting the macrocyclization step. Thus, stirring a mixture of the dialdehyde 295, TiCI4 and a three fold excess of the diamine 321 in toluene overnight furnished the imino macrocycle 334. This route, however, suffered from inconsistent yields and it was wasteful of the amines. Two factors which soon proved critical with the subsequent difficulties encountered with the attempted isolation and oxidation of this cyclic bisimine macrocycle. Introducing a hindered non-nucleophilic (e.g. N,N-diisopropylethylamine, Hiinig's base) base as an HCI acceptor, and thus conserving the diamine 321, prevented the macro-cyclization altogether. Moreover, these macrocyclic products were severely contaminated by side-products (as inferred from the 1 H NMR). Through reevaluation of our earlier efforts with furfuryl amine and benzyl 5-formyl-3-ethyl-4-methylpyrole-2-carboxylate a more efficient approach to these cyclic imines was subsequently developed. Here, refluxing the diamines 321 or 332 in toluene containing small amounts of methanol, to improve the solubility of the aldehyde, in the presence of catalytic p-TsOH, nicely furnished macrocycles 334 and 335 respectively. Analogously, this methodology has since been generalized to provide the homologous macrocycle 340 derived from diformylbipyrrole 116 1 0 1 and diamine 321 as reviewed in scheme 59. 149 336 X=0 338 X=0, S 337 X=S Scheme 58. Synthesis of bisimine macrocycles 334 and 335, and the partially oxidized macrocycles 336 and 337. Despite being isolated as solids, these bisimine macrocycles proved difficult to purify by recrystallization in various solvent combinations. In fact, to this date we have been unable to obtain reasonably crystalline samples for a more detailed structural characterization by single crystal X-ray analyses. Furthermore, chroma-tographic separations on silica gel or deactivated alumina were severely hampered by the high basicity of the macrocycles, and the ease at which they undergo hydrolysis to their respective acyclic components. However, it is expected that 150 oxidation of these cyclic imines (by six electrons), and the accompanying aromati-zation, would not only enhance their stability, but almost certainly improve the potential utility of these systems as ligands. 339 R = CH=C(CN)C02Me ~ N KOH, MeOH, 116 R = CHO - / A Scheme 59. Synthesis of macrocycle 340. 16.3 Spectroscopic and Chemical Properties The critical structural feature to note of these systems (i.e. 334, 335, 340) is that they are nonconjugated and therefore nonaromatic. As such, they are structurally related to both the porphyrinogens11a"c ,3° and heteroporphyrinogens,11e,f respec-tively. This resemblance is reflected in their chemical and even more so in their spectroscopic properties. For instance, the bridging methylene protons of the dipyrr-olic halves of both macrocycles 334 and 335 appear as singlets at 8 3.7 and 3.8 ppm respectively in their 1 H NMR spectra. The analogous signals appear at 8 3.8 ppm in the 1 H NMR spectrum of the N',N",N"',N""-tetramethylporphyrinogen, prepared by Franck and Wegner. 1 1 a The equivalent groups linking the furan and 151 thiophene subunits of 334 and 335 resonate at 6 3.89 and 4.25 ppm respectively. These values correlate well with those of the corresponding bridging methylene groups in the tetraoxa-1 1 6 and tetrathiaporphyrinogens11f (cf. 5 3.83 and 4.14 respec-tively), of Vogel. This comparison in 1 H NMR signals can be extended to the respective p-protons of the furan and thiophene rings. Thus, for macrocycle 334 these protons appear as a multiplet at 6.02, whilst the analogous protons of the dithienyl macrocycle 335 resonate at 8 6.7, (cf. 8 5.95 and 6.61 for the p protons of the tetraoxa-1 1 6 and tetrathiaporphyrinogens11' respectively). The correspondence is even greater in the 1 3 C NMR spectra. In accord herewith, in the 1 3 C NMR of the 27,28-dithiamacrocycle 335, for example, these bridging carbon signals appear at 8 22.1 ((pyrr)2-CH2) and 31.7 ((thio)2-CH2), whilst those of the true porphyrinogens are located at 22.3 1 1 a and 31.77 1 1 1 respectively. The imine carbons and the methylene carbons bridging the thiophene and imine subunits in this system resonate at 150.077 and 59.96 ppm respectively. Compounds 334 and 335 are, of course, by no means porphyrinogens. They, thus display singlets at 8 8.0 and 8.1 (respectively) ascribable exclusively to the imine protons, in addition to singlets integrating for the four methylene protons adjacent to the two imine nitrogens at 8 4.59 and 4.69 ppm in their respective 1 H NMR's, and C=N stretching bands in the IR region of the electromagnetic spectrum. Interestingly, however, the pyrrolic NH protons are not observed for either of these macrocycles in the field range scanned in these NMR experiments. Evidently, this was not limited to the present examples, but was also found to be typical of both the other cyclic 340 and acyclic (see later sections) imines prepared by us. Macrocycle 340, which incidentally, may be regarded as an expanded corrole,96 shows similar spectroscopic properties in the 1 H NMR. However, the resonance peaks in these spectra were severely broadened. Nevertheless, it was possible to 152 assign the key features, from these spectra. Thus, as with the congeneric 25,28-dioxamacrocycle 334 the bridging methylenes appear as broad singlets at 3.8 ((furyl)2-CH2) and 4.5 (furan-CH2-N=C), and the imine protons resonating at 5 8.05 ppm respectively. The formation of these macrocycles (i.e. 334, 335, and 340), as opposed to higher oligomers, is fairly evident in their 1 H NMR spectra, where peaks corresponding to either terminal formyl or amino groups are not observed. This is further corroborated in their mass spectra, by the presence of the molecular ion peaks. Moreover, as is true for the porphyrinogens11a~c,3° and heteroporphyrinogens,11ef macrocycles 334, 335, and 340 show absorbances only in the UV and not in the visible portion of the electromagnetic spectrum. The dioxa and dithia macrocycles 334 and 335 display intriguing chemical reactivity, in which, the dipyrrolic and the dioxa or the dithia subunits (respectively) seemingly behave as two independent molecules. Nonetheless, in analogy to the porphyrino-gens, they are thermodynamically unstable in air. However, unlike the latter systems where air or chemical oxidation leads rapidly to porphyrin formation, n a" c , 3° com-pounds 334 and 335 react only slowly with air to give as yet uncharacterized decomposition products. In marked contrast, however, when the acid catalyzed cyclization between the dialdehyde 297 and either of the diamines 321 or 332 was effected over a 12h period in the presence of air, under anhydrous conditions, the partially oxidized macrocycles 336 and 337 were now isolated as the sole products, as shown in scheme 58. This reaction was serendipitously realized during an initial attempt at accessing the dithia compound 335. The formation of 337 is evident in its 1 H NMR, where the bridging methylene signal of the dipyrrolic subunit of 335 (at 8 3.8 ppm) was replaced by a singlet at 6.8 ppm, ascribable to the bridging methine 153 proton of the dipyrromethene subunit. Furthermore, the UV-visible spectrum of 337 contains a broad absorption band at ^ m a x = 472, typical of dipyrromethenes- cf. 480 nm, 1 3 8 with a shoulder at 504 nm. The difuryl macrocycle 336 exhibited similar spectral features. Unfortunately, despite this exciting breakthrough, attempts to fully oxidize these macrocycles 336 and 337 to their fully conjugated congeners 338 by prolonging the above refluxing conditions (up to 48h) were unsuccessful. Indeed, to date, standard organic oxidations of macrocycles 334-337 have provided no evidence for the formation of the desired aromatic product 338. Compound 340, on the other hand, is somewhat even more unstable, decomposing in a matter of weeks when left unprotected, and even more rapidly when in solution. For instance, the 1 H NMR of a crude sample 340 in deuterochloroform exposed to the atmosphere over a three day period at room temperature revealed a rather complex mixture, with no evidence of its aromatic congener 341. Electronic spectra of the NMR samples corroborate this contention. Nevertheless, these preliminary results do indicate that the degradation of these imino-macrocycles may involve chemical processes other than simply just the expected hydrolysis of the imine bridges. Furthermore, the unusually low stability observed for compound 340, may be indicative of the steric strain imposed on the difuryl subunit (specifically, lone pair-lone pair interactions) in order to accommodate the smaller bipyrrolic portion of the macrocycle, in its reduced-form. Molecular modelling1 3 6 of this species, as shown figure 53, show that the two furan oxygens are indeed forced above the general macrocyclic plane. Similarly, substantial ruffling of the bipyrrolic portion of the molecule is predicted. The figure on the right depicts the oxidized form of the macrocycle. Here, the flexibility of this system is more restricted, and thus a more planar geometry is anticipated. However, 154 the molecular mechanical calculations do predict some deviation from planarity around the bipyrrolic nitrogens, and one of the imine nitrogens. In the case of the former, the deviation of one of the pyrrolic subunits js such that, clearly an efficient overlap of the Tc-framework in a planar array does not appear achievable, and may be (amongst other factors) one of the reasons why the aromatic species has not yet been attained. Figure 53. Computer generated, energy minimized structures of 340 (left) and the aromatic species 341 (right). Unfortunately, the oxidation of macrocycles 334-337, 340, has thus far proved to be an extremely difficult feat; reminiscent of the early texaphyrin chemistry.21 However, whilst in the latter case, Sessler et al. have shown that stirring the reduced-form macrocycles in air-saturated chloroform/methanol solutions containing N,N,N',N'-tetramethyl-1,8-diaminonaphthalene effected the desired 4-electron oxidation,32 these conditions did not result in oxidation to the aromatic species with the present "expanded porphyrinogens". In fact, under a range of conditions and in the presence of a variety of oxidants, as summarized in scheme 60, the reduced-form macrocycles 334-337, 340 have failed to yield even trace amounts of their 22TC-155 electron aromatic analogues 338 and 341, respectively. In most instances decomposition of the initial macrocycle resulted. 340 341 Scheme 60. Some typical oxidants utilized: (a) DDQ; (b) Chloranil; (c) l2; (d) Br2; (e) Ce(NH4)2(N03)6; (f) l2/p-benzoquinone;133 (g) N,N,N',N'-Tetramethyl-1,8-naphthalenendiamine/02;32 (h) Mn02, PhMe, reflux;171 (i) Ph 3CBF 4, CHCI3; (j) Ba(N03)2/02. The failure to effect this key oxidative transformation is undoubtedly a reflection of the instability of these macrocyclic systems, more specifically due to the presence of the imine bridges. Moreover, molecular models predict that these nonconjugated systems do not assume highly distorted conformations which would prevent the oxidation to the planar, fully conjugated aromatic species. Figure 54 depicts the computer-generated, energy-minimized136 structures of macrocycles 334 and 336 respectively. As can be seen from the diagrams below, in the hexahdyro species, 334, the major distortions occur around the dipyrrolic subunits such that the internal NH's of the two pyrroles are directed above and below the mean macrocyclic plane. The furans, on the other hand, are twisted in the same direction (in respect of the general macrocyclic plane), but to a lesser extent. Interestingly, the partially oxidized compound 336, is already in an almost planar conformation. In fact, this structure is virtually identical to that of the fully oxidized compound (cf. figure 55 below, depicting 156 the energy-minimized structure of 338 (X = S), which contains the even more sterically demanding sulphur atoms). Figure 54. Molecular models of macrocycles 334 (left) and 336 (right), depicting the expected energy minimum conformation. Nevertheless, it is highly plausible that with the oxidants employed here, macrocy-cles 334 and 335 are initially oxidized to the dipyrromethene-containing system 336-337 (respectively), and it is these compounds which are highly unstable to oxidation. It should be noted, however, that Vogel et al., have oxidized tetraoxa- and tetrathia-porphyrinogens to their porphyrinoid dicationic counterparts in the presence of strong oxidants, such as DDQ, Ce i v , H N 0 3 and Brg,1 1 6'''1 all of which, with the exception of nitric acid, have failed to yield the same effect with compounds 334-337 and 340. Thus, collectively these findings serve to indicate that introducing imino bridges into heteroatom-containing porphyrinoids substantially destabilizes the resultant expanded porphyrinogens towards oxidation. These results, however, do not indicate whether the bridging groups between the furans or thiophene subunits and/or the methylene groups attached to the imine nitrogen are in fact dehydrogenated with these oxidants. Attempted in situ 1 H NMR experiments provided very little insight into the nature of the products of these oxidation reactions. 157 Figure 55. Computer simulated structure of the fully conjugated macrocycle 338, showing the expected planar conformation, and some of the calculated key inter atomic distances. The difficulties encountered in the above work, thus prompted us to investigate their metallochemistry instead. With six potential chelation sites, we anticipated a reasonably varied coordination chemistry wherein, perhaps, at least one or two first-row transition metals, or a single lanthanide cation may be inserted into the central core. This metal complexation may further engender greater stability upon the resultant adducts, and thus aid the subsequent isolations. Initial attempts to form coordination complexes between a variety of transition metals and lanthanide(lll) cations, and with barium salts, however, proved to be unsuccessful. This by and large, is a direct consequence of the instability of these reduced-form macrocycles. Preliminary indication of this phenomena came during one attempt to isolate the copper (II) adduct of 334. Instead, the corresponding copper tetralkylporphyrin complex was isolated in extremely minuscule quantities. The above findings, were to some extent in striking parallel to the reported early difficulties encountered with the organometallic chemistry of the texaphyrins.2 1 , 2 6 158 Since our macrocyclic systems (334, 335, 340) contained structural features analogous to the texaphyrins, and displayed similar chemical reactivity with respect to oxidation, we attempted to adapt some of this work into the current synthesis. In marked contrast to the texaphyrins, 2 8 , 3 2 , 3 3 , 3 7 ' 3 8 however, metallation reactions of these macrocycles under aerobic conditions in chloroform/methanol mixtures with Cd(ll) and Ln(lll) salts, in the presence of Et 3N or N,N,N',N'-tetramethyl-1,8-diamino-naphthalene, at ambient or reflux temperatures did not proceed with concomitant metal insertion and oxidation. These systems appear to readily decompose instead under these conditions. Indeed, thus far, no stable metallo complexes of macrocy-cles 334, 335, or 340 have been either detected spectroscopically or isolated. Clearly, as mentioned previously, it is apparent that in order to fully exploit the potential of the heteroporphocyanines as ligands, the critical oxidation step to provide their presumably more stable aromatic congeners is a major requirement. 16.4 Attempted Preparations of 5-Oxoporphocyanine Analogues The most noteworthy point that can be deduced from the above chemistry, is that the dipyrrolic parts of these macrocycles display reactivity expected of dipyrrometh-anes. On the other hand, the difuryl and dithienylmethane containing portions appear to be fairly "stable" to the oxidation conditions employed. With this view in mind, we now sought to structurally modify the framework of these systems. The main idea here was to introduce a group at this methylene bridge, which could be more easily oxidized, or chemically abstracted. It was further envisaged, that such a reaction may trigger a similar effect on the saturated methylenes adjoining the imine nitrogens and these heterocyclic subunits. Thus, driving the oxidation reaction towards their 22jt-electron congeners. The most obvious group that came to mind, 159 was that of a bridging carbonyl moiety, thus giving the 5-oxo-porphocyanine analogue 343 (scheme 61). o o OR 342 343 344 Scheme 61. Schematic representation of 5-oxo-bisimine macrocycle 342, its cross-conjugated congener 343, and the meso-hydroxy tautomer 344. Although, others have shown that in both the oxophlorins172 and their 23,24-dioxa analogues, 1 7 3 the keto tautomers (cf. 343) prevail in neutral organic solvents, the meso-hydroxy tautomer (i.e. structures corresponding to 344 , R = H) is favoured by the dications and metal complexes of these porphyrin derivatives.174 Moreover, this hydroxy species can be trapped as the acetate. 1 7 2 Nevertheless, to us the cross-conjugated system 343 and its aromatic congener 344, were both of interest as ligands and more importantly, may give access to the elusive aromatic hetero-porphocyanines. Incidentally, compound 343, which may also be described as an annulenone-type structure,7 contains an odd number of conjugated endocyclic double bonds. Therefore, by definition, 4 6 , 1 7 5 , 1 7 6due to the polarization of the carbonyl bond, this macrocycle may potentially be an aromatic system, provided a planar geometry can be attained. 160 342 x= o, s o 346 O 347 Scheme 62. Retrosynthesis of 5-oxoporphocyanine derivatives. Synthetically, however, incorporating these structural modifications into the porpho-cyanine periphery proved to be a substantial challenge. As shown retrosynthetically above, the major problem arises with the portion of the molecule containing the amino groups. Whilst the appropriate precursor 5,5'-dialdehydes 347 (X = O, S) are readily available, 5 6 , 1 7 3 0 the ensuing manipulation of these groups plays a pivotal role in the synthesis. Specifically, would the key condensation reaction between 347 and hydroxylamine proceed, as anticipated, with a high degree of regioselectivity, thereby attacking the formyl carbonyl carbons only? Thus, such a reaction relies heavily on the successful exploitation of the differing reactivity between these two types of carbonyl groups. Although, generally aldehydes are more reactive than ketones towards nucleophiles, selectivity between the two groups is often difficult. However, with compound 347 we were fairly confident that the desired regio-selectivity could be attained by virtue of steric hindrance of the ketone by the two flanking heterocyclic rings. Indeed, such reasoning has been invoked by Cresp and Sargent to explain the lack of reactivity of the triepoxy[17]-annulenone 348 towards hydroxylamine and other nucleophilic reagents.1 7 6 Additionally, one should not 161 preclude mesomeric interactions between the heterocycles and the bridging carbonyl, which will further serve to reduce its reactivity towards nucleophilic reagents. Finally, the overall stability of the diamine 346 needs to be addressed. Would the above suppositions still hold for this system, with two reactive amino group, and prevent oligermerization? o 348 Figure 56. 2,5:8,11:14,17-Triepoxy[17]annulenone (348). The attempted synthesis of this key fragment is outlined in the scheme 63. The initial steps were adapted from well established chemistry with a few minor modific-ations. 1 7 6 Thus, refluxing a solution of 2-furaldehyde and 2,2-dimethyl-1,3-propane-diol in toluene in the presence of catalytic pyridinium tosylate (PPTS) 1 7 7 gave the cyclic acetyl 349b in 87% yield. Metallation of the acetal 349b with xBuL\ at -78 °C over approximately 1.5h and subsequent addition of ethyl N,N-dimethylcarbamate176 whilst allowing the temperature to rise to -35 °C, and then to -10 °C over a ca 50 min period, furnished the difurylketone 350b in 20% yield. Surprisingly, the acetal groups of the latter compound resisted all attempts to effect the acid catalysed hydrolysis. Thus under various conditions with mildly acidic reagents (e.g. PPTS, 10% HCI) and even stronger acids (e.g. 6N and cone. HCI, H 2 S 0 4 , 70% HCI04), either no reaction occurred or extensive decomposition resulted. The lack of reactivity observed here can be interpreted in terms of the steric congestion arising from the presence the furan rings and the bulky alkyl substituents of the dioxolanes. 162 On the other hand, under strongly acidic aqueous conditions the furan rings can be cleaved to furnish the respective 1,4-diones and higher polymers.1 7 8 This approach based on this dioxolane was therefore abandoned. o 353 352 351 Scheme 63. Attempted synthesis of the key diamine fragment 353. Conditions: (i) PPTS, PhMe\reflux; (ii) a. *BuLi, THF, -78 °C; b. Me2NC02Et; c. H 30 +; (iii) HCI, H20/acetone, RT; (iv) NH2OH.HCI, NaOAc, EtOH; (v) H 2\ Pd-C, THF; (vi) H 2\ R-Ni, NH4OH, EtOH; a, R = -CH 2 CH 2 - ; b, R = -CH2C(CH3)2CH2—. Ironically, this acetal was specifically chosen over that of the simpler 1,3-dioxolane 349a because of its greater stability. Furthermore, we chose to avoid synthesizing the latter compound, in view of pervious accounts by Hinz et. al, who reported that acetal 349a was extremely difficult to handle, let alone synthesize, with extensive resinification being commonplace.1 7 9 However, a more recent account indicated that when PPTS is used as the acid catalyst, and toluene as solvent, this previously believed intractable dioxolane can be prepared in excellent yields. 1 8 0 Indeed, subjecting 2-furaldehyde to the same conditions used in the synthesis of 349b 163 (scheme 63) provided the requisite dioxolane 349a in 75% yield, after distillation from the inevitable tars. In an identical reaction sequence to that described above for the synthesis of 350b, 1,3-dioxolane 349a was elaborated to the analogous difurylketone 350a in 47% yield. Subsequent acid catalysed hydrolysis furnished the desired dialdehyde 351, which was carried forward without further purification. Treatment of this dialdehyde with a slight excess of hydroxylamine hydrochloride and sodium acetate in ethanol at ambient temperature furnished the bisoxime 352. The formation of 352 was tentatively elucidated from its spectroscopic data. The 1 H NMR spectrum of the major product from this reaction (i.e. the more polar compound) obtained after chromatography on silica gel contained numerous other proton signals which were not consistent with the given structure. These uncharac-teristic peaks included a singlet at 8 3.5 ppm, a complex multiplet at 7.8 ppm, and a broad band at 11.5 ppm. However, the bands ascribable to those of the furyl p-protons, the imino-like proton, and the oxime OH's respectively all integrate well for the expected product. Additionally, the parent mass ion of bisoxime 352 is detected in the mass spectrum, with no evidence of the trisoxime; i.e. the product resulting from reaction at the bridging carbonyl group as well as the two formyl positions. The absence of the aldehyde proton signals in the 1 H NMR of 352, however, provides a clear indication that the formyl groups are indeed preferentially attacked. Further-more, TLC analyses of this product failed to reveal the presence of any other contaminants. Examination of the crude reaction mixture by 1 H NMR led to the same conclusions, thereby indicating that the source of degradation was not a side-reaction catalysed by the silica gel during purification. Incidentally, similar discrepancies were noted in the 1 H NMR spectra of other formylfurans, when treated with hydroxylamine. 164 Despite these unusual 1 H NMR spectra, the detection of the parent ion in the mass spectrum prompted us to carry this material forward to the next step. Catalytic hydrogenation of the bisoxime 353 over 10% Pd-C or Raney Nickel at atmospheric pressure failed to provide the desired bisaminomethyl compound 353. Preliminary examinations of these reactions by 1 H NMR indicate that degradation of these systems result. Evidently, it is worth noting that whilst addition of hydrogen to the furan nucleus is known to occur, providing a range of cyclic and acyclic products,181 under the conditions used here, such reactivity is highly unlikely. Analogously, the bridging keto group is not expected to react under these mildly reducing conditions.182 Finally, some chemistry pertaining to the difurylketone 350b, is worthy of mention here. Originally, this work was carried out in order to access the 5,5'-bisformyl-difurylmethane, a key precursor required in our earlier syntheses of the biscyano adduct 322, in sufficient quantities. The precedent for this approach was based on the greater degree of control over the regioselectivity in the coupling reaction, since metallation occurs exclusively at the free a-position of the furan ring; henceforth eliminating the resinification processes encountered in the alternative acid catalysed condensations with formaldehyde. This goal was, however, not fully achieved as, the yields obtained here at best reached 47% (for 350a), whilst that for 350b (37%) was identical to the acid catalysed procedure (cf. scheme 54). The low yields realized in these coupling reactions more than likely result from the temperature dependency of the metallation equilibrium, which incidentally has been shown to shift towards the a-lithio furan species at temperatures below -60 °C. 1 8 0 Nevertheless, this work did provide an insight into the reactivity of this group. For instance, reduction with LiAIH4 in ether at ambient temperature, or with NaBH 4 and 165 KOH in refluxing ethanol, cleanly provided the "meso" hydroxy derivative 354 as shown in scheme 64. On the other hand, with LiAIH4 in the presence of AICI3 at temperatures below -15 °C, or with borane-THF complex at room temperature, 350b furnished the difurylmethane system 355 instead. However, it was the hydroxy derivative 354 which attracted our attention, in light of our inability to oxidize macro-cycles 334-337 and 340, since we now had a system with a fairly labile substituent attached to the bridging group. Unfortunately, the utility of the latter species was severely hampered by the unusually high stability of the cyclic dioxolanes which to this date, we have failed to cleave. In addition, with 354 this was further compounded by the reactivity of the hydroxyl moiety which now imparts a greater propensity for fragmentation, particularly under acidic conditions. In fact, even in the presence of the mildly acidic PPTS catalyst1 8 3 fragmentation of 354 occurred. Thus, strategies based on this intermediate (354) were abandoned. Scheme 64. Reduction of the bridging keto group of 350b. 166 16.5 Some Alternative Approaches to Heteroporphocyanines The various strategies towards the heteroporphocyanines presented in the previous account have so far all relied on a "2 + 2" type condensation employing highly symmetric dipyrrolic, furyl, and thienyl precursors, with the appropriate functional groups built into the molecules. Although, this route has been successful in providing the reduced-form macrocycles, there is, however, an alternative, shorter pathway towards these systems as summarized retrosynthetically in scheme 65. In addition to providing an opportunity for the isolation of two other novel porpho-cyanine isomers (359 and 360), this strategy presents a means for introducing meso substituents into the macrocyclic skeleton (scheme 66), via a well established route used in the syntheses of meso aryl porphyrins. 1 5 7 1 3 283 X= 0 356 357 284 X= S Scheme 65. Schematic representation of an alternative "2 + 2" coupling route to porphocyanines. The pivotal step in this sequence, involves an acid catalysed condensation of the imine 357 with an appropriate aldehyde. Although, the imines are expected to hydrolyse to their constituent aldehyde and amine precursors in the presence of an acid catalyst, we believed that if the latter condensation was carried out in the absence of water, the hydrolysis may be averted. Synthesis of the key imino 167 building block (357) turned out to be a remarkably facile procedure, requiring simple, readily available heterocyclic precursors. The only limitation of this method is that it is only effective for water soluble amine and aldehyde derivatives. Thus, imine 357 was obtained by briefly heating a mixture of furfurylamine and 2-formylpyrrole in water, until a homogeneous mixture results. Upon cooling to ambient temperature the product precipitated out of the solution. The simplicity of this reaction, however, was marred by the disappointing results in the subsequent transformations. 359 R = H 283 R = H 360 R = Ph 361 R = Ph Scheme 66. Synthesis of the precursor imine 358, and the attempted coupling of this intermediate. Conditions: (i) (EtO)3CH, CCI 3C0 2H, CHCI3; (ii) PhCHO, CH2CI2, TFA; [O] = 0 2 , DDQ, etc. Unfortunately, to date, all attempts to self-condense this intermediate, under rigor-ously anhydrous conditions, with triethylorthoformate and catalytic trichloroacetic acid in chloroform, initially at room temperature then at reflux, failed to yield any macrocyclic products, let alone the desired aromatic congeners 359-361. Similar results were obtained with benzaldehyde and TFA in dichloromethane, and there-fore this route was abandoned. 168 o o ZnCI2, CH2CI2 (67%) (CHO)n, HCI(g) •R + 330 O O 328 362 R = CI 363 R = OAc 364 R = OEt NaOAc, AcOH reflux (78%) O 310, p-TsOH, THF, reflux, 24h o o 365 Scheme 67. Synthesis of 2-chloromethyl and 2-acetoxymethyl furans 362 and 364. The scheme above summarizes our attempts at another route towards obtaining the elusive bisaminomethyldipyrromethanes, albeit, in this instance via the mixed phthalimide derivative 365. Thus, furan 328 was initially chloromethylated184 in dichloromethane at ambient temperature, then treated with sodium acetate in refluxing acetic acid to furnish the 2-acetoxymethyl derivative 363. Interestingly, the chloromethyl compound 362 demonstrated extremely high reactivity towards nucleophilic substitution. For example, during an initial attempt to chromato-graphically separate (with ethyl acetate:hexane, 3:7 as eluent) this material from the dimer 330, the ethoxy derivative 364 was instead isolated, as the only monomeric product of this reaction; compounds 328 and 362 have identical R f values on silica gel with dichloromethane eluent. The formation of this product, which at first appeared puzzling, apparently arose from a substitution reaction between the ethyl acetate solvent and the chloromethyl derivative 362. Subsequent experiments in 169 which neat dichloromethane was used as eluent in the purification step yielded the expected chloromethyl derivative. If, on the other hand, the chloromethylation is effected with paraformaldehyde in refluxing carbon tetrachloride, whilst bubbling a steady stream of HCI through the solution, markedly different results occur. In this instance, the difurylmethane 330, was isolated in 14% yield, together with unreacted starting material, and, a decomposition product as the major component. Initially, we anticipated that this product was that of the chloromethyl derivative 362. However, the 1 H and 1 3 C NMR spectra of this compound did not contain any peaks which could be assigned to those of the furan (3-hydrogens. In fact, this data and the mass spectrum definitively suggested that the furan ring was indeed no longer intact. This approach was in fact investigated as a improved method for synthesizing the difurylmethane system 330, on account of the high degree of regioselectivity achieved with the chloro-methylation step. Interestingly, under these conditions 2-furoic esters are reported to predominantly yield the expected 5-chloromethylfurans and the corresponding difurylmethane.184 With the acetoxymethylfuran 363 in hand, we now sought to couple this compound with the corresponding phthalimidomethylpyrrole 310. Unfortunately, despite several attempts, under varying conditions in non-nucleophilic solvents, we were unsuccessful in achieving this transformation. The unreacted starting compounds were simply recovered in each of these experiments. These results coupled with our previous attempts to obtain the bisphthalimido dipyrromethane 311 (scheme 52) leads us to believe that the inclusion of a phthalimidomethyl substituent onto an ex-position of the pyrrole ring, significantly deactivates the other a-position towards electrophilic attack. Further evidence for this deduction, may be inferred from 170 Sessler's work with heterosapphyrins, wherein they successfully coupled an a-free pyrrole-2-ester with 2,5-bisacetoxymethylfuran, under conditions similar to that described in the scheme above. 1 0 1 Perhaps, effecting this coupling with the chloromethylfuran 362 and pyrrole 310 in the presence of SnCI 4, may yield this desired product (365). Incidentally, this bisphthalimidomethyl intermediate raises an interesting question. Specifically, upon deprotection of the amines, on account of the inherent lability of the aminomethylpyrrole subunit, would the resultant diamine spontaneously self-condense in a "2 + 2" fashion to yield the corresponding bisimine macrocycles, in a sequence analogous to that of scheme 50? These suppositions, however, can only be answered by future work in this area. 17.0 Porphycene Analogues of Porphocyanine N 18TC N N 20TC N N 22TC N 366 367 285 Figure 57. Generalized structures of the three possible electronic states accessible by the porphycene analogue of porphocyanines. In order to extend these studies, we attempted to synthesize the porphycene congeners 285-287 and their expanded variants 288-291 (fig. 52). The tetrapyrrolic compound 285 was of particular interest in that it would represent the first example of a "vertically" expanded porphycene; to date, only the "laterally" expanded 171 variants (e.g. 77, 5 1 78, and 91 ) have been synthesized. More significantly, this product may provide an insight into the extent to which the strong N H - H bonding contribute to the remarkable stability of these expanded variants, and the porphy-cenes in general. 9 , 1 0 Finally, assuming that this macrocycle were to form, the further question arises as to which oxidation state(s) such a system would posses. In other words, would an aromatic structure such as 366 or 285 prevail (fig. 57)? Or, would the nonaromatic isomer 367 result? 370 (9%) Scheme 68. Ullmann coupling of iodopyrrole 300. The initial Ullmann coupling reaction of iodopyrrole 300 (scheme 68), required for the synthesis of the bipyrrolic intermediates, yielded some interesting results. Here, in addition to the expected by-product 368, i.e. the deiodinated pyrrole, and the desired bipyrrole 370, albeit in a mere 9% yield, the N,2'-bipyrrole 369 was also isolated as a minor component. This astonishing result prompted further investiga-172 tions into the reaction conditions employed in this coupling sequence, since, earlier literature accounts 1 0 1 , 1 4 6 of exemplar Ullmann reactions only identified the a-free pyrroles and the 2,2'-bipyrroles as the principal products. 1 8 5 , 1 8 6 However, in these instances, the desired bipyrrolic derivatives were isolated in yields approaching 80%, depending on the electron withdrawing nature of the a- and p-substi-tuents. 1 0 1 , 1 4 6 Interestingly, attempts to derivatize bipyrrole 369, by refluxing in acetic anhydride in the presence of catalytic DMAP, failed to introduce an acetyl group at either the free pyrrolic nitrogen or at the other free pyrrole 2-postion. The unreacted starting molecule was simply recovered. The reason for the failure of this bipyrrole to react with the acetic anhydride is probably accountable in terms of steric hindrance by the P-ethyl side chain and the ester group of the second pyrrole. Nevertheless, subsequent work in this area revealed that the product ratio was highly sensitive to the quality of the copper powder used in the reaction mixture. Thus, when activated copper bronze 1 8 7 was employed a three-fold increase in the reaction yields (for the desired bipyrrole 370) resulted (i.e. 24-26%), with pyrrole 268 still being the major by-product. However, the increased yields of the 2,2'-bipyrrole greatly simplified the purification step. Recrystallization from petroleum ether was, in fact, sufficient to provide pure 370. The mediocre yields still realized here are an inherent problem with the Ullmann coupling, and can be remedied by increasing the number of electron withdrawing substituents attached to the precursor 5-iodopyrroles. 1 4 6 , 1 8 6 , 1 8 8 Despite this, the bipyrrole 370 was obtained in sufficient quantities for further elaboration. 173 Scheme 69. Attempted preparation of the 18K macrocycle 374. The 5,5'-dicyano-2,2'-bipyrrole 372 was then prepared in a sequence (scheme 69) involving an initial saponification and decarboxylation with NaOH in refluxing ethylene glycol, to furnish the di-cc-free bipyrrole 371, followed by biscyanation with CSI in DMF/CH 3 CN under the conditions described previously.152 Hydride reduction of this material and subsequent oxidation (with air, DDQ, or chloranil) of the presumed initially produced cyclic expanded "porphyrinogen-like" intermediate 373, 174 possible oxidation states- fig. 57). Upon addition of a methylene chloride solution of DDQ or chloranil to the reaction mixture, after destruction of the excess LiAIH4, a deep bottle green solution rapidly developed. However, the UV-visible spectrum of this compound did not contain typical porphyrin-type absorption bands, particularly the strong Soret bands in the visible region of the electromagnetic spectrum, characteristic of all porphyrin-type macrocycles. Examination of the 1 H NMR spectra of the crude reaction samples of the initial reduction product have so far failed to provide sufficient evidence for either formation of the 5,5'-bisaminomethyl bipyrrole derivative or the nonaromatic macrocycle 373. Similar investigations following the subsequent dehydrogenation attempts have also, unfortunately, revealed very little useful structural information. In view of this, the question still remains as to whether the macrocyclization does actually occur during the reductive process of this dicyano intermediate. Or, is the dicyano bipyrrole 370 reduced to its bisaminomethyl derivative, and if so, does the latter product display the same reactivity as its dipyrromethane counterparts (cf. compound 306, scheme 50), thus undergoing an analogous disproportionation and cyclization sequence yielding the "porphyrinogen-like" macrocycle 373? Or, on the other hand, does the hydrogen bonding play an integral role in the stabilization of the expanded porphycenes, such that in their absence even the expected resonance energy gains from aromatization are insufficient to arrest the molecule in a stable planar conformation? The lack of physical data here therefore prompted a theoretical evaluation of these macrocyclic systems. Molecular mechanical calculations predict that the initial imine macrocycle 373 adopts a non-planar pyrrolophanediene-type10f structure (fig. 58), in which the two bipyrrolic subunits are almost parallel to each other. This predicted 175 conformation of 373, incidentally, bears a striking resemblance to that observed in the crystal structure of the corresponding N,N'-dihdyroporphycene intermediates.101 However, in the latter examples, the unusual geometry adopted by these intermedi-ates has little effect on its oxidation to the fully aromatic species. In fact, the N,N'-dihydroporphycenes were only isolable when bulky substituents were introduced onto the vinyl bridges between the bipyrrole subunits.10' In the absence of these groups, this intermediate spontaneously dehydrogenates to the aromatic porphycene.10 Figure 58. Predicted conformation of bisimine macrocycle 373; left: side view, right: top view. The aromatic compound 374, on the other hand, is predicted to assume a dome-shape in its preferred conformation, as shown in the figure below. This distortion of the macrocyclic framework observed here can largely be accounted for in terms of non-bonding interactions between the ethyl side chains at the 3 and 6, and, 14 and 17 positions. Consequently, these steric factors may now destabilize the resultant macrocycle. On the other hand, X-ray crystallography has revealed that in the solid state the porphycenes (with comparable p substituents) do indeed adopt a planar conformation.1 0 a'b , 9 , i These observations thus suggests that steric factors may not be 176 the sole cause of the puckering of the ring skeleton in the present compound but, both a combination of the above factors and unfavourable lone pair interactions. Interestingly, however, the calculated N-N' distances within the bipyrrolic portions of the macrocycle 374, are in good agreement with those determined by X-ray crystallography for the parent porphycene systems. Thus, it is fairly plausible that the additional strain imposed on 373, upon aromatization, may be a key factor for its inability to dehydrogenate to the aromatic congener 374. However, one cannot rule out the possibility that this anticipated reductive-macrocyclization sequence may not be applicable for the smaller bipyrrolic dicyano derivatives as is the case with the dipyrromethanes, on account of the greater rigidity of the former systems. Figure 5 9 . Computer generated structure of a bipyrrolic derivative ( 3 7 4 ) of porphocyanine, in its most stable conformation. Thus, in order to evaluate the above suppositions, we focused our efforts on the tetraoxa derivative 286, since we could access the precursor cyclic imine 375 via a more conventional approach; i.e. through a condensation reaction between the appropriate acyclic diformyl and diamine bifuryl intermediates. Moreover, molecular mechanical calculations predict that unlike the tetrapyrrolic analogue 374, macro-cycle 286 is expected to adopt a more planar conformation. In addition, X-ray crystallographic studies of the tetraoxaporphycene dication and its precursor oxa-177 bridged [20]annulene10e revealed that, unlike the tetrapyrrolic N,N'-dihdyropor-phycene,1 0 1 this nonaromatic intermediate adopts a planar geometry. We thus envisaged that oxidation of the intermediate bisimine 375 may now be a more readily achievable goal on the basis of the above observations. Furthermore, by virtue of our experience with the dioxaporphocyanine derivatives coupled with similar observations by Vogel's group with tetraoxaporphyrinogens,11 we strongly believed that this key cyclic intermediate (375) would also display reasonably good chemical stability. 286 375 Figure 60. Schematic representation of the porphocyanine derivative (286), and its bisimine counter-part (375), of Vogel's tetraoxaporphycene. The bifuran 185, thus served as the pivotal acyclic intermediate, since, using basic carbonyl chemistry, we envisaged smooth transformation of this species to the second component (i.e. bisaminomethyl 379) required in this cyclization. During our initial exploratory work towards bipyrrole 370, we became aware of an alternative coupling reaction for the preparation of hetero-biaryl systems. 1 8 9 This single step procedure involved refluxing 2-furaldehyde with palladium (II) acetate in acetonitrile (scheme 70). The simplicity of this reaction was unfortunately overshadowed by the disappointingly low yields of the requisite diformyl bifuran 185, the inconsistent results upon scale-up, and, lastly, the extensive chromatographic separative proce-dures required, after which the product was still found to contain undesirable side 178 products. What was even more puzzling, despite several attempts, under the exact conditions described in Takajo's paper, we were basically unable to mirror their results. Preliminary 1 H NMR data of these reactions, and TLC examinations, served to indicate that >90% of these experiments simply contained unchanged 2-furalde-hyde, in addition to the dimer 185 and decomposition by-products. Extending the reflux times, or conducting the reactions under anhydrous, inert conditions had little effect on the yields of 185 and product ratios. Incidentally, the recovery of a substantial amount of the starting material was anticipated from the onset, as was reported in the literature,189 however, it was the one-step synthesis and the use of commercially available reagents that attracted us to this route. 376 377 Scheme 70. Synthetic routes to 5,5'-diformyl-2,2'-bifuran (185). . The difficulties encountered in the above palladium catalysed couplings prompted us to approach the target bifuran 185 via the more traditional Ullmann coupling route. As outlined in scheme 70 above, 2-furaldehyde was functionalized at the 5-position by treating with Br 2 in refluxing 1,2-dichloroethane, according to the procedure of Chadwick era/ . 1 9 1 Slightly improved yields were obtained when 0.01% (w/w) each of sulphur and hydroquinone were added to the reaction 1 9 2 but, a large percentage of the products still comprised intractable black tars. In spite of this, 179 however, the 5-bromo-2-furaldehyde 376 was readily isolable from this material by steam distillation. Reaction of 376 with Kl furnished the 5-iodofuran 377 1 9 3 in good yield, which was subsequently dimerized to the requisite diformyl bifuran 185 in the presence of activated copper bronze. 379 Scheme 71. Attempted preparation of diamine 379. Conditions: (i) NH2OH.HCI, NaOAc, MeOH; (ii) R-Ni/H2, NH4OH, RT, 1 atm. With the above-described intermediate in hand, the next stage in the synthesis centred around the second component of the target tetraoxa bisimine macrocycle, i.e. the bisaminomethyl bifuran 379. Transformation of 379 to the corresponding bisoxime 378 was accomplished by reaction with hydroxylamine, which was carried forward without any further purification. Catalytic reduction of this compound over Raney nickel at atmospheric pressure led to the key 5,5'-bisaminomethyl-2,2'-bifuran 379, in extremely low yield (< 15%). Evidence for the formation of this product has tentatively been deduced from the 1 H NMR of the crude reaction mixture. However, this NMR data also indicates the presence of several other decomposition products in substantial quantities. In view of the small quantities obtained in these experiments, we were neither able to purify this intermediate nor derivatize this expected diamino compound (as its acetate, for example). 180 Although, as mentioned earlier, under the current conditions this reductive step is not generally expected to lead to complete hydrogenolysis and subsequent fission of the heterocyclic ring systems, 1 6 9 ' 1 8 1 , 1 9 4 we strongly believe that the diminished yields observed here result from the high activity of the nickel catalyst. For example, in a model study with 2-furaldehyde, employing the same conditions described above, we were unable to isolate furfurylamine. Perhaps, therefore, either with a lower grade Raney nickel catalyst or by partial poisoning of the present system with elemental sulphur we may realize this goal more efficiently. Or, even a palladium based catalyst may give more promising results. Unfortunately, due to time constraints we were unable to pursue this matter further. The alternative route to 379 via hydride reduction of 5,5'-dicyano bifuran, fared no better in that the initial dehydration step of bisoxime 378 proceeded with decomposition of >60% of the starting material. Another notable discrepancy observed in the chemistry of this intermediate (379, and also with the difurylketone 351) is the unusually complex 1 H NMR data of these products. Whilst, isomeric mixtures pertaining to the orientation of the hydroxyl group were expected, the presence of several peaks in the high field end of the spectrum, particularly in the range between 1.0 and 5.5 ppm, and around the vinylic and aromatic resonance frequencies were clearly inconsistent with the given structures, and appear to indicate some form of decomposition may be taking place as well. Generally, however, this reaction is known to proceed quantitatively and cleanly, with formylfurans.1 9 4 , 1 9 5 This anomaly is indeed surprising, in view of (i) the mild conditions employed in this nucleophilic reaction; and (ii) furans do not generally react with nucleophiles at the heterocyclic ring by addition or by substitution. 181 Perhaps further work in this area may reveal the source of the above atypical results. During the course of this work, we also directed some attention towards even larger macrocyclic systems representing the next logical progression in this series. These systems basically expand on the above-described ligands by the introduction of a third heterocyclic ring between the two bi-heterocyclic subunits of compounds 285-287 giving 288 and 289, or introducing an additional heterocyclic ring at one half and a methylene bridge at the other as depicted by the generalized structures 290 and 291 (fig. 52). At present, the chemistry in this area is still in the preliminary stages. Nevertheless, a brief account of the chemistry leading towards these systems will be presented here. Scheme 72. Synthesis of dipyrrolylfuran 382. Conditions: (i) Divinyl sulphone, [3,4-dimethyl-5-(2-hydroxyethyl)thiazolium iodide], Et3N, p-dioxane, 84 °C; (ii) H 2S0 4 , EtOH/reflux; (iii) H2- Pd/C, THF, Et3N, 1 atm.; (v) DMF, N2, reflux. 182 The intermediate dipyrrylbutanedione 381 was prepared, as shown above, in accordance with methodology reported by Ibers.64 Interestingly, the choice of thiazolium catalyst employed in the initial coupling of the formyl pyrrole 380 with divinyl sulphone proved crucial. For instance, when 3-benzyl-5-(2-hydroxyethyl)-4-methylthiazolium chloride was used, only unchanged starting material was recovered, whereas 3,4-dimethyl-5-(2-hydroxyethyl)thiazolium iodide provided the desired dione in 48% yield. This clearly is a consequence of the bulky benzyl group preventing the initial formation of the critical p-hydroxy enamine intermediate 384, between pyrrole 380 and the former thiazolium salt. The dione 381 was then cyclized with catalytic sulphuric acid in refluxing ethanol. Subsequent reduction in the reaction mixture volume, and cooling to 0 °C cleanly provided the tetraester dipyrrolylfuran 382 as a yellow solid, that exhibits blue fluorescence. This compound is fairly photosensitive when in solution. For instance, during the acquisition of the 1 H NMR spectrum, the resultant spectrum displayed signals which were more consistent with a 5,5"-dibenzylester structure, wherein the pyrrolic p-ethyl ester side chains had undergone a saponification and concomitant decarboxylation sequence. This photo-catalytic side-reaction more than likely proceeds through a free radical-type mechanism. It should be noted, however, that the photosensitivity of a similar tetraester dipyrrolylfuran, has been reported by the Ibers' group at Northwestern University,65 but, no indication as to the nature of the degradation products have been presented. Unfortunately, due to the photolability of these compounds we have yet to conclusively confirm its absolute structural configuration, as continued exposure to light undoubtedly led to even further decomposition. This photo-instability, may additionally have contributed to our failure to isolate the desired di-a-free compound 383, through a sequential hydrogenolysis of the benzyl groups and then thermal decarboxylation in boiling DMF. 183 384 Figure 61. The p-hydroxy enamine intermediate formed from between the thiazolium salt catalyst and pyrrole 380. Incidentally, this pyrrole containing two different esters was chosen at the onset of this work in order to determine the solubility of the final dipyrrolylfuran 383. It should further be noted that ester groups at the 3 and 5-positions of the formylpyrrole are a prerequisite for these coupling reactions. 6 6 , 1 9 6 These two electron withdrawing substituents principally serve to stabilize the key p-hydroxy enamine intermediate (fig. 61), initially formed between the aldehyde and the thiazolium salt catalyst, which would other-wise be destabilized by the electron-rich 7t-system. Recent developments in pyrrole chemistry have, however, resulted in an alternative procedure towards dipyrrolylfurans and terpyrroles in general, 1 2 3 which circumvents this restrictive p-substituent pattern dictated by the Stetter methodology67 used in our synthesis. Thus, future work in this area may now prove more fruitful. In a parallel sequence, we have also prepared the terfuran 386, 1 9 7 in two steps via the Stetter procedure,6 6 6 as reviewed in the scheme below. In this instance, with the furan being a less electron-rich system, the above restrictions do not apply and thus the synthesis provides the intermediate without any a substituents. Preliminary attempts at biscyanating this intermediate with CSI under our standard low 184 temperature conditions have given some promising results. However, due to time constraints further detailed investigations were not possible. Nevertheless, subject to obtaining the corresponding diformyl terfuran, we can anticipate elaboration of these intermediates to the corresponding bisimine macrocycle. X C T X H O (53%) L _ 0 X O O \QJI (22%) ^ _ / ° \Q_J 385 386 Scheme 73. Synthesis of terfuran 386. Conditions: (i) Divinyl sulphone, [3-benzyl-5-(2-hydroxyethyl)-4-methylthiazolium chloride], Et3N, NaOAc, EtOH/reflux, 22h; (ii) HCI, Ac 20, RT. 18.0 Texaphyrin Analogues In the past decade, the texaphyrins have emerged as one of the most intriguing and promising expanded porphyrins in terms of their potential biomedical applications and as ligands for the trivalent lanthanide cations.8 The structural similarities of these compounds to the porphocyanines, and the parallel difficulties encountered in oxidizing the heteroporphocyanine variants to their aromatic congeners prompted us to attempt synthesizing the analogous tripyrrane macrocycles 294 and 295, in which the o-phenylene diamine subunit of texaphyrin is replaced by a 2,5-bisaminomethyl furan or thiophene. Synthetically, we anticipated that we could achieve this goal via a typical acid catalysed coupling between a diformyl tripyrrane (e.g. 387) and the bisaminomethyl furan 388 or thiophene 389 respectively, as depicted retrosynthetically in scheme 74. Thus, we did not envisage any major drawbacks associated with this macro-cyclization step and in the synthesis of the tripyrrane portion. Therefore, we set 185 about devising a shorter, more efficient, route towards these key diamino hetero-cycles. Scheme 74. General synthetic strategy towards tripyrrane bisimine macrocycles 294-295. The most logical approach to the target bisaminomethyl furan 388 was through the known 2,5-diformylfuran 391. 1 8 0 This intermediate was in turn synthesized from the acetal 349a in two steps, by reaction of the 5-lithium salt of 349a with an excess of DMF at -78 °C followed by an acid catalysed hydrolysis. However, treatment of diformylfuran 391 with hydroxylamine, and subsequent hydrogenation of the resultant bisoxime, which was carried forward without further purification, at atmospheric pressure over Raney-Ni failed to provide the desired bisaminomethyl-furan 388. At this stage, we were unsure as to whether our failure to isolate this key intermediate diamine was due to its instability, or due to over-reduction of the furan nucleus itself. These observations, together with our previous results with the 5,5'-diformyl-2,2'-bifuran 185, collectively indicate that due to the activity of the Raney-Ni catalyst over-reduction of the furan nucleus more than likely occurs here. However, since the resultant reaction mixtures comprised of gross mixtures, no attempts were 294 x= o 295 X= S 387 186 made to isolate the various by-products. Furthermore, it was fairly apparent that extensive decomposition of the starting materials had resulted. The alternative procedure, which involved dehydration of the bisoxime to the corresponding dicyanofuran, and subsequent, reduction turned out to be even worse, for when we attempted the initial dehydration step in refluxing acetic anhydride, only trace amounts of the required 2,5-dicyanofuran could be isolated. We thus investigated alternative procedures. 388 391 Scheme 75. Attempted preparation of 2,5-bisaminomethylfuran (388). Perhaps, one of the critical issues that needed to be addressed in regards to the above synthesis, was the overall stability of the desired diamines. Since we had no information on these compounds, we henceforth sought to introduce the amino functionality in a more stable, latent form. Here, our pervious successes with the bisphthalimidomethyl difurylmethane 330 led us to adopt a similar approach. Indeed, this proved to be the most effective strategy. Thus, treating the chloro-methyl derivative 362 with potassium phthalimide in anhydrous benzene at reflux, cleanly furnished the 2,5-bisphthalimidomethylfuran 370, albeit in a mere 9% crude yield. 187 o o o 362 392 X= O 395 X= S v (35%) (27%) II OHC CHO (85%) in R. 93 393 R = OH 394 R = CI Scheme 76. Preparation of the 2,5-bisphthalimidomethyl furan (392) and thiophene (395). Conditions: (i) Potassium phthalimide, C6H6\reflux. (ii) a. nBuLi/TMEDA/hexane/reflux, 30 min.; b.THF, DMF, -40 ° C -> RT; c. HCI/ice-H20; (iii) NaBH4, THF/Et20; (iv) SOCI2, CH2CI2; (v) Potassium phthalimide, DMF, 90 °C. The analogous thiophene compound 395 was obtained in a similar fashion as shown in scheme 76. Following the method of Garrigues, 1 9 9 2,5-diformylthiophene 93 1 8 0 was first reduced with sodium borohydride to give the bishydroxymethyl-thiophene 393 in 85% yield. This crude material was subsequently converted to the bischloromethyl compound 394 in excellent yield by treatment with an excess of SOCI 2 in anhydrous dichloromethane. Heating the latter intermediate with potassium phthalimide in dry DMF then provided the N-protected diamine 395, as a tan coloured solid. Once these key precursors were in hand, we began to synthesize the tripyrrane subunit, which was obtained in four steps from simple monomeric pyrroles. Thermal saponification and decarboxylation of pyrrole 396 1 4 9 gave the di-oc-free pyrrole 118 1 4 4 (scheme 77). Condensation of this pyrrole with the acetoxymethyl pyrrole 15 188 under acidic conditions furnished the tripyrrane 397 in 72% yield. Hydrogenolysis of the benzyl esters provided the intermediate diacid 398 in 88% yield. Exposing this moderately unstable diacid to the Clezy formylation sequence, 2 4 a procedure involving the sequential treatment of 398 with trifluoroacetic acid to effect decarboxylation, followed by treatment with triethylorthoformate at 0 °C, and finally a basic hydrolysis, produced the requisite diformyl adduct 399 in a meagre 20% yield. The low yield realized here, were largely due to the formation of an intractable green gum (with a metallic hue) prior to the hydrolysis step. 118 15 397 R = B n ^ H 2 , Pd/C 398 R = H «*^THF, Et3N A Etylene gycol Scheme 77. Synthesis of diformyltripyrrane 399. The spectroscopic data of this product, however, provided some rather surprising results. For instance, the signals of the ethyl side chains display splitting patterns which were not consistent with the expected symmetric structure of 399, as depicted in the scheme above. Furthermore, three distinct pyrrolic NH signals were observed. This initially led us to believe that perhaps two limiting conformations of the this 189 tripyrrane were possible, in which, the two terminal pyrroles are directed in opposite directions. On the other hand, there is no apparent evidence for this restriction of rotation of the terminal pyrroles around the bridging methylene groups. Moreover, there is only one signal ascribable to the two pairs of bridging methylenes at 5 3.99 ppm. However, the mass spectral data does not contain any evidence for the molecular ion peaks, nor for any fragmentation products thereof. Consequently, this has halted our efforts in coupling this product with the diamines 388-389 derived from their corresponding phthalimido derivatives. Nevertheless, after several attempts, we have very recently isolated the desired diformyl tripyrrane 399 in 34% using slightly modified conditions to those used in our initial attempts. Here, effecting the decarboxylation in TFA at room temperature followed by formylation with triethyl orthoformate at -20 °C, gave 399 as a dark brown solid after recrystallization form EtOH , which gave the correct spectroscopic data. Thus, with these precursors in hand, one can anticipate the synthesis of these tripyrrane macrocycles in the near future. 190 Chapter 3 Conclusion and Recommendations 191 19.0 Conclusions and Recommendations The major aim of this project was to synthesize new expanded porphyrins based on the novel porphocyanine macrocycle, which has recently emerged from our laboratories. Furthermore, we were also interested in determining the mechanistic aspects of the rather intriguing macrocyclization step which formed the central part of our synthesis of these systems. Since, we believed that the bisaminomethyl dipyrrolic intermediate 306 was the key precursor species in this cyclization we therefore set out to synthesize this intermediate via an alternative pathway. The most logical approach here was to prepare this molecule from a closely related precursor. Thus, attempts at obtaining the phthalimide derivative 311, and variants thereof, were carried out, though only to meet with very little success. Whilst this intermediate was obtained in one instance as a very minor component, several attempts to reproduce this result failed. This in turn has now led us to believe that the failure in accessing this dipyrromethane and the "mixed" furan-pyrrole compound 365 is likely due to the fact that the phthalimidomethyl group is not sufficiently electron withdrawing to stabilize these products or the mono pyrrolic (and/or furyl) intermediates of the coupling reaction. Despite this, our work in this area of the porphocyanines resulted in the discovery of an efficient method for introducing the requisite dicyano groups at the free-a positions of dipyrromethanes.152 This strategy clearly surpassed the original route employing the 5,5'-diformyl compounds, which encompassed transforming these groups to their corresponding oximes, and finally effecting a rather inefficient dehy-dration step on these intermediates.88 ,89 Synthetically, this cyanation route with the low temperature CSI/DMF/CH 3CN combination has indeed turned out to be extremely versatile, enabling the introduction of cyano groups at the free-a positions 192 of both bipyrrolic and difurylmethane intermediates. The successes realized here has henceforth prompted extension of this work towards larger macrocycles based on ter-pyrrolic and furyl systems 383 and 386, respectively. Presently, this work has reached the point where this biscyanation procedure needs to be effected on the terfuran 386. Thus, one can anticipate the synthesis of the corresponding aromatic macrocycle within the near future. Work towards the dipyrrolylfuran based system 383, was unfortunately prematurely terminated due to an unexpected side-reaction which eventually led to decomposition of this linear intermediate. However, recent advancements in this area of heterocyclic chemistry1 2 3 has rekindled our interest in this system and the expanded porphyrin derived herewith. To date, with the exception of the recently reported amethyrins,123 compounds of this type have generally found very little use as ligands probably due to their rectangular shape, and were thus largely of photophysical and scientific importance. With these porphocyanine analogues, on the other hand, it is highly plausible that the presence of the additional bridging nitrogens may endow these expanded porphyrins (e.g. compounds 288 and 289) with interesting metal binding properties. From our work with the corresponding 5,5'-biscyanodifurylmethane 322, it was fairly apparent that this reductive-macrocyclization sequence was unique to pyrrole-based acyclic precursors. In view of this supposition we duly attempted a similar strategy with an appropriate dicyanobipyrrole, e.g. 372. However, unlike the previous example this compound failed to yield the desired aromatic macrocycle. We have thus concluded that a combination of steric and electronic factors contribute to the ruffling of the molecular framework observed in molecular mechanical calculations, and thus give rise to perturbation of this expanded porphyrin. Therefore, in order to 193 alleviate these steric factors, future attempts in this area should be based on 2,7,13,18-tetraalkyl derivatives 400 (fig. 62), where, due to the distances between the substituents, the planarity of these compounds should now essentially remain unchanged. Accordingly, further developments in the chemistry pertaining to these expanded ligands depends on the availability of macrocycles possessing solubilizing alkyl or aryl substituents. Therefore, choice in the size of these substituents will largely determine the solubility properties of the resultant macrocycles. However, whilst being a critical requirement, this can only be decided by future experiments. Figure 62. Generalized structure of a tetra-p-substituted porphocyanine analogue of porphycene. Another compound of interest, which is related to the above discussion, is the 2,5-dicyanopyrrole 401. As mentioned earlier, the key question surrounding this intermediate remains to be answered. Thus, would hydride reduction of this species under the usual conditions furnish the tetraimino macrocycle 403 (scheme 78)? If so, would this synthesis proceed in an identical manner to the biomimetic sequence employed by Franck's group in their syntheses of the expanded [26]porphyrins 160-163 8 3 , 8 4 ? Furthermore, would one achieve the same level of selectivity for the desired tetrapyrrolic macrocycle as was obtained by Franck in the latter syntheses? Clearly, this will almost certainly be determined by the size of the 3 and 4-alkyl substituents on the precursor dicyanopyrrole 401. N N 400 194 R R 403 Scheme 78. Proposed route to the tetraimino macrocycle 403. We have also shown that incorporating heterocycles other than pyrrole into the porphocyanine framework can be achieved relatively easily, and thereby prepared three novel analogues 334, 335 and 340. These new macrocycles, however, display markedly different chemical properties to their tetrapyrrolic congeners. For instance, they do not readily oxidize to the corresponding aromatic species in either air or with a range of common oxidants. Similarly they have thus far resisted all attempts to undergo metal insertion. These shortcomings are undoubtedly linked to the poor thermal stability of the imine linkages. Therefore, we believe that in order to ensure the long stability, and further exploit the photophysical and chemical properties of these heteroporphocyanine analogues, this oxidation of these bisimine macro-cycles to their conjugated aromatic congeners remains a fundamental prerequisite. Surprisingly, partial oxidation of these systems have been achieved, however, the resultant macrocycle still resisted further oxidation to the fully conjugated isomer. Through this we determined that it was the difuryl and dithienyl (respectively) subunits which were responsible for preventing total oxidation of these bisimine compounds. In this regard, the 5-oxo-derivative is of particular interest since, the 195 keto group can be tautomerized to yield the corresponding "meso"-hydroxy group, which now provides an excellent opportunity for abstracting such a labile moiety. Thus, it should be stressed that considerable future efforts in this area ought to be directed towards, either finding suitable conditions to effect this critical oxidation step; or, isolating a non-labile metallo complex of these reduced form macrocycles. Nonetheless, current efforts are being devoted to isolating X-ray quality crystals of these non-conjugated macrocycles. Similarly, the "mixed" macrocycle 359 would also be an ideal target for future oxidation experiments. Here, perhaps further evaluation of other coupling reactions which we initially attempted (scheme 66), with the imine 358 may provide this novel macrocycle. However, the uncertainty of the stability of the precursor acyclic imine 358 under these acidic conditions has now led us to propose two other routes shown in the scheme below. Here again, though, questions on the stability of the di-heterocyclic compounds 407 and 408 still remain. Despite this, the second approach appears to be somewhat more feasible in light of a reported procedure towards porphobilinogen lactam methyl ester. 1 5 5 Here, we believe that under the correct dilution, in a parallel fashion to the latter synthesis, cleavage of both the methyl ester and the phthalimide groups of 405a will occur when treated with the NH 2OH/NaOMe combination, revealing intermediate 408, which will then proceed to cyclize yielding the bisamide 409. Finally, dehydration and oxidation of the resultant cyclic bisimine will now reveal the aromatic compound 358. This route is additionally appealing in that the initial bisamide, once formed, is expected to be more stable than the corresponding bisimine compounds. 196 197 Finally, our recent work with the chloromethyl furans and thiophenes in providing the respective bisphthalimidomethyl compounds now leads us to propose another alternative route towards the key diamine 379. As reviewed in scheme 80, we believe that this system can be synthesized via the bisphthalimide derivative 411, which in turn may be obtained from the bischloromethyl derivative 410, prepared via the procedure of Jouillie 2 0 0 The advantages of this route no doubt stem from (i) a common intermediate (i.e. the diformyl bifuran 185) which serves to provide both precursors required for the final cyclization, and (ii) the phthalimide derivatives of amines are generally stable solids. The alternative approach to this strategy would be a biscyanation of the known di-a-free bifuran, 1 9 0 , 2 0 1 which is expected to provide the 5,5'-dicyano-2,2'-bifuran, followed by LiAIH4 reduction of the latter species. In any event, only further experimental work will decide which of these would be the most conducive towards the desired macrocycle. r\r\ H2N 1/ \JI X 379 1. LiAIH, OHC "O" "O CHO 2.SOCI2, Py. 185 Et20 NH, NH2OH CI II \w/ \ 410 CI Potassium phtalimide, C 6H 6, A 411 Scheme 80. Proposed derivatization of bifuran 185, in the synthesis of the diamine intermediate 379. 198 20.0 Future Outlook The potential of the bisimine macrocycles 334-337, 340 will undoubtedly only be fully realized once the elusive conditions required for the oxidation step are found. In spite of our failure to achieve this critical reaction, the synthesis of these macro-cycles has opened up a new area for a wide array of related expanded porphyrins. Indeed, with many of the key acyclic precursors already in hand, it is highly conceivable that we will soon be able to present many of these target compounds. To conclude, the most dramatic point to note in the preceding work was the marked difference in chemical properties endowed on the resultant macrocycles, upon substitution of the pyrrole subunits of porphocyanine with furan and thiophene entities. 199 Chapter 4 Experimental 200 2 1 . 0 Experimental 2 1 . 1 General and Instrumental All reactions, unless otherwise stated, were routinely carried out under an atmosphere of nitrogen. Organic extracts were dried over anhydrous MgS04 and evaporated, after filtration, using a Buchi rotary evaporator. Tetrahydrofuran (THF), 1,4-dioxane and diethyl ether were dried over sodium/benzophenone ketyl and freshly distilled immediately prior to use. Anhydrous dichloromethane and toluene were freshly distilled from calcium hydride; acetonitrile, triethylamine, diisopropylamine and DMF were distilled and stored over 3 A sieves. Absolute methanol and ethanol were distilled from magnesium and stored under N 2 over 3 A sieves. Deuterochloroform was passed through a column of basic alumina and stored over 3 A sieves. Furan, thiophene, furfuryl alcohol, furfurylamine, 2-furaldehyde, and thionyl chloride were all distilled immediately prior to use. 2-Furoic acid and Proton Sponge® were recrystallized from CCI 4 and petroleum ether respectively prior to usage. Cu-bronze was activated by suspending it with iodine in acetone for 10 min then in 1:1 (v/v) concentrated HCI in acetone, 1 8 7 filtered off then dried and stored in a vacuum dessicator. PPTS was prepared according to the procedure of Miyashita et al.UJ All other commercial reagents and in-house mono pyrroles were used without further purification. Melting points were determined on a Thomas Model 40 hot stage melting point apparatus and are uncorrected. IR spectra were recorded on a Perkin-Elmer Model 834 FTIR instrument, Mass spectra (low and high resolution) were recorded on a Kratos-AEI MS-50 mass spectrometer. 1 H NMR (200 MHz) and 1 3 C NMR (50MHz) spectra were recorded in the solvents indicated on a Bruker AC200 spectrometer, 2 0 1 1 H NMR 300 MHz and 400 MHz spectra were recorded on Varian XL300 and Bruker WH400 spectrometers respectively, with all chemical shifts are reported in 8 ppm with residual chloroform (8 = 7.24) as internal standard. UV/visible spectra were recorded on Hewlett Packard Model 8452A diode array spectrophotometer. Experiments were monitored using TLC analysis performed on Merck Kieselgel 60 F 2 5 4 silica gel or, aluminum oxide 150 F 2 5 4 neutral (type T) 0.2 mm coated aluminum plates with fluorescent indicator and developed with bromine vapour and vanillin solution where necessary, whilst column chromatography was effected under medium pressure employing silica gel 60 (70-230 or 230-400 mesh), supplied by E. Merck Co. Aromatic macrocycles were purified on neutral alumina, Brockman Activity Grade 1 (80-200 mesh), supplied by Fischer Scientific, which was deactivated with 4% MeOH. 202 Benzyl 5-acetoxymethyl-4-ethyl-3-methylpyrrole-2-carboxylate (15) Benzyl 4-ethyl-3,5-dimethylpyrrole-2-carboxylate (292,139 4.11 g, 16 mmol) was dissolved in glacial acetic acid (15 mL) containing a little acetic anhydride (ca. 0.5 ml_). To the resultant orange solution was added lead tetracetate (7.16 g, 17.1 mmol) and the suspension stirred at room temperature. After 5 min, an exothermic reaction set in producing a deep orange-red solution, 10 min later, a cream precipitate was observed and the mixture was then heated at 100 °C until completion of the reaction. Ethylene glycol (ca 3 mL) was added to destroy excess lead (IV) salts, and the mixture left to cool. The resultant precipitate (addition of water to the reaction mixture at this point enables complete precipitation of the product in an easily filtered form) was filtered off, thoroughly washed with water and dissolved in dichloromethane (50 mL). The organic solution was successively washed with water and saturated brine (15 mL of each), and the solvent removed in vacuo to yield a cream coloured solid (4.71 g, 93%), which was carried forward without further purification. A small sample (250 mg) was recrystallized from dichloromethane-hexane providing fine white needles. m.p.: 122.5-124 °C; Lit.2 3 122 °C 203 1 H NMR (400 MHz, CDCI3): 6 1.08 (3H, t, 4-CH 2CH 3, J = 7.4 Hz); 2.05 (3H, s, CH 3C(0)0-); 2.39 (3H, s, 3-CH3); 2.48 (2H, q, 4-CH 2CH 3 , J = 7.4 Hz); 5.00 (2H, s, 2-CH2OAc); 5.29 (2H, s, -0-CH 2Ph); 7.3-7.43 (5H, m, CH2Ph); 8.98 (1H, bs, NH). Dibenzyl 3.3'-diethyl-4.4'-dimethyl-2.2'-dipyrromethane-5.5'dicarboxylate (295) A solution of the acetoxymethyl pyrrole 15 (4.46 g, 14.1 mmol) in 80% aq. acetic acid (30 mL) was heated in a hot water bath for 25 min. The dark brown solution was allowed to cool, then ice-water was carefully added to precipitate the product. This material was extracted into dichloromethane and the solvent removed under reduced pressure yielding a dark brown oil. Recrystallization from dichloromethane-methanol furnished the title compound as an off-white solid (2.09 g, 59%). m.p.: 129-130 °C; Lit.2 3 126-127 °C 1 H NMR (400 MHz, CDCI3): 5 1.05 (6H, t, 3,3'-CH 2CH 3, J = 7.4 Hz); 2.30 (6H, s, 4,4'-CH3); 2.40 (4H, q, 3,3'-CH 2CH 3, J = 7.4 Hz); 3.8 (2H, s, bridge CH2); 5.25 (4H, s, -O-CH2Ph); 7.28-7.37 (10H, m, CH2Ph); 8.84 (2H, bs, NH). 293 204 3.3'-Diethyl-5.5'-diformyl-4.4'-dimethyl-2.2'-dipyrromethane (295) A solution of dibenzyl 3,3'-diethyl-4,4'-dimethyldipyrromethane-5,5'-dicarboxylate (293, 2.43 g, 4.9 mmol) in THF (65 mL), containing 10% Pd/C (253 mg) and triethylamine (6 drops) was hydrogenated at room temperature and 1 atm. over-night. The catalyst was filtered off over a celite pad, and the filtrate concentrated to a rose-white solid. This diacid 294 was briefly dried under high vacuum. The diacid was dissolved in DMF (20 mL) and heated at reflux under an N 2 stream for 2 h monitoring the disappearance of the absorbance band at A,MAX = 285 nm in the UV spectrum. After cooling to ambient temperature, the reaction vessel was immersed in an ice-salt slurry and cooled to -5 °C. To this cold solution was added (dropwise) an excess of benzoyl chloride (2.5 mL, 21.5 mmol) whilst maintaining the internal temperature below 0 °C. A brown paste resulted 5 min after complete addition of the benzoyl chloride. This slurry was allowed to reach room temperature, then diluted with DMF (5 mL) and stirred for an additional hour at this temperature. The solids were filtered off and washed with cold DMF and diethyl ether, before dissolving in water (ca 50 mL). This aqueous solution was neutralized with sodium bicarbonate, and then heated on a hot plate for 1 h- precipitation of the product began after 10 min of heating. This solid material was filtered off and recrystallized from aq. ethanol to afford the title compound as small tan-coloured needles (869.3 mg, 62%). 205 m.p.: 205-206 °C; Lit. 218-220 °C 1 4 1 ' 2 0 2 1 H NMR (200 MHz, CDCI3): 5 1.10 (6H, t, 3,3'-CH 2CH 3 J = 7.6 Hz); 2.28 (6H, s, 4,4'-CH3); 2.55 (4H, q, 3,3 '-CH 2CH 3 J = 7.6 Hz); 3.98 (2H, s, bridge CH2); 9.5 (2H, s, 5,5'-CHO); 10.9 (2H, bs, NH). LRMS (El) m/e (rel. intensity): 286 (65.3, M+), 271 (37.3), 257 (56.3), 229 (12.9), 163 (22.4), 149 (100), 136 (58), 120 (36.9). 5.5'-Dicyano-3.3'-diethyl-4.4'-dimethyl-2.2'-dipyrromethane (297) Method A: Via the bisoxime 296 To a degassed suspension of 295 (45.86 mg, 0.16 mmol) in ethanol (5 mL) was added hydroxylamine hydrochloride (23.94 mg, 0.34 mmol) and sodium acetate (60.8 mg, 0.74 mmol), and the mixture heated at 60 °C for 3h, under an inert atmosphere. Removal of the solvent in vacuo provided the bisoxime 296 as a purple solid, which was neither purified nor characterized. This solid was dissolved in acetic anhydride and N 2 bubbled through the solution for 45 min and then stirred at 100 °C for an hour. The reaction mixture was allowed to cool and the solvent removed under reduced pressure furnishing a black powder (120.9 mg). This material was purified by flash chromatography on silica gel (230-H H 297 206 400 mesh) eluting with 4% methanol in dichloromethane, collecting the most polar fraction, to give 297 (22.7 mg) in 50% yield, as a dark brown gum. IR (CHCI3): 2225, 2200 cm" 1 1 H NMR (200 MHz, CDCI3): 5 0.99 (6H, t, 3,3'-CH 2CH 3 J = 7.5 Hz); 2.10 (6H, s, 4,4'-CH3); 2.3 (4H, q, 3,3'-CH 2CH 3 J = 7.5 Hz); 3.79 (2H, s, bridge CH2); 8.6 (2H, bs, NH). HRMS: calcd. for C 1 7 H 2 0 N 4 280.1688, found 280.1689. Method B: Via diacid 294 A sample of the dibenzyl ester 293 (105.93 mg, 0.21 mmol) in THF (10 mL) containing triethylamine (3 drops), was hydrogenated at atmospheric pressure and room temperature over 10% Pd/C (20 mg), for 17.5 h. After this time the catalyst was filtered off, and the excess solvent removed in vacuo. The diacid so obtained, was dissolved in DMF (5 mL) and thermally decarboxylated at 145 °C, under an N 2 stream, until the reaction was complete (2.5 h). The reaction was left to reach ambient temperature, then diluted with DMF (5 mL) and acetonitrile (5 mL) and cooled to -78 °C. At this point chlorosulphonyl isocyanate (150 LLL, 1.72 mmol) was added. The mixture was then allowed to reach an internal temperature of -10 °C over 1.5 h. Finally, the solution was carefully poured into a mixture of 10% NaOH (5 mL) and 5% NaHC0 3 (15 mL) and ice. This aqueous mixture was extracted with dichloromethane (3 x 10 mL). The combined organic extracts were successively washed with water and saturated brine (10 mL of each), dried and evaporated to dryness on the rotary evaporator. The resultant dark brown gum (34.4 mg) was purified on silica gel (230-400 mesh, 5 g, 4% MeOH in CH 2 CI 2 eluent), to furnish the 5,5'-dicyano-2,2'-dipyrromethane 295 (23.7 mg, 40%) as a dark brown solid. The product produced in this manner proved identical with that prepared by procedure A. 207 5.5'-Dicyano-3.3'-diethyl-4.4'-dimethyl-1.1'-(1.1-vinyl)-2.2'-dipyrromethane (299) The title compound was isolated in a preliminary dehydration experiment of the bisoxime 296 (prepared from 295 (203.4 mg, 0.71 mmol) as described in method A above) in acetic anhydride over a 2 h period, after two successive chromatographic purifications of the black residue on silica gel (230-400 mesh), initially with a 0.1% MeOH in CH 2 CI 2 eluent, and subsequently with an EtOAc:CH 2CI 2 (1:25, v/v) eluent system (respectively). Removal of the solvent from the first fraction gave the product 299 as a tan solid (64.4 mg, 30%). 1 H NMR (200 MHz, CDCI3): 8 1.1 (6H, t, 3,3'-CH 2CH 3 J = 7.5 Hz); 2.20 (6H, s, 4,4'-CH3); 2.42 (4H, q, 3,3'-CH 2CH 3 J = 7.5 Hz); 3.9 (2H, s, bridge CH2); 5.5 (2H, s, C=CH2). LRMS (El) m/e (rel. intensity): 304 (38.8, M+), 289 (21.8), 275 (100). 299 208 1-Acetyl-5.5'-dicyano-3.3'-diethyl-4.4'-dimethyl-2.2'-dipyrromethane (298) The title compound 298 was isolated as the second fraction in the final chromato-graphic purification of the residues from the previous experiment, in 37% yield as a brown gum (85.9 mg). 1 H NMR (200 MHz, CDCI3): 6 1.05 (6H, 2t, 3,3'-CH2CH3); 2.18 (3H, s, 4,4'-CH3); 2.02 (3H, s, 4-CH3); 2.4 (4H, 2q, 3,3'-CH2CH3); 2.8 (3H, s, CH 3C(0)-N 4.08 (2H, s, bridge CH2); 8.7 (1H, bs, 1 '-NH). 2-Cyano-5-ethoxycarbonyl-3-ethyl-4-methylpyrrole (301) 2-Ethoxycarbonyl-4-ethyl-5-iodo-3-methylpyrrole (300, 1 4 6 55.5 mg, 0.18 mmol) and CuCN (19.5 mg, 0.22) in DMF (5 mL) were refluxed for 4 h. The resultant dark purple solution was briefly allowed to cool (to about 65-70 °C) then poured into a solution of ethylenediamine (40 mL) in water (100 mL), and warmed for a few minutes. The purple mixture was extracted with EtOAc (3 x 15 mL), and the o' 298 301 209 combined organics then successively washed with 10% aq. potassium cyanide solution, water and saturated brine (50 mL of each), and dried. Removal of the solvent on the rotary evaporator left a tan coloured solid which was recrystallized from CH2CI2/petroleum ether (35-60) to give a pink crystalline material (29.4 mg, 79%)- 301. m.p.: 127.5-129 °C IR: 2220, 1770 cm - 1 1 H NMR (200 MHz, CDCI3): 6 1.16 (3H, t, OCH 2CH 3); 1.38 (3H, t, 3-CH2CH3); 2.25 (3H, s, 4-CH3); 2.55 (2H, q, 3-CH 2CH 3); 4.3 (2H, q, OCH 2 CH 3 ) ; 9.8 (1H, bs, NH). 1 3 C NMR (CDCI3): 8 10.019 (4-CH3), 14.344 (OCH2CH3), 14.724 (3-CH2CH3), 18.149 (3-CH2CH3), 61.227 (OCH 2CH 3), 102.347 (C-2), 113.193 (C=N), 123.548 (C-4), 125.313 (C-3), 137.478 (C-5), 160.875 (C=0). LRMS (El) m/e (rel. intensity): 206 (73.1, M+), 191 (38.2), 177 (53.7), 159 (46), 145 (100). -tetramethylporphocyanine (305) 305 To a stirred suspension of LiAIH4 (19.4 mg, 0.51 mmol) in THF (5 mL) at 0 °C was added (dropwise) a solution of the dicyanodipyrromethane 297 (22.71 mg, 0.0793 210 mmol) in THF (3 mL). The reaction mixture was stirred at this temperature for 30 min, then quenched with water and diluted with an equal volume of CH 2CI 2 . The lithium salts were filtered off, and the filtrate treated with solution of DDQ in CH 2CI 2 , until no changes were observed in the Soret band at A,m a x = 450 nm, in the electronic spectrum. The resulting dark solution was concentrated to a dark black gum which was then passed through a short alumina column (eluting with initially 10% EtOAc in CH 2CI 2 , then increasing the polarity to 30% EtOAc in CH 2 CI 2 , and finally switching to 0.1% MeOH in CHCI3) until the fluorescent green fraction ceased to come off the column. Removal of the solvent from these fractions provided 305 (3.8 mg, 18%) as a dark green solid with metallic hue. 1 H NMR (300 MHz, CDCI3): 8 2.0 (12H, t, CH 2CH 3); 3.98 (12H, s, pyrr.-CH3); 4.36 (8H, q, pyrr.-CH2CH3); 10.4 (2H, s, meso-H); 12.7 (4H, bs, CH=N-CH=pyrr.) LRMS (El) m/e (rel. intensity): 532 (100, M+), 517 (30.4), 503 (37.6). UV-vis [kmax, nm, CH 2CI 2]: 454, 592, 632, 798. Benzyl 4-ethyl-3-methyl-5-phthalimidomethylpyrrole-2-carboxylate (309) A solution of 5-acetoxymethylpyrrole 15 (1.66 g, 5.25 mmol) in anhydrous DMSO (20 mL) was added to a solution of potassium phthalimide (1.25 g, 6.77 mmol) in dry DMSO (70 mL). After stirring at room temperature for 2 h, a white precipitate of potassium acetate was observed. Removal of the solvent under reduced pressure 211 provided a solid, which was dissolved in CH 2 CI 2 (100 mL) and washed with water (2 x 50 mL). (NB at this point an emulsion formed). The organic solution was thus washed with saturated brine (50 mL), and the solvent stripped off on a rotary evaporator. Recrystallization of the solid residue from EtOH furnished 1,28 g (60%) of 309 as shiny white plates. m.p.: 123.8-125 °C 1 H NMR (200 MHz, CDCI3): 8 1.08 (3H, t, 4-CH 2CH 3, J = 7 Hz); 2.26 (3H, s, 3-CH3); 2.57 (2H, q, 4-CH 2CH 3 , J = 7 Hz); 4.76 (2H, s, 5-CH2-N-phthaloyl); 5.3 (2H, s, O-CH2Ph); 7.3-7.5 (5H, m, Ph); 7.65 (2H, m, phthaloyl Hb); 7.9 (2H, m, phthaloyl Ha); 9.2 (1H, bs, NH). 1 3 C NMR (CDCI3): 8 10.350 (3-CH3), 15.892 (4-CH2CH3), 17.014 (4-CH2CH3), 32.169 (5-CH2-N-phthaloyl), 65.494 (C0 2CH 2Ph); 118.672, 123.510 (phthaloyl C-a and b), 126.149 (phenyl o, m, p-C's), 126.894 (C-3), 127.965 (C-4), 128.494 (C-5), 131.942 (phenyl-C-1), 134.165 (phthaloyl-C), 136.538 (C-2), 161.009 (C0 2Bn); 168.161 (phthaloyl C=0). 3-Ethyl-4-methyl-2-phthalimidomethylpyrrole (310) To a vigorously stirred solution of the foregoing pyrrole 309 (1.13 g, 2.8 mmol) in anisole (100 mL) was rapidly poured 10% (v/v) cone. H 2 S 0 4 in C F 3 C 0 2 H (100 mL). The orange coloured solution was stirred at ambient temperature for 3 h, then 310 212 diluted with CH 2 CI 2 (125 mL) and ice-cold water. The organic layer was separated off and washed with water (6 x 100 mL) and then with saturated brine (100 mL) before removing the excess solvent in vacuo. The resulting solid was recrystallized from ethanol to yield a fine, pale yellow solid (310, 509.1 mg, 68%). m.p.: 174-176 °C (dec) 1 H NMR (200 MHz, CDCI3): 8 1.13 (3H, t, 3-CH 2CH 3, J = 7.6 Hz); 2.03 (3H, s, 4-CH3); 2.57 (2H, q, 4-CH 2CH 3 , J = 7.6b Hz); 4.755 (2H, s, 2-CH2-N-phthaloyl); 6.45 (1H, s, pyrr. H-5); 7.65 (2H, m, phthaloyl Hb); 7.8 (2H, m, phthaloyl Ha); 8.4 (1H, bs, NH). 1 3 C NMR (CDCI3): 8 10.081 (4-CH3), 16.134 (3-CH2CH3), 17.256 (3-CH2CH3), 32.503 (2-CH2-N-phthaloyl), 115.672 (C-5), 122.405 (phthaloyl C-a and b), 123.280 (C-3 and C-4),132.103 (C-2), 133.999 (phthaloyl C), 168.161 (phthaloyl C=0). 4.4'-Diethyl-3.3'-dimethyl-5.5'-diphthalimidomethyl-2.2'-dipyrromethane (311) 311 A mixture of pyrrole 310 (73.4 mg, 0.27 mmol) and paraformaldehyde (5.4 mg) in EtOH (5 mL) containing CHCI 3 (1.5 mL) was brought up to reflux. Cone. HCI (4 drops) was then added, and the reaction refluxed for 2 h. The cooled solution was poured into a mixture of water (10 mL) and CH 2 CI 2 (10 mL) and the aqueous phase extracted once more with an equal volume of CH 2CI 2 . The combined organic extracts were successively washed with 10% aq. NaHC0 3 , water and saturated 213 brine (10 mL of each), dried and the solvent removed to furnish a shiny brown solid. This material was purified by chromatography on silica gel (1% MeOH in CH 2 CI 2 as eluent) to afford the title dipyrromethane (311, 6 mg, 4%) as a yellow gum. 1 H NMR (200 MHz, CDCI3): 8 1.15 (6H, t, 4,4'-CH2CH3); 1.95 (6H, s, 3,3'-CH3); 2.6 (4H, q, 4,4'-CH2CH3); 3.70 (2H, s, bridge CH 2); 4.75 (4H, s, 5,5'-CH2-N-phthaloyl); 7.65 (4H, m, phthaloyl Hb); 7.78 (4H, m, phthaloyl Ha); 8.15 (2H, bs, NH). NB all attempts to repeat this procedure have failed to give this product. 2-Furoyl chloride (316) o 316 2-Furoic acid (53.48 g, 0.48 mol) in neat thionyl chloride (160.0 g, 1.35 mol) was heated at reflux for 3 h. The bulk of the excess thionyl chloride was distilled off, and the residual solvent removed in vacuo. The crude product was distilled at 25 mmHg collecting the fraction boiling at 74-75 °C (lit.195 176 °C, at 760 mmHg), yielding the acid chloride as a colourless liquid (51.81 g, 83%). 1 H NMR (200 MHz, CDCI3): 8 6.7 (1H, m, H-4); 7.5 (1H, d, H-5); 7.75 (1H, d, H-3). 214 Ethyl 2-furoate (317) OEt O 317 The above acid chloride 316 (51.81, 0.4 mol) was slowly added to ice-cold absolute EtOH (125 mL), then allowed to reach room temperature after completion of the addition. After heating at reflux for 3 h, the cooled reaction mixture was partitioned between water (10 mL) and ether (20 mL). The organic phase was washed with dilute aqueous solution of potassium carbonate, water and saturated brine (10 mL of each) and dried. Removal of the solvent under reduced pressure gave a colourless oil, which solidified upon cooling in an ice bath, to afford 49.31 g (89%) of ethyl 2-furoate as a white solid in high purity- judged on the basis of TLC analysis and 1 H NMR. m.p.: 36-38 °C; lit.204 35-37 °C 1 H NMR (200 MHz, CDCI3): 6 1.40 (3H, t, C H 3 C H 2 0 , J = 7 Hz); 4.35 (2H, q, C H 3 C H 2 0 , J = 7 Hz); 6.5 (1H, dd, H-4, J 4 5 = 1.6 Hz, J 3 4 = 3 Hz); 7.2 (1H, dd, H-3, J 4 5 = 1.6 Hz, J 3 4 = 3 Hz); 7.58 (1H, m, H-5). LRMS (El) m/e (rel. intensity): 140 (15, M+), 112 (45), 95 (100), 68 (16.3), 39 (39.6), 32(41.7). 215 5.5'-Dicarboethoxy-2.2'-difurylmethane (318) 318 Ethyl 2-furoate (317, 49.3 g, 0.35 mol) was dissolved in cone, sulphuric acid (145 mL), and the solution cooled to 0 °C. To this mixture was added a solution of paraformaldehyde (6.4 g) in cone, sulphuric acid (40 mL) whilst maintaining the temperature at 0 °C, over a 25 in period. The internal temperature increased to 8 ° C, at one point, thus the reaction mixture was recooled to -7 °C, and stirred for a further 15 min after complete addition of the aldehyde. The viscous dark green mixture was subsequently poured into a vigorously stirred ice-water mixture, and the resultant sludge extracted into ether (1000 mL in total). The ether extracts were successively washed with aqueous K 2 C 0 3 solution, water and saturated brine (200 mL of each), dried, and concentrated to a viscous, dark red oil. The latter material was distilled at 0.15 mmHg (collecting the fraction boiling at 145-150 °C, lit.159 b.p. 204 °C at 4 mmHg) furnishing the diester as a viscous yellow oil (18.78 g, 36%). Incidentally, a minor more polar difuryl impurity which was not mentioned in the literature also distilled over with 318. This material, however, did not effect the ensuing reactions, therefore no attempts were made to further purify the above product. An analytical sample was nonetheless purified on silica gel with CH 2 CI 2 eluent. m.p.: 38-39 °C; Lit. 1 5 9 39 °C. 1 H NMR (200 MHz, CDCI3): 8 1.35 (6H, t, 5,5'-CH 3CH 20, J = 7 Hz); 4.12 (2H, s, bridge CH 2); 4.31 (4H, q, 5,5'-CH 3CH 20, J = 7 Hz); 6.235 (2H, d, H-3,3', J = 3 Hz); 7.075 (2H, d, H-4,4', J = 3 Hz). 216 1 3 C NMR (CDCI3): 5 14.134 (5,5'-CH 3CH 20); 27.957 (bridge CH 2); 60.876 (5,5'-CH 3 CH 2 0) ; 109.588 (C-3,3'); 118.882 (C-4,4'); 144.129 (C-5,5'); 154.578 (C-2,2'); 158.634 (C=0) LRMS (El) m/e (rel. intensity): 292 (31.7, M+), 247 (16.4), 219 (100), 147 (23.5), 119 (20.8), 91(51.6). Anal. Calcd for C 1 5 H 1 6 0 6 : C, 61.62, H, 5.52. Found: C, 61.53, H, 5.54. 5.5'-Dicarboxy-2.2'-difurylmethane (319) Method A: From diester 318 The above diester 318 (7.74 g, 26.5 mmol) was dissolved in the minimum quantity of EtOH, and the resultant solution added to 20% aqueous NaOH (50 mL) in an Erlenmeyer flask. This basic reaction mixture was then heated on a hot plate until the volume had been reduced down to ca < 30 mL, and a yellow precipitate was apparent. The resultant solution was diluted with hot water and allowed to cool before acidification with cone. HCI giving a yellowy-tan precipitate, which was recrystallized from water to afford the title compound as a tan powder (2.94 g, 47%). m.p.: 238-239 °C; Lit. 1 5 9 238 °C. 1 H NMR (300 MHz, Acetone-d6): 6 4.26 (2H, s, bridge CH 2); 6.45 (2H, d, H3,3'); 7.2 (2H, d, H-4,4') -10-11 (2H, bs, OH). 1 3 C NMR (Acetone-d6): 5 110.309 (C-3,3'); 119.74 (C-4,41); 145.059 (C-5,5'); 155.897 (C-2,2'); 159.158 (C0 2H). O o 319 217 Method B: From the diamide 320 Difurylmethane-5,5'-dicarboxamide (320, 1.04 g, 4.45 mmol) was added to 10% aqueous solution of NaOH (35 mL), and the suspension refluxed for 2 h. After this time, the mixture was cooled then immersed in an ice-water bath and carefully acidified with cone. HCI to yield a tan paste. Recrystallization of this material from aqueous ethanol furnished a tan powder (319, 716 mg, 68%) which gave identical spectroscopic properties as that of the above sample. 2.2'-Difurylmethane-5.5'-dicarboxamide (320) The foregoing diacid 319 (1.78 g, 7.55 mmol) was added to SOCI 2 (50 mL), and the resulting suspension heated at 120 °C for 2.5 h. The excess SOCI 2 was removed in vacuo to afford a dark brown solid. To this material was added ether (15 mL), and the mixture cooled to 0 °C. The cold mixture was subsequently saturated with ammonia over a 15 min period, during which the product immediately precipitated out. The solids were filtered off and thoroughly washed with water and recrystallized from ethanol giving the diamide 320 as an off white powder (1.27 g, 72%). 1 H NMR (200 MHz, DMSO-d 6): 8 4.15 (2H, s, bridge CH 2); 6.38 (2H, d, H-3,3'); 7.1 (2H, d, H-4,4'); 7.35 (2H, bs, CONH 2); 7.78 (2H, bs, CONH 2). 1 3 C NMR (DMSO-d6): 8 27.137 (bridge CH 2); 109.276 (C-3,3'); 114.722 (C-4,4'); 147.279 (C-5,5'); 152.995 (C-2,2'); 159.385 (C=0). LRMS (El) m/e (rel. intensity): 234 (100, M+), 218 (12.2), 190 (66.7), 145 (32.5), 119 (30.9), 103 (22.9), 91 (81.4). O o 320 218 2.2'-Difurylmethane (314) 314 Method A: From furan and furfuryl alcohol. Following the procedure of Brown and Sawatzky, 1 5 8 to a rapidly stirred solution of furan (18.1 mL, 0.25 mol) and furfuryl alcohol (313, 21.6 mL, 0.25 mol) at 12 °C was added 38% HCI (5 mL). The ensuing white suspension was stirred at this temperature for 4 h (turned dark brown after an hour), then neutralized with Na 2 C0 3 , and steam distilled to afford a yellow oil (2.7 g). This material was distilled at 11 mmHg, collecting the fraction boiling at 73-75 °C as a colourless oil (1.81 g). 1 H NMR (200 MHz, CDCI3) showed this product to be consistent with a 1.6:1 mixture of 313:314. Despite the differences in boiling points, this mixture could not be enriched in the desired difurylmethane by repeated distillation, or derivatization of the alcohol, due to further polymerization. The product turned black on standing. 1 H NMR (200 MHz, CDCI3): 5 2.15 (1H, bs, OH*); 4.0 (2H, s, bridge CH 2); 4.8 (2H, s, 2-CH 2OH*); 6.09 (2H, dd, H-3,3', J 4 5 = 1 Hz, J 3 4 = 3 Hz); 6.26-6.345 (4H, m, H-p*, H-4,4'), 7.30 (2H, d, H-5,5'); 7.4 (2H, H-oc*); *= furfuryl alcohol 313 LRMS (El) m/e (rel. intensity): 148 (76.7, M+), 120 (31.8), 98 (74), 91 (100), 81 (79). Method B: From the diacid 319. 5,5'-Dicarboxy-2,2'-difurylmethane (319, 2.9 g, 12.3 mmol) and activated Cu-bronze (7.5 g) were intimately mixed, and ground to a fine powder. This solid mixture was then heated with a hot flame (under N2) until the product ceased to distill over. The 219 distillate was diluted with ether (15 mL), and washed with a dilute aqueous solution of Na 2 C0 3 . The aqueous phase was washed with a portion of ether (10 mL), and the combined organics washed with saturated brine, dried and concentrated to an oil which rapidly darkens on standing. This material was redistilled at 1.5 mmHg, collecting the colourless fraction boiling at 38 °C (lit.159 b.p. 66 °C at 5 mmHg), giving difurylmethane 314 in 22% (408.3 mg) yield. 1 H NMR (200 MHz, CDCI3): 8 4.0 (2H, s, bridge CH 2); 6.08 (2H, m, H-3,3'); 6.30 (2H, m, H-4,4*), 7.38 (2H, m, H-5,5'). 1 3 C NMR (CDCI3): 8 27.396 (bridge CH 2); 106.444 (C-3,3'); 110.386 (C-4,4'); 141.583 (C-5,5'); 151.565 (C-2,2'). LRMS (El) m/e (rel. intensity): 148 (53.6, M+), 120 (24.4), 91 (100). Anal. Calcd for C 9 H 8 0 2 : C, 72.96; H, 5.44. Found: C, 73.12; H, 5.39. 5-Carboxy-2.2'-difurylmethane (323) 323 Acidification of the aqueous washings from the preceding experiment furnished the mono acid as fluffy grey needles (51 mg, 2%). m.p.:117-118 °C; Lit. 1 5 9118 °C. 1 H NMR (200 MHz, CDCI3): 84.10 (2H, s, bridge CH 2); 6.15 (1H, m, H-3'); 6,28 (1H, m, H-4'); 6.39 (1H, m, H-3), 7.25 (1H, m, H-4); 7.39 (1H, m, H-5'); 10.35 (1H, bs, C0 2 H). 220 2.2-Dithienylmethane (324) 324 Thiophene (6.66 g, 79.2 mmol) was slowly added (over a 10 min period) to a vigorously stirred solution of ZnCI2 (7.12 g, 50 mmol) in cone. HCI (10 mL) at -10° C. Formaldehyde (4.8 mL, 64 mmol, 37% aq. solution) was then slowly added to this suspension over 45 min, whilst maintaining the temperature below -10 °C, and the mixture stirred for a further 1.5 h at between -5° and -15 °C. The solution was diluted with water (25 mL), extracted with ether (3 x 25 mL), and the organic solution washed with 10% NaHC0 3 (30 mL), water and saturated brine (20 mL of each), dried and the ether removed in vacuo. The resultant yellow oil was distilled under reduced pressure to yield the dithienylmethane 324 as a colourless oil (1.65 g, 29%), which turned yellow on standing. b.p. 85-88 °C at 0.2 mmHg, lit.56 158-163 °C at 15 torr. 1 H NMR (200 MHz, CDCI3): 5 4.35 (2H, s, bridge CH 2); 6.9-7.00 (4H, m, H-3,3' and H-4,4'); 7.2 (2H, m, H-5,5'). 5.5'-Dicyano-2.2'-difurylmethane (322) To a solution of 2,2'-difurylmethane (314, 115.4 mg, 0.78 mmol) in CH 3 CN (2 mL) and DMF (2 mL) at -78 °C, was added a solution of CSI (575 u.L, 6.6 mmol) in 322 221 CH 3 CN (1.5 mL). The reaction solidified after complete addition of the CSI reagent thus CH 3 CN (2 mL) was added and the resultant viscous slurry allowed to warm up to -20 °C. After stirring for 45 min at this temperature, the mixture was diluted with a further 2 mL of CH 3 CN, and the temperature allowed to reach 10 °C. This product was then worked up as described above for 297. The crude mixture so obtained was flashed chromatographed on silica gel (230-400 mesh, eluting initially with EtOAc:hexane,1:6; then switching to EtOAc:hexane,1:4, collecting the most polar fraction) to give108.5 mg (70%) of 322 as a red oil. IR (CDCI3): 2230 cm" 1 1 H NMR (300 MHz, CDCI3): 5 4.1 (2H, s, bridge CH 2); 6.28 (2H, d, H-3,3', J= 3.3 Hz); 7.02 (2H, d, H-4,4', J = 3.3 Hz). 1 3 C NMR (CDCI3):5 27.609 (bridge CH 2); 109.320 (C-3,3'); 111.213 (C-4,4'); 123.104 (C=N); 154.852 (C-2,2'). LRMS (El) m/e (rel. intensity): 198 (100, M+), 181 (2.3), 170(12.8), 155 (2), 143 (66.9). HRMS calcd. for C ^ H ^ O , , : 198.0429. found 198.0429. Attempted co-reduction of 5.5'-dicyanodifurylmethane (322) with 5.5'-dicyano-dipyrromethane 279 To a vigorously stirred suspension of LiAIH4 (39.4 mg, 1.0 mmol) in THF (5 mL) at -10 °C was slowly added a solution of 297 (23.0 mg, 0.12 mmol) and 322 (23.26 mg, 0.08 mmol) in THF (10 mL). The resultant mixture was stirred at 0 °C for 40 min, then quenched with water and diluted with CH 2CI 2 . The aluminum salts were filtered off, and to the filtrate was added a solution of DDQ in CH 2CI 2 . The resultant crude 222 mixture displayed a typical porphyrin-type UV-visible spectrum, and was therefore passed through an alumina column eluting with 10% EtOAc in CH 2 CI 2 , then gradually increasing the solvent polarity to 30% EtOAc in CH 2 CI 2 , to yield a brown solid. 1 H NMR (300 MHz, CDCI3) failed to provide any evidence of the expanded porphyrin, however the UV-visible spectrum was identical to that of 305. UV-vis [ ? i m a x , nm, CH 2CI 2]: 452, 590, 630, 798. 2-Phthalimidomethylfuran (328) o 328 A mixture of furfurylamine (23.33 g, 0.24 mol) and phthalic anhydride (43 g, 0.29 mol) in glacial acetic acid (160 mL) was heated at reflux for 50 min. Subsequent cooling of the reaction mixture afforded white needles, which were filtered off and thoroughly washed well with water, and recrystallized from aqueous ethanol to yield 328 (47.4 g, 87%) as white needles. A second crop (455 mg) was scavenged from the mother liquors. m.p.: 116-117 °C •" . 1 H NMR (200 MHz, CDCI3): 5 4.85 (2H, s, furan-CH2-N-phthaloyl); 6.25 (1H, m, H-3); 6.38 (1H, m, H-4); 7.35 (1H, d, H-5); 7.7 (2H, m, H-5',6'); 7.85 (2H, m, H-4',7'). 223 1 3 C NMR (CDCI3): 5 34.282 (furan-CH2-N-phthaloyl); 108.7 (C-4); 110.455 (C-3); 123.429 (C-4',7'); 132.056 (C-5',6'); 134.033 (C-3',8'); 142.421 (C-2); 149.283 (C-5); 167.586 (C=0). LRMS (El) m/e (rel. intensity): 227 (100, M+), 198 (33.4), 182 (6.6), 170 (21.1), 143 (9.9), 133(21.1), 115(12.5), 104(35). Anal. : Calcd. for C 1 3 H 9 N 0 3 , C, 68.72; H, 3.99; N, 6.16; found: C, 68.71; H, 4.00; N, 6.05. 2-(Aminomethyl)thiophene (327) To a vigorously stirred suspension of LiAIH4 (5.75 mg, 0.152 mole) in THF(80 mL) at 0 °C was added (dropwise) a solution of 2-thiophenecarbonitrile (5 g, 0.046 mole) in THF 5 mL). The resulting olive solution was stirred for 2 h, then quenched with potassium tartrate. (50 g) in water (250 mL) and bxtracted into ethyl acetate (3 x 200 mL). The organic phase was successively washed with water and saturated brine (100 mL of each), dried and concentrated to furnish a green oil (4.2 g, 83%).This material was carried forward without further purification. 1 H NMR (200 MHz, CDCI3): 5 1.6 (2H, bs, NH2); 4.0 (2H, bs, CH 2NH 2) ; 6.90-6.99 (2H, m, H-3,4);7.2(1H, d, H-5). 327 224 2-Phthalimidomethylthiophene (328) o o 329 2-(Aminomethyl)thiophene (327, 4.2 g, 37.0 mmol) and phthalic anhydride (7.9 g, 53.3 mmol) in acetic acid (25 mL) was heated at reflux for 1 h. The reaction was cooled to room temperature, and the precipitate filtered off and washed well with water. Recrystallization from ethanol furnished 328 as large tan-coloured needles (7.41 g, 6 8 % ) . m.p.: 126-127 °C; lit.168 124-125 °C. 1H NMR (200 MHz, CDCI 3, 6): 5.0 (2H, s, ArCH2N-phthaloyl); 6.9 (1H, m, H-4); 7.1-7.25 (2H, m, H-3,5); 7.7 (2H, m, H-5',6'); 7.85 (2H, m, H-4',7'). 1 3 C NMR (CDCI 3): 5 35.685 (thio-CH2-N-phthaloyl); 123.425 (C-4',7'); 125.866 (C-5); 126.872 (C-3); 127.724 (C-4); 132.063 (C-5',6'); 134.042 (C-3',8'); 138.061 (C-2); 167.554 (C=0). 225 5.5'-Bisphthalimidomethyl-2.2'-difurylmethane (330) o o Method A: To a rapidly stirred solution of furan 328 (17.29 g, 76.1 mmol) in TFA (60 mL) at -8 ° C was added a suspension of paraformaldehyde (1.3 g) in TFA(20 mL), while maintaining the reaction temperature below 0 °C through out the addition. After stirring at 0 °C for 45 min, the dark olive solution was carefully poured into an ice-water mixture (100 mL), and extracted into CH 2 CI 2 (3 x 75 mL). The combined organic extracts were washed with 10% aqueous NaOH (2 x 30 mL), water (2 x 100 mL) and saturated brine (100 mL), dried and the solvent removed in vacuo yielding a golden yellow foam. This material was purified by flash chromatography on silica gel, eluting with CH 2 CI 2 , to provide the title compound as a white solid (4.05 g, 25%). In addition, 1.8 g of unreacted 328 was recovered. On larger scale, the product can be more efficiently separated from 328 and the higher oligomers by passing through a plug of silica gel, eluting with CH 2CI 2 . An analytical sample was recrystallized from EtOH, furnishing white needles. m.p.: 177-179.3 °C. 1 H NMR (200 MHz, CDCI3): 5 3.86 (2H, s, bridge CH 2); 4.775 (4H, s, furan-CH2-N-phthaloyl); 5.94 (2H, d, H-3,3', J = 3 Hz); 6.21 (2H, d, H-4,4', J = 3 Hz); 7.62 (4H, m, H-5",5"'; H-6",6m); 7.85 (4H, m, H-4",4"'; H-7",7"'). 226 1 3 C NMR (CDCI3): 5 27.459 (bridge CH 2); 34.407 (furan-CH2-N-phthaloyl); 107.499 (C-4,4'); 109.606 (C-3,3'); 123.401 (G-4 , , I4"'; C-7",7m); 132.095 (C-5",5"'; C-6",6"'); 133.981 (C-3".3"'; C-8",8'"); 148.109 (C-2,2');151.213 (C-5,5'); 167.550 (C=0). LRMS (El) m/e (rel. intensity): 466 (76.9, M+), 319 (93.2), 306 (68.3), 239 (100), 226 (68.7). Anal. : Calcd. for C 2 7 H 1 8 N 2 0 6 , C, 69.52; H, 3.89; N, 6.01; found: C, 69.16; H, 3.9; N, 5.75. Method B: A mixture of 328 (114.6 mg, 0.5 mmol), ZnCI2 (72 mg, 0.53 mmol) and paraformal-dehyde (8.9 mg) in anhydrous CCI 4 (3.5 mL) was heated between 75-80 °C. HCI gas was then slowly bubbled through the reaction mixture over a 2 h period. A further portion of formaldehyde (4.5 mg, 0.15 mmol) added and reaction continued for 30 min, then the HCI source was removed and the mixture stirred at 80 °C for 17 h (tic still showed unreacted 328). After cooling, the suspension was poured into water (10 mL), and the organic layer separated off. The latter was then successively washed with 10% aq NaOH (2x10 mL), water and saturated brine (.10 mL of each), dried and concentrated to a straw-yellow oil (213 mg). Flash chromatography of this material gave 16.8 mg (14%) of 330, which gave identical 1 H NMR spectrum to that produced by method A. The major component of this reaction, however was a ring-opened decomposition product. 227 5.5'-Bisphthalimidomethyl-2.2'-dithienylmethane (331) o o 331 An aqueous solution of formaldehyde (1.4 mL, 17 mmol, 37% aq. solution) was carefully added to a mixture of 329 (7.37 g, 30 mmol) and ZnCI2 (2.47 g, 18 mmol) in TFA (30 mL) at -13 °C. This mixture was stirred for 2 h, whilst gradually allowing the temperature to reach -5 °C, and then poured into an ice-water mixture. Removal of the solvent, after working up as described for 330 (method A), and subsequent purification of the resultant yellow foam through a short plug of silica gel (CH 2CI 2 eluent) gave the title product (1.94 g, 27%). Recrystallization from EtOH gave the product as a white solid. m.p.: 183-185 °C 1 H NMR (200 MHz, CDCI3): 5 4.1 (2H, s, bridge CH 2); 4.84 (4H, thio-CH2-N-phthaloyl); 6.55 (2H, d, H-3,3'); 6.9 (2H, d, H-4,4'); 7.65 (4H, m, H-5",5"'; H-6",6m); 7.8 (4H, m, H-4",4"'; H-7",7m). 1 3 C NMR (CDCI3): 8 30.453 (bridge CH 2); 35.932 (thio-CH2-N-phthaloyl); 123.4219 (C-4",4"'; C-7",7m); 125.086 (C-3,3'); 127.565(0-4,4'); 132.067 (C-5",5"'; C-6",6"'); 133.998 (C-3".3"'; C-8",8"'); 136.834 (C-2,2');143.369 (C-5,5'); 167.556 (C=0). LRMS (El) m/e (rel. intensity): 498 (32, M+), 351 (32.8), 338 (100), 318 (8). 228 5.5'-Bis(aminomethyl)-2.2'-difurylmethane (321) 321 To a stirred solution of hydroxylamine hydrochloride (150.6 mg, 2.17 mmol) in absolute MeOH (3.5 mL) was added 1N NaOMe in MeOH (6.0 mL), a fine white precipitate of NaCI soon became apparent. After stirring for 5 min, this suspension was then added to a well-stirred solution of 5,5'-bisphthalimidomethyl-2,2'-difuryl-methane (330, 166.3 mg, 0.36 mmol) in THF (10 mL), at this point the solution turned orange in colour. The suspension was stirred at room temperature for 45 min, then poured into water (25 mL), extracted with CHCI 3 (5 x 20 mL), and the organic solutions washed with saturated brine (10 mL), dried, and concentrated on a rotary evaporator giving a brown oil (321, 71 mg) in 96% crude yield. 1 H NMR (300 MHz, CDCI3): 5 1.55 (4H, bs, NH2); 3.75 (4H, s, 5,5'-CH 2NH 2); 3.9 (2H, s, bridge CH 2); 5.95 (2H, d, H-3,3'); 6.05 (2H, d, H-4,4'). 1 3 C NMR (CDCIg): 6 27.538 (bridge CH 2); 39.378 (5,5'-CH2NH2); 105.868 (C-4,4'); 106.980 (C-3,31) 150.660 (C-2,2'; C-5,5'). LRMS (El) m/e (rel. intensity): 206 (0.4, M+), 189 (100), 161 (11), 109 (17); 96 (53.7). FAB LRMS (thioglycerol matrix) m/e (rel. intensity): 207 (71,[M + H]+), 190 (100), 173 (32), 161 (19.6); 133 (12.5), 81 (17.9). 229 5.5'-Bis(aminomethyl) -2.2'-dithienylmethane (332) 332 Prepared by the foregoing procedure from 5,5'-bisphthalimidomethyl-2,2'-dithienyl-methane (331,107.5 g, 0.22 mmol). Upon removal of the solvent, the diamine 332 (50 mg, 90%) was isolated as a brown gum, in sufficient purity (by 1 H NMR) for subsequent elaboration. 1 H NMR (200 MHz, CDCI3): 8 1.47 (4H, bs, NH2); 3.89 (4H, bs, 5,5'-CH 2NH 2); 4.16 (2H, s, bridge CH 2); 6.64 (4H, bs, H-3,3'; H-4,4'). 1 3 C NMR (CDCI3): 8 30.785 (bridge CH 2); 41.645 (5,5'-CH2NH2); 123.237 (C-4,4'); 124.783 (C-3,3') 141.934 (C-2,2'; C-5,5"). LRMS (El) m/e (rel. intensity): 238 (87.2, M+), 221 (33.6), 205 (8), 192 (22.4), 160 (11), 147 (6), 112 (100), 97 (22.6). N.N.N'-Triacetyl-5.5'-Bis(aminomethyl)-2.2'-difurylmethane (333) 5,5'-Bis(aminomethyl)-2,2'-difurylmethane (321, 22.6 mg, 0.11 mmol) was dissolved in A c 2 0 (1 mL), and the solution refluxed for 30 min. After cooling to ambient temperature, water (ca 1 mL) was added and the mixture heated for 10 min. The resultant opaque brown reaction mixture was diluted with CH 2 CI 2 (10 mL), and the organics washed successively with 5% aqueous NaHC0 3 and saturated brine (5 mL of each), dried, and the solvent removed on the rotary evaporator to give a gold 333 230 gum (10 mg). This material was flashed chromatographed on silica gel (4% MeOH in CH 2 CI 2 eluent) to yield the title compound as a green gum (333, 4.6 mg, 12%). 1 H NMR (300 MHz, CDCI3): 8 2.05 (3H, N'-acetyl CH3); 2.5 (6H, s, N-acetyl CH3); 3.9 (2H, s, bridge CH 2); 4.39 (2H, d, 5'-CH2NHAc); 4.85 (2H, s, 5-CH 2NAc 2); 5.95 (1H, bs, NH); 6.0 (2H, m, H-3,3'); 6.19 (2H, m, H-4,4'). LRMS (El) m/e (rel. intensity): 333 (4, M+), 332 (20), 289 (44.8), 247 (51.1), 231 (43.2), 203 (4.8),189 (100), 176 (9.4), 161 (6.5), 136 (17.1), 109 (18.6), 94 (10.9), 79 (5.5), 43 (41.4), 30 (8.2). 12.16-Diethyl-11.17-dimethyl-[ 6.11013.11518]octacosa- (334) 3 3 4 Method A: To a solution of 3,3'-diethyl-5,5'-diformyl-4,4'-dimethyldipyrromethane (295, 18.7 mg, 0.063 mmol) and 5,5'-bis(aminomethyl)-2,2'-difurylmethane (321, 46.6 mg, 0.225 mmol) in anhydrous toluene (10 mL) at 0 °C was added a 1M solution of TiCI4 in CH 2 CI 2 (66 (iL, 0.066 mmol). The reaction was allowed to warm up to room temperature and stirred for 17 h. Filtration of the crude mixture and subsequent removal of the solvent under reduced pressure yielded the crude product 334 (33.4 mg) as a solid, which gave satisfactory 1 H NMR data. 231 NB on some occasions this method required warming (70-100 °C) for 1-2 h, in order to complete the reaction-. Method B: (preferred method) A solution of the diamine 321 (100 mg, 0.48 mmol) and the diformyldipyrromethane 295 (125.9 mg, 0.44 mmol) in anhydrous toluene (100 mL) containing absolute MeOH (ca 1-5 mL, until a clear solution results) and p-TsOH was heated at 115 °C for 20 h, under an N 2 atmosphere. The solution rapidly turns green after 5 min, and progressively changes to a dark turquoise. After cooling, potassium carbonate was added and the mixture filtered through Mg 2 S0 4 . The solvent was then removed in vacuo and the resultant dark green solid dried under high vacuum for 24 h, giving the title compound (236.2 mg). IR(CDCI3): 1631 (C=N) 1 H NMR (400 MHz, CDCI3): 8 0.965 (6H, t, CH 2 CH 3 , J = 7.5 Hz); 2.03 (6H, s, pyrr.-CH3); 2.34 (4H, q, CH 2 CH 3 , J = 7.5 Hz); 3.7 (2H, s, pyrr.2-CH2); 3.875 (2H, s, furan2-CH2); 4.44 (4H, s, furan-CH2N); 5.99-6.05 (4H, 2d, H-3,3', H-4,4', J 3 4 = 3.0 Hz); 7.9 (1H, bs, NH); 8.0 (2H, s, HC=N). LRMS (El) m/e (rel. intensity): 456 (22.4, M+), 441 (2.3), 307 (9.8), 109 (100), 80 (50.3), 53 (18.5) 39 (7.7). UV-vis: contains a broad band between 200 and 300 nm, no other bands beyond 400 nm. 232 12.16-Diethyl-11.17-dimethyl-[ 6.11013.11518]octacosa- (335) This compound was prepared by using the foregoing procedure (method b) to prepare 334. From 332 (400 mg, 1.68 mmol) and 295 (479 mg, 1.67 mmol) was produced 905 mg of crude 335. 1 H NMR (200 MHz, CDCI3): 8 1.05 (6H, t, CH 2 CH 3 , J = 7.6 Hz); 2.095 (6H, s, pyrr.-CH3); 2.405 (4H, q, CH 2 CH 3 , J = 7.6 Hz); 3.79 (2H, s, pyrr.2-CH2); 4.27 (2H, s, thio.2-CH2); 4.675 (4H, s, thio.-CH2N); 6.69-6.75 (4H, 2d, H-3,3', H-4,4', J 3 4 = 3.2 Hz); 7.9 (1H, bs, NH); 8.1 (2H, s, HC=N). 1 3 C NMR (CDCI3): 8 15. 684, 17.242, 22.113, 31.696, 59.956, 122.99, 123.597, 124.362, 124.878, 128.219, 129.032, 143.559, 144.809, 150.077. LRMS (El) m/e (rel. intensity): 488 (4.3, M+), 340 (3.6), 97, 7.2), 91 (100), 77 (8), 65 (20.8), 51 (18.5) 39 (20.8), 32 (61.6). UV-vis: contains a broad band between 200 and 300 nm, no other bands beyond 400 nm. 335 233 12.16-Diethyl-11.17-dimethyl-[ 6.110,13/L15,181octacosa- (336) 336 To a solution of 321 (110 mg, 0.53 mmol) and 295 (150 mg, 0.52 mmol) in anhydrous toluene (50 mL) and MeOH (2 mL) was added p-TsOH (ca 10 mg). After stirring at reflux temperature for 18 h in dry a i r , ;K 2 C0 3 was added, and the mixture filtered through MgS0 4 . The solvent was removed in vacuo giving a red-brown solid (230 mg). 1 H NMR (200 MHz, CDCI3): 5 1.13 (6 H, t, CH2CH3); 2.19 (6H, s, CH3); 4.0 (2H, s, furan2-CH2); 4.8 (4H, s, furan-CH2N); 6.16 (4H, m, H-3,3', H-4,4*); 6.7 (1H, s, CH); 8.10 (2H, s, HC=N). LRMS (El) m/e (rel. intensity): 454 (91, M+), 439 (12.8), 331 (22.4) 306 (100), 149 (70.4) 105 (38.1), 97 (33.6), 83 (100), 71 (49.6), 57 (84.8), 43 (88), 32 (70.4). UV-vis : contains a broad band centred around 450 nm, no other bands beyond this absorption. 234 12.16-Diethyl-11.17-dimethyl-[ 6.110,13/L15,18]octacosa- (337) 337 Condensation of 332 (30 mg, 0.126 mmol) and 297 (36.1 mg, 0.126 mmol) under the conditions described in the foregoing preparation (of macrocycle 336), furnished 15.6 mg of 337 as a red-brown oil. 1 H NMR (200 MHz, CDCI3): 1.05 (6H, t, CH2CH3); 2.1 (6H, s, CH3); 2.55 (4H, q, CH 2CH 3); 4.3 (2H, s, thio.2-CH2); 5.0 (4H, s, thio.-CH2N); 6.7 (3H, m, H-3,3', CH); 6.8 (2H, d, H-4,4'); 8.19 (2H, s, HC=N). UV-vis [ ? i m a x , nm, CHCI3]: 472, 504 (sh). 4.4'-Diethyl-5.5'-diformyl-3.3'-dimethyl-2.2'-bipyrrole (339) 116 To a hot suspension of 5,5'-di(2-cyano-2-methoxycarbonylvinyl)-4,4'-tetraethyl-3,3'-tetramethyl-2,2'-bipyrrole (339,149 900 mg, 2.1 mmol) in MeOH (50 mL) was added KOH (8.1 g, 0.14 mmol) in water (10 mL). The resultant purple mixture was heated 235 until the majority of the MeOH had been boiled off, giving a yellow precipitate. The residual MeOH was then removed in vacuo, and the product filtered off, once the mixture had cooled. The crystalline product (431.7 mg, 75%) was sufficiently pure for use in the ensuing macrocyclization reaction. m.p.: 240.5-241 °C, Lit. 1 0 1 241-242 °C. 1 H NMR (200 MHz, CDCI3): 8 1.28 (6H, t, CH 3CH 2); 2.1 (6H, s, CH3); 2.78 (4H, t, CH3CH2); 9.65 (2H, s, CHO), 9.75 (2H, bs, NH). 11.16-Diethyl-12.15-dimethyl-[ 6.11013.11417]heptacosa-,23-decaene (340) 340 Diamine 322 (100 mg, 0.48 mmol) and the foregoing 5,5'-diformyl-2,2'-bipyrrole 116 (129.7 mg, 0.48 mmol) were refluxed in a mixture of toluene (30 mL) and MeOH (5 mL), in the presence of p-TsOH, for 48 h. The cooled solution was neutralized with K 2 C 0 3 , and filtered through MgS0 4 . Subsequent removal of the solvent furnished the title macrocycle 340 (197 mg) in 90% crude yield. An effort to recrystallize this material from chloroform/hexane provided a dark brown insoluble precipitate. 1 H NMR (200 MHz, CDCI3): 5 1.15 (6H, t, CH 3CH 2); 2.02 (6H, s, CH3); 2.5 (4H, q, CH3CH2); 3.9 (2H, s, furan2-CH2); 4.5 (4H, furan-CH2-N); 5.95-6.05 (4H, m, furan (3-H); 8.05 (2H, s, CH=N). 236 NB the peaks in the NMR were all broadened fairly substantially, several attempts to improve the resolution with numerous samples on both 200 MHz and 400 MHz NMR instruments met with little success. These experiments, however, did indicate that the macrocycle had begun to decompose. Incidentally, these results mirrored an initial attempt at synthesizing 340 via the titanium tetrachloride method described for 334 (method B). LRMS (El) m/e (rel. intensity): 442 (1, M+), 189 (100), 161 (1.33), 109 (26), 96 (76), 78 (31), 68 (19), 53 (8), 41 (24), 30 (54). UV-vis : contains a broad band between 200 and 300 nm, no other bands beyond 400 nm. 2-(2-furanyh-1.3-dioxolane (349a) 349a 2-Furaldehyde (13.19 g, 0.14 mol), ethylene glycol (12 mL, 0.22 mol), and pyridinium tosylate (PPTS; 1 8 3 3.44 g, 0.014 mol) in toluene (100 mL), were heated at reflux (whilst azeotropically removing the water produced) for 2.5 h. After cooling, ether (100 mL) was added to the dark purple mixture, and this was then washed with 10% aqueous NaHC0 3 (2 x 50 mL), water and saturated brine (50 mL of each), dried and the solvents removed in vacuo yielding a dark reddish-purple liquid. This material was distilled under reduced pressure (water pump: ca 40 mmHg, collecting the fraction boiling between 108-115 °C; lit.205 91-93 °C at 16 mmHg) to give the 14.62 g (75%) of the acetal 349a as a colourless liquid. 237 1 H NMR (300 MHz, CDCI3): 5 3.98 (4H, m, 0-CH 2 CH 2 -0) ; 5.91 (1H, s, CH); 6.34 (1H, dd, H-4, J 3 4 = 3.6 Hz, J 4 5 = 1.5 Hz); 6.43 (1H, d, H-3, J 3 4 = 3.6 Hz); 7.40 (1H, d, H-5, J 4 > 5 = 1.5 Hz). 2-(2-furanyl)-5.5-dimethyl-1.3-dioxane (349b) Prepared as described above for 349a from 2-furaldehyde (10.29 g, 0.11 mol) and 2,2-dimethyl-1,3-propanediol (17.2 g, 0.165 mol) in 87% (16.1 g) pure yield; obtained as a colourless liquid after distillation. b.p.: 75-77 °C at 0.2 mmHg 1 H NMR (200 MHz, CDCI3): 8 0.8 (3H, s, CH 3); 1.3 (3H, s, CH 3); 3.6-3.8 (4H, m, O-CH 2C); 5.50 (1H, s, CH); 6.38 (1H, m, H-4); 6.48 (1H, m, H-3); 7.40 (1H, m, H-5). 1 3 C NMR (CDCI3): 8 21.846 (CH3); 22.949 (CH3); 30.376 (C[CH3]2); 77.495 (CH 20); 96.112 (-OCHO-); 107.370 (C-4); 110.158 (C-3); 142.5 (C-5); 150.977 (C-2). LRMS (El) m/e (rel. intensity): 182 (41.9, M+), 154 (5.7), 128 (36.4), 113 (3.9), 97 (100), 81 (4.4), 69 (16.1), 56 (11.3), 41 (5.4). Ethyl N.N-dimethylcarbamate An aqueous 40% (w/v) solution of dimethylamine (228 mL, 2 mol) was poured into water (77 mL), and cooled to 0 °C. To the latter solution was added (dropwise) ethyl 349b 238 chloroformate (95 mL, 1 mmol) over a 1 h period. After stirring at room temperature for a further 5.5 h, the reaction was extracted with ether (3 x 150 mL), and the combined extracts washed with saturated brine (150 mL), then dried. The solvent was the partially removed under vacuum, and the residue (ca 200 mL) was fractionally distilled using a Vigreux column, collecting the colourless fraction (95.11g, 81%) boiling between 136-144 °C; (lit.176 b.p.145 °C), at 760 mmHg. 1 H NMR (200 MHz, CDCI3): 6 1.2 (3H, t, OCH 2 CH 3 ) ; 2.85 (6H, s, CH 3N); 4.05 (2H, q, OCH 2 CH 3 ) . 5.5'-Bis(1.3-dioxolan-2-yl)-2.2'-difurylmethanone (250a) nBuLi (56 mL, 0.0896 mol, 1.6N solution in hexane) was slowly added to a solution of 349a (10.36 g, 0.0735 mol) in THF (100 mL) at -78 °C. The resulting yellow solution was then allowed to warm up to -20 °C and then a solution of the foregoing carbamate (4.73 g, 0.04 mol) in THF (10 mL) was added, whilst maintaining the temperature below -20 °C. The reaction was then allowed to warm up to -10 °C over 1.5 h, then quenched with a saturated NH4CI solution to yield a brown precipitate. This material was filtered off, washed well with water, then dissolved in CH 2CI 2 . The organic solution was washed with saturated brine, dried and the volume reduced (to half its original total volume) in vacuo. To this hot residue was added petroleum ether, and the flask stored in the freezer for a few hours. The 350a 239 resulting brown powder was then collected by filtration and dried under high vacuum to yield 350a (5.28 g, 47%). m.p.: 128-129 °C; Lit. 1 7 6 129-131 °C 1 H NMR (300 MHz, CDCI3): 8 3.98 (8H, m, 0-CH 2 CH 2 -0) ; 6.0 (2H, s, CH); 6.65 (2H, d, H-3,3'); 7.55 (2H, d, H-4,4'). 5.5'-Bis(1.3-dioxan-2-yl-5.5-dimethyl)-2.2'-difurylmethanone (350b) o Treatment of the 5-lithium salt of 349b (13.92 g, 0.076 mol) with ethyl N,N-dimethyl-carbamate (4.65 g, 0.0397 mol) as described above in the synthesis of 350a, provided 5.49 g (37%) of 350b as a yellow solid, after recrystallizatipn form CH 2CI 2 / petroleum ether. m.p.: 235-236 °C. 1 H NMR (200 MHz, CDCI3): 8 0.8 (6H, s, CH 3); 1.3 (6H, s, CH 3); 3.6-3.8 (8H, m, O-CH 2C); 5.56 (2H, s, CH); 6.65 (2H, d, H-3,3', J = 3.6 Hz); 7.50 (2H, d, H-4,4', J = 3.6 Hz). 1 3 C NMR (CDCI3): 8 21.819 (CH3); 22.929 (CH3); 30.440 (C[CH3]2); 77.410 (CH 20); 95.626 (-OCHO); 109.666 (C-3,3'); 120.185 (C-4,4'); 150.732 (C-2,2'); 155.474 (C-5,5'); 168.359 (bridge C=0). 240 LRMS (El) m/e (rel. intensity): 390 (4.8, M +) ,305 (3.3), 181 (100), 123 (9.6), 95 (23.8), 69 (20.6), 56 (6.3), 41 (5.4), 32 (6.3). 5.5'-Diformyl-2.2'-difurylmethanone (351) 351 A solution of 350a (1.11 g, 5.61 mmol) in acetone (50 mL) and 10% aqueous HCI (5 mL) was stirred at room temperature for 10 h, the starting material precipitated out therefore, 6N HCI (5 mL) was added and the mixture refluxed for 3 h. The acetone was then removed in vacuo, and the residue taken up in CHCI 3 and successively washed with water, 10% aqueous NaHC0 3 , water and saturated brine (50 mL of each), and dried. Removal of the solvent provided the dialdehyde as a tan coloured powder (603 mg, 77%), in good purity for further elaboration. m.p.: 184-185 °C; Lit. 1 7 3 182-184 °C 1 H NMR (200 MHz, CDCI3): 8 7.38 (2H, d, H,3,3'); 7.75 (2H, d, H-4,4'); 9.8 (2H, s, CHO). 1 3 C NMR (CDCI3): 8 119.660, 120.927 (4 x (3-furyl); 153.972 (2 x oc-furyl); 178.799 (CHO). 241 Attempted preparation of 5.5'-dioxime-2.2'-difurylmethanone (352) o To a stirred suspension of the foregoing dialdehyde 351 (145 mg, 0.66 mmol) in degassed EtOH (70 mL) was added hydroxylamine hydrochloride (111 mg, 1.59 mmol) and sodium acetate (131 mg, 1.6 mmol). The resulting solution was stirred for 28 h, then diluted with water (50 mL) and extracted with CH 2 CI 2 (3 x 30 mL). The combined organic extracts were washed with saturated brine (25 mL), dried and concentrated in vacuo. The residue was purified by flash chromatography on silica gel eluting with 2% MeOH in CH 2 CI 2 to afford a golden gum (106 mg) as the major product (second fraction). 1 H NMR (200 MHz, Acetone-d6): 5 6.99 (2H, d, H-3,3'); 7.4 (2H, d, H-4,4'); 8.2 (2H, s, CH=N); 11.08 (2H, s, N-OH). In addition the following two unaccountable peaks were observed: 8 3.5 (s);11.5 (bs). LRMS (El) m/e (rel. intensity): 248 (10.3, M+), 230 (17.6), 212 (25.6), 184 (4.8), 149 (8.8), 138 (14.4), 120 (100), 95 (14.4), 79 (16), 64 (33.6), 58 (10.4), 51 (13.6), 43 (35.2), 39 (18.4), 32 (79.2). The first fraction yielded a yellow gum (16.7 mg) upon removal of the solvents, which from 1 H NMR spectroscopy is a non-heterocyclic decomposition by-product. 242 5.5'-Bis(1.3-dioxan-2-yl-5.5-dimethyl)-2.2'-difurylmethanol (354) OH I 354 I Method A: To a stirred suspension of LiAIH4 (101.1 mg, 2.66 mmol) and AICI3 in dry ether (50 mL) at 0 °C was added 350b (53 mg, 0.136 mmol). The mixture was stirred at this temperature for 30 min, then at room temperature for 45 min. The reaction was cooled in an ice-water bath, and quenched with saturated NH4CI solution until the precipitate coagulated, and these aluminium salts were subsequently filtered off. The filtrate was diluted with ether (50 mL), and successively washed with water and saturated brine (20 mL of each), dried, and the solvent removed at the water pump, giving a yellow gum (40 mg). This material flash chromatographed on silica gel (230-400 mesh, 2.5% MeOH in CH 2 CI 2 as eluent) giving the meso-hydroxydifuryl-methane 354 (33 mg, 62%) as a golden yellow gum. 1 H NMR (200 MHz, CDCI3): 8 0.8 (6H, s, CH 3); 1.25 (6H, s, CH 3); 2.7 (1H, bd, OH); 3.5-3.78 (8H, m, OCH 2C); 5.50 (2H, s, OCHO); 5.80 (1H, bs, (furan)2CHOH); 6.30 (2H, d, H-3,3'); 6.40 (2H, d, H-4,4'). 1 3 C NMR (CDCI3): 8 21.841, 22.959 (2 x CH 3); 30.368 (C[CH3]2); 77.405 (CH 20); 96.405 (-OCHO); 108.84 (C-3,3'); 109.988 (C-4,4'); 151.143 (C-2,2'); 153.116 (C-5,5'). LRMS (El) m/e (rel. intensity): 392 (9.2, M+), 376 (6.9), 305 (5.3), 289 (6.1), 203 (12.9), 191 (11.1), 181 (100), 115 (90.3), 95 (46.7), 81 (8.8), 69 (100), 56 (21), 41 (61.3), 32 (85). 243 In a second attempt, with a newer batch of AICI3, treating 350b with LiAIH4 at temperatures below -10 °C provided the difurylmethane 355 only. Method B: To a solution of difurylketone 350b (261 mg, 0.67 mmol) in EtOH (25 mL) was added KOH (50 mg, 0.87 mmol) and THF (ca 15 mL), and the mixture briefly refluxed under a nitrogen stream. Once the bulk of the solids had dissolved, NaBH 4 (4 x 50 mg, 5.3 mmol) was added to the mixture in portions over 1.5 h. After complete addition of the borohydride reagent, the mixture was heated at 75 °C for 19 h. (NB, a further portion of KOH (1 pellet) and NaBH 4 (100 mg, 2.15 mmol) was added after 2 h). After cooling, the mixture was then poured into water (50 mL) and extracted with ether (3 x 25 mL). The ethereal extracts were washed with water and saturated brine (25 mL of each), dried and concentrated in vacuo down to a cream solid (253.1 mg, 96%). This crude material was identical to that obtained in the previous experiment (by 1 H NMR), and required no further purification. 5.5'-Bis(1.3-dioxan-2-yl-5.5-dimethyl)-2.2'-difurylmethane (355) To an ice-cold 1N solution of borane-THF complex (10 mL, 10 mmol) in THF, was added 350b (56.5 mg, 0.14 mmol). The reaction was then allowed to reach room temperature and left stirring for 2.5 h. The excess borane was destroyed by addition 244 of MeOH (ca 10 mL) and once the fizzing stopped, the solvent was removed on the rotary evaporator. The residue was taken up into CH 2 CI 2 (15 mL), and the organic solution washed with 10% aqueous NaHC0 3 , water and saturated brine (10 mL of each), then dried. Removal of the solvent furnished a deep maroon solid (39.3 mg) which was chromatographed on silica gel to afford 355 (20 mg, 38%) as a colourless solid. A small sample was recrystallized from EtOH giving a fluffy, white solid. m.p.: 135.5-137 °C. 1 H NMR (200 MHz, CDCI3): 8 0.78 (6H, s, CH 3); 1.3 (6H, s, CH 3); 3.58-3.79 (8H, m, OCH 2C); 4.0 (2H, s, bridge CH 2); 5.40 (2H, s, OCHO); 6.05 (2H, d, H-3,3'); 6.40 (2H, d, H-4,4'). 1 3 C NMR (CDCI3): 8 21.839, 22.955 (2 x CH 3); 27.466 (bridge CH 2); 30.350 (C[CH3]2); 77.447 (CH 20); 96.148 (-OCHO); 107.394 (C-3,3'); 108.549 (C-4,4'); 151.426 (C-2,2'); 153.122 (C-5,5'). LRMS (El) m/e (rel. intensity): 376 (9.6, M+), 290 (8.1), 205 (14.5), 181 (100), 147 (4.8), 109 (12.9), 95 (50), 81 (11.3), 69 (46.7), 56 (9.7). 2-(2-Pyrrolylmethyleniminomethyl)furan (358) A mixture of furfurylamine (1.29 g, 13.2 mmol) and pyrrole-2-carboxaldehyde (1.25 g, 13.1 mmol) in water (10 mL) was briefly heated with a heat gun, whilst stirring, until a homogeneous solution results. The reaction was then allowed to cool, then briefly placed in an ice-water bath to ensure complete precipitation. The off white 358 245 solid was collected at the pump, and dried over-night in vacuo to furnish the title compound 358 (2.00 g) in 87% yield. m.p.: 76.5-77 °C. 1 H NMR (300 MHz, CDCI3): 8 4.66 (2H, s, furan-CH2N=); 6.22 (3H, m, pyrr. H-4, fury I H-3, NH); 6.32 (1H, m, furyl H-4), 6.51 (1H, m, pyrr. H-3); 6.82 (1H, m, pyrr. H-5); 7.36 (1H, m, furyl H-5), 8.11 (1H, s, CH=N). 1 3 C NMR (CDCI3): 8 56.482 (CH2N); 107.854 (furyl C-3), 109.760 (furyl C-4); 110.361 (pyrr. C-4); 114.970 (pyrr. C-3); 122.137 (pyrr. C-5); 129.914 (pyrr. C-2); 142.155 (furyl C-5); 152.734 (furyl C-2);153.450 (CH=N). LRMS (El) m/e (rel. intensity): 174 (97.6, M+), 157 (2.4), 145 (20), 118 (4.8), 93 (12), 81 (100), 68 (8), 53 (27.2). 2-Chloromethyl-5-phthalimidomethylfuran 362 o o 362 To a solution of 2-phthalimidomethylfuran 328 (3.86 g, 17 mmol) in dry CH 2 CI 2 (150 mL) was added ZnCI2 (948 mg) and paraformaldehyde (814 mg). HCI was then bubbled through the rapidly stirred mixture for 30 min. The reaction was subsequently washed with water (2 x 30 mL), 10% aqueous NaHC0 3 , water and finally saturated brine (30 mL of each). After drying the organic solution, the solvent was removed under reduced pressure to provide the chloromethyl derivative 362 (3.13 g, 67%) in sufficient purity for further manipulations. 246 m.p.: 118-119 °C. 1 H NMR (200 MHz, CDCI3): 8 4.45 (2H, s, CH2CI); 4.79 (2H, s, CH2N-phthaloyl); 6.2 (2H, m, H-3, H-4); 7.62 (2H, m, phthaloyl H-5', H-6'j; 7.70 (2H, m, phthaloyl H-4', H-7')-1 3 C NMR (CDCIg): 8 34.048 (CH2N-phthaloyl); 37.052 (CH2CI); 109.559 (C-4); 110.363 (C-3); 123.156 (phthaloyl C-4', C-7'); 131.687 (phthaloyl C-5', C-6'); 133.771 (phthaloyl C-3',C-8'); 149.673 (C-2); 149.866 (C-5); 167.132 (C=0). LRMS (El) m/e (rel. intensity): 275 (11.5, M+), 239 (100), 226 (6.4), 212 (3), 160 (6.4), 130 (12.9), 104 (14.5), 92 (20), 65 (16.1). 2-Acetoxymethyl-5-phthalimidomethylfuran (363) o The foregoing furan 362 (391.8 mg, 1.42 mmol) and anhydrous NaOAc (417 mg, 5.74 mmol) in glacial acetic acid (5 mL) containing acetic anhydride (0.25 mL) was heated at reflux. After an hour, the reaction was allowed to cool, then diluted with ether (30 mL) washed twice with water and 10% aqueous NaHC0 3 , then once each with water and saturated brine (15 mL of each) and dried. Removal of the solvent furnished a pale yellow solid. Recrystallization from CH2CI2/hexane gave 332.3 mg (78%) of the target compound 363, as brilliant yellow needles. m.p.: 116-118 °C. OAc o 363 247 1 H NMR (200 MHz, CDCI3): 5 1.95 (3H, s, CH 3C=0); 4.7 (2H, s, CH 2OAc); 4.8 (2H, s, CH2N-phthaloyl); 6.18 (2H, m, H-3, H-4); 7.62 (2H, m, phthaloyl H-5', H-6'); 7.78 (2H, m, phthaloyl H-4', H-7'). 1 3 C NMR (CDCI3): 8 20.581 (CH3C=0); 34.148 (CH2N-phthaloyl); 57.789 (CH2OAc); 109.625 (C-4); 111.341 (C-3); 123.205 (phthaloyl C-4', C-7'); 131.775 (phthaloyl C-5', C-6'); 133.825 (phthaloyl C-3',C-8'); 149.06 (C-2); 149.730 (C-5); 167.132 (C=0); 171 (acetate C=0). LRMS (El) m/e (rel. intensity): 299 (4, M+), 257 (8), 239 (100), 160 (10), 110 (27.2). 2-Ethoxymethyl-5-phthalimidomethylfuran (364) o Obtained as a by-product from the reaction used to prepare 362, during initial exploratory work. It was isolated as the major component after chromatographic purification of the reaction mixture on silica gel with EtOAc/hexane (30:70) eluent as a white crystalline material (502.7 mg, 25%); from 1.56 g (6.85 mmol) of 328. m.p.: 85 °C. 1 H NMR (200 MHz, CDCI3): 8 1.17 (3H, t, CH 3 CH 2 , J = 7 Hz); 3.50 (2H, q, OCH 2 CH 3 , J = 7 Hz); 4.36 (2H, s, CH 2 OCH 2 CH 3 ) ; 4.83 (2H, s, CH2N-phthaloyl); 6.2 (1H, d, H-3, J = 3 Hz); 6.28 (1H, d, H-4, J = 3 Hz); 7.7 (2H, m, phthaloyl H-5', H-6'); 7.85 (2H, m, phthaloyl H-4', H-7'). 1 3 C NMR (CDCI3): 8 15.061 (CH 3CH 2); 34.510 (CH2N); 64.542 (OCH 2CH 3); 65.722 (OCH2-furan); 109.399 (C-4); 109.888 (C-3); 123.409 (phthaloyl C-4', C-7'); 132.085 o 364 248 (phthaloyl C-5', C-6'); 134.023 (phthaloyl C-3', C-8'); 149.085 (C-2); 152.053 (C-5); 167.513 (C=0). LRMS (El) m/e (rel. intensity): 285 (4.2, M+), 239 (100). Diethyl 3.3'-diethyl-4.4'-dimethyl-2.2'-bipyrrole-5.5'dicarboxylate (370) Ethyl 4-ethyl-5-iodo-3-methylpyrrole-2-carboxylate (300, 6 10.11 g, 33 mmol) was dissolved in DMF (50 mL). Activated copper bronze (9.68 g, BDH) was added and the mixture heated at reflux for 2 h, during which the solution colour changed from dark rose-pink to olive. After cooling, the copper bronze was filtered off through a celite pad, and the solids washed several times with hot CHCI 3 until the washings were colourless. The filtrate was then washed with 1N HCI (2 x 100 mL), water (2 x 100 mL), and saturated brine (100 mL). After drying over MgS0 4 , the solvent was removed on the rotary evaporator to leave a dark brown oil, which was placed under high vacuum to remove all traces of the DMF for 2 days, at which point the material solidified. The solid residue was then triturated with petroleum ether to yield the bipyrrole (370, 1.55 g, 26%) as scintillating off-white crystals. An analytical sample was recrystallized from EtOH to furnish colourless thick needle-like crystals. m.p.: 180-181 °C; Lit. 1 4 6 182-183 °C. o o 370 249 1 H NMR (200 MHz, CDCI3): 5 1.1 (6H, t, 3,3'-CH 2CH 3, J = 7.5 Hz); 1.38 (6H, t, OCH 2 CH 3 , J = 7 Hz); 2.3 (6H, s, 4-CH3); 2.45 (4H, q, 3,3'-CH 2CH 3, J = 7.5 Hz); 4.36 (4H, q, OCH 2CH 3 , J = 7 Hz); 8.7 (2H, bs, NH). LRMS (El) m/e (rel. intensity): 360 (61, M+), 314 (35.2), 286 (100), 267 (19.6), 253 (12.3), 240 (34.9), 225 (15.5), 212 (34.3), 197 (16.3). Ethyl 3-ethyl-4-methylpyrrole-2-carboxylate (368) Obtained as the major product from the Ullmann coupling procedure of 300 (1.03 g, 3.35 mmol) as describe in the foregoing preparation when normal copper bronze was employed. Flash chromatography (silica gel, 230-400 mesh, eluting with CH 2CI 2) of the crude reaction mixture provided 368 (193 mg, first fraction), as red oil which solidified on standing. m.p.: 23-24 °C; Lit. 2 0 6 25 °C. 1 H NMR (200 MHz, CDCI3): 5 1.1 (3H, t, 3-CH 2CH 3, J = 7 Hz); 1.28 (3H, t, OCH 2 CH 3 , J = 7 Hz); 2.25 (3H, s, 4-CH3); 2.40 (2H, q, 3-CH 2CH 3 , J = 7 Hz); 4.25 (2H, q, OCH 2 CH 3 , J = 7 Hz); 6.6 (1H, s, H-5); 8.8 (2H, bs, NH). LRMS (El) m/e (rel. intensity): 181 (44.2, M+), 166 (39.4), 152 (15.4), 134 (25.1), 120 (100), 108 (12.4), 92 (12.4), 79 (13.9), 65 (10.8), 53 (7.3), 39 (12.0), 32 (12.9). o 368 250 Diethyl 3.3'diethyl-4.4'-dimethyl-N.1'-bipyrrole-2.2'-dicarboxylate (369) Isolated from the second fraction in the foregoing attempted synthesis of bipyrrole 370, during the course of column chromatographic separation (silica gel, CH 2 CI 2 as eluent). Obtained as a colourless gum (54.3 mg, 9%), m.p.: 90-91 °C. 1 H NMR (200 MHz, CDCI3): 8 0.9 (3H, t, 3 '-CH 2 CH 3 , J = 7 Hz); 1.055 (3H, t, 3-CH 2 CH 3 , J = 7 Hz); 1.125, 1.28 (6H, 2t, OCH 2 CH 3 , J = 7 Hz); 2.14 (2H, q, 3'-CH 2 CH 3 , J = 7 Hz); 2.25, 22.6 (6H, 2s, 4,4'-CH3); 2.39 (2H, q, 3-CH 2 CH 3 , J = 7 Hz); 4.05 (2H, q, 5 ' -C0 2 CH 2 CH 3 , J = 7 Hz); 4.23 (2H, 5 -C0 2 CH 2 CH 3 , J = 7 Hz); 6.54 (1H, s, H-5); 8.75 (1H, bs, NH). 1 3 C NMR (CDCI3): 8 10.560, 10.896 (CH3); 14.058, 14.175 (OCH 2CH 3); 14.514, 14.667 (CH 2CH 3); 16.560, 18.071 (CH 2CH 3); 59.595, 59.859 (OCH 2CH 3); 118 (C-5'); 121.664, 126.334 (pyrr. p-C); 127.046 (C-5'); 129.160 (C-2,2'); 163 (C=0). LRMS (El) m/e (rel. intensity): 360 (30.6, M+), 314 (5.5), 299 (8.5), 285 (4.5), 269 (11.2), 253 (5.5), 241 (22.9), 225 (7.9), 213 (7.6), 197 (6.9), 180 (100), 166 (7.1), 134 (31.1). EtO 369 251 3.3'-Diethyl-4.4'-dimethyl-2.2'-bipyrrole (371) and 5.5'-Dicyano-3.3'-diethyl-4.4'-dimethyl-2.2'-bipyrrole (372) H H H H 371 372 Diethyl S.S'-diethyl^'-dimethyl^'-bipyrrole-S.S'dicarboxylate (370, 255 mg, 0.71 mmol) was saponified and decarboxylated by heating in ethylene glycol (4 mL) in containing NaOH (70 mg, 1.75 mmol) at 194 °C under N 2 . After 2.5 h, the mixture was allowed to cool and diluted with water (4 mL), then extracted with CHCI 3 (3x10 mL). The organic extracts were washed with saturated brine (10 mL), dried and concentrated in vacuo to reveal the di-a-free bipyrrole 371 (114 mg, 74%) as a deep blue-grey coloured solid. 1 H NMR (200 MHz, CDCI3): 5 1.1 (6H, t, CH 2 CH 3 ) ; 2.15 (6H, s, CH 3), 2.45 (4H, q, CH 2 CH 3 ) ; 6.6 (2H, s, H-5,5'); 7.7 (2H, bs, NH). After briefly drying under high vacuum, the bipyrrole 371 was dissolved in DMF (8 mL) and CH 3 CN (5 mL), and the solution cooled to -80 °C. CSI (370 uL, 4.25 mmol) was subsequently added (dropwise) to this cold solution, and the mixture allowed to slowly reach room temperature over a 1 h period. The reaction mixture was then poured into a solution of 10% aqueous NaOH (6 mL) and 10% aqueous NaHC0 3 (19 mL), extracted into CHCI 3 (3 x 15 mL). The combined organic extracts were washed with water and saturated brine (10 mL of each), dried and the solvent removed on the rotary evaporator to give a dark purple solid. Purification of the latter 252 by flash chromatography on silica gel (70-230 mesh, EtOAc/hexane: 1:4 eluent) gave 82.5 mg (43%) of 372, as a faint yellow solid. IR (CDCI3); 2210 cm- 1 1 H NMR (200 MHz, CDCI3): 0.95 (6H, t, CH2CH3); 2.1 (6H, s, CH 3), 2.35 (4H, q, CH 2CH 3); 10.55 (2H, bs, NH). 1 3 C NMR (CDCI3): 5 10.078 (CH3); 14.799 (CH3CH2); 17.670 (CH3CH2); 99.707 (C-5,5'); 124.924 (C=N); 125.227 (pyrr. p-C); 129.837 (C-2,2'). LRMS (El) m/e (rel. intensity): 266 (100, M+), 251 (89.9), 237 (27.3), 222 (16), 149 (17.6). 5-Bromofuran-2-carboxaldehyde (376) 376 A solution of bromine (41.1 g, 0.257 mol) in anhydrous 1,2-dichloroethane (30 mL), was added during 45 min to a well stirred solution of 2-furaldehyde (23.45 g, 0.24 mol) in anhydrous 1,2-dichloroethane (60 mL) held at reflux. After refluxing for a further 4 h, the dark reaction mixture was steam distilled. The distillate was then extracted with ether (3 x 200 mL), and the combined extracts washed with saturated brine (200 mL) and dried. Removal of the solvent gave an oil, to which was added EtOH before placing in the freezer for 2 days. The product was subsequently collected at the water pump, washed with cold EtOH and dried in vacuo to give 376 (2.5 g, 6%) as cream coloured needles. m.p.: 82-83 °C; Lit. 1 9 181-82 °C 253 1 H NMR (200 MHz, CDCI3): 8 6.5 (1H, d, H-3); 7.2 (1H, d, H-4); 9.78 (1H, s, CHO). 1 3 C NMR (CDCI3): 8 114.697 (C-3,4); 122.406 (C-2); 130.968 (C-5); 176.312 (CHO). LRMS (El) m/e (rel. intensity): 175 (100), 174 (80, M+), 119 (25.8), 95 (8), 81 (6.4), 77 (11.3), 69 (22.5), 57 (29), 44 (77). 5-lodofuran-2-carboxaldehyde (377) 5-Bromofuran-2-carboxaldehyde (376, 2.5 g, 14.3 mmol) and Kl (2.5 g, 15 mmol) in glacial acetic acid (20 mL) were heated at reflux for 1 h. The mixture was left to cool over-night, during which time the product began to crystallize out from the solution. The reaction mixture was diluted with water to ensure complete precipitation before collecting the product, which was then washed well with water. Recrystallization from EtOH furnished tan coloured needles (377, 2.35 g, 74%). m.p.: 126-128 °C; Lit. 1 9 1 127-129 °C 1 H NMR (200 MHz, CDCI3): 8 6.77 (1H, d, H-3, J = 3.6 Hz); 7.095 (1H, d, H-4, J = 3.6 Hz); 9.5 (1H, s, CHO). 1 3 C NMR (CDCI3): 8 99.09 (C-2); 122.294 (C-4); 123.2 (C-3); 158.0 (C-5); 176.161 (CHO). LRMS (El) m/e (rel. intensity): 222 (100, M+); 165 (12.9), 39 (30.6). O' CHO 377 254 5.5'-Diformyl-2.2'-bifuran 185 185 Method A: (Ullmann Coupling) A mixture of 377 (1.41 g, 6.37 mmol) and activated copper bronze (1.47 g, BDH) in DMF (10 mL) was refluxed for 48 h. The suspension was the cooled to ca 50 °C, and subsequently filtered through a celite pad. The solids were washed with hot CHCI 3 until the washings were colourless. The combined filtrate and washings were sequentially washed with 1N HCI (2 x 50 mL), water and saturated brine (50 mL of each), dried and concentrated in vacuo to give a dark brown solid. The latter was taken up into the minimum quantity of CHCI 3 and hexane added, before placing in the freezer for several hours to give 185 as a brown powder (192 mg, 32%). m.p.: 261-262 °C (dec); Lit. 1 8 9 262-265 °C. 1 H NMR (200 MHz, CDCI 3/DMSO-d 6): 5 6.97 (2H, d, H-4,4', J = 4 Hz); 7.28 (2H, d, H-3,3', J = 4 Hz); 9.60 (2H, s, CHO). LRMS (El) m/e (rel. intensity): 190 (100, M+), 162 (5.7), 133 (56), 105 (15.1). Method B: (Pd2 + promoted coupling) Following the procedure of Itahara et a / . 1 8 9 2-furaldehyde (10 g, 0.1 mol) and palladium (II) acetate (675 mg, 3 mmol) in CH 3 CN (40 mL) was stirred at reflux temperature in air for 20 h. After the required time, the reaction was concentrated on the rotary evaporator to give a black oil. This material was purified by flash chromatography on silica gel eluting with 2.5% MeOH in CH 2 CI 2 in order to remove the unreacted aldehyde, then subjected to a second chromatographic purification (silica gel), eluting with EtOAc/hexane (2:1), to provide the tittle compound as a 255 yellow solid (55 mg, 0.6%). Despite the extensive purification attempts, this product was still found to be contaminated (by 1 H NMR). Attempted preparation of 5.5'-di(aminomethyl)-2.2'-bifuran (379) 379 To a suspension of 5,5'-diformyl-2,2'-bifuran (185, 52 mg, 0.27 mmol) in MeOH (10 mL), was added anhydrous NaOAc (56 mg, 0.68 mmol) and NH2OH.HCI (47.9 mg, 0.68 mmol). The resulting homogeneous solution was stirred for 24 h at room temperature, then diluted with water (10 mL), and extracted with CHCI 3 (3 x 20 mL). The organic extracts were then washed with saturated brine, dried and the solvent removed on the rotary evaporator to give a yellow solid (88 mg). The latter was dissolved in MeOH (10 mL) with NH 4OH (5 drops) and Raney nickel and hydrogenated at room temperature under 1 atm. of hydrogen. The catalyst was filtered off, washed with CH 2 CI 2 and the filtrate washed with water and saturated brine, dried, and the solvent removed under reduced pressure, furnishing a yellow solid (8 mg). The 1 H NMR (200 MHz, CDCI3) of this material contained the following signals, which were tentatively assigned as indicated: 3.8 (2H, bs, NH); 4.4 (2H, s, CH 2); 6.2, 6.4 (4H, 2m, (3-H). However, in addition to the above, the spectrum contained several singlets between 8 1-2.2 ppm respectively. 256 1.4-Bis(5-benzyloxycarbonyl-3-ethoxycarbonyl-4-methyl-2-pyrrolyl)-1,4-butanedione 381 2-Benzyl 4-ethyl 5-formyl-3-methylpyrrole-2,4-dicarboxylate 380146 (2.04 g, 6.47 mmol) was dissolved in anhydrous 1,4-dioxane (25 mL). 3,4-Dimethyl-5-(2-hydroxy-ethyl)thiazolium iodide (274 mg, 0.96 mmol) was added, followed by triethylamine (370 |iL, 2.8 mmol) and divinyl sulphone (360 |iL, 3.59 mmol). After heating at 84 °C for 48 h, the reaction mixture was allowed to cool to ca 50 °C, and then filtered, washing the solids with 1,4-dioxane. Removal of the solvent from the filtrate gave a dark orange oil which was dissolved in CH 2 CI 2 (50 mL), washed with water and brine, dried and the volume reduced to approximately 25 mL. Hexane was added to this solution, and the flask subsequently stored in the freezer for several hours. The resulting precipitate was collected at the water pump to yield 381 (1.02 g, 48%) as a pale orange-tan coloured solid. m.p.: 139-140 °C. 1 H NMR (200 MHz, CDCI3): 5 1.56 (6H, t, C H 3 C H 2 0 , J = 7 Hz); 2.65 (6H, s, CH 3); 3.51 (4H, s, 0=CCH 2CH 2C=0); 4.55 (4H, q, OCH 2 CH 3 , J = 7 Hz); 5.5 (4H, s, OCH 2Ph); 10.1 (2H, bs, NH). 1 3 C NMR (CDCIg): 5 11.033, 13.794, 34.635, 60.780, 66.305, 119.444, 121.431, 127.947, 127.983, 128.065, 128.236, 129.45, 131.629, 134.948, 159.780, 164.336, 190.570. 381 257 LRMS (El) m/e (rel. intensity): 656 (9.4, M+), 638 (1.7), 610 (1.8), 564 (5.1), 530 (1.5), 519 (2.4), 473 (2.9), 456 (1.7), 396 (1.4), 342 (2.2), 314 (6.3), 296 (8.4), 91 (100). 2.5-Bis(5-benzyloxycarbonyl-3-ethoxycarbonyl-4-methyl-2-pyrrolyl)furan 382 382 A solution of 381 (900 mg, 1.37 mmol) in ethanol (35 mL) containing cone. H 2 S 0 4 (2 drops) was heated at reflux for 24 h. The mixture was cooled to room temperature, and the solution volume reduced to 10 mL, and the flask placed in the freezer for a few hours. Filtration and washing with cold hexane afforded a yellow solid (382, 594 mg, 68%). m.p.: 129-131 °C. 1 H NMR (300 MHz, CDCI3): 5 1.39 (6H, t, C H 3 C H 2 , J = 7 Hz); 2.57 (6H, s, CH 3); 4.33 (4H, q, OCH 2 CH 3 ) ; 5.3 (4H, s, OCH 2Ph); 7.38 (10H, m, Ph); 7.5 (2H, s, furan p-H); 10.32 (2H, bs, NH). 258 1.4-Bis(2-furyl)-1.4-butanedione 385 385 To a solution of 2-furaldehyde (1.3 mL, 15.7 mmol) in absolute EtOH (10 mL), was added 3-benzyl-5-(2-hydroxyethyl)-4-methyl-1,3-thiazolium chloride (632 mg, 2.3 mmol), anhydrous sodium acetate (384 mg, 4.68 mmol) followed by triethylamine (650 LLL, 4.66 mmol). The mixture was heated to reflux temperature, then divinyl sulphone (785 LLL, 7.82 mmol) was added dropwise, and the reaction heated at reflux for 22 h. This mixture was poured into water (20 mL) and extracted with CHCI 3 (3 x 20 mL), the organic extracts were washed with saturated brine (20 mL), dried, and the solvent evaporated in vacuo to furnish a yellow solid. Recrystallization from CH2CI2/hexane furnished 385 (916 mg, 53%) as an off-white powder. m.p.: 128-129 °C; Lit. 6 7 b 131 °C. 1 H NMR (200 MHz, CDCI3): 5 3.175 (4H, s, 0=CCH 2CH 2C=0); 6.425 (2H, dd, H-4,4', J 3 4 = 3 Hz, J 4 5 = 1.6 Hz); 7.125 (2H, d, H-3,3', J 3 4 = 3 Hz); 7.47 (2H, d, H-5,5', J = 1.6Hz). 1 3 C NMR (CDCI3): 8 31.79 (CH2); 112.13 (C-4,4'); 116.996 (C-3,3'); 146.243 (C-5,5'); 152.373 (C-2,2'); 187.426 (C=0). LRMS (El) m/e (rel. intensity): 218 (23, M+), 123 (49.6), 95 (100). 259 2.2':5'. 2"-Terfuran (386) 386 A solution of 385 (253 mg, 1.16 mmol) in acetic anhydride (7.5 mL) containing cone. HCI (0.4 mL) was stirred at room temperature for 6 days, then poured into water (200 mL) and neutralized with Na 2 C0 3 . The mixture was extracted with CCI 4 (2 x 100 mL), and the combined organic extracts washed with brine, dried, and the solvent evaporated off giving a dark solid. Subsequent purification of this material by flash chromatography on silica gel, eluting with CCI 4, furnished the terfuran 386 as a cream coloured solid (50.87 mg, 22%), which darkens on standing. m.p.: 62-63 °C; Lit. 1 9 7 63 °C. 1 H NMR (200 MHz, CDCI3): 8 6.4 (2H, dd, H-4,4", J 4 5 = 1.4 Hz, J 3 4 = 3.6 Hz); 6.53-6.55 (4H, m, H-3,3',3",4'); 7.35 (2H, m, H-4,4"). 1 3 C NMR (CDCI3): 5 105.429 (C-3,3"); 106.92 (C-3',4'); 111.477 (C-4,4'); 141.968 (C-5,5'); 145.726 (C-2,2"); 146.262 (C-2',5'). 2-(5-formyl-2-furanyl)-1.3-dioxalane (390) A solution of nBuLi (41 mL, 65.6 mmol, 1.6N) in hexane was added to a solution of anhydrous diisopropylamine (9.4 mL, 67 mmol) in THF (50 mL) at -78 °C, and the mixture stirred at this temperature for 20 min. To this mixture (at -78 °C), was then added (dropwise) a solution of 349a (7.66 g, 54.7 mmol) in THF (25 mL). After 390 260 stirring for 40 min, an excess of anhydrous DMF (51 mL) was added to the solution whilst maintaining the internal temperature at -78 °C, throughout the procedure. The reaction was left to stir for 12 h, allowing the temperature to gradually rise to 20 °C over this period. The mixture was diluted with ether (200 mL), and washed with water (3 x 150 mL) then saturated brine (100 mL) and dried. Removal of the solvent in vacuo furnished a liquid, which was distilled to provide 390 (1.855 g, 20%) as a yellow oil. b.p.: 100-104 °C at 0.02 mmHg; Lit. 1 8 0 120 °C at 0.06 mbar. 1 H NMR (200 MHz, CDCI3): 5 3.93-4.15 (4H, m, OCH 2 CH 2 0) ; 5.96 (1H, s, CH); 6.60 (1H, d, H-4, J 3 4 = 3.2 Hz); 7.18 (1H, d, H-3, J 3 4 = 3.2 Hz); 9.70 (1H, s, CHO). Furan-2-5-dicarboxaldehyde (2.5-diformylfuran) (391) 391 A solution of the foregoing acetal 390 (516 mg, 3 mmol) in acetone (35 mL) containing 6N HCI (3 mL) was heated at reflux temperature for 2 h. After cooling, the acetone was removed on the rotary evaporator and the solid residue dissolved in CH 2 CI 2 (50 mL). The solution was sequentially washed with 10% aqueous N a 2 C 0 3 (3 x 25 mL), water (25 mL) and brine (25 mL), dried and the solvent removed under reduced pressure. The product was recrystallized from CHCI3/cyclo-hexane to furnish a yellow solid (121 mg), and the filtrate concentrated and passed through a silica gel column eluting with 2% MeOH in CH 2CI 2 , to give a white solid. Total yield: 224 mg, 60%. 261 m.p.: 107-109 °C; Lit. 1 8 0 109 °C. 1 H NMR (200 MHz, CDCI3): 6 7.30 (2H, s, H-3,4); 9.90 (2H, s, CHO). 1 3 C NMR (CDCI3): 5 119.211 (C-3,4); 254.218 (C-2,5); 179.167 (C=0). 2.5-Bis(phthalimidomethyl)furan (392) o o o o 392 A suspension of 362 (201 mg, 0.73 mmol) and potassium phthalimide (203 mg, 1.1 mmol) in benzene (5 mL) was heated at reflux for 2 h, then anhydrous DMSO (2 mL) added and the solution heated at 90 °C for a further hour. After cooling to room temperature, CHCI 3 (10 mL) was added and the solution washed with water (2x10 mL) and brine (10 mL), dried and concentrated in vacuo to give a brown solid (180 mg, 64%). m.p.: 250 °C (dec). 1 H NMR (200 MHz, CDCI3): 8 4.80 (4H, s, furan-CH2N); 6.25 (2H, s, H-3,4); 7.7-7.9 (8H, m, phthaloyl). LRMS (El) m/e (rel. intensity): 386 (2.7, M+), 306 (2.5), 255 (38.7), 239 (100), 226 (23.8), 198 (9.6), 160 (16.1), 147 (76.8), 130 (15.2), 115 (9.6), 104 (85.6), 76 (84.8), 50 (36.8), 39 (9.6), 32 (19.3). 262 Thiophene-2-5-dicarboxaldehyde (93) 93 To a solution of thiophene (4 mL, 50 mmol) and TMEDA (9.2 mL, 60 mmol) in anhydrous hexane (15 mL) at room temperature was added a 1.6N solution of "BuLi (70 mL, 112 mmol) in hexane. The resultant suspension was heated at reflux for 30 min, the mixture was allowed to cool to room temperature before adding THF (60 mL). The solution was then cooled to -40 °C and dry DMF (19.5 mL) was added over 15 min. The solution was gradually allowed to reach room temperature, then left to stir over the weekend. The suspension was subsequently poured into a vigorously stirred solution of cone. HCI (100 mL) and water (800 mL) at -5 °C, and to the latter was then added a saturated solution of NaHC0 3 until the aqueous layer had reached pH 6. The resultant mixture was extracted with ether (7 x 25 mL), and the combined ether solutions washed with saturated brine (30 mL), dried and the solvent stripped off at the water pump. The solid residue was recrystallized from THF/Et 20 (4:1) to afford a tan powder (393, 1.87 g, 27%). m.p.: 110-112 °C; Lit. 1 8 0 109-112 °C. 1 H NMR (200 MHz, CDCI3): 5 7.90 (2H, s, H-3,4); 10.09 (2H,s, CHO). 1 3 C (NMR CDCI3): 5 135.021 (C-3,4); 149.182 (C-2,5); 183.359 (C=0). 263 2.5-Bis(hydroxymethyl)thiophene (394) 393 To a solution of thiophene-2-5-dicarboxaldehyde (393, 1.70 g, 12 mmol) in THF (25 mL) at room temperature was added a solution of NaBH 4 (1.00 g, 26 mmol) in water (25 mL)-an exothermic reaction set in. After stirring for 24 h, the organic phase was separated off, and the aqueous phase saturated with NaCI. The latter was then extracted with ether (3 x 20 mL). The combined organic solutions were washed with saturated brine before drying and removal of the solvent to give the 394 as a golden oil (1.47 g, 85%), which was carried forward without further purification. 1 H NMR (200 MHz, DMSO-d 6): 6 4.55 (4H, d, t h i o . - C H 2 O H , J = 5.2 Hz); 5.29 (2H, t, OH, J = 5.2 Hz); 6.76 (2H,s, H-3,4). 1 3 C NMR (DMSO-d6): 6 58.553 (CH2); 123.607 (C-3,4); 145.246 (C-2,5). 2.5-Bis(chloromethyl)thiophene (395) 394 Thionyl chloride (2.1 mL, 28.7 mmol) was added to a stirred solution of 2,5-bis-(hydroxymethyl)thiophene (394, 1.35 g, 9.4 mmol) in dry CH 2 CI 2 (20 mL); rapid evolution of HCI observed. After stirring at room temperature for 24 h, the excess solvent and thionyl chloride were removed in vacou, furnishing the title compound 395 (1.51 g, 89%) as a brown oil which required no further purification. 264 1 H NMR (200 MHz, CDCI3): 5 4.78 (2H, thio.-CH2CI); 6.95 (2H, s, H-3,4). 1 3 C NMR (CDCI3): 5 40.326 (CH2); 127.360 (C-3,4); 141.685 (C-2,5). 2.5-Bis(phthalimidomethyl)thiophene (396) o o 395 A suspension of 2,5-bis(chloromethyl)thiophene 395 (1.44 g, 7.9 mmol) and potas-sium phthalimide (3.2 g, 17.3 mmol) in dry DMF (10 mL) was heated at 90 °C for 2 h, then left to stand at room temperature over-night. The mixture was diluted with water (10 mL) and extracted with CHCI 3 (3 x 30 mL). The combined organic solutions were washed with brine (15 mL), dried and the solvent removed on the rotary evaporator to provide the title compound as a tan solid (1.11 g, 35 %). m.p.: 200 °C (dec). 1 H NMR (200 MHz, CDCI3): 6 4.90 (4H, s, thio.-CH2N); 6.99 (2H, s, C-3,4); 7.7, 7.85 (8H, m, phthaloyl). 1 3 C NMR (CDCI3): 5 40.51 (CH2N); 128.099 (4 x phthaloyl aromatic C); 132.095 (C-3,4); 136.624 ((4 x phthaloyl aromatic C); 138.799 (4 x phthaloyl aromatic C); 143.301 (C-2,5); 172.5 (C=0). LRMS (El) m/e (rel. intensity): 402 (7.7, M+), 255 (91.2), 242 (100). 265 3.4-Dimethylpyrrole (118) H 118 Ethyl 3,4-dimethylpyrrole-2-carboxylate 396 1 4 9 (1.13 g, 6.7 mmol) and NaOH (453 mg, 11.4 mmol) in ethylene glycol (6 mL) were heated at 190 °C for 1 h. The reaction was allowed to cool to room temperature then partitioned between CH 2 CI 2 (20 mL) and water (10 mL). The aqueous phase was then extracted with CH 2 CI 2 (2 x 20 mL). The combined CH 2 CI 2 solutions were washed with brine (20 mL), dried, and the solvent removed in vacuo to give a dark brown oil. Distillation under reduced pressure (at the water pump, collecting the fraction boiling at 89 °C) then provided 3,4-dimethylpyrrole 118 as a colourless liquid which solidified upon standing; Yield: 396 mg, 62%. (NB, this material rapidly turns blue-black upon standing in air). b.p.: 89 °C at ca 50 mmHg. Lit. 1 4 4 164-166 °C at 760 mmHg. 1 H NMR (200 MHz, CDCI3): 5 1.95 (6H, s, CH 3); 6.39 (2H, s, H-2,5); 7.6 (1H, bs, NH). 1 3 C NMR (CDCI3): 9.867 (CH3); 115.426 (C-3,4); 118.07 (C-2,5). 266 2.5-Bisff5-(benzyloxycarbonyl)-3-ethyl-4-methylpyrrol-2-yl1methyl]-3.4-dimethylpyrole (397) The foregoing pyrrole 118 (333.7 mg, 3.5 mmol), benzyl-5-(acetoxymethyl)-4-ethyl-3-methylpyrrole-2-carboxylate 15 (2.15 g, 6.82 mmol) and p-TsOH (ca 10 mg) were dissolved in absolute EtOH (40 mL) and heated at 65 °C for 17 h. The resulting suspension was then reduced in volume to ca 20 mL, and cooled to afford a precipitate, which was collected and subsequently recrystallized from CH 2 CI 2 / EtOH to give a pale orange powder (399, 1.54 g, 72%). 1 H NMR (200 MHz, CDCI3): 8 0.95 (6H, t, C H 3 C H 2 , J = 7 Hz); 1.985 (6H, s, CH 3); 2.225 (6H, s, CH 3); 2.33 (4H, q, CH 2 CH 3 , J = 7 Hz); 3.52 (4H, s, pyrr.2-CH2); 4.4 (4H, s, OCH 2Ph); 6.98-7.05 (4H, m, Ph); 7.2-7.3 (6H, m, Ph); 8.7 (1H, bs, NH); 10.95 (2H, bs, NH). 1 3 C NMR (CDCI3): 6 9.654, 11.3, 16.042, 17.531, 22.485, 65.595, 117.542, 123.109, 123.4, 126.7, 127.065, 127.102, 127.576, 127.612, 128.44, 128.473, 133.219, 137.375, 162.942. 397 267 2.5-Bisf(3-ethyl-5-formyl-4-methylpyrrol-2-yl)methyl]-3.4-dimethy (399) A solution of 397 (4.38 g, 7.23 mmol) in dry THF (500 mL) containing triethylamine (1 drop) was hydrogenated over 10% Pd/C (250 mg) at atmospheric pressure and room temperature. After 24 h, the catalyst was filtered off and washed with warm CHCI 3. Removal of the solvent from the filtrate furnished a faint rose coloured solid (2.92 g), which was briefly (3 h) dried under high vacuum. The diacid was then dissolved in freshly distilled TFA (6 mL), whereupon evolution of C 0 2 immediatley became apparent. The reaction mixture was stirred at RT for ht 15 min then cooled to -20 °C, and freshly distilled triethyl orthoformate (6 mL, 36 mmol) was slowly added to the vigorously stirred solution. After stirring for 20 min at -20 °C, the mixture was allowed to reach RT, and stirred at this temperature for an additional 30 min. The dark solution was poured into an ice-water mixture (ca 100 mL), and the resulting precipitate was collected by filtration and washed well, with water. This solid product was suspended in a mixture of EtOH (60 mL) and 10% aqueous NH 4OH (350 mL) and the mixture stirred well for an hour. The resulting mixture was then extracted with CH 2 CI 2 (5 x 100 mL), and the organic solutions subsequently washed with water and saturated brine (100 mL of each), dried and finally concentrated in vacuo to a small volume. EtOH was added to this solution, and the volume reduced to ca 25 mL in vacuo. The dark red solution was then stored in the freezer for serai months. The product was collected at the water pump, washed with ice-cold ethanol to give dark brown solid (962 mg, 34%). 399 268 m.p.: 210-212 °C. 1 H NMR (200 MHz, CDCI3): 8 1.01 (6H, t, CH 3 CH 2 , J = 7.4 Hz); 1.97 (6H, s, CH 3); 2.19 (6H, s, CH 3); 2.40 (4H, q, CH 2 CH 3 , J = 7.4 Hz); 3.80 (4H, s, pyrr.2-CH2); 9.15 (2H, s, CHO); 9.65 (1H, bs, NH); 10.3 (2H, bs, NH). 1 3 C NMR (CDCI3) 5 8.725 (CH3); 9.3 (CH3); 15.168 (CH 3CH 2); 16.941 (CH 2CH 3); 22.65 (pyrr.2CH2); 114.241, 121.921, 124.594, 128.031, 138.159 (6 x pyrr. (3 and 6 x pyrr. a); 175.448 (CHO). Spectrocopic and physical data for the crystalline products isolated from initial; experiments: This material was obtained via the same procedure to that described above, with the following exceptions. After dissolving the diacid 398 in TFA, the reaction mixture was heated to reflux over 20 min, then allowed to cool to RT, and subsequently reheated to reflux over a 5 min period. The resulting dark mixture was then cooled to 0 °C, and treated as described above. After stirring in the presence of triethyl ortho formate (5 mL) for 20 min, the mixture was poured into ice-water (300 mL), and the resulting metallic, bottle green paste was extracted into CH 2CI 2 . The organic solution was concentrated in vacuo and the residue supended in a mixture of EtOH (25 mL), NH 4OH (50 mL), and water (50 mL) and stirred at RT for an hour. The tan precipitate was extracted into CH 2CI 2 . (5 x 100 mL), then washed with saturated brine (100 mL), dried and concentrated under reduced presuure to ca 5 mL. EtOH was added and the solution stored in the freezer for 2 days, to furnish a white powder (282 mg). m.p.: 201-203 °C. 269 1 H NMR (200 MHz, CDCI3): 8 1.055 (3H, t, C H 3 C H 2 , J = 7.6 Hz); 1.215 (3H, t, CH 3 CH 2 , J = 7.6 Hz); 2.05, 2.075 (6H, 2s, CH 3); 2.27 (6H, s, CH 3); 2.50 (1H, q, CH 2 CH 3 , J = 7.6 Hz); 2.73 (3H, q, C H 2 C H 3 , J = 7.6 Hz); 3.99 (4H, s, pyrr.-CH2); 9.5 (2H,s, CHO);10.79, 10.9 (2H, 2bs, NH); 11.5 (1H, bs, NH). LRMS (El) m/e (rel. intensity): 354 (33.6), 340 (100), 326 (55.2), 311 (40), 297 (11), 271 (10.4), 257 (112), 243 (17.6), 229 (12.8), 217 (35.2), 203 (44.8), 149 (31.7), 136 (33), 122 (20.8), 106 (16), 12.8 (94), 77 (12.8). 270 Chapter 5 Bibliography 271 22.0 References 1 See for instance: (a) The Porphyrins; Dolphin, D., Ed.; Academic Press: New York, 1978; Vols. 1-8. (b) Porphyrins and Metalloporphyrins; Smith, K. M., Ed.; Elselvier: Amsterdam, 1976. 2 For an overview of PDT see (a) Brown, S. B.; Truscott, T. G. Chem. Brit. 1993, 29, 955-958. (b) van den Bergh, H. Chem. Brit. 1986, 22, 430-439. (c) Gomer, C. J . Photchem. Photobiol.; 1987, 46, 561-562. (d) Marcus, S. L. Proceedings of the IEEE 1992, 80, 869-889. (e) Henderson, B.; Dougherty, T. J . , Eds. Photodynamic Therapy: Basic Principles and Clinical Applications; Marcel Dekker: New York, 1992. 3 For a review of MRI see: (a) Lauffer, R. B. Chem. Rev. 1987, 87, 901-927. See also for a brief clinical overview: (b) Bell, J . D.; Sadler, P. J . Chem. Brit. 1993,29,597-600. 4 (a) Grigg. R. J Chem. Soc, Chem. Commun. 1967, 1238-1239. (b) Callot, H. J. ; Tschamber, Th. Tetrahedron Lett. 1974, 3155-3158. (c) Callot, H. J . ; Tschamber, Th. J. Am. Chem. Soc. 1975, 97, 6175-6180. (d) Callot, H. J . ; Tschamber, Th. J. Org. Chem. 1977, 42,1567. (e) Callot, H. J . ; Schaeffer, E. Tetrahedron 1978, 34, 2295-2300. 5 Callot, H. J . ; Tschamber, Th.; Schaeffer, E. Tetrahedron Lett. 1975, 2919-2922. 6 Louati, A.; Schaeffer, E.; Callot, H. J . ; Gross, M. Nouv. J. Chim. 1979, 3, 191-194. 7 Sessler, J . L; Burrel, A. K. Top. Curr. Chem. 1991, 767,177-273. 8 Sessler, J . L; Hemmi, G.; Mody, T. D.; Murai, T.; Burrel, A. K.; Young, S.W. Acc. Chem. Res. 1994, 27, 43-50. 9 Vogel, E. Pure Appl. Chem. 1993, 65, 143-152. 10 (a) Vogel, E.; Kocher, M.; Schmickler, H.; Lex, J . Angew. Chem. Int. Ed. Eng. 1986, 25, 257-259. (b) Vogel, E.; Balci, M.; Pramod, K.; Koch, P.; Lex, J . ; Ermer, O. Angew. Chem. Int. Ed. Eng. 1987, 26, 928-931. (c) Vogel, E.; Kocher, M.; Balci, M.; Teichler, I.; Lex, J . ; Schmickler, H.; Ermer, O. Angew. Chem. Int. Ed. Eng. 1987, 26, 931-934. (d) Wehrle, B.; Limbach, H.-H.; Kocher, M.; Ermer, 0.;Vogel, E. Angew. Chem. Int. Ed. Eng. 1987, 26, 934-936. (e) Vogel, E.; Sicken, M.; Rohrig, P.; Schmickler, H.; Lex, J . ; Ermer, O. Angew. Chem. Int. Ed. Eng. 1988, 27, 411-414. (f) Vogel, E.; Grigat, I.; Kocher, M.; Lex, J . Angew. Chem. Int. Ed. Eng. 1989, 28, 1655-1657. (g) 272 Vogel, E. Pure Appl. Chem. 1990, 62, 557-564. (h) Waluk, J . ; Muller, M.; Swiderek, P.; Kocher, M.; Vogel, E.; Hohlneicher, G.; Michl, J . J. Am. Chem. Soc 1991, 7 73, 5511-5527. (i) Vogel, E.; Koch, P.; Hou, X.-L;l_ex, J . ; Lausmann, M.; Kisters, M.; Aukauloo, M. A.; Richard, P.; Guilard, R. Angew. Chem. Int. Ed. Eng., 1993, 32, 1600-1604. (j) De Munno, G.; Lucchesini, F.; Neidlein, R. Tetrahedron 1993, 49, 6863-6872. 11 (a) Franck, B.; Wegner, C. Angew. Chem. Int. Ed. Eng. 1975, 14, 424. (b) von Schnering, H. G.; Sawitzki, G. Angew. Chem. Int. Ed. Eng. 1976, 75, 552. (c) Franck, B. Angew. Chem. Int. Ed. Eng. 1982, 21, 343-353. (d) Vogel, E.; Haas, W.; Knipp, B.; Lex, J . ; Schmickler, H. Angew. Chem. Int. Ed. Eng. 1988, 27, 406-409. (e) Haas, W.; Knipp, B.; Sicken, M.; Lex, J . ; Vogel, E. Angew. Chem. Int. Ed. Eng. 1988, 27, 409-411. (f) Vogel E.; Rohrig, P.; Sicken, M.; Knipp, B.; Herrmann, A.; Pohl, M.; Schmickler, H.; Lex, J . Angew. Chem. Int. Ed. Eng. 1989, 28, 1651-1655. (g) Pohl, M.; Schmickler, H.; Lex, J . ; Vogel E. Angew. Chem. Int. Ed. Eng. 1991, 30, 1693-1697. (h) Bachmann, R.; Gerson, G.; Gescheidt, G.; Vogel E. J. Am. Chem. Soc. 1992,774, 10855-10860. (i) Vogel E.; Dorr, J . ; Herrmann, A.; Lex, J . ; Schmickler, H.; Walgenbach, P.; Gisselbrecht, J . P.; Gross, M. Angew. Chem. Int. Ed. Eng. 1993, 32, 1597-1600. 12 Will, S.; Rahbar, A.; Schmickler, H.; Lex, J . ; Vogel, E.Angew. Chem. Int. Ed. Eng. 1990, 29, 1390-1393. 13 Berlin. K.; Breitmaier, E. Angew. Chem. Int. Ed. Eng. 1994, 33, 1246-1247. 14 (a) Acholla, F. V.; Mertes, K. B. Tetrahedron Lett. 1984, 25, 3269-3270. (b) Acholla, F. V.; Takusagawa, F.; Mertes, K. B. J. Am. Chem. Soc. 1985, 707, 6902-6908. 15 Sessler, J . L.; Brucker, E. A.; Weghorn, S. J . ; Kisters, M.; Schafer, M.; Lex, J . ; Vogel, E. Angew. Chem. Int. Ed. Eng. 1994, 33, 2308-2312. 16 (a) Marks, T. J . ; Stojacovik, D. R. J Chem. Soc, Chem. Commun. 1975, 28. (b) Marks, T. J . ; Stojacovik, D. R. J. Am. Chem. Soc. 1978, 700, 1695-1705. (c) Marks, T. J . ; Cuellar, E. A. Inorg. Chem. 1981, 20, 3766-3770. 17 Day, V. W.; Marks, T. J . ; Wachter, W. A. J. Am. Chem. Soc. 1975, 97, 4519-4527. 18 Curtis, N. F. Coord. Chem. Rev. 1968, 3, 3-47. 19 (a) Cook, D. H.; Fenton, D. E.; Drew, M. G. B.; McFall, S. G.; Nelson, S. M. J. Chem. Soc, Dalton Trans. 1977, 446-449. (b) Drew, M. G. B.; Rodgers, A.; McCann, M.; Nelson, S. M. J Chem. Soc, Chem. Commun. 1978, 415-416. (c) Busch, D. H. Acc. Chem. Res. 1978, 7 7, 392-400. (d) McKee, V.; Smith J . 273 J Chem. Soc, Chem. Commun. 1983, 1465-1467. (e) McKee, V.; Shepard, W. B. J Chem. Soc, Chem. Commun. 1985, 158-159. (f) Adams, H.; Bailey, N. E.; Fenton, D. E.; Moss, S.; Rodriguez de Barbarin, C. O.; Jones, G. J . J. Chem. Soc, Dalton Trans. 1986, 693-699. (g) Chung, L.-Y.; Constable, E. C ; Khan, M. S.; Lewis, J . Inorg. Chim. Acta. 1991, 185, 93-96. (h) Sessler, J . L ; Mody, T. D.; Lynch, V. Inorg. Chem. 1992, 31, 529-531. 20 (a) Nelson, S. M. Pure App. Chem. 1980, 52, 2461-2476. (b) Nelson, S. M.; Knox, C. V.; McCann, M.; Drew, M. G. B. J. Chem. Soc, Dalton Trans. 1981, 1669-1677. (c) Drew, M. G. B.; Nelson, J . ; Nelson, S. M. J. Chem. Soc, Dalton Trans. 1981, 1678-1684. (d) Fenton, D. E. J. Chem. Soc, Dalton Trans. 1986, 58, 1437-1444. 21 Sessler, J . L; Johnson, M. R.; Lynch, V. J. Org. Chem. 1987, 52, 4394-4397. 22 (a) Barton, D. H. R.; Zard, S. Z. J Chem. Soc, Chem. Commun. 1985, 1098-1100. (b) Sessler, J . L.; Mozaffari, A.; Johnson, M. R. Org. Synth. 1991, 70, 68-77. 23 Johnson, A. W.; Kay, I. T.; Markham, E.; Price, P.; Shaw, K. B. J Chem. Soc. 1959, 3416-3424. 24 Clezy, P. S.; Liepa, A. J . Aust. J. Chem. 1972, 25, 1979-1990. 25 Sessler, J . L.; Mody, T. D.; Hemmi, G. W.; Lynch, V. Inorg. Chem. 1993, 32, 3175-3187. 26 Sessler, J . L.; Johnson, M. R.; Lynch, V.; Murai, T. J. Coord. Chem. 1988, 18, 99-104. 27 Maiya, B. G.; Mallouk, T. E.; Hemmi, G.; Sessler, J . L. Inorg. Chem. 1990, 29, 3738-3745. 28 Sessler, J . L.; Mody, T. D.; Ramasamy, R.; Sherry, A. D. New J. Chem. 1992, 16, 541-544. 29 (a) Sessler, J . L.; Mody, T. D.; Ford, D. A.; Lynch, V. Angew. Chem. Int. Ed. Eng. 1992, 31, 452-455. (b) Sessler, J . L; Mody, T. D.; Lynch, V. J. Am. Chem. Soc. 1993, 7 75, 3346-3347. 30 Mauzerall, D. In ref 1a, Vol II, Chapter3. 31 Sessler, J . L.; Cyr, M.; Furuta, H.; Krai, V.; Mody, T.; Morishima, T.; Shionoya, M.; Weghorn, S. Pure App. Chem. 1993, 65, 393-398. 274 32 Sessler, J . L; Murai, T ; Lynch, V.; Cyr, M. J. Am. Chem. Soc. 1988, 110, 5586-5588. 33 Sessler, J . L.; Murai, T.; Lynch, V. Inorg. Chem. 1989, 28, 1333-1341. 34 Whitlock, H. W., Jr.; Buchanan, D. H. Tetrahedron Lett. 1969, 42, 3711-3714. 35 (a) Janson, T. R.; Katz, J . J . ; in ref. 1a. Vol IV, Chapter 1. (b) Scheer, H.; Katz, J . J . ; in ref 1b, Chapter 10. 36 (a) Gouterman, M., in ref. 1a. Vol III, Chapter 1. (b) Becker, R. S.; Allison, J . B. J. Phys. Chem. 1963, 67, 2669-2675. 37 Sessler, J . L.; Murai, T.; Hemmi, G. Inorg. Chem. 1989, 28, 3390-3393. 38 Sessler, J . L.; Mody, T. D.; Hemmi, G. W.; Lynch, V. Inorg. Chem. 1993, 32, 3175-3187. 39 See reference 8, and references cited therein. 40 Hoard, J . L., in ref. 1 b, Chapter 8. 41 Kennedy, M. A.; Sessler, J . L.; Murai, T.; Ellis, P. D. Inorg. Chem. 1990, 29, 1050-1054. 42 (a) Buchler, J . W.; De Cian, A.; Fischer, J . ; Kihn-Botulinski, M.; Paulus, H.; Weiss, R. J. Am. Chem. Soc. 1986, 108, 3652-3659. (b) Buchler, J . W.; Scharbert, B. J . J. Am. Chem. Soc. 1988, 110, 4272-4276. (c) Buchler, J . W.; De Cian, A.; Fischer, J . ; Kihn-Botulinski, M.; Weiss, R. Inorg. Chem. 1988, 27, 339-345. (d) Buchler, J . W.; Kihn-Botulinski, M.; Loftier, J . ; Wicholas, M. Inorg. Chem. 1989, 28, 3770-3772. (e) Schaverien, C. J . ; Orpen, A. G. Inorg. Chem. 1991, 30, 4968-4978. (f) Buchler, J . W.; Loftier, J . ; Wicholas, M. Inorg. Chem. 1992, 31, 524-526 and references cited therein. 43 Shannon, R. D. Acta Crystallogr. 1976, A32, 751-767. 44 Liksowski. J . ; Sessler, J . L.; Lynch, V.; Mody, T. D. J. Am. Chem. Soc. 1995, 7 77,2273-2285. 45 For a brief overview see : Sessler, J . L. Angew. Chem. Int. Ed. Eng. 1994, 33, 1348-1350. See also, references 9 and 10 cited therein. 46 Sondheimer, F. Acc. Chem. Res. 1972, 5, 81-91. 47 (a)Garratt, P. J . , Aromaticity, Wiley, New York 1986, pp 52-55. (b) Chapter 4 in reference 47a. 275 48 (a) Sondhiemer, F. Pure App. Chem. 1963, 7, 363-388 and references cited therein, (b) Fukui, K.; Nomoto, T.; Nakatsuji, S.; Nakagawa, Tetrahedron Lett. 1972, 3157-3160. (c) lyoda, M.; Nakagawa, M. Tetrahedron Lett. 1972, 3161-3164. (d) Nakagawa, M. Pure App. Chem. 1975, 44, 885-924. (e) M. Kabuto, C ; Kitahara, Y.; lyoda, M.; Nakagawa, M. Tetrahedron Lett. 1976, 2787-2790, 2791-2794. (f) Nakagawa, M. Angew. Chem. Int. Ed. Eng. 1979, 18, 202-214. 49 (a) McMurry, J . E. Chem. Rev. 1989, 89, 1513-1524. (b) Lenoir, D. Synthesis 1989, 883-897. 50 (a) Kato, N.; Nakanishi, K.; Takeshita, H. Bull. Chem. Soc. Jpn. 1986, 59, 1109-1123. (b) Brudermuller, M.; Musso, H. Angew. Chem. Int. Ed. Eng. 1988, 27, 298-299. (c) McMurry, J . E.; Rico, J . G.; Shih, Y. Tetrahedron Lett. 1989, 30, 1173-1176. 51 Jux, N.; Koch, P.; Schmickler, H.; Lex, J . ; Vogel, E. Angew. Chem. Int. Ed. Eng. 1990, 29, 1385-1387. 52 Martire, D. Gv; Jux, N.; Aramendfa, P. F.; Negri, R. M.; Lex, J . ; Braslavsky, S. E.; Schaffner, K.; Vogel, E. J. Am. Chem. Soc. 1992, 114, 9969-9978. 53 Emsley, J . Chem. Soc. Rev. 1980, 9, 91-124. 54 Hu, Z.; Cava, M. P. Tetrahedron Lett. 1994, 35, 3493-3496. 55 Hu, Z.; Atwood, J . L; Cava, M. P. J. Org. Chem. 1994, 59, 8071-8075. 56 Armiger, Y. L. S-T.; Lash, T. D. J. Heterocycl. Chem., 1992, 29, 523-527. 57 Kawase, T.; Enomoto, T.; Wei, C ; Oda, M. Tetrahedron Lett. 1993, 34, 8143-8146. 58 Markl, G.; Striebl, U. Angew. Chem. Int. Ed. Eng. 1993, 32, 1333-1335. 59 Mullen, K. Chem. Rev. 1984, 84, 603-646. 60 Vogel, E.; Jux, N.; Rodriguez-Val, E.; Lex, J . ; Schmickler, H. Angew. Chem. Int. Ed. Eng. 1990, 29, 1387-1390. 61 Markl, G.; Sauer, H.; Kreitmeier, P.; Burgemeister, T.; Kastner, F.; Adolin, G.; Noth, H.; Polborn, K. Angew. Chem. Int. Ed. Eng. 1994, 33, 1151-1153. 62 Johnson, M. R.; Miller. D. C ; Bush, K.; Becker, J . J . ; Ibers, J . A. J. Org. Chem. 1992, 57, 4414-4417. 276 63 Hu, Z.; Scordilis-Kelley, C ; Cava, M. P. Tetrahedron Lett. 1993, 34, 1879-1882. 64 Miller. D. C ; Johnson, M. R.; Becker, J . J . ; Ibers, J . A. J. Heterocyclic Chem. 1993, 30, 1485-1490. 65 Miller. D. C ; Johnson, M. R.; Ibers, J . A. J. Org. Chem. 1994, 59, 2877-2879. 66 Merrill, B. A.; LeGoff, E. J. Org. Chem. 1990, 55, 2904-2908. 67 (a) Stetter, H. Angew. Chem. Int. Ed. Eng. 1976, 15, 639-712. (b) Stetter, H.; Bender, H.-J. Angew. Chem. Int. Ed. Eng. 1978, 17, 131. 68 Baum, L. F. The Wonderful Wizard of Oz, G. M. Hill Co., Chicago, 1900 69 (a) Latos-Grazynsky, L ; Lisowsky, J . ; Olmstead, M. M.; Balch, A. L. J. Am. Chem. Soc. 1987, 109, 4428-4429. (b) Broadhurst, M. J . ; Grigg, R. ; Johnson, A. W. J. Chem. Soc.(C) 1971, 3681-3690. (c) Ulmann, A.; Manassen, J . J. Am. Chem. Soc. 1975, 97, 6540-6544. (d) Latos-Grazynsky, L.; Lisowsky, J . ; Szterenberg, L.; Olmstead, M. M.; Balch, A. L. J. Org. Chem. 1991, 56, 4043-4045. 70 Sessler, J . L.; Weghorn, S. J . ; Lynch, V. J . ; Johnson, M. R. Angew. Chem. Int. Ed. Eng. 1994, 33, 1509-1512. 71 Berger, R. A.; LeGoff, E. Tetrahedron Lett. 1978, 4225-4228. 72 Corriu, R. J . P.; Bolin, G.; Moreau, J . J . E.; Vernhet, C. J Chem. Soc, Chem. Commun. 1991, 211 -213. 73 Carre, F. H.; Corriu, R. J . P.; Bolin, G.; Moreau, J . J . E.; Vernhet, C. Organometallics, 1993, 12, 2478-2486; and references 1-4 cited therein. 74 Berlin. K.; Breitmaier, E. Angew. Chem. Int. Ed. Eng. 1994, 33, 219-220. 75 (a) Takenaka, A.; Sasada, Y.; Omura, T.; Ogoshi, H.; Yoshida, Z.; I. J Chem. Soc, Chem. Commun. 1973,792-793. (b) Takenaka, A.; Sasada, Y.; Ogoshi, H.; Omura, T.; Yoshida, Z.; I. Acta Crystallogr. 1975, B31, 1-6. 76 LeGoff, E.; Weaver. O. G. J. Org. Chem. 1987, 52, 710-711. 77 Haddon, R. C. J. Am. Chem. Soc. 1979, 101, 1722-1728. 78 Gosmann. M.; Vogt, A.; Franck, B. Liebigs Ann. Chem. 1990, 163-168. 277 79 Arsenault, G. P.; Bullock, E.; MacDonald, S. F. J. Am. Chem. Soc. 1960, 82, 4384-4389. 80 Beckmann, S.; Wessel, T.; Franck, B.; Honle, W.; Bormann, H. Angew. Chem. Int. Ed. Eng. 1990, 29, 1395-1397. 81 Konig, H.; Eickmeier, C ; Moller, M.; Rodewald, U.; Franck, B. Angew. Chem. Int. Ed. Eng.1990, 29, 1393-1395. 82 Wessel, T.; Franck, B.; Moller, M.; Rodewald, U.; Lage, M. Angew. Chem. Int. Ed. Eng. 1993, 32, 1148-1151. 83 Gosmann. M.; Franck, B. Angew. Chem. Int. Ed. Eng. 1986, 25, 1100-1101. 84 Franck, B.; Nonn, A.; Fuchs, K.; Gosmann, M. Liebigs Ann. Chem. 1994, 503-510. 85 Schermann, G.; Schmidt, R.; Volcker, A.; Brauer, H.-D.; Mertes, H.; Franck, B.; Photchem. Photobiol.; 1990, 52, 741-744. 86 Kniibel, G.; Franck, B. Angew. Chem. Int. Ed. Eng.1988, 27, 1170-1172. 87 Dewar, M. J . S., Gleicher, G. J . J. Am. Chem. Soc. 1965, 87, 685-692, 692-696. 88 Dolphin, D.; Rettig, S. J . ; Tang, H.; Wijesekera, T.; Xie, L. Y. J. Am. Chem. Soc. 1993, 7 75, 9301-9302. 89 Dolphin, D.; Xie, L. Y. J. Chem. Soc, Chem. Commun. 1994, 1475-1476. 90 Dolphin, D.; Xie, L. Y. unpublished results. 91 Boyle, R. W.; Xie, L. Y.; Dolphin, D. Tetrahedron Lett. 1994, 35, 5377-5380. 92 First reported by R. B. Woodward, in Aromaticity: An International Symposium, Sheffield, U.K., 1966, special publication No. 21, The Chemical Society London, 1966. (See ref. 97) 93 King, M. M. Ph.D. Dissertation, Harvard University, Cambridge, MA, 1970. 94 Broadhurst, M. J . ; Grigg, R. J Chem. Soc, Chem. Commun. 1969, 23-24. 95 Broadhurst, M. J . ; Grigg, R.; Johnson, A. W. J. Chem. Soc, Perkin Trans 1, 1972,1124-1135. 278 96 (a) Grigg, R. In ref. 1a. Vol. II. Ch 10.; (b) Johnson, A. W. In ref. 1b. pp 750-751. 97 Broadhurst, M. J . ; Grigg, R.; Johnson, A. W. J. Chem. Soc, Perkin Trans 1 1972,2111-2116. 98 Bauer, V. J . ; Clive, D. L. J . ; Dolphin, D.; Paine, J . B.; Harris, F. L; King, M. M.; Loder, J . ; Wang, S.-W. C. Woodward, R. B. J. Am. Chem. Soc. 1983, 105, 6429-6436. 99 Sessler, J . L ; Cyr, M. J . ; Lynch, V.; McGhee, E.; Ibers, J . A. J. Am. Chem. Soc. 1990, 7 72, 2810-2813. 100 Sessler, J . L; Cyr, M. J . ; Burrell, A. K. SynLett 1991, 127-134. 101 Sessler, J . L; Cyr, M. J . ; Burrell, A. K. Tetrahedron 1992, 48, 9661-9672. 102 Sessler, J . L; Brucker, E. A. Tetrahedron Lett. 1995, 36, 1175-1176. 103 Broadhurst, M. J . ; Grigg, R. J Chem. Soc, Chem. Commun. 1970, 807-809. 104 Maiya, B. G.; Cyr, M.;Harriman, A.; Sessler, J . L. J. Phys. Chem. 1990, 94, 3597-3601. 105 Shionoya, M.; Furuta, H.; Lynch, V.; Harriman, A.; Sessler, J . L. J. Am. Chem. Soc. 1992, 7 74, 5714-5722. 106 Levanon, H.; Regev, A.; Michaeli, S.; Galili, T.;Cyr, M.; Sessler, J . L. Chem. Phys. Lett. 1990, 774, 235-240. 107 Regev, A.; Michaeli, S.; Levanon, H.; Cyr, M.; Sessler, J . L. J. Phys. Chem. 1991, 95, 9121-9129. 108 Furuta, H.; Cyr, M. J . ; Sessler, J . L. J. Am. Chem. Soc. 1991, 7 73, 6677-6678. 109 Burrell, A. K.; Sessler, J . L.; Cyr, M. J . ; McGhee, E.; Ibers, J . A. Angew. Chem. Int. Ed. Eng. 1991, 30, 91-93. 110 (a) Abeysekera, A. M.; Grigg, R.; Trocha-Grimshaw, J . ; Viswanatha, V. J. Chem. Soc, Perkin Trans 11977, 36-44. (b) Abeysekera, A. M.; Grigg, R.; Trocha-Grimshaw, J . ; Viswanatha, V. J. Chem. Soc, Perkin Trans 11977, 1395-1403. 111 Abeysekera, A. M.; Grigg, R.; Trocha-Grimshaw, J . ; Viswanatha, V.; King, T. J . J. Chem. Soc, Perkin Trans 11979, 2184-2192. 279 112 Burrell, A. K.; Cyr, M. J . ; Lynch, V.; Sessler, J . L. J Chem. Soc, Chem. Commun. 1991, 1710-1713. 113 (a) Calano, M. M.; Crossley, M. J . ; King, L. G. J Chem. Soc, Chem. Commun. 1984, 1537-1538; (b) Gong, L.-C; Dolphin, D. Can. J. Chem. 1985, 63, 406-411. (c) Crossley, M. J . ; King, L. G.; Pyke, S. M. Tetrahedron 1987, 43, 4569-4577. 114 Krai, V.; Sessler, J . L; Furuta, H. J. Am. Chem. Soc 1992, 114, 8704-8705. 115 Iverson, B. L.; Shreder, K.; Sessler, J . L. J. Am. Chem. Soc. 1993, 115, 11022-11023. 116 Iverson, B. L.; Thomas, R. E.; Krai, V.; Sessler, J . L. J. Am. Chem. Soc. 1994, 116, 2663-2664. 117 Sessler, J . L.; Ford, D. A.; Cyr, M. J . ; Furuta. H.J Chem. Soc, Chem. Commun. 1991, 1733-1735. 118 Chmielewski, P. J . ; Latos-Grazynski, L.; Rachlewicz, K.; Glowiak, T. Angew. Chem. Int. Ed. Eng. 1994, 33, 779-781. 119 Furuta, H.; Asanp, H,; Ogawa, T. J. Am. Chem. Soc. 1994, 116, 767-768. 120 Chmielewski, P. J . ; Latos-Grazynski, L.; Rachlewicz, K.; Chem. Eur. J. 1995, 1, 68-73. 121 (a) Rothemund, P. J. Am. Chem. Soc. 1936, 58, 625-627. (b) Rothemund, P.; Menotti, A. R.J. Am. Chem. Soc. 1941, 63, 267-270. (c) Treibs, A.; Haberle, N. Liebigs Ann. Chem. 1968, 718, 183-207. (d) Doplhin, D. J. Heterocyclic Chem. 1970, 7, 275-283. (e) Lindsey, J . S.; MacCrum, K. A.; Tyhonas, J . S.; Chuang, Y.-Y. J. Org. Chem. 1994, 59, 579-587 and references cited therein. 122 Broadhurst, M. J . ; Grigg, R. J. Org. Chem. 1969, 1480-1482. 123 Sessler, J . L; Weghorn, S. J . ; Hiseada, Y.; Lynch, V. Chem. Eur. J. 1995, 7, 56-67. 124 Rexhausen. H.; Gossauer, A. J Chem. Soc, Chem. Commun. 1983, 275. 125 Gossauer, A. Bull. Soc. Chim. Belg. 1983, 92, 793-795. 126 Gossauer, A. Chimia 1984, 37, 341-342. 127 Gossauer, A. Chimia 1984, 38, 45-46. 280 128 Burrell, A. K.; Hemmi, G.; Lynch, V.; Sessler, J . L. J. Am. Chem. Soc. 1991, 113, 4690-4692. 129 Schumacher, K.-H.; Franck, B. Angew. Chem. Int. Ed. Eng. 1989, 28, 1243-1245. 130 Sessler, J . L.; Weghorn, S. J . ; Morishima, T.; Rosingana, M.; Lynch, V.; Lee, V. J. Am. Chem. Soc. 1992, 114, 8306-8307. 131 Barkigia, K. M.; Berber, M. D.; Fajer, J . ; Medforth, C. J . ; Renner, M. W.; Smith, K. M. J. Am. Chem. Soc. 1990, 112, 8851-8857. 132 Sessler, J . L.; Morishima, T.; Lynch, V. Angew. Chem. Int. Ed. Eng. 1991, 30,977-980. 133 Charriere, R.; Jenny, T. A.; Rexhausen, H.; Gossauer, A. Heterocycles 1993, 36,1561-1575. 134 Bell, T. W.; Cragg, P.J.; Drew, M. G. B.; Firestone, A.; Kwok, A. D.-l.; Liu, J . ; Ludwig, R. T.; Papoulis, A. T. Pure App. Chem. 1993, 65, 361-366, and references cited therein. 135 Newkome, G. R.; J , Y. J . ; Evans, D. W.; Pappalardo, S.; Fronczek, F. R. J. Org. Chem. 1988, 53, 786-790. 136 Biosym Insight II, San Diego, CA. using a combination of Steepest Descent, Conjugate Gradient, and VA09A algorithims. 137 Paine, J . B., III.; Woodward, R. B.; Dolphin, D. J. Org. Chem. 1976, 41, 2826-2835. 138 See Paine, J . B., III. in ref. 1 a, Vol. I. Chapter 4. 139 Johnson, A. W.; Markham, E.; Price, R.; Shaw, K. B. J. Chem. Soc. 1958, 4254-4257. 140 Clezy, P.S.; Liepa, A. J . Aust. J. Chem. 1970, 23, 2443-2459. 141 Chong, R.; Clezy, P.S.; Liepa, A. J . ; Nichol, A. W. Aust. J. Chem. 1969, 22, 229-230. 142 Bringmann, G.; Scheider, S. Synthesis 1983,139-141. 143 Campagna, F.; Carotti, A.; Cosini, G. Tetrahedron Lett. 1977, 21, 1813-1816. 281 144 Badger, G. M.; Ward, A. D. Aust. J. Chem. 1964, 77, 649-660. 145 Pichon, C.; Clemens, K. R.; Jacobson, A. R.; Scott, A. I. Tetrahedron 1992, 23, 4687-4712. 146 Grigg, R.; Johnson, A. W.; Wasley, J . W. F. J. Chem. Soc. 1963, 359-366. 147 Friedman, L ; Shecter, H. J. Org. Chem. 1961 26, 2522-2524. 148 Lash, T. D.; Cataterello, J . J . Tetrahedron 1993, 49, 4159-4172. 149 Paine, J . B., III. Ph.D. Dissertation, Harvard University, Cambridge, MA, 1973 150 Fischer, H.; Halbig, P.; Walach, B. Justus Liebigs Ann. Chem. 1927, 452, 268. 151 (a) Lohaus, G. Chem. Ber. 1967, 100, 2719-2729. (b) Vorbruggen, H. Tetrahedron Lett. 1968, 13, 1631-1634. (c) Mehta, G.; Dhar, D. N.; Suri, S. C. Synthesis 1978, 374-376. (d) Barnett, G. H.; Anderson, H. J . ; Loader, C. E. Can. J. Che/77.1980, 58, 409-411. (e) Loader, C. E.; Anderson, H. J . Can. J. Chem. 1981, 59, 2673-2676. 152 Boyle, R. W., Karunaratne, V.; Jasat, A.; Mar, E. K.; Dolphin. D. Synlett, 1994, 939-940. 153 Cookson, G. H.; Rimmington, C. Biochem. J. 1954, 57, 476-484. 154 Mar, E. K.; Dolphin. D. unpublished results. 155 For a review of this methodology, see Malek, J . Org. React. 1988, 367, 249-590. 156 Fuhrhop, J.-H.; Smith, K. M. in ref. 1b, p 760. 157 (a) Vigmond, S. J . ; Chang, M. C ; Kallury, K. M. R.; Thompson, M. Tetrahedron Lett. 1994, 35, 2445-2458. (b) Lee, C.-H.; Lindsey, J . S. Tetrahedron 1994, 39, 11427-11440. 158 Brown, W. H.; Sawatzky, H. Can. J. Chem. 1956, 34, 1147-1153. 159 Dinelli, D.; Marini, G. B. Gazz. Chim. Ital. 1937, 67, 312-317. Chem. Abst. 1938, 32, 154-155. 160 Clezy, P. S.; Parnis, C. Aust. J. Chem. 1972, 25, 2055-2060. 161 Brown, H. C ; Rao, C. G.; Kulkarni, S. U. Synthesis 1979, 704-705. 282 162 Cha, J . S.; Kim, J . E.; Yoon, M. S.; Kim, Y. S. Tetrahedron Lett. 1987, 28, 6231-6234. 163 Cha, J . S.; Kwon, S. S. J. Org. Chem. 1987, 52, 5486-5487. 164 Rickborn, B.; Jensen, F. R. J. Org. Chem. 1962, 27, 4608-4610. 165 (a) Kanaoka, Y.; Kuga, T.; Tanizawa, K. Chem. Pharm. Bull. 1970, 18, 397-399. (b) Imamoto, T.; Takaoka, T.; Yokoyama, M. Synthesis 1983, 142-143. 166 Vorbruggen, H.; Krolikiewicz, K. Tetrahedron 1994, 50, 6549-6558, and references cited therein. 167 Vogels Textbook of Practical Organic Chemistry; John Wiley & Sons Inc.; New York: 5th ed. 1989, p.1276. 168 Cymerman-Craig, J . ; Loder J . W. J. Chem. Soc. 1954, 237-242. 169 The Furans, Dunlop, A. P.; Peters, F. N.; Reinhold Publishing Corporation: New York, 1953. 170 Dinelli, D.; Marini, G. B. Gazz. Chim. Ital. 1941, 71, 117-128. Chem. Abst. 1942, 36, 1928-1930. 171 Bonnaud, B.; Bigg, D. C. H. Synthesis, 1994, 465-467. 172 (a) Jackson, A. H.; Kenner, G. W.; Smith, K. M. J. Chem. Soc, Chem. C 1968, 294-302. (b) Clezy, P. S.; Looney, F. D.; Nichol, A. W.; Smythe, G. A. Aust. J. Chem. 1966, 19, 1481-1486. 173 (a) Clezy, P. S.; Diakiw, V. Aust. J. Chem. 1971, 24, 2665-2677. (b) Lash, T. D.; Motta, Y. G. Heterocycles 1983, 20, 2343-2346. (c) Lash, T. D.; Armiger, S.-T., Y. L. J. Heterocycl. Chem., 1991, 28, 965-970. (d) Lash, T. D. J. Org. Chem. 1992, 57, 4312-4315. 174 Clezy, P. S. in ref. 1a, Vol II, pp 103-130. 175 Garrat, P. J . , Aromaticity, Wiley, New York 1986, Chapter 7. 176 Cresp, T. M.; Sargent, M. V. J. Chem. Soc, Perkin Trans 1,1973, 2961-2971. 177 Miyashita, M.; Yoshikoshi, A.; Grieco, P. A. J. Org. Chem. 1977, 42, 3772-3774. 178 Bosshard, P.; Eugester, C. H. Advances in Heterocyclic Chemistry 1966, 7, 436-442. 283 179 Hinz, A.; Meijer, G.; Schucking, G. Ber. Dtsch. Chem. Ges. 1943, 76, 676. 180 Feringa, B. L ; Hulst, R.; Rikers, R.; Brandsma, L. Synthesis, 1988, 316-318. 181 Bosshard, P.; Eugester, C. H. Advances in Heterocyclic Chemistry 1966, 7, pp 413-421. 182 Ballatine, J . A.; Jackson, A. H.; Kenner, G. W.; McGillivray, G. Tetrahedron 1966, 22, Suppl. 7, 241-259. 183 Sterzycki, R. Synthesis 1979, 724-725. 184 Bremner, J . G. M.; Jones, D. G. Chem. Abst. 1947, 41, 6581. 185 To the best of our knowledge, this represents the first such example of an N,2'-bipyrrly isolated from the Ullmann coupling of a 5-iodopyrrole. However, N.N'-bipyrroles are known, but in these instances, the products were obtained via two successive cyclo-condensations; for example, see Chang, C ; Adams, R. J. Am. Chem. Soc. 1931, 53, 2353-2357. 186 A recent short paper by Sessler and Hoehner, which presents a more efficient modification of the basic Ulmann reaction (See: Sessler, J . L; Hoehner, M. C. Synlett 1994, 211-212), does briefly mention the possible formation of these N,2'-coupled bipyrrolic side products. 187 Vogels Textbook of Practical Organic Chemistry; John Wiley & Sons Inc.; New York: 5th ed, 1989, p. 426. 188 Webb, J . L. A.; Threlkeld, R. R. J. Org. Chem. 1953, 18, 1406-1412. 189 Itahara, T.; Hasimoto, M.; Yumisashi, H. Synthesis 1984, 255-256. 190 Grigg, R.; Knight, J . A., Sargent, M. V. J. Chem. Soc, Chem. (C) 1966, 976-981. 191 Chadwick, D. J . ; Chambers, J . ; Meakins, G. D.; Snowden, R. L. J. Chem. Soc, Perkin Trans 1,1973, 1766-1773. 192 Nazarova, Z. N. Chem. Abst. 1955, 49, 10261g (Zh. Obshch. Khim., 1954, 24, 575.). 193 Nazarova, Z. N. Chem. Abst. 1956, 50, 10261g (Zh. Obshch. Khim., 1955, 25, 509-513; J. Gen. Chem. U.S.S.R. 1955, 25, 509-513.). 194 Danso-Danequah, R. E.; Scott, A. I.; Becker, D. Tetrahedron 1993, 49, 8195-8210. 195 Gilman, H.; Hewlett, A. P. Iowa State College J. of Science 1929, 4, 27-33. 284 196 Hinz, W.; Jones, R. A.; Patel, S. U.; Karatza, M.-H. Tetrahedron 1986, 42, 3753-3758. 197 El-Hajj, T.; Martin, J . - C ; Descqtes, G. J. Heterocyclic Chem. 1983, 20, 233-235. 198 Finan, P. A. J. Chem. Soc. 1963, 3917-3918. 199 Garrigues, B. Phosphorus, Sulphur, and Silicon 1990, 53, 75-79. 200 Divald, S.; Chun, M. C ; Jouille, M. M. J. Org. Chem. 1976, 17, 2835-2846. 201 Katritzky, A. R.; Li, J . ; Gordeev, M. F. J. Org. Chem. 1993, 58, 3038-3041. 202 Jackson, A. H.; Kenner, G. W.; Warburton, D. J. Chem. Soc. 1965, 1328-1337. 203 Holdren, R. F.; Barry, W. T. J. Am. Chem. Soc. 1947, 69, 1230-1231. 204 Aldrich Catalogue/Handbook, 1992-93, 592. 205 The Furans, Dunlop, A. P.; Peters, F. N.; Reinhold Publishing Corporation: New York, 1953, p. 345. 206 Die Chemie des Pyrrols; Fischer, H.; Orth, H.; Vol. I, 1968, p. 241. 285 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items