UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Synthesis and photochemistry of AZO-type derivitized porphyrins with crosslinking applications Desjardins, Angèla M. 2000

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

Item Metadata

Download

Media
831-ubc_2000-0647.pdf [ 5.24MB ]
Metadata
JSON: 831-1.0061435.json
JSON-LD: 831-1.0061435-ld.json
RDF/XML (Pretty): 831-1.0061435-rdf.xml
RDF/JSON: 831-1.0061435-rdf.json
Turtle: 831-1.0061435-turtle.txt
N-Triples: 831-1.0061435-rdf-ntriples.txt
Original Record: 831-1.0061435-source.json
Full Text
831-1.0061435-fulltext.txt
Citation
831-1.0061435.ris

Full Text

SYNTHESIS AND PHOTOCHEMISTRY OF AZO-TYPE DERIVITIZED PORPHYRINS WITH CROSSLINKING APPLICATIONS by A N G E L A M . DESJARDINS B.Sc, University of Western Ontario, 1998 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES DEPARTMENT OF CHEMISTRY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A September 2000 © Angela M . Desjardins, 2000 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that 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 for reference and study. I further agree that permission for extensive copying of t h i s thesis for s c h o l a r l y purposes may be granted by the head of my department or by h i s or her representatives. I t i s understood that copying or p u b l i c a t i o n of t h i s t h e s i s for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of Ctt6rA\STflJf The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada ABSTRACT The objective of this work was to synthetically modify porphyrins such that they would possess a functionality that would crosslink within its biological environment upon photoactivation, in photodynamic therapy applications. The reaction of diazomethane and protoporphyrin dimethylester (9) yielded a series of three novel pyrazoline porphyrins (10 - 12) that were isolated and characterized. These were investigated for their ability to extrude molecular nitrogen on thermal and photochemical activation with both near-UV and long wavelength light to yield a 1,3-biradical intermediate that may provide a site for potential crosslinking. Upon activation three cyclopropane derivatives (15 - 17) were formed in high yield, presumably through the 1,3 biradical intermediate. These novel cyclopropane porphyrin derivatives were purified and characterized. The optimized diazomethane reaction was then carried out on a chlorin (18), a photosensitizer currently undergoing clinical trials for PDT, to yield the first example of a pyrazoline derivatized methylpyropheophorbide (MePPP, 19). Activation with heat, U V ii light or long wavelength light yielded the cyclopropane product of MePPP (20), analogous to the photoreaction described for the model PPDME pyrazoline reactions. These studies represent the first example of long wavelength activation of a photoactive functionality at the porphyrin periphery. Attempts to trap the biradical intermediates produced on photoactivation of the pyrazolines were carried out in solution. Unfortunately, no trapped products were isolated, presumably because the lifetimes of these intermediates proved to be too short due to efficient formation of the cyclopropane. The second photoactive functionality that was pursued is anticipated to yield a much longer lived carbene intermediate that should be more effective in cross-linking studies. The synthetic target was the diazirine of TPP (25). Five of the eight reactions planned in the synthesis of (25) were optimized, and scaled-up. Each of the intermediates (21, 22, 23, 24) were isolated and characterized. The reactions leading up to the novel trifluoroacetyl-TPP (23) provide the first example of introducing a trifluoroacetyl group at the (3 position of the porphyrin. 23 in TABLE OF CONTENTS Abstract i i Table of Contents iv List of Figures vi List of Schemes viii List of Tables ix Acknowledgements x 1 Introduction 1.1 Tetrapyrrolic Macrocycles 1 1.1.1 Background 1 1.1.2 Nomenclature 3 1.1.3 Structural Characteristics 4 1.1.4 Optical Absorption Spectra 6 1.1.5 Preparation of Porphyrins 9 1.2 Photodynamic Therapy 12 1.2.1 Mechanism of Photosensitization 13 1.2.2 Type I Reactions 16 1.2.3 Type II Reactions 17 1.3 Photoaffinity Labeling 18 1.4 Research Objective 24 iv 2 Results and Discussion 2.1 Synthesis and Charaterization of Protoporphyrin Pyrazolines 28 2.2 Thermal and Light Activated Chemistry of Protoporphyrin Pyrazolines.. 47 2.2.1 Thermal Activation 48 2.2.2 Photochemical Activation 51 2.3 Synthesis and Characterization of Methylpyropheophorbide Pyrazoline.. 58 2.4 Photochemistry of Methylpyropheophorbide Pyrazoline 64 2.5 Progress Towards the Synthesis of Diazirine Tetraphenylporphyrin 81 3 Conclusions and Suggestions for Further Studies 96 4 Experimental 4.1 Instrumentation and Materials 99 4.2 Photochemical Studies-General Procedures 103 4.3 Preparation of Protoporphyrin Pyrazolines 10,11, and 12 104 4.3.1 Photoproducts of Protoporphyrin Pyrazolines 15,16, and 17 109 4.4 Preparation of Methypyropheophorbide Pyrazoline 19 113 4.4.1 Photoproduct of Methylpyropheophorbide Pyrazoline 20 116 4.5 Progress Towards the Synthesis of Diazirine Tetraphenylporphyrin 25... 117 Appendix: Crystallographic Analysis of 17 125 References 131 V LIST OF FIGURES Figure 1.1: Important tetrapyrrolic backbones 1 Figure 1.2: Protoporphyrin IX 3 Figure 1.3: Porphyrin numbering and nomenclature 3 Figure 1.4: The 187t-electron pathway 4 Figure 1.5: Q-band porphyrin spectra for metal-free porphyrins 7 Figure 1.6: Wavelength dependent penetration of light through tissue 13 Figure 1.7: Modified Jablonski diagram for a typical photosensitizer 15 Figure 1.8: Schematic of photoaffinity labeling 19 Figure 1.9: Photoaffinity taxol analogue 21 Figure 1.10: Photoactive functionalities 26 Figure 2.1: BPD-diacid 28 Figure 2.2: 1,3-Dipolar cycloaddition 31 Figure 2.3: Estimated energy of frontier n orbitals for conjugated vinyl group and diazomethane 32 Figure 2.4: Orbital coefficients for diazomethane 32 Figure 2.5: 'H NMR of B-ring pyrazoline of PP-DME, with expansions of pyrazoline signals 34 Figure 2.6: Expansion of HMQC spectrum, illustrating l 3 C - ' H associations within the pyrazoline of 10 36 Figure 2.7 Expansion of COSY spectrum, highlighting correlations within the pyrazoline 39 Figure 2.8 Expansion of NOESY spectrum, illustrating meso proton correlations 42 Figure 2.9: 1 H NMR spectra of 9 and 12 44 Figure 2.10: Course of thermal cycloelimination of nitrogen 48 vi Figure 2.11: Schematic of T L C plate: Thermolysis of di-adduct pyrazoline 50 Figure 2.12: UV-vis spectra of A-ring pyrazoline of PP-DME and its photoproduct 52 Figure 2.13: Changes in 'H NMR spectra of pyrazoline (11) and photoproduct (16) 53 Figure 2.14: *H NMR spectrum of 16 54 Figure 2.15: X-ray crystal structure of 16 56 Figure 2.16: Transmittance of red light filter 57 Figure 2.17: UV-vis spectra of 18 and 19 in CH 2 C1 2 60 Figure 2.18: 'H NMR spectra of 19 and 20 66 Figure 2.19: UV-vis spectra of pyrazoline (19) and its photoproduct (20) 67 Figure 2.20: UV-vis spectra of chlorin bands, monitoring photoconversion of 19 to 20 69 Figure 2.21: Emission spectra of 19 and 20 70 Figure 2.22: Course of photochemical cycloelimination of nitrogen 71 Figure 2.23: Energy level diagram of porphyrin and pyrazoline 73 Figure 2.24: Cyclic voltammogram of 19 measured in CH2CI2 75 Figure 2.25: Energy diagram for the formation and dissociation of pyrazoline.... 77 Figure 2.26: Foscan® 81 Figure 2.27: Diazirine-TPP 81 Figure 2.28: Predicted photochemistry of diazirine-TPP and diazirine-DPP 83 Figure 2.29: 'H NMR spectra of trifluoroacetyl-TPP and intermediates 94 Figure 2.30: Coupled 1 9 F spectra of 22 and 23 95 vii LIST OF SCHEMES Scheme 1.1: Protoporphyrin derived from blood 10 Scheme 1.2: Chlorins derived from leaves 11 Scheme 1.3: Synthesis of TPP 11 Scheme 1.4: Photochemistry of nitrenes 20 Scheme 1.5: Photochemistry of diazirines 22 Scheme 2.1: Reaction of PP-DME and diazomethane 28 Scheme 2.2: Numbered B-ring pyrazoline 35 Scheme 2.3: Temperature Ranges required for thermolysis in cyclic azo compounds 47 Scheme 2.4: Photochemistry of pyrazoline functionality 51 Scheme 2.5: Numbered pyrazoline of MePPP 59 Scheme 2.6: Formation of 20 from 19 64 Scheme 2.7: Preparation of Fluorinated Porphyrins 84 Scheme 2.8: Synthesis of 3-trifluoromethyl-3-phenyl diazirine 84 Scheme 2.9: Progress towards the synthesis of diazirine-TPP (25) 85 Scheme 2.10: Formation of trifluoromethanol-TPP 91 V l l l LIST OF TABLES Table 2.1: HMQC correlations within 10 37 Table 2.2: Pyrazoline methylene coupling constants 38 Table 2.3: NOESY correlations for the A and B-ring pyrazolines of PP-DME.. 41 Table 2.4: Thermolysis of pyrazolines 49 Table 2.5: H M Q C correlations for 19 61 Table 2.6: Summary of redox potentials measured in CH2CI2 at 22°C vs. Fc.. . . 74 Table 2.7: Results of crosslinking attempts in various reagent solutions 79 Table 2.8: TPP metallation conditions 87 Table 2.9: Formylation of NiTPP 88 I X A C K N O W L E D G E M E N T S For the opportunity to work in such a dynamic laboratory, I would like to thank Professor David Dolphin. The last two years have been a very exciting time in the Dolphin labs, and I count myself among the most fortunate graduate students to have been able to work on a great project in this unique learning environment. Every member of the Dolphin research group has contributed to my learning experience at UBC, and I am sincerely grateful. Jill MacAlpine extended her wisdom and friendship to me even before my arrival in Vancouver, and I thank her. Alison Thompson was an invaluable sounding board who offered encouragement both in my laboratory and academic pursuits, not to mention being an exquisite dinner partner! Ethan Sternberg was a mentor, a fountain of knowledge and creative suggestions, and I count him among my true friends. This page would not be complete without mentioning my classmates. To Michael Fenster, Roger Linington, Jeffery Flemming and Jomi Samuel, thank you for the late study nights and the great laughs! To my family, who has always offered me unconditional love and support, thank you for caring for me half a country away and for being so proud of me. And to Mark, who has read and re-read every word of this thesis, who was my best critic and is my biggest fan. This is for you L O M L . It represents the end of one stage of my life, and the beginning of our life together.. .1 can't wait! Introduction 1 Introduction 1.1 Tetrapyrrolic Macrocycles 1.1.1 Background It is somewhat surprising, but true, that a single family of pigments is responsible for making our blood red and the grass green. These pigments are based on a macrocyclic ring of carbon and nitrogen atoms, some of the best examples of which are the iron-containing porphyrins found in heme (of hemoglobin) and the magnesium-containing reduced porphyrin (or chlorin) found in chlorophyll.1 The macrocycles are made up of four pyrrolic moieties, often joined together via methine bridges (Figure 1.1). H •H Porphyr in Ch lor in H H H H-H "H Bacter ioch lor in Isobacter iochlor in Figure 1.1 Important tetrapyrrolic backbones 1 Introduction The ubiquity and diversity of tetrapyrrolic macrocycles have prompted much research into their structural determination, isolation, total synthesis and function. Since their elucidation early in the 20th century2, porphyrins have come a long way; particularly in their diverse applications. Over sixty years ago, the first porphyrins were isolated from oil shale.3 These 'petroporphyrins' now serve as chemical markers for new oil reserves, and recent advances continue to fuel this field of investigation, known as organic geochemistry.4 Today, research into metalloporphyrins has led to their integration in many fields of study, including medicinal chemistry. These efficient chelators have the ability to harness radioactive metals and can thus be used in diagnostic pharmaceuticals.5 One need only scan through recent porphyrin abstracts to recognize their wide range applications, including molecular wires in nanotechnology, catalysis, light emitting diodes,8 enzyme mimetics,9 industrial paint dyes, and optical recording materials.10 The Dolphin lab has been particularly interested in those characteristics of porphyrins that make them well-suited for use in photodynamic therapy (PDT). Both non-metallated (free base) and a variety of metallated porphyrins have shown promise as potential photosensitizers.11 It is the ability of these extended aromatic molecules to absorb long wavelength light that enables them to sensitize oxygen, which can then incite the degradation of tumors and other hyperproliferative tissue. Section 1.2 will elaborate on PDT and the use of porphyrins as photosensitizers. 2 Introduction 1.1.2 Nomenclature The nomenclature of porphyrins and related systems is often difficult to understand, due to the lingering presence of trivial naming schemes. The initial naming scheme was developed by Fischer and others12'13'14 in the 1930's and utilizes a large number of trivial names in conjunction with a basic Figure 1.2 Protoporphyrin IX numbering system. For example, protoporphyrin IX (Figure 1.2), one of the naturally occurring porphyrins, was assigned its numeral designation as it happened to be the ninth isomer discovered in a series. Other important trivially named porphyrins are etioporphyrin, rhodoporphyrin, and phylloporphyrin, on which the classification of their optical spectra is based (see Section 1.1.4). A revised form of porphyrin nomenclature was proposed in I96015 and adopted in 198816 by an IUPAC-IUB joint commission. It accounts for all substituents, including the pyrrolic nitrogen in a systematic way. The numbering system is as shown, for the simplest porphyrin, named porphine (Figure 1.3). Positions 1, 4, 6, 9, 11, 14, 16, and 19 are the a-positions, 2, 3, 7, 8, 12, 13, 17, and 18 are the (3-positions and 5, 10, 15, and 20 Figure 1.3 Porphyrin numbering and nomenclature 3 Introduction are the meso-positions. The pyrrole rings are lettered clockwise, starting from the top left, A through D. Alongside the systematic IUPAC nomenclature, for the sake of brevity and perhaps tradition, the old type nomenclature has been retained in some cases. IUPAC allows a semi-systematic nomenclature based on a dozen trivial names. Any compound that has the basic structure of protoporphyrin keeps protoporphyrin as the root, and modifications are systematically added to the name. Thus, the compound resulting from methyl esterification of the propionic acid chains of protoporphyrin IX is called protoporphyrin dimethylester (PP-DME), and not 2, 7, 12, 18-tetramethylporphyrin-3,8-divinyl-13,17-dimethylpropranoate. 1.1.3 Structural Characteristics Porphyrins are a unique class of molecules thanks to their structural and inherent electronic characteristics. The feature that accounts for most of the porphyrins unique behaviour is their aromaticity. They are planar (for the most part) and possess 22 71-electrons, 18 of which participate in a conjugated pathway (Figure 1.4). Thus, these molecules are very stable, have reduced heats of combustion, characteristic Figure 1.4 The 18Tt-electron pathway bond lengths, and a large ring current. 17 4 Introduction The two remaining exocyclic double bonds can be reduced without compromising the aromaticity of the molecule, but do markedly change some of the molecules' properties (see section 1.1.4). The reduction of one of these bonds leads to the formation of a chlorin. When both double bonds are reduced, the result is either a bacteriochlorin -opposite rings, or an isobacteriochlorin- adjacent rings. Figure 1.1 depicts each of these classes of reduced porphyrins. The effect of large ring current is best observed in 'H-NMR spectra of porphyrins. Both the raeso-protons and (3-pyrrole protons are deshielded, and are shifted downfield (5=8 to 10 ppm). The inner N-H protons find themselves inside the ring current, are shielded, and appear far upfield (5 = -1 to -5 ppm)18 relative to tetramethylsilane (TMS). It is important to note that the protons on these nitrogens are not localized on any two nitrogen atoms, but exchange rapidly between all four nitrogens. The resultant signal visible in NMR spectroscopy is a single peak representative of both central N-H protons. The central imino nitrogen atoms are basic, and can be protonated to form the dication in strong acid. The central amine groups can be deprotonated, but strong base is required to form the dianion. The pKi and p K 2 values of the N-H protons are both approximately 16, while pK 3 is 5 and pIQ is 2.1 9 These properties of the central nitrogen atoms, together with their shape and size make porphyrins excellent chelators. Introduction of divalent metals replaces the central amine hydrogen atoms, and the porphyrin ligand exists formally as its dianion in a metal complex. Such changes in the core of the macrocyle can alter the porphyrin's reactivity at its periphery substantially. 5 Introduction Porphyrin reactivity can be simplified, perhaps naively, by returning to the fundamental aromaticity of the macrocyle. Porphyrins undergo many of the electrophilic substitution reactions characteristic of aromatic compounds including nitration, halogenation, formylation, acylation and deuteration. Porphyrins differ from molecules such as benzene, in that there are two different sites on the macrocycle where electrophilic substitution can take place with different reactivities; 2 1 the meso-position and the pyrrole (3-position. Which of these sites is activated depends on how electronegative is the porphyrin. This can be controlled by the choice of metal to coordinate to the central nitrogen atoms. 1.1.4 Optical Absorption Spectra Functionalized porphyrins come in a variety of shapes and sizes.. .and colours! The vivid green, purple, and red hues of these molecules stem from their absorption of near-ultraviolet and visible light in the electromagnetic spectrum. The most intense band, between 390 and 425 nm, 2 2 is called the Soret band. The Q bands, situated between 480 and 700 nm, vary in number and intensity depending on the substitution pattern of the porphyrin and whether it is metallated or protonated. Metal-free (free base) porphyrins have four Q bands, and are labeled IV , III, II, and I, in increasing order of wavelength. There are four classifications for porphyrins of this nature, etio-type, rhodo-type, oxo-rhodo type and phyllo-type. Figure 1.5 depicts the appearance of each type of spectrum, and the nature of the porphyrin's substitution. 6 Introduction When the relative intensities are such that IV>III>II>I, the spectrum is said to be etio-type after the etioporphyrins in which the (3-substituents are all alkyl groups. When electron withdrawing-groups such as carbonyl or vinyl groups are found in the (3-positions, a subtle change is observed such that III > IV > II > I. This is called a rhodo-type. spectrum because of the slight red shift that is also observed. When two such electron-withdrawing groups are situated on opposite pyrrole rings in the macrocyle, the rhodofying effect is intensified to give an oxo-rhodo-lype spectrum in which III > II >IV IV alkyl substituents in [3-positions electron withdrawing groups (carbonyl, vinyl) in (3-positions electron withdrawing groups are on opposite pyrrole units (3-positions are unsubstituted, or meso positions are occupied 500 600 Figure 1.5 Q-band porphyrin spectra for metal-free porphyrins 7 Introduction > I. When some of the pyrrole p-positions are left unsubstituted, or when a meso-position is occupied, the phyllo-type spectrum is obtained, and the intensities of the Q bands are such that IV > II > III > I. Addition of strong acid to the porphyrin protonates the central imino nitrogen atoms. The macrocyle is more symmetrical (than the free base) as a result, and the Q band spectrum is simplified by the collapse of four bands to two. By the same token, addition of a metal increases symmetry, thus metalloporphyrins generally have two Q bands. Chlorin spectra are characterized by a marked increase in intensity in the longest wavelength Q band, as well as a bathochromic shift (>25 nm). The phenomenon is somewhat counterintuitive, as one would expect decreased conjugation to result in a higher energy transition, or lower wavelength shift. However, selection rules and the allowedness of this transition contribute extensively in the chlorin spectra. Section 2.3 contains the UV- visible spectrum of one such chlorin, methylpyropheophorbide. Many theoretical chemists have elucidated the origins of these spectra. In the 1960's, Gouterman developed the four-orbital model theory23 that simplifies the task because it considers only the two highest occupied molecular orbitals (HOMOs), and the two lowest unoccupied molecular orbitals (LUMOs) of a porphyrin. A comprehensive presentation of this theory requires an extensive explanation nonetheless, and is beyond the scope of this work. 8 Introduction The differences in electronic spectra of tetrapyrroles can serve as an important diagnostic tool. The nature of a compound can thus be identified, and the spectra are invaluable in monitoring the progress of reactions. Finally, it is these optical properties of porphyrins that make them so valuable as phototherapeutic agents. Their absorption of long wavelength visible light, in regions where most living tissue does not absorb light, enables them to photosensitize oxygen, and kill surrounding cells in desired regions. This property of tetrapyrrolic macrocyles, as well as an overview of PDT will be further described in Section 1.2. 1.1.5 Preparation of Porphyrins None of the porphyrins used as starting materials in this thesis were synthesized in house. Each of the porphyrins that has been synthetically modified was chosen because of its inherent potential as a photosensitizer, and its availability. Although many elegant total syntheses of porphyrins have been reported over the years,18 these are invariably multi-step, time consuming, and low yielding. Because the targets herein may have applications as pharmaceutical agents, the starting materials have to be inexpensive, and the transformations kept to a minimum. In recognition of these facts, and the diversity of the three classes of porphyrins used, the preparation of PP-DME, methylpyropheophorbide (MePPP), and tetraphenylporphyrin (TPP) will now be described briefly. In the case of PP-DME and MePPP, nature has done most of the work for us. Starting from mammalian blood which has been strained 9 Introduction and heparinized (decoagulated), hemin (Fe(III) protoporphyrin IX chloride) is separated. The blood is poured into hot acetic acid containing sodium chloride, and the hemin separates on cooling.24 Several isolation techniques exist25'26, all which lead to a porphyrin that can be converted to various other derivatives by chemical manipulation of its substituents27 (Scheme 1.1). BLOOD hematoporphyrin IX Scheme 1.1 Protoporphyrin derived from blood Chlorophyll a and b can be extracted from leaves simply by boiling them in methanol or acetone. The two chlorophylls are separated by column chromatography, with the a-conformer being in a 3:1 excess. Removal of magnesium metal, and transesterification yields the methylpheophorbide, which is subsequently converted to the MePPP (Scheme 1.2). 10 Introduction LEAF TISSUE o = u!°phli'm „ methylpheophorbide methylpyropheophorbide R = CHO, chlorophyll b ' a:b = 3:1 Scheme 1.2 Chlorins derived from leaves TPP is the simplest porphyrin to make (Scheme 1.3). The synthesis involves refluxing pyrrole with benzaldehyde in proprionic acid for a half hour. The purple solid is filtered off after cooling, and then washed to free the porphyrin from soluble poly-pyrrolic impurities. The low yield (20-25% based on pyrrole) is offset by the ease of synthesis. The mechanism is based on the formation of a carbonium ion, the result of a pyrrole attack (a-position) on a protonated aldehyde. Another pyrrole then attacks the carbonium ion to give a meso-substituted dipyrrylmethane. Chain building continues until tetrapyrrylcarbinols are formed. Ring closure follows to give porphyrinogens, which oxidize in air to porphyrins. Scheme 1.3 Synthesis of TPP 11 Introduction 1.2 Photodynamic Therapy Photodynamic therapy is a minimally invasive, two-step medical procedure that uses light-activated drugs called photosensitizers to treat a range of diseases involving rapid cell growth, such as cancerous tumors or abnormal blood vessels." First, the photosensitizer is administered intravenously and, once in the bloodstream, associates with lipoproteins. Rapidly dividing (neoplastic) cells require more lipoproteins for membrane construction than normal cells, thus a higher concentration of the drug accumulates in these tissues. The photosensitizer drug is then activated by exposure to a pre-calculated dose of light at a particular wavelength. Once activated, the drug converts oxygen found in the cells into highly energized singlet oxygen. Singlet oxygen can react with subcellular components such as proteins and lipids, which disrupts normal cellular function and results in killing the cells. Lasers and fiber optics are used to deliver light when treating deeper internal organs. Photodynamic therapy is a relatively safe treatment, since without exposure to light the drug has no effect. Furthermore, the drug accumulates primarily in diseased cells as described above, thus upon irradiation the effects on surrounding healthy tissue are minimized. Patients treated with photodynamic therapy will experience photosensitivity- skin sensitivity to bright light - for anywhere from 24 hours up to six weeks depending on the photosensitizer used. 12 Introduction 1.2.1 Mechanism of Photosensitization Photosensitizers are generally porphyrogenic compounds with strong absorption coefficients at wavelengths in the (red) region (> 630 nm) of the electromagnetic spectrum. Namely, the Q bands must be used as a means of photosensitization since the presence of endogenous chromophores such as hemoglobin and light scattering result in poor penetration of tissue at wavelengths below 630 nm. Between 630 nm and 750 nm, light penetration through human tissue doubles (Figure 1.6), and allows efficient Q . CD T3 £Z o "5 <D cz CD Q . CD > . O CD CD CD > CD 550 630 700 Wavelength (nm; 800 Figure 1.6 Wavelength dependent penetration of light through-tissue excitation of the photosensitizer drug, causing the most phototoxic effect.30 Compounds absorbing energy above 800 nm do not have a large enough gap between the triplet and ground states to be able to generate singlet oxygen. 13 Introduction These properties of light penetration through skin are superfluous to note to any child who has succumb to the temptation of holding a flashlight in his mouth. Although we might have felt as though the eerie red hue given off by our cheeks was because of our blood, our skin and various biomolecules were in fact acting to filter the light. Thus it is obvious that the red light would penetrate most effectively! A modified Jablonski diagram, shown in Figure 1.7, best illustrates the principles of photosensitization. Non-radiative processes are depicted by curly arrows. Energy absorption (photons) is denoted by straight upward arrows, while radiative processes such as fluorescence and phosphorescence are shown with straight downward arrows. The porphyrin photosensitizer absorbs radiant energy, and this photon promotes an electron from the porphyrin's outer shell ground state (So), to an excited state (Sn). The molecule is then likely to lose energy via internal conversion (IC), until it reaches the first excited state (Si). The lifetime of the singlet excited state is usually not more than a few nanoseconds, since the electron can readily undergo internal conversion to its ground state (Si—>So + heat), or fluoresce (Si—>So + hv'). However, under certain circumstances it is possible for one of the electrons to undergo the 'forbidden' process of flipping its spin (known as intersystem crossing, ISC). Now both electrons have parallel spin, which results in the triplet excited state of the sensitizer (Si—>Ti + heat). Before the triplet excited state can lose its energy, the electrons must first pair their spins again. As these processes have low probability (spin-forbidden), compared with the singlet excited state, the triplet excited state has a much longer lifetime (microseconds to 14 Introduction seconds). One of the possible routes the first triplet state may take to the ground state is via phosphorescence ( T | - » S 0 + hv"). Triplet excited states can undergo many types of reactions. In considering those important to PDT, the other possibilities involve non-radiative interactions of the triplet excited state with an easily oxidized or reduced substrate ("Type I" reactions), or spin-exchange with ground state triplet exchange. The latter "Type II" process allows the excited triplet state of the sensitizer to pass its excitation energy (not an electron) to the triplet ground sate of the oxygen. The sensitizer thus returns to its singlet ground state, while oxygen is converted to a singlet excited state. Both "Type I" and "Type II" reactions are responsible for the cytotoxicity of the drug. 1 UJ PHOTOSENSITIZER OXYGEN Figure 1.7 Modified Jablonski diagram for a typical photosensitizer 15 Introduction The efficiency with which this internal system crossing to the triplet state occurs is quantifiable, and essentially determines the limit of effectiveness of a photosensitizer. The triplet quantum yield is the ratio of the amount of triplet sensitizer formed to the amount of singlet sensitizer initially formed. 1.2.2 Type I Reactions Photoprocesses of the Type I nature involve reaction of the sensitizer (SENS) in its triplet excited state (3SENS*), with other molecules in its environment. Because of its unpaired electrons (the promoted electron and the residual electron left by excitation), the triplet state is more readily oxidized and reduced. A photosensitizer in its triplet excited state can abstract an electron or a hydrogen atom from a substrate (SUB) as shown: 3SENS* + SUB -» (SENS*)" + (SUB*)+ 3SENS* + SUB-H -> SENS-H + SUB* The resultant, highly reactive, radicals can behave in a number of ways. Among these, the semi-reduced sensitizer is capable of reacting with molecular oxygen to generate the superoxide radical (02»)~. The free radicals formed upon hydrogen abstraction can also react with molecular oxygen, in this case forming peroxides, also capable of initiating radical chain reactions. 16 Introduction 1.2.3 Type II Reactions This second series of reactions stems from the triplet excited state of the sensitizer shunting its energy to triplet ground state oxygen. Formation of highly reactive singlet oxygen by this means is very efficient because it is 'spin allowed'. It is worthy to note that in water (the biological medium) the lifetime of singlet oxygen is only 4 p:s31, thus reactions attributed to it must be localized near the site of its generation. Singlet oxygen reacts with biological substrates in a variety of ways32. 1) Ene Reaction. Ethenes that have allylic hydrogens readily yield hydroperoxides in photo-oxygenation processes by reaction with '0 2 . The l02 ene reaction involves abstraction of the allylic hydrogen, and a shift of the ethene bond. Tryptophan and derivatives of this amino acid are susceptible to singlet oxygen attack: 2) [2+2] Cycloaddition. Electron rich ethenes, such as enol ethers, vinyl sulfides, and enamines which do not have allylic hydrogens, undergo photo-oxygenation to give dioxetanes. These unstable species undergo decomposition to give two carbonyl compounds: o o A*X Introduction 3) [2+4] Cycloaddition to 1,3-dienes. Singlet oxygen behaves as a powerful dienophile in the Diels-Alder reaction and cyclic endoperoxides can be produced in this way, as the following reaction in the formation of ergosterol peroxide illustrates: 4) The photooxidation of sulfides to sulfoxides was originally reported by Schenck and Krauch in 196233. It is now generally accepted that the reaction proceeds via the lowest excited state of molecular oxygen, singlet oxygen. The major product is the sulfoxide, with varying amounts of sulfone, depending on the substrate and reaction conditions.34 The following example depicts the oxidation to methionine sulfoxide: 1.3 Photoaffinity Labeling Historically, photoaffinity labeling has been used to locate active sites and binding sites in enzyme-substrate complexes (Figure 1.8). As a result of covalent bonding between substrates and a variety of cellular components, enzymes, membranes35, protein structures," neural receptors , and RNA or DNA structures' have been successfully 18 Introduction labeled. Each of these successes yields information about its respective target. Vigorous application of the technique can reveal the sequence of amino acids involved in binding, and effectively paint a picture of the active site in an enzyme. Enzyme Substrate Enzyme-Substrate Photolabeled Labeled Complex Enzyme Fragment Figure 1.8 Schematic of photoaffinity labeling Photoaffinity labeling typically requires chemical modification of the substrate so that irradiation of the resultant complex will produce a reactive intermediate that will bond covalently to the enzyme.39 A critical element of photoaffinity labeling is synthesis of the modified substrate, and often the substrates have radioactivity incorporated to allow tracking of the tagged binding site. Synthetic methods for introduction of radiolabels and choice of radiolabeling reagents are also important aspects of this technique40, but do not relate to the present research topic. Photolysis of the modified substrate generates an intermediate whose behaviour will dictate the success of its labeling, thus the intermediate precursor must be chosen wisely. Several classes of reactive intermediates have been exploited in photoaffinity labeling; the predominant ones being nitrenes, carbenes and radicals. Each of these will be briefly introduced in this section. 19 Introduction 3r recombine ISC H atom abstraction radical coupl ing 1 2 Scheme 1.4 Photochemistry of nitrenes Nitrenes stemming from aryl azides have been studied extensively and appear throughout the literature. In spite of the frequent use of azides, it has been pointed out recently that this moiety "rarely gives stoichiometric labeling."41 In fact, yields of azide photolabeling are generally less that 30%.42 Low cross-linking yields in aryl azide applications can result for several reasons. The most probable complication arises from the chemistry that follows irradiation of the aryl azide (1) (Scheme 1.4), which shows that irradiation produces a singlet nitrene that intersystem crosses to the ground state triplet. (At low temperatures, some chemistry from the singlet is observed to form 2.) The triplet nitrene is expected to behave like a diradical, and is hoped to effect useful binding by hydrogen abstraction then radical coupling.43 In addition to the low reactivity of the ground state triplet, simple phenyl azides suffer from complications of instability (they are easily reduced) and high energy irradiation wavelengths. Most of these chromophores have to be excited at 254nm, a wavelength that can substantially damage biological systems39. A representative example of the use of azides in photoaffinity labeling is found in the work of Swindell et. al.44 Taxol 4 5, a promising anti-cancer drug, is known to interfere 20 Introduction with the normal function of microtubules in rapidly dividing cells.46 To understand the interaction between taxol and microtubules, the researchers efforts were directed toward determining the structure of the binding site(s) of the drug on microtubules. This was accomplished by the synthesis of a photoaffinity taxol analogue that bears an azide photoreactive moeity in the A-ring side ACC; ,O OH chain of taxol (Figure 1.9). The tritiated [~H] version of this taxol analogue (3), was also prepared, carrying tritium on the P h C ° 2 OAc aromatic rings of the N-acyl subtituents for TAXOL X = PhCONH 3 X = 4-(N3)-C6H4CONH ease of detection. The results of this work determined that 3 labels the TV-terminal Figure 1.9 Photoaffinity taxol analogue domain of (3-tubulin36 with specificity, and that this specificity coupled with the efficiency of labeling make it an attractive candidate for further characterization of the taxol binding site(s) on microtubules. In general, radical intermediates are ideally suited for photoaffinity labeling.39 They are known to abstract hydrogen atoms from virtually any site, are more reactive with C-H bonds than are nitrenes and have less propensity for intramolecular rearrangements than carbenes. Unfortunately, many free radical reactions including intramolecular hydrogen abstraction and bond formation between the initial radical pair can minimize their ability to crosslink. Single electron transfer to generate long -lived charged species, can either minimize effectiveness or be conducive to covalent bonding between nucleophilic and electrophilic sites. Examples of radical generating photoprobes include: benzophenone47 21 Introduction (whose triplet diradical has a lifetime of 300ns in the absence of easily abstractable AQ hydrogen atoms), enones (which have a higher propensity for 2+2 cycloaddition with alkenes), and various diazo/diazonium compounds whereby loss of N 2 results in the reactive intermediate49. The first series of potential crosslinking photosensitizers appearing in this thesis are pyrazolines, which are expected to behave as biradicals upon photoexcitation. Diazirines were first proposed as photochemical reagents by Smith and Knowles in 197350, but most of the recent work with diazirines exploits the photochemical reactivity of the trifluoroethyldiazirinephenyl group51. Irradiation of a substituted diazirine (4) (Scheme 1.5) has been shown to give carbene (5) and the corresponding diazo compound (6). Carbenes have the ability to insert into carbon-hydrogen bonds (Sol-H) to yield products like 7. The diazirine unit is small, non-bulky, and lipophilic. It has a chromophore that extends significantly into the 300 nm range. Many applications of photo cross-linking have been reported for this functionality. Some of the disadvantages that have been reported include the triplet carbene reacting with molecular oxygen to give the corresponding ketone, and the photoactive diazo intermediate having a different 22 Introduction excitation wavelength than the diazirine.53 Despite these potential drawbacks, its efficient insertion capabilities coupled with its stability in mild acidic or basic environments54 make it a most attractive photocrosslinking functionality. COO-Researchers at Columbia University55 reported the synthesis, radioisotopic labeling, and resolution of a phenylalanine analog, 3-[p-[3-(trifluoromethyl)-3H-dizirin-3-yl]phenylalanine (8), contain-ing the 3-(trifluoromethyl)-3H-diazirinyl group. Like all diazirines, 8 absorbs in the near U V (^max 350; e 265), at wavelengths which do little damage to most biological samples. In this case, they hoped the analog would be useful in preparing peptide and polypeptide photoaffinity reagents. Understandably, receptors for peptides and polypepetides, including various hormones, neurotransmitters, and toxins, are of considerable interest. In summary, the literature abounds with examples of successfully labeled binding sites in a wide range of supramolecular systems.56'57'58 That the field of photoaffinity labeling is immense and growing rapidly testifies to the utility of this analytical procedure.39 It is clear from the information in this section that the advantages and disadvantages of any given photoactive moiety must be weighed carefully in each application, since many variables play a role in the efficiency and selectivity of photolabeling. Although the objective of the work in this thesis does not involve labeling of a specific binding site, the intent is to capitalize on the mechanism of photoaffinity labeling. Futhermore, the design 23 Introduction of the crosslinking photosensitizers that are envisioned herein is based on literature precedence for the crosslinking ability of the functionalities. 1.4 Research Objective As the search for more effective and less costly photosensitizers continues, so does the investigation of their characteristics and capabilities. There is a great need to understand the critical structural and electronic factors that control these molecules' reactivities and bioactivities. In addition, the mechanisms by which phototoxins biodistribute are just beginning to be understood, and it would be very desirable to increase their selectivity for diseased tissue. Photosensitizers are generally porphyrogenic compounds with strong absorption coefficients at wavelengths in the 650 nm (red) region of the electromagnetic spectrum. At this wavelength, human tissue is the most transparent to light, and allows efficient excitation of the photosensitizer drug, causing the most phototoxic effect.59 Unfortunately, this strength of light is present in ambient daylight and patients of PDT have experienced varying degrees of skin photosensitivity resulting from residual photosensitizer in healthy tissue. Although photosensitizers tend to accumulate in cancerous tissue, their transport by lipoproteins results in their presence throughout systemic circulation. This 'passive selectivity' is a result of the pharmacokinetic behaviour of the drug. Results have shown60'61 that photosensitizers are taken up and released rapidly by all cell types. All cells retain a proportion of the drug taken up, and 24 Introduction cancerous cells have demonstrated an ability to take up increased amount of photosensitizer. This equilibrium requires that enough drug be administered to ensure effective photosensitization of the target cells, but is also to blame for residual skin photo-sensitivity. This side-effect of PDT has prompted an exploration of possible structural modifications of the photosensitizers, and new ways of using visible light for their photoactivation. One approach is to develop a method of binding the drug once it has reached the diseased tissue, creating an 'active selectivity'. If the drug could be fixed in the desired cells, the possibility of multiple irradiations and new approaches to PDT are envisioned. Such a process might allow for the use of much lower doses and also be more sparing to healthy tissue. Photoreactive crosslinkers offer an efficient means of conjugation with proteins and various other cellular components. Many azo-type (N=N) functionalities are known to form biradical62, carbene63, or nitrene64 intermediates upon irradiation. Photoaffinity labeling techniques have long capitalized on these highly reactive intermediates that can crosslink carbon-carbon multiple bonds, or insert into carbon-hydrogen (C-H) bonds within the targeted cells.6 5'6 6'6 7 Figure 1.10 depicts a generalized scheme for some of the reactivity of the aforementioned reactive intermediates. The objective of this work is to synthetically modify porphyrins such that they possess a functionality that could crosslink within their chemical environment. Thus, the synthetic targets will be designed according to precedence for the crosslinking moiety's aptitude, as well as the known photosensitizing ability of the porphyrin starting material. With the 25 Introduction Pyrazoline-modified Drug N=N Diazirine-modified Drug hv -N, hv —) -N, Reaction of Biradical h b - ^ C H Reaction of Carbene Drug bound to cellular component Drug bound to cellular component Azide-modified Drug hv -N, Reaction of Nitrene Figure 1.10 Photoactive functionalities c m Drug bound to cellular component potential for these molecules to become PDT agents of the future, it is imperative that transformations be kept to a minimum and that scale-up procedures be considered. Each of the novel synthetic targets will be isolated and fully characterized. More fundamentally, this work presents an opportunity to examine the photochemistry involved in activating such a crosslinking moiety through absorption of light by the porphyrin chromophore. Each of the successfully modified porphyrins will be 26 Introduction photoactivated, and their photochemistry observed both at long and near-UV wavelengths. The photochemical products will be isolated and fully characterized. An attempt will be made to understand the photochemical phenomena in question, and experiments carried out to further explain the results. In addition, a series of crosslinking experiments will be performed, to assess the modified porphyrins abilities to crosslink within their chemical environment. 27 Results and Discussion 2 Results and Discussion 2.1 Synthesis and Characterization of Protoporphyrin Pyrazolines A serendipitous discovery in the Dolphin lab prompted much of the research described in this chapter. An esterification reaction involving diazomethane and benzoporphyrin diacid (Figure 2.1) was left overnight erroneously, and resulted in a myriad of products in addition to the expected esters of the 0 H H 0 proprionic acids. The identity of the products was not known, Figure 2.1 BPD-diacid but it was rationalized that some of the products could be due to diazomethane reacting with the activated olefins of the benzoporphyrin. In order to explore this possibility further, a simpler porphyrin was investigated. It was discovered that the reaction of diazomethane and protoporphyrin dimethylester, 9 (PP-DME, the dehydrated version of the naturally occurring starting material for many photosensitizers,) generated three addition products. The three pyrazoline porphyrins (10-12) shown in Scheme 2.1 depict one set of the many structural identities that were initially hypothesized for the products. N Scheme 2.1 Reaction of PP-DME and diazomethane 28-Results and Discussion Initial research in this project involved the optimization and full characterization of the aforementioned reaction and products. The yields obtained are respectable (77% recovery for reaction Scheme 2.1), but in order to isolate sufficient amounts of the mono-adduct pyrazolines, the reaction must be stopped before completion. Thus the overall conversion of starting material to products must be high, but if a huge excess of diazomethane is present and left to react, the mono adducts will go on to form the diadduct. The reaction was not as simple as first anticipated. Gaseous diazomethane is formed on base catalyzed decomposition of N-methyl-N-nitroso amines68, and is carried by an ethereal distillate. The low solubility of porphyrins in most solvents was further reduced by addition of ether. In fact, the ether resulted in the precipitation of the porphyrins from solution throughout the course of the diazomethane addition. Methylene chloride was found to be the best solvent for PP-DME. Under dilute porphyrin solution conditions, the conversion to products was very poor. The most successful reactions resulted from a compromise of the conditions: dissolving the nitroso compound in a minimum amount of ether, and adding more methylene chloride to the stirring reaction mixture upon complete addition of diazomethane. Most references for diazomethane reactions69 on small molecules cite usage of an excess of approximately 10:1 diazomethane: substrate. With high molecular weight porphyrins and their inherent low solubility it was found through successive increases that a molar ratio of 50:1, diazomethane:porphyrin was required for successful conversion. In 29 Results and Discussion addition, because the reaction is sluggish and must be left for 12 hours, a method of containing the gaseous diazomethane in the reaction mixture was required. On several occasions, built-up pressure resulted in septa that were either blown off the reaction flask, or torn open. An alternate approach was found in sealing a balloon over the reaction flask, thus containing the reactant, but allowing for expansion. Once these problems were overcome and successful conversion was achieved, the task of isolation and characterization began. The three products were observable as red bands by T L C (5% THF in CH2CI2 eluent). All three bands were more polar than the starting material. The slowest moving band was established as the di-adduct, not only for its increased polarity (Rf u = 0.09), but also because the other two bands disappeared on completion of the reaction. The mono-adduct structural isomers differed only slightly by chromatography (Rfn = 0.43 and Rf 10 = 0.33), rendering separation extremely difficult. The aforementioned R f values would probably not present much of a separation challenge in small molecule chemistry. High molecular weight porphyrins however, are notorious for tailing and unexpected diffusion behaviour. Large scale separation was only achieved using the Chromatotron®. Identifying the product pyrazolines was challenging for several reasons. To our knowledge, this presents the first example of pyrazoline-modified protoporphyrins. There is precedence for the regioselectivity of diazomethane reactions with many familiar substrates, but none could be found involving a conjugated vinyl group of a porphyrin. 30 Results and Discussion Before embarking on the task of deciphering extensive NMR spectra, some thought was given to the reaction mechanism. The pyrazolines are formed as a result of a concerted [47U+2TX] cycloaddition.70 The 1,3 dipolar molecule is diazomethane, and the vinyl group is the dipolarophile. Most examples in the literature cite diazomethane reactions with electron-poor species68, or with alkenes conjugated to strong electron-withdrawing substituents. It was thus quite surprising to find that these vinyl groups, conjugated to the electron-rich porphyrin system, reacted to form the products. The stereochemistry of the 1,3-dipolar cycloaddition reaction is known to be stereospecific syn addition71, but in the present case of addition to a vinyl group, need not be addressed. While the products in Scheme N Q N : II NI-^ r \\. J 3 C H 2 of the pyrazolines, there were Vs. N II OR II O N O A . N 2.1 show only one arrangement ©ftO 0 ( l H ( E J ^ ^ r C H 2 two to consider. When both the ..^~~^OCH2 [ CH, — ^ \\ © N ^ OR N © N j - I N (U) 1,3 dipole and the dipolarophile are unsymmetrical, as they are in F i g u r e 2 . 2 1,3-dipolar cycloaddition, where 'S' = symmetrical conformation and 'IT = unsymmetrical the formation of the protoporphyrin pyrazolines, there are two possible orientations for addition (Figure 2.2). Both steric and electronic factors play a role in determining the regioselectivity of the addition, and can be predicted considering orbital coefficients and by calculation or estimation of the relative energies of the interacting orbitals.72 The most favorable Results and Discussion 1 .8 L U M O 1 .0 H O M O -9.1 -9.0 C H 2 N = N R = c o n j u g a t e d g roup with m o d e s t e l e c t r o n - r e l e a s i n g c a p a c i t y Figure 2.3 Estimated energy of frontier n orbitals for conjugated vinyl group and diazomethane (Relative energies are given in eV) orientation is that which involves complimentary interaction (smallest energy gap) between the frontier orbitals of the dipole and dipolarophile. Figure 2.3 depicts the estimated energy of the frontier TZ orbitals, examined by experimental means in the early work of K. H. Houk, 7 2 for the interacting molecules. According to these relative energies, we are lead to conclude that the L U M O of the conjugated alkene is more likely to react with the HOMO of the diazomethane. The last piece of the puzzle to answer the question of regioselectivity involves the orbital coefficients of the reactants. Diazomethane has several contributing resonance structures, the two most prominent ones shown HOMO LUMO reacting in Figure 2.2, but which of the 0 78 0 7 4 orbitals will react with which is not ' f") 0.13p,0.61 f~) f%0-50 obvious. The HOMO and L U M O % MOs of the most common 1,3-dipoles were Figure 2.4 Orbital coefficients for diazomethane calculated by Houk et. al. , and those of 32 Results and Discussion diazomethane are shown in Figure 2.4. If the vinyl group of the porphyrin is likened to one on styrene, the literature72 cites the larger orbital coefficient to be on the terminal carbon of the double bond. Thus the atom with the larger terminal coefficient, in this case the carbon atom of the HOMO, will become bonded preferentially in the transition state to the terminal carbon of the vinyl group of the porphyrin to yield the unsymmetrical pyrazoline ('U', in Figure 2.2). Of course the best way to reinforce this conclusion based on theoretical calculations, is to prove it experimentally. The characterization of each of the isolated protoporphyrin pyrazolines was a challenging and rewarding process. Initially, the stability of the pyrazolines was questioned, because *H NMR spectra could only be obtained immediately upon preparation of the sample. It has since been determined that the small amount of acid present in CDC1 3 (and reagent grade C H 2 O 2 ) is enough to cause the degradation of these porphyrins, which are otherwise stable at room temperature. The first series of experiments will address the structural assignment of the pyrazoline (ie; 'S' vs. 'U' , Figure 2.2). To simplify matters, the spectroscopic investigation of only one of the mono-adducts will be presented. Following the assignment of the pyrazoline, the A and B-ring adducts will be differentiated by spectroscopic means. Finally, the characterization of the di-adduct will be discussed briefly. 33 Results and Discussion Results and Discussion The spectra obtained, albeit first order, are complex because of the number of signals, and particularly due to the long-range coupling within the heterocycle. A spectrum of 10 (Scheme 2.2), and expansions of the pyrazoline heterocycle signals are shown in Figure 2.5. There is ample precedence for Scheme 2.2 Numbered B-ring the assignment of those signals that are not labeled, pyrazoline thus only those particular to the novel pyrazoline moeity will be discussed here. The distinctive multiplet at 6.80 ppm that integrates for a single proton can be assigned to the 'ipso' position (HL-81) of the pyrazoline. The coupling pattern was difficult to decipher and was initially thought to be a triplet of triplets, indicating a symmetrical pyrazoline ('S', in Figure 2.2). This being the case, the splitting patterns and shifts of the four methylene proton signals were even more puzzling, and hindered rotation of the azo moiety was considered. The orientation of the pyrazoline was further investigated using HMQC (Heteronuclear Multiple Quantum Coherence) spectra, to determine the assignment of the pairs of methylene signals to their respective carbon atoms. Thus, if one of the multiplets around 5 ppm and another around 2 ppm correlated to a single carbon, the difference in shift within a pair of methylene signals could be attributed to some shielding effect, resultant from hindered rotation. The results however, explicitly countered this theory and point to the unsymmetrical arrangement of the pyrazoline ('U', in Figure 2.2). The methylene protons that are deshielded and downfield (H 0-8 3: 5.45 ppm and Hp-83: 4.74 ppm) 35 Results and Discussion Results and Discussion correlate to a single carbon atom, presumably because they are adjacent to the nitrogen. The pair of signals further upfield (HM-8 2: 2.74 ppm and H N -8 2 : 2.11 ppm) similarly correlate to their own carbon signal (Figure 2.6). Assignment H-x 'H NMR (500 Mhz) 5 ppm (mult., J (Hz)a) H M Q C I J C Correlations (125.8 Mhz) H-10 9.95 (s, 1H) 97.62 H-15 9.79 (s, 1H) 96.86 H-20 9.69 (s, 1H) 96.66 H-5 9.57 (s, 1H) 95.89 H-31 8.05 (m, 1H, JT R A N S= 17.8, JC I S= 11.5) 129.99 H L-8' 6.81 (m, 1H, JL_M, L-N= 9.8, J L . G , L-P= 2.6) 85.89 H-3 2 6.23 (m, 1H, JT R A N S= 17.8, J G E M = 1.2) 120.57 H-3 2 6.10 (m, 1H, JC I S= 11.5, J G E M = 1.2) 120.57 H 0 - 8 3 5.46 (m, 1H, J0-P= 17.9, J 0 - M = 9.8, J0-N.O-L= 2.6) 77.81 Hp-83 4.75 (m, 1H, JP_o= 17.9, J P _ M . P.N= 9.3, JP.L=2.8) 77.81 H-131 4.28 (t, 2H, J = 7.8) 21.65 or 21.62 H-17' 4.24 (t, 2H, J = 7.8) 21.65 or 21.62 H-133 3.66 (s, 3H) 51.69 or 51.67 H-173 3.65 (s, 3H) 51.69 or 51.67 H-71 3.49 (s, 3H) 11.42 H-21 3.46 (s, 3H) 12.41 H-121 3.43 (s, 3H) 11.83 H-181 3.42 (s, 3H) 11.51 H-132 3.21 (t, 2H, J = 7.8) 36.79 (2xC) H-172 3.18 (t, 2H, J = 7.8) 36.79 (2xC) HM-8 2.72 (m, 1H, JM-N=13.0) 27.72 HN-8 2.10 (m, 1H, JM-N=13.0) 27.72 MH -4.20 (s, 2H) a- Only those J values that could be unambiguously assigned are recorded Table 2.1 H M Q C correlations within 10. By pairing the proton and carbon signals of 10, each of the primary, secondary and tertiary carbon atoms were assigned (Table 2.1). This exercise would prove invaluable to further characterization. The multitude of overlapping quaternary signals were not assigned. 37 Results and Discussion In order to understand the complex and long-range coupling within the pyrazoline, COSY (1H, 1H Correlation Spectroscopy) spectra were acquired. Figure 2.7 is an expansion of the most informative region of the spectrum. It is clear that protons 'O' and 'P' are strongly coupled, which reinforces the HMQC data indicating their attachment to the same carbon. The same holds true for ' M ' and 'N'. Also noteworthy is the strong correlation between protons ' M \ 'N' and 'L ' . This supports the earlier conclusion that the remaining methylene signals ('O' and 'P') are downfield as a result of their proximity to the electron withdrawing nitrogen atom. The obvious question that follows is: 'Why then, is proton 'L ' so far downfield?'. Proton 'L' finds itself directly outside the ring current of the porphyrin. Like the meso protons, it is deshielded by the current and gets shifted to the aromatic region of the spectrum. With a clear picture of the coupling partners, the J values could thus be deduced, and the spatial orientation of the protons assigned (Table 2.2). (These preliminary assignments would later be confirmed by NOE spectroscopy). The Karplus relationship and a molecular model kit provided a way to estimate the stereochemistry along the methylene-methylene bond of the pyrazoline. JO-P 17.9 Hz JM-N 13.0 Hz JOM 9.8 Hz Jp_M, Jp-N 9.3 Hz Jo-N 2.6 Hz Table 2.2. Pyrazoline methylene coupling constants 38 Results and Discussion 39 Results and Discussion With the knowledge acquired from the COSY, an explanation for the long range (4-bond) coupling remained. Thinking the pyrazoline might be puckered, one might propose a 'W' arrangement of these protons to explain the large J values of 2.6 Hz (L-O, transoid) and 2.8 Hz (L-P, cisoid) respectively. Brief experimentation with molecular modeling (MM) negated this possibility. As it were, there is abundant precedence in the literature for the coupling phenomena occurring within the pyrazoline, entitled homoallylic coupling.74 Nuclear spin-spin coupling constants over four single bonds and one double bond are said to be of the homoallylic type between protons H A and HB in the system H A — C ] — C 2 ^ = C 3 — C 4 — H B with carbon atoms C] and C4 having tetrahedral hybridization. Many examples76 can be found in the literature for which C2 and/or C3 have been replaced by sp2-hybridized nitrogen atoms. Furthermore, it has been shown 77 that in cyclic systems such as the 5-membered pyrazoline ring, the interaction of the protons in question is enhanced because of a double homoallylic path78; namely, across the methylene bridge and across the azo bond. Thus the mislabeled 'triplet of triplets' at 6.81 ppm (Figure 2.5), is actually a doublet of doublet of doublet of doublets, and reaffirm the asymmetric orientation of the pyrazoline. The consistency among the spectra of each of the mono-adduct pyrazolines required that this extensive spectroscopy to be carried out on one of the structural isomers only. Outside of small shift differences, the spectra of the respective isomers are indistinguishable. An additional task however, presented itself in assigning the respective 'A' and 'B' ring adducts. NOESY (Nuclear Overhauser Effect Spectroscopy) correlations were used in these assignments, and were carried out on each of the isomers 40 Results and Discussion to positively identify them. The exercise was fundamental, but interesting enough to briefly discuss the strategy involved. A-ring B-ring H-x NOESY Correlations" (500 MHz) H-x NOESY Correlationsb (500 MHz) H-10 Med. H-8 2, str. H-8 1, str. H-121 H-5 Str. H-7 1, w. H-3 1 H-15 Str. H-131, str. H-171 H-20 Str. H-181, str. H-2 1 H-20 Str. H-181, str. H-21 H-15 Str. H-17', str. H-131 H-5 Str. HL-3 1 , str. H-71 H-10 Str. H L - 8 1 , str. H-121 H-81 Str. H-10, str. H-8 2 H-3' w. H-5 HL-3 1 Str. H-2 1, str. H-5, str. H M -3 2 , med. Hp-3, w. H N -3 2 HL-8 1 Str. H-7', str. H-10, str. H M -8 2 , w. Hp-83, med. H N -8 2 H-8 2 w. H-7 1, med. H-10, str. H-81 H-3 2 Str. H-2' Ho-33 Str. Hp-33, str. H N -3 2 , med. H M - 3 2 H 0 -8 3 Str. Hp-83 Hp-33 Str. H 0 -3 3 , str. H M -3 2 , med. HL-3' Hp-83 Str. H 0 -8 3 , med. H M -8 2 , w. H L -8 ' H-131 Med. H-121, str. H-132, str. H-15 H-13' Med. H-12', str. H-132, str. H-15 H-171 Str. H-15, str. H-172, med. H-181 H-171 Str. H-15, str. H-172, med. H-181 H-7' Str. H-5, w. H-8 2 H-121 Str. H-10, med. H-13' H-21 Str. HL-3', str. H-20 H-181 Str. H-20, med. H-17' H-121 Str. H-10.med.H-131 H-71 Str. H-5, str. H L-8' H-181 med. H-171, med H-172, str. H-20 H-21 Str.H-32, str. H-20 H-132 str. H-131 H-132 Str. H-13' H-172 str. H-171 H-172 Str. H-17' HM-3 2 str. H N -3 2 , str. H L-3', str. H P-3 3 , med. Ho-33 HM-8 2 Str. H N -8 2 , str. H L - 8 \ med. H P -8 3 HN-3 2 str. H M -3 2 , str. H 0 -3 3 , w. HL-3' HN-8 2 str. H M -8 2 , med. H L-8' b- str. = strong, med. = medium, w. = weak Table 2.3 NOESY correlations for the A and B-ring pyrazolines of PP-DME In the case of the B-ring isomer for example (Figure 2.8 -NOESY), signals from the vinyl group, pyrazoline, and proprionate side chains are readily identified. The meso protons (H-5, H-10, H-15, H-20) resonate downfield, between 9.57 and 9.95 ppm. The methyl groups (CH3-71, CH3-I21, CH3-I81, CH3-21), occur as four sharp singlets between 3.42 and 3.49 ppm. The meso proton H-20 is the only one in contact (through space) with two 41 Results and Discussion Results and Discussion porphyrin methyl groups (CH 3-2' and CH3-I8 1 ) and is assigned at 9.69 ppm. The C H 3 -181 is distinguished from the CH 3 -2' by its NOE to the methylene groups of one of the proprionate side chains (CH.2-171). From the H-20 meso proton, the following connectivities are therefore traced around the ring: CH3-I8 1 <— H-20 meso —> CH3-2 1 —> vinyl -31 -> H-5 meso - » CH 3 -7' - » pyrazoline-81 - » H-10 meso -> CH 3 -12' - » proprionate-131. The H-15 meso proton, between the two proprionate side chains, resonates at 9.79 ppm and exhibits the expected NOE correlations to each. Thus the two structural isomers, indistinguishable by mass spectrometry, UV-vis, l 3 C NMR, and 'H NMR spectroscopies, and barely separable by chromatography were assigned their respective conformations using 2D NMR spectroscopy . These findings can be related back to the initial chromatographic results; the faster moving regioisomer (Rf = 0.43) is 11, and the slower moving one (Rf = 0.33) is 10. This conclusion is supported by the crystal structure of the photoproduct 11, which is discussed in the next section. The di-adduct pyrazoline of protoporphyrin-DME (12) did not present the isolation challenges of the A and B-ring adducts. It was much more polar than either of the mono-adducts, thus distinguishable by T L C , as well as by MS and UV-vis spectroscopy. The extensive 2D NMR experiments to assign the regiochemistry of the pyrazoline rings were carried out on the single adducts, and could be extrapolated to similarly assign those of the di-adduct pyrazolines. There are however, some interesting features of the di-adduct that should be noted. Figure 2.9 displays the proton NMR spectrum of the starting 43 Results and Discussion Results and Discussion material, PP-DME, and 12. On first observation, one of the most startling features of the di-adduct spectrum was the presence of two central amine peaks at -3.75 ppm (unfortunately, only a large expansion would illustrate this effectively). Although it is known that there are two central amine protons, normally their rapid exchange on the NMR time scale does not allow them to be resolved separately. The peaks were indicative of something far more fundamental; the presence of more than one porphyrin in solution. On reflection, this makes perfect sense because in the case of the di-adduct, two stereocenters are created at each of the pyrazolines, and the result is a pair of diastereomers. The presence of these diastereomers would have made the assignment of the pyrazoline signals far more difficult, because of the signal overlap in each of the multiplets (HL: 6.91, H 0 : 5.55 and 5.50, H P : 4.82 and 4.78, H M : 2.84, H N : 2.24 and 2.22). Some J values could be assigned unambiguously, but only thanks to the fact that they had been determined previously in the mono-adduct pyrazolines. Unique to the di-adduct, in comparing it with the starting material PP-DME (9), is the complete disappearance of the vinyl signals (a-vinyl: 8.2 ppm, (3-vinyl:6.32 ppm). The last observation made of the di-adduct proton NMR spectrum is perhaps more subtle. The meso protons, in the aromatic region of the spectrum (around 10 ppm) have a different pattern in the di-adduct. This could be a result of two factors. First, the presence of diastereomers creates new signals which may not overlap exactly. It is more likely however, that the change in symmetry brought about in the A and B rings of the porphyrin alters the shift of the meso protons. 45 Results and Discussion In summary, Scheme 2.1 accurately depicts the products formed in the reaction of diazomethane and protoporhyrin dimethylester. The spectroscopic (and crystallographic) evidence supports the assumptions made in adopting the theoretical calculations from the literature. The mystery of the overnight diazomethane reaction is solved, and three novel pyrazoline porphyrin derivatives have been identified. 46 Results and Discussion 2.2 Thermal and Light Activated Chemistry of Protoporphyrin Pyrazolines It is well known that cyclic azo compounds including heteroaromatics with a N=N bond of strong double-bond character, such as pyrazolines, can eliminate molecular nitrogen on addition of sufficient thermal energy or on electronic excitation.79 Some substrates decompose very quickly, while others are moderately or even extremely resistant to N 2 -elimination. In fact, the range of activation energies for the pyrrolytic extrusion of molecular nitrogen covers values as low as 15-20 kcal/mol to as high as 40-45 kcal/mol60. The observed rates of decomposition and temperatures vary dramatically. For example, 2,3-diazabicyclo[2.2.2]octa-2,5-diene (13)80 decomposes at -78 °C, whereas the gas-phase pyrolysis of benzotriazole (14)81 (Scheme 2.3) occurs at 1100 °C. In order for the porphyrin pyrazolines described in this work to be applied as crosslinking agents in PDT, they would have to be stable at physiological temperatures. As a result, both the thermal and photochemical activation of the protoporphyrin pyrazolines will be examined in this chapter. Scheme 2.3 Range in temperatures required for thermolysis in cyclic azo compounds 47 Results and Discussion 2.2.1 Thermal Activation It was anticipated that the protoporphyrin pyrazolines in this work would also undergo a thermal cycloelimination reaction through the extrusion of nitrogen. The course of the cycloelimination of nitrogen on addition of thermal energy is shown in Figure 2.10 (adapted from literature ). In the rate determining step, substrate (A) cleaves a single bond of the azo function either homolytically or heterolytically, giving rise to a generally energy-rich diradical (C) or zwitterionic species (only the former will be illustrated here for clarity), (A—>B—>C). Nitrogen is eliminated in a second activated step, and the primary fragment, namely a diradical (E), is formed (C—>D—>E). This second diradical intermediate subsequently forms the cyclopropane (F) by ring closure (E—>F). Reaction coordinate • Figure 2.10 Course of thermal cycloelimination of nitrogen 48 Results and Discussion Each of the product pyrazolines was investigated for its thermal stability towards cycloelimination by looking for decomposition products at a variety of temperatures in various refluxing solvents. The fear that these pyrazolines were not stable at or near room temperature was very real, and was intensified when it seemed as though they were breaking-down before NMR spectra could be obtained (Section 2.1). It was reassuring to find that the pyrazolines could sustain relatively high temperatures for extended periods of time with little or no decomposition. In fact, even in refluxing toluene they undergo conversion to their thermal products relatively slowly (Table 2.4). In addition, it was encouraging to observe that both the A and B-ring pyrazoline adducts reacted cleanly and gave a single product each, which were presumed to be their respective cyclopropanes. The thermal products were visible by T L C as faster moving pink spots (Rf 15 = 0.55, and R f i 6 = 0.53, 10% EtAce in CH2CI2 eluent) above their respective pyrazolines (Rf 10= 0.18 and R f 11 = 0.21). Pyrazoline Time required for complete thermolysis (reflux in toluene) to cyclopropanes B-ring Isomer (10) 23 h A-ring Isomer (11) 21 h Di-Adduct (12) 47 h Table 2.4 Thermolysis of pyrazolines The di-adduct (12), on the other hand, formed three products in the course of the reaction. For the di-adduct it was thought that the three pink bands visible by T L C (Figure 2.11) represented the two possible products formed by cycloelimination from the pyrazoline moieties on the A and B ring, and a third product from cycloelimination from both (17). 49 Results and Discussion Thus the two bands directly above the di-adduct pyrazoline represented the products whereby one ring-pyrazoline had reacted while the other remained intact, and the fastest moving band represented the porphyrin where both pyrazolines reacted to give the di-cyclopropane product (17). Disappearance of the two middle spots in the T L C as the reaction proceeded helped to verify this suspicion. The intermediates were not isolated. The thermal reactions described above were performed on relatively small scales, and the products were not fully characterized. The identities of the thermal products (15, 16, 17) were confirmed by MS, and by comparing the Rf values and UV-vis spectra of the products in question to their photochemical counterparts, which were characterized extensively and are discussed next. In summary, these simple thermal studies verified that the pyrazolines are stable to relatively high temperatures and have an extremely long half-life at physiological temperature, which made them suitable for the photochemical studies described next. 50 Figure 2.11 Schematic of T L C plate: Thermolysis of di-adduct pyrazoline (8% EtAce/CH2Cl2eluent) Results and Discussion 2.2.2 Photochemical Activation The protoporphyrin pyrazolines were then investigated for their susceptibility to eliminate nitrogen on electronic excitation, in accordance with the thrust of the project, to create a reactive intermediate that will crosslink cellular components. The spectral characteristics82 of the cw-azoalkanes (pyrazolines) reveal that the photochemical excitation process is n-7t* derived. The n-it* absorption of a pyrazoline lies in the 300-400 nm range. Although the n-7t* transition is much weaker (ca. 100-fold) than the K-K* transition (e = 104-105), the lower energy light of the former is convenient for the photoextrusion of nitrogen from azoalkanes. " Scheme 2.4 Photochemistry of the pyrazoline functionality Irradiation of degassed, dilute solutions [3x10_ M] of the protoporphyrin pyrazolines 10, 11 and 12 in benzene with 350 nm light produced a single photoproduct in each case. The photochemistry is thought to proceed initially through a diazenyl biradical intermediate84 (11a), followed by extrusion of nitrogen, yielding a biradical (lib), which reacts intramolecularly to form the cyclopropane (16). Scheme 2.4 illustrates the proposed reaction scheme for the A-ring pyrazoline (11). Analogous reactions would 51 Results and Discussion lead to the formation of the respective cyclopropanes (15, 17) of the other mono-adduct (10) and the di-pyrazoline (12) as well. 350 400 450 500 550 600 650 700 W a v e l e n g t h ( n m ) Figure 2.12 UV-vis spectra of A-ring pyrazoline of PP-DME and its photoproduct Unfortunately, the absorption spectra of the photo-products did not differ markedly from that of the pyrazolines, and the reaction could not be monitored conveniently by UV-vis spectroscopy (Figure 2.12). The photoproducts were visible by T L C , and appeared as a faster moving pink bands above the pyrazoline (identical R f values as the thermal products, Section 2.2.1). Additional characterizations of the product cyclopropanes (15, 16, 17) were carried out by mass spectrometry, elemental analysis and 'H NMR spectroscopy, and are outlined in detail in the experimental section. Figure 2.13 compares the proton NMR spectra of the pyrazoline (11), and its photoproduct (16) to illustrate the changes that are observed. What is readily apparent is the disappearance of the pyrazoline signals, and the appearance of distinctive signals in the upfield region of 52 Results and Discussion 7 .0 6 .5 6 .0 5 .5 5 .0 4 . 5 4 . 0 3 .5 3 .0 2 . 5 2 . 0 1.5ppm Figure 2.13 Changes in [ H NMR spectra of pyrazoline (11) and photoproduct (16) the spectrum (H-31: 3.06 ppm, H-32: 1.66 and 1.45 ppm) that can be attributed to the cyclopropane. Figure 2.14 shows expansions of these cyclopropane signals. Regrettably, the signal at 1.45 ppm is partially obscured by residual water in the sample. Normally, one would expect to find these signals even higher upfield (between 0 and 0.5 ppm)8 5, but what cannot be dismissed is the large deshielding effect of the porphyrin ring. Similar features were found for the cyclopropane photoproduct 15. A notable spectroscopic distinction between the di-adduct pyrazoline (12) and its photoproduct di-cyclopropane (17) is the absence of diastereomers in the latter. Formation of the cyclopropanes removes the stereocenters in the pyrazolines and the 53 Results and Discussion Results and Discussion result is a single product visible in the NMR spectrum (complete signal assignment can be found in the experimental section of this work). With all these pieces of the puzzle in place, there was not much doubt with respect to the identity of the photoproducts. Throughout the course of the characterization however, an attempt was made to grow single crystals of each of the products. Single crystals of free-base porphyrins, as it happens, are very difficult to grow and present a diffraction challenge to the crystallographer. Their planar nature yields very thin crystals and they tend to stack and twin. The twinning in particular makes it difficult to isolate single crystals, thus it was a fortuitous event that the structure of the photoproduct of the A-ring pyrazoline was obtained (Figure 2.15). As mentioned in Section 2.1, this X-ray structure also helped to confirm the NOE assignment of the A versus B-ring adducts. Unfortunately, suitable crystals for X-ray analysis of the other cyclopropane products could not be isolated. In Section 1.2.1, the phenomena of long wavelength light penetration through tissue was explained. Out of curiousity, and as a result of the findings detailed in the next section of this work, the possibility of activating the protoporphyrin pyrazolines with long wavelength light was investigated. With the prospect that these compounds may someday play a role as PDT photosensitizers, red light activation would be very beneficial. The first challenge in this experiment was finding a suitable filter to allow for selective long wavelength irradiation and restrict short wavelength irradiation. The protoporphyrin family of tetrapyrroles have but a weak absorption at long wavelengths, 55 Results and Discussion Figure 2.15 X-ray crystal structure of 16 56 Results and Discussion thus the window for activation was not very big; a suitable filter was found and its transmission spectrum is shown in Figure 2.16. The results of the long wavelength photochemical activation were encouraging. Each of the A and B-ring pyrazolines (10, 11), as well as the di-adduct pyrazoline (12) formed their respective cyclopropanes (15, 16, 17) on 70 60 0 5 0 o 1 40 I 30 20 10 0 10 to Corion 600 \ 1 1 I 1 ! 1 1 I 1 ' 1 1I 1 1 1 1 I 1 1 1 1 i 1 l _ r n I 1 1 1 1I 1 1 ' 1 ! 350 400 450 500 550 600 650 700 750 Wavelength (nm) Figure 2.16 Transmittance of red light filter irradiation with filtered long-wavelength red light. Degassed dilute benzene solutions of the pyrazolines with similar concentrations to those irradiated with 350 nm light were exposed to the red light, and complete conversion to corresponding cyclopropane(s) (identical to those obtained with 350 nm irradiation) was achieved in approximately 48 hours. The fact that this low energy light was sufficient to bring about the extrusion of molecular nitrogen was surprising since the pyrazoline chromophore is known to absorb between 300-350 nm. Thus, the efficient nitrogen extrusion to form the cyclopropanes using long wavelength irradiation is not expected and these results raise many interesting photophysical questions. These, in addition to a series of experiments exploring the crosslinking abilities of this series of pyrazolines will be discussed in Section 2.4. 57 Results and Discussion 2.3 Synthesis and Characterization of Methylpyropheophorbide Pyrazoline As mentioned in Section 1.2.1, the efficacy of a porphyrin-based PDT drug depends on its ability to absorb light in the red or near infrared region of the absorption spectrum. The nature of the optical spectrum is determined by the level of saturation of the tetrapyrrolic macrocycle. Consequently, research towards new photosensitizers (so-called second-generation photosensitizers) in recent years has focused on chlorins and bacteriochlorins. I Q Methylpyropheophorbide (MePPP, 18) and its derivatives meet many of the important criteria for a suitable PDT agent. MePPP is an effective producer of singlet oxygen. A chlorin, this chromophore has an appreciable absorption in the visible region at wavelengths greater than 630 nm (Figure 2.17). OCH3 Derived from a natural product, the drug consists of a single isomer. Several derivatives have also demonstrated an ability to localize in desired tissue, and clear from the body in a relatively short time post-irradiation.88 In addition to these chemical, photochemical and biological features, practical considerations such as ease, yield and cost of production (it comes from spinach!) also contribute to its strengths. These attributes, and the presence of a single vinyl substituent on the A-ring of the macrocycle prompted the present investigation of a cycloaddition reaction similar to that 58 Results and Discussion performed on PP-DME described in the previous section. Initially, it was feared that multiple products might result from the reaction of diazomethane and 18. The reaction of diazomethane with ketones is known and has been studied for some time.68 In the case of cyclic ketones, such as the one in the exocyclic ring fused with the C-ring of 18, the dominant reaction pathway is expected to be ring expansion. This can be complicated by the conflicting migratory aptitude of the involved carbon atoms89, and reaction of the product with excess diazomethane to produce undesired higher homologues.90 Fortunately, the reaction of diazomethane and 18 formed a single product. On one occasion, when the reaction was left for an extended period of time, a second product was visible by T L C (Rf = 0.2 vs. R f 19 = 0.44 and R r 18=0.67, 10% EtAce/Ch 2Cl 2 eluent), but not enough was present for isolation or characterization. From the beginning, the main product of the reaction was expected to be the pyrazoline of MePPP (19, Scheme 2.5) based on the experience with diazomethane and PP-DME. The increased solubility of the starting material in CFLCL (compared with PP-DME) and the experience gained from the previous reactions facilitated this procedure greatly. A 19 large excess of diazomethane: porphyrin (50:1) was still Scheme 2.5 Numbered pyrazoline of MePPP required, but the fact that only one product was formed and that the reaction could go to completion, simplified the isolation. The product was purified by column chromatography to give a final yield of 78% yield. 59 Results and Discussion 350 400 450 500 550 600 650 700 750 Wavelength (nm) Figure 2.17 UV-vis spectra of 18 and 19 in CH2CI2. The spectrum of 19 is scaled by a factor of three for ease of comparison. As mentioned on several occasions, the absorption characteristics of the chlorin distinguish it from other types of tetrapyrrolic macrocycles. The first indication that 19 had in fact been formed was given by the UV-vis spectrum of the product (Figure 2.17). The expected cycloaddition of diazomethane would bring about a reduction in conjugation. As a result, we might expect to see a hypsochromic (higher energy) shift in spectrum of the product pyrazoline. It is not an extremely large shift, but the difference in the first Q-band is noteworthy; Q{ (MePPP, 18): 669 nm vs. Qj (19): 663 nm. The spectra are very similar, indicating the overall electronics of the chlorin have not been altered drastically by the pyrazoline functionality. Slight variations in the Q-bands can be quite informative nonetheless. Overall, it was a relief to find that the intense Q-band of 19 had not shifted to much lower wavelengths, as it is hoped that this chlorin will be photoactive at long wavelengths. 60 Results and Discussion Characterization of 19 was also facilitated by the extensive NMR spectroscopy experiments carried out previously on the pyrazolines of PP-DME. Figure 2.18 (p.66) displays the NMR spectrum of 19, and the familiar pyrazoline signals are present: HL-3': 6.67 ppm, H 0 -3 3 : 5.47 ppm, H P-3 3: 4.76 ppm, H M -3 2 : 2.75 ppm and H N -3 2 : 2.11 ppm. Assignment ! H NMR (500 MHz) H M Q C 1 3 C H-x 8 ppm (mult., J (Hz)a) Correlations (125.8 MHz) H-10 9.51 (s, 1H) 104.08 H-20 8.96 (s, 1H) 96.69 H-5 8.57 (s, 1H) 93.01 H L-3' 6.67 (m, 1H, J L . M , L-N= 9.7, J L . 0 . L-P= 2.2) 85.32 H 0 - 3 3 5.47 (m, 1H, J0-P= 17.6, J 0 . M = 10.1, J0-N,O-L= 2.5) 78.01 H-151 5.26 (d, 1H, J = 19.6) 48.01 H-151 5.11 (d, 1H, J = 19.6) Hp-33 4.76 (m, 1H, J P . 0 = 18.2, J P . M , P.N= 9.1, J P - L = 3.1) 78.01 H-18 4.48 (m, 1H, J d o u b = 7.4, J q u a r t = 7.1) 51.01 H-17 4.29 (m, 1H, Jdoub = 8.1, Jd O U b= 2.2, J d 0 U b = 2.2) 51.75 H-172 3.68 (t, 2H, J = 7.4) 19.38 H-175 3.65 (s, 3H) 51.61 H-7 1 3.59 (s, 3H) 11.94 H-21 3.25 (s, 3H) 11.45 H-121 3.17 (s, 3H) 11.11 HM-3 2.75 (m, 1H, JM-N=19.7, JM-p= 7.6, JM-o, M-L= 2.4) 27.32 H-171 2.68 (m, 1H) 29.85 H-81 2.55 (m, 1H) 30.93 H-81 2.28 (m, 1H) H-171 2.28 (m, 1H) 29.85 HN-3 2 2.11 (m, 1H, JN-M=19.6, JN-L,N-O=9.7 27.32 H-181 1.80 (d, 3H, J = 7.4) 23.10 H-8 2 1.67 (t, 3 H , J = 7.7) 17.33 NR -1.77 (s, 2H) a- Only those J values that could be unambiguously assigned are recorded Table 2.5 H M Q C correlations for 19. 61 Results and Discussion The consistency within the signals of the pyrazoline functionality for PP-DME and MePPP are quite remarkable in fact. Comparing the pyrazoline signals of the mono-adduct 11 with 19 for example, the difference in shift varies only ±0.14 ppm in the most extreme case, and the coupling constants differ by less than 0.4 Hz in all cases. The complications that arose in characterizing 19 were a result of two factors. In forming the pyrazoline, a new stereocenter is created at C 3 ' (Scheme 2.5). Relative to the fixed natural stereocenters at Q 7 and Ci8, a diastereomeric mixture results. Expansion of the NR signal at -1.77 ppm reveals a broad peak (broader than that which results from the rapidly exchanging amine protons), indicative of two species in solution. Other features such as fine splitting in the meso protons (H-5: 8.57 ppm, H-10: 9.51 ppm, and H-20: 8.96 ppm) also point to the presence of diastereomers. The other difficulty in completing the assignment of signals of 19 was a result of the large number of signals found in the region of the H N and H M pyrazoline signals (2.11 -2.75 ppm). Also found in this region, and overlapping with the pyrazoline signals, are the methylene signals from the ethane moiety (H-81) and from the methyl propyl ester (H-171). H M Q C spectroscopy was used to resolve the signals in question, by assigning each of the proton signals to their respective carbon signals (Table 2.5). Thus the pairs of proton signals were resolved, J values were extracted, and each of the 38 protons signals of 19 was assigned. It was reassuring to observe the proof of concept in forming the resultant pyrazoline of MePPP. The reaction of diazomethane with the vinyl group of a chlorin, known for its 62 Results and Discussion PDT potential, was successfully carried out. Furthermore, the predictions based on theoretical calculations in Section 2.1 were once again supported experimentally in forming the unsymmetrical pyrazoline functionality. With the characterization complete, the more interesting task of exploring the photochemistry of this novel chlorin can now be explored in the next section. 63 Results and Discussion 2.4 Photochemistry of Methylpyropheophorbide Pyrazoline The pyrazoline of methylpyropheophorbide (19) makes for an interesting photochemical study. A chlorin, this molecule has a strong Q-band absorption at 665 nm (e = 52, 700, in benzene) whose intensity is almost half that of its Soret band (411 nm, e = 107, 100) (Figure 2.19). As mentioned previously in Section 1.2.1, this region of the visible spectrum is particularly useful for light penetration through tissue. In Section 2.2.2, the photochemistry of the PP-DME pyrazolines was investigated. In this section, the photochemistry of 19 will be discussed. In particular, an attempt will be made to rationalize the photochemistry that takes place both from irradiation in the near-U V region (350 nm), and at long wavelengths (672 nm). To the best of our knowledge, this is the first report of photochemical activation of functionalities at the periphery a chlorin derivative. Since chlorins are known to be efficient singlet oxygen sensitizers, irradiations were performed in Ar or N 2 saturated solutions. Irradiation of a dilute [3 x 10'4M], oxygen purged solution of 19 in Scheme 2.6 Formation of 20 from 19 b e n z e n e w i t h 3 5 0 n m H § h t produced a single photoproduct. It was expected that similar chemistry to the pyrazolines of PP-DME (Scheme 2.6) would be observed. Thus, via the proposed diazenyl biradical 64 Results and Discussion intermediate, extrusion of nitrogen, and intramolecular closure of the resulting biradical, a cyclopropane (20) would be formed. The photoproduct was visible by T L C as a faster moving (Rf = 0.61, 10% EtAce in CH 2 C1 2 eluent), single light green spot above the pyrazoline (Rf = 0.48). This decrease in polarity in the photoproduct was the first good indication that the cyclopropane had been formed. NMR spectroscopy was relied upon again in the identification of 20. Figure 2.19 compares the spectra of the pyrazoline (19) and photoproduct (20). Particularly notable is the disappearance of the pyrazoline signals (HL-3', Ho-33, H P -3 3 , H M -3 2 , HN-3 2). The methylene cyclopropyl signals (H-32: 1.57 ppm, and H-32: 1.30 ppm) are somewhat difficult to observe due to some residual ethyl acetate (q: 3.65, s: 2.00, t: 1.25 ppm) and water (s: 1.5 ppm) in the sample. Their J values (1.9 and 1.6 Hz respectively) were extracted from expansions however, and correlate with those expected for cyclopropyl protons. The distinctive multiplet observed at 2.80 ppm, seems to couple with the methylene cyclopropane signals at 1.57, and 1.30 ppm. (J ~ 7 Hz); although the ambiguity of the splitting made it difficult to assign a definitive value. Other differences in the spectrum, such as the shifting meso protons between 8 and 10 ppm all indicate that the macrocyle has been altered, but it is difficult to rationalize each of these subtle changes. A comprehensive assignment of each of the signals can be found in the 65 Results and Discussion Results and Discussion experimental section. Coupled with MS, and UV-vis spectroscopy data, the proton NMR is nonetheless very telling and confirms the formation of the cyclopropyl photoproduct 350 400 450 500 550 600 650 700 750 Wavelength (nm) Figure 2.19 UV-vis spectra of pyrazoline (19) and its photoproduct (20) PDT agents are activated at long wavelengths. It is obvious from the absorption characteristics of 19, that it too absorbs low energy light. What was not known was whether the same photochemistry, if any, could be observed by irradiation in this low energy absorption. Thanks to a light emitting diode (LED) panel that emits at a single wavelength (672 nm) in the red region of the optical spectrum, this aspect of the photochemistry was explored. Irradiation of an oxygen purged, dilute [3 x 10'4M] solution of 19 in benzene with 672 nm light produced a single photoproduct. The T L C and the spectral characteristics of this (20). MePPP Pyrazoline MePPP Cyclopropane 67 Results and Discussion product were identical to that obtained on 350 nm excitation; suffice to say that isolation and all characterization led to the identification of 20 again. Fundamentally, we did not expect this chemistry to be possible at such a long wavelength activation. This is based on the fact that only 43 kcal/mol of energy are produced by 672 nm light and may not be sufficient to initiate a reaction that involves the cleavage of a C-N covalent bond in the pyrazoline. In order to ensure that the product observed was due to photochemistry and not a thermal reaction that was taking place as a result of the heat given off by the LED, some simple control experiments were carried out. The temperature of the photoreaction was monitored by placing a thermometer in the solution (inserted through the rubber septum, which was then sealed). At no time did the temperature of the solution rise above 30 °C. By way of comparison, after heating 19 in various solvents it was found that efficient conversion to 20 required over 42 hours of refluxing in toluene. In addition, concurrent with every photoreaction, a 'dark' reaction was carried out. A foiled vial containing the same solution as that which was irradiated was placed in front of the light source, and its products monitored. No products were ever observed from any of the dark solutions. The evidence herein suggests that photochemistry at both long (672 nm) and short (350 nm) wavelengths leads to efficient extrusion of nitrogen from the pyrazoline moiety to yield the cyclopropane derivative 20. As mentioned in Section 2.2.2, from the spectral characteristics of cz's-azoalkanes82 (pyrazolines) it is clear that the excitation process in light activation of simple systems 68 Results and Discussion involves the n-71* state which lies in the 300-400 nm range. Since pyrazolines are known to absorb in the near U V they would not be expected to be directly excited at such low energy wavelengths. This poses an interesting mechanistic and photophysical question. The UV-vis spectra of 19 and 20 are very much alike (Figure 2.19). With the intensity of the chlorin bands (Qi 19: 665 nm, 52,700 vs. Qi 20: 658 nm, 46,400) and the observed subtle blue shift (~7 nm) between reactants and products, the reaction progress was monitored. By watching the spectrum change as the photoconversion proceeded, we gained insight into the mechanism of this reaction and got an idea of the rate of conversion. The following spectra (Figure 2.20) were taken at timed intervals (-5-10 minutes), and show the distinct disappearance of the pyrazoline, and formation of the 1.2-1 1 9 0.8 o c .Q 0 0.6 CO . a < 0.4 0.2-0 Isosbestic Point 1 1 1 I 700 1 , 1 1 j , , 1 r—, 1 1 1 1 1 1 1 1 ! 1 1 1 1 1 1 r 640 650 660 670 680 690 Wavelength (nm) Figure 2.20 UV-vis spectra of chlorin bands, monitoring photoconversion of 19 to 20 69 Results and Discussion cyclopropyl MePPP. The isosbestic point was identified at 662 nm. Identifying an isosbestic point was informative, because it meant that one product (19) was being converted to another (20), without an intervening product. Slight disparity between UV-vis measurements is apparent in Figure 2.20. This is the result of the fact that the L E D could not be installed inside the UV-vis chamber, so there was a temperature discrepancy between the lamp and the recording instrument which might affect peak position. While the reaction proceeded quite efficiently with long wavelength excitation, the rate of conversion decreased as the reaction proceeded and complete conversion of solutions of 19 in benzene (monitored by TLC) took several hours. This is most likely due to an inner filter effect of the product competing for light absorption as the reaction proceeds. However, it is noteworthy that this photoreaction could indeed be taken to completion cleanly, with no secondary photolysis products evident. The emission spectra of the pyrazoline of MePPP (19) and its cyclopropane (20) were also obtained. Irradiation at both the Soret band at 350 nm and at longer wavelength of CD O c CD 8 -I s o 666 669 / i i i i: Cyclopropane Pyrazoline 600 620 640 660 680 700 Wavelength (nm) 720 740 Figure 2.21 Emission spectra of 19 and 20. 70 Results and Discussion 665 nm yielded identical fluorescence spectra (Figure 2.21). It is therefore assumed that fluorescence decay competes with reaction from the pyrazoline. The emissions appear in the region where they would be expected; each show only a minor Stake's shift (19: 658—>666, A = 8 nm , 20 665—>669, A = 4 nm respectively.) Relative quantum yields of emission of 19 and 20 were not estimated. Extensive research of a wide variety cyclic azo compounds by Herbert Meier 7 9, has contributed to a more thorough understanding of their photochemical behaviour. Figure 2.22 illustrates graphically what is known about the photochemical cycloelimination of nitrogen and reviews the proposed mechanism. Cyclic azo compounds like the pyrazoline can be activated by direct irradiation, usually in the longest wavelength band s 0 J = ^ = Figure 2.22 Course of the photochemical cycloelimination of nitrogen. Adapted from Ref. #79 71 Results and Discussion (350 nm), and often also by triplet-sensitization (Sens.)- It is rarely possible to reach the T| level by intersystem crossing (ISC) without a sensitizer, so most reactivity ensues from the singlet state, ^-elimination can occur directly from Si or T| , by crossing of a repulsion term (predissociation), or by internal conversion (IC) to a highly vibrationally excited electronic ground state S 0 V. According to many theoretical studies,91 the conversion from the biradical resultant from nitrogen extrusion to cyclopropane involves a very small energy barrier. Therefore, it is not a kinetically relevant intermediate and the rate-determining step involves the initial homolysis of the C-N bond in the pyrazoline. The question that remains is why does the long wavelength excitation of 19, which is presumably excitation of the porphyrin, lead to efficient elimination of nitrogen from the pyrazoline unit? Of the possible explanations, intramolecular energy transfer from the porphyrin to the pyrazoline is tempting because of the porphyrins known sensitizing abilities. Thus the porphyrin moiety (M*) would absorb the incident light and transfer its excitation energy to the pyrazoline moiety (Q), affecting the photochemistry without the latter having to absorb the incident light. M * + Q -> M + Q* Energy transfer of this type could be ruled out if we consider the intermolecular case of an excited state chlorin reacting with pyrazoline. This is based on the fundamental requirement for energy transfer that the energy level of Q be lower than (or at the very 72 Results and Discussion 92 most equal to) that of M * . Thus if we approximate the energy levels available to the porphyrin, and consider that the singlet states of the n-Tt* transition of pyrazolines have excitation energies Es of ca. 70-90 kcal/mol and for the triplet states E T is ca. 50-60 kcal/mol (illustrated in Figure 2.23), energy transfer is an extremely endothermic process. 80 kcal/mol A 34.4 kcal/mol 672 nm Figure 2.23 Energy level diagram of porphyrin and pyrazoline. Another possible mode of action for extrusion of nitrogen, in particular in vivo, might be a photoinduced electron transfer, although there is no precedence for this is reaction with pyrazolines. Again, if we consider the porphyrin and pyrazoline as separate molecules one can estimate the feasibility of an electron transfer process. Generally, a photochemical electron transfer occurs only when it is exergonic or <5 kcal mof1 endergonic.93 The free energy change for a photoinduced electron transfer can be determined from the oxidation potential of the electron donor (E°D) , the reduction potential of the electron acceptor ( E ° A ) and the excited state energy of the sensitizer 73 Results and Discussion (E0,o), using the relationship A G E T = 23.06 kcal mol - 1 [ (E° D -E° A ) - E 0, 0] - C ; where C is a term that accounts for solvation. In order to estimate the feasibility of electron transfer using this relationship, the redox properties of the molecules must be known. Each of the starting material MePPP (18), the pyrazoline (19), and the photoproduct cyclopropane (20) were examined by electrochemical methods. Table 2.6 lists both the reduction and oxidation potentials of the methylpyropheophorbide series, while Figure 2.24 depicts the cyclic voltammogram of the pyrazoline. The C V reveals the reversible nature of the reduction waves of 19 (18 and 20 are also reversible). The radical anions and dianions are therefore stable on the time scale of these experiments, with no chemistry occurring on addition of electrons to these molecules. If any chemical reactions were occurring from the radical anion or dianion, the CV's would be irreversible. The reduction potentials do not vary a great deal, which tells us that these particular functionalizations of MePPP did not change the overall electronics of the molecule to a great extent. On reduction, two consecutive electrons are accepted by the porphyrin macrocycle in each of 18, 19 and 20. The oxidations are also reversible; the only notable feature of the oxidation potentials is that 19 and 20 are somewhat easier to oxidize than MePPP. Compound Reduction 1 Reduction 2 Oxidation 1 Methylpyropheophorbide (18) -1.47 -1.73 0.61 Pyrazoline of MePPP (19) -1.48 -1.76 0.45 Cyclopropane of MePPP (20) -1.52 -1.83 0.43 Table 2.6 Summary of redox potentials measured in dichloromethane at 22°C versus Fc. Details in the experimental section. 74 Results and Discussion R2 Current R l 02 J N R4 01 R3 1 . 5 1 . 0 0 . 5 0 . 0 - 0 . 5 - 1 . 0 - 1 . 5 - 2 . 0 E vs. S C E Figure 2.24 Cyclic voltammogram of 19 measured in methylene chloride containing 0.2M Bu 4 NPF 6 on a Pt electrode at 0.2V s"1. With the estimates for the reduction potential of the porphyrin from the electrochemistry and the E0 io = 1.8 eV (or 43 kcal mol"1) available from the absorption and fluorescence data, the feasibility for an electron transfer can be estimated. In order for electron transfer from the pyrazoline to the porphyrin to be exergonic the oxidation potential of the pyrazoline would have to be less than 0.3 V. Even accounting for the solvation term, the oxidation potentials of pyrazolines are too high to make electron transfer feasible in this case. Nonetheless, the electrochemical data acquired for these compounds may be useful in future applications. In biological systems, donors and acceptors are ubiquitous. When 75 Results and Discussion the question of whether or not electron transfer will occur with this drug and a specified substrate, this redox information will enable the investigator to predict the outcome. All these investigative experiments, UV-vis product study, emission spectra, and the electrochemistry, provide valuable information about the pyrazoline of MePPP (19) and its photoproduct cyclopropane (20). The one message that is clear, is that it is wrong to think of the pyrazoline as a distinct entity from the porphyrin. Initially, the lack of conjugation between the pyrazoline and the porphryin led to this assumption, but many examples exist of molecules where two chromophores are separated by far more than a single carbon, and interaction between the groups is still observed.94' 9 5 All the evidence suggests that this functional group is an integral part of the chromophore, and is intact with the overall electronic structure of the macrocyle. Throughout this quandary, a statement made early in our amazement that long wavelength light was able to effect the observed chemistry kept resurfacing: "About 75 kcal/mol are required to break the C-N covalent bond, and only 43 kcal/mol of energy are produced by 672 nm light." Although this is still true, what may have been neglected is the overall free energy of the reaction. In a theoretical study by Branchadell,91 the picture becomes more clear. They investigated the mechanism of addition of diazomethane to ethylene, as well as the nitrogen elimination reaction from the cycloadduct. In considering entropy and enthalpy values96 the exergonicity of these reactions is highlighted. Figure 2.25 is a free energy diagram for the reactions of the pyrazoline. 76 Results and Discussion N; N + C H 2 N I 23.5 1 2 23.0 f 1 . N W N ••N N 35.2 43.4 7^ Figure 2.25. Energy diagram for the formation and dissociation of pyrazoline (kcal/mol) In summary, upon absorption of a photon of long wavelength light, 19 forms an excited state with sufficient energy (672 nm = 43 kcal/mol) to extrude nitrogen from its pyrazoline moeity (E a = 35.2 kcal/mol). While the mechanism involved in long wavelength irradiation is not known, it is be feasible to imagine a mechanism similar to that invoked by the short wavelength irradiation The fact that this low energy light is enough to bring about the observed photochemistry at the periphery of a porphyrin is very encouraging for the design of future potential crosslinking photosensitizers. 77 2.4.1 Crosslinking Attempts in Solution Results and Discussion Until now, the photochemistry that has been discussed was carried out in benzene; an 'inert' solvent in terms of its potential to react with radicals in solution. The title of this thesis reminds us of the impetus for synthesizing these pyrazoline porphyrins: "Synthesis and Photochemistry of Azo-Type Derivitized Porphyrins with Crosslinking Applications." It is hoped that these derivitized PDT agents will have the ability to crosslink within their biological environment. As mentioned in Section 1.3, these functionalities were chosen based on their ability to form reactive intermediates on photoactivation. Before attempting in vitro assays on venerable cell lines with this series of compounds, several experiments were carried out in solution where we attempted to trap the 1,3 biradical intermediate. Literature precedence was found that stated that 1,3-diradical species like the intermediate created from photolysis of the pyrazolines can be trapped by alkenes97, isocyanates98'99, and enamines100, which led to some of the first attempts in Table 2.7. The results were disappointing to say the least. Only the cyclopropane was observed, except for those cases where degradation of the starting material or polymerization of the solution made it impossible to recover any products. A paper by Jerome Berson gave hope to these crosslinking efforts citing their experience with various substrates: "The efficiency of the cycloaddition, as measured by the competition between adduct formation and cyclization, is greater with conjugated olefins than with simple ones. 1,3 -Cyclohexadiene, 2,4-hexadiene, methyl acrylate, acrylonitrile, fumaronitrile and crotononitrile all form adducts in high yield, whereas 78 Results and Discussion cyclohexene, dihydropyran, and 1,2-diethoxyene do not...conjugated acetylenes do not seem to be good trapping agents...decomposition of cyclic azo X in the presence of dimethyl acetylenedicarboxylate gives no adduct and a substantial amount of a polymeric product." 9 7 ( c ) Even though a lesson was learned from these previous attempts -not to try to crosslink with D M A D or unconjugated olefins- efforts to react the porphyrin pyrazolines with conjugated olefins were unsuccessful. Substrate Concentration Result Styrene (solution in benzene) 0.28M No reaction Styrene (neat) 8.70M No reaction Dimethylacetylene Dicarboxylate (neat) 8.10M No reaction/ Polymerization Tributyltinhydride (neat) 3.72M N o reaction/ Polymerization Diethylamine (neat) 9.70M Degradation of S M Ethyl Acrylate 9.23M No Reaction Cyclohexadiene 10.57M -Table 2.7 Results of crosslinking attempts in various reagent solutions (no reaction is synonymous with cyclopropane formation and recovery) The conclusion, it seemed, was that the only hope for trapping this incredibly short-lived reactive intermediate would be in a very concentrated environment. However, even upon irradiation in neat solutions of the solvents listed in Table 2.7 only the cyclopropane was formed. The lifetimes for the biradicals mentioned in the Berson papers were on the order of x = 5-20 ns. A quick calculation comparing the resultant rate of intramolecualr cyclisation with the bimolecular crosslinking reaction leads us to believe that the lifetime of the biradical created from the porphyrin pyrazolines is much shorter than originally anticipated. Other factors, including sterics, may also be inhibiting the bimolecular reaction. 79 Results and Discussion Solution studies alone may not reveal the modified photosensitizers' crosslinking abilities. In vivo reactions of this nature would likely occur more readily due to the proximity of the crosslinking agent and closely associated cellular components. A variety of in vitro testing regimes are available to assess these crosslinking systems in a biological environment,61 and will be discussed in Chapter 3. Since the lifetimes of biradical intermediates are very short, and because these can undergo hydrogen atom abstraction and other competitive reactions, their efficacy as crosslinking agents may be limited. For this reason, other (longer lived) reactive intermediates will also be investigated, and the progress towards one such functionality is discussed in the next section. On a positive note, were nothing to come of the pyrazoline as a crosslinking agent, it is nonetheless a fortuitous event to have discovered a means of functionalizing porphyrins with cyclopropyl moieties. The cyclopropane ring, due to its unusual bonding and inherent ring strain (27.5 kcal/mol) is unique among carbocycles in both its properties and reactions. Thus cyclopropane derivatives provide building blocks of unprecedented synthetic potential. Moreover, natural and synthetic cyclopropanes bearing simple functionalities are endowed with a large spectrum of biological properties ranging from enzyme inhibition to antibiotic, antiviral, antitumor and neurochemical properties.101 These novel cyclopropyl derivitized photosensitizers could therefore hold promise in areas not yet considered. 80 Results and Discussion 2.5 Progress Towards the Synthesis of Diazirine Tetraphenylporphyrin The second series of compounds of interest is not dependent upon a naturally occurring starting material. This becomes an important consideration for potential PDT agents. With the fear of contaminants causing illness such as 'mad cow disease' threatening natural blood sources, it OH seems judicious to explore alternative porphyrin Figure 2.26 Foscan® families. As mentioned in Section 1.1.5, tetraphenylporphyrin (TPP) is the simplest and least expensive porphyrin to make. These attributes combined make such a parent molecule very attractive to the pharmaceutical industry. In addition, the phenyl groups of TPP are easily modified, and many derivatives are available. An example of one such TPP-derived photosensitizer exists in 'Foscan' 1 0 2' 1 0 3 (Figure 2.26), a promising drug that is currently undergoing clinical trials in Europe. The modification to TPP envisioned for a model crosslinker herein is inspired by the efficient crosslinking abilities of the diazirine functionality (Figure 2.27). As mentioned in Section 1.3, 25 photochemically induced extrusion of nitrogen from Figure 2.27 Diazirine-TPP diazirine yields a relatively long-lived and reactive 81 Results and Discussion carbene species that is capable of insertion into C-H bonds or addition to C-C multiple bonds. From the wide variety of functionalities able to produce reactive intermediates including phenyl azides, diazo groups, and benzophenone derivatives, the diazirine was chosen because of its particular attributes. Recent literature accounts also emphasize the functionality's ability in photolabeling of biological substrates.63'104'105 The diazirine unit is small, non-bulky, and lipophillic. The carbene intermediate derived from photolysis of the diazirine has a lifetime in the milliseconds50, unlike the biradical intermediate derived from the pyrazolines, increasing its propensity to react with its surroundings. Its chromophore extends significantly into the 300 nm range, which means that its activation is non-destructive to biological tissue and also increases the chances of its activation by the porphyrin's absorption. The discoveries in the previous sections, which indicate the porphyrin is capable of long wavelength absorption and consequent activation of functionalities at its periphery, are encouraging for the predicted photochemistry of this target molecule. At this time, it should be noted that there exists a parent molecule that might be more appropriate for this model photocrosslinking compound; one based on diphenylporphyrin (DPP). Figure 2.28 illustrates the predicted photochemistry for a diazirine-modified DPP (either in the (3 or meso position) versus that expected from diazirine-TPP. Thus in the case of diazirine-TPP, there is a great likelihood that the reactive intermediate carbene will react intramolecularly with an adjacent phenyl ring to form a stable exocyclic six-membered ring. Functionalized DPP, on the other hand, could only react with a carbon-hydrogen bond four bonds away, resulting in a highly strained four-membered ring. 82 Results and Discussion Additionally, the envisioned functionalizations of DPP would likely be easier because the (3 and meso positions are less sterically hindered than in TPP. Ph Ph Ph Figure 2.28 Predicted photochemistry of diazirine-TPP and diazirine-DPP This work however, describes the progress towards diazirine-TPP and not diazirine-DPP. Quite simply, we have chosen TPP over DPP because TPP can be purchased, whereas the synthesis of DPP is lengthy, arduous, and has low yields.106 Synthesis and subsequent derivatization of DPP may still prove to be worth the time investment in the future, however it was felt that proof of concept could be illustrated with the TPP derivative, and the photochemistry could nonetheless be explored upon successful synthesis of the diazirine functionality. 83 Results and Discussion An additional point of interest regarding the trifluoromethyl diazirine functionality is the presence of fluorine atoms. In porphyrin and chlorin systems, it has been shown that overall lipophilicity of the molecules plays an important role in PDT efficacy. Coupled with the interesting solubility properties brought about by the incorporation of fluorine functionalities, there exists another intriguing aspect. I 9 F NMR can be used to provide the pharmacokinetic profiles of these photosensitizers, thus greatly increasing the ability to study mechanistic aspects of in vivo photodynamic therapy.107 The synthesis of 25 is based on a combination of two published synthetic strategies. The transformations that lead to the trifluoroacetyl-TPP (23) are based on a modified procedure by Ravindra Pandey108, who worked on octaethylporphyrin and functionalized it at he meso position (Scheme 2.7). From the trifluoroacetyl towards the diazirine, the reactions are planned according to a series of reactions that have previously been carried out on aryl systems51 (Scheme 2.8) Et Et Et Et Et Et Et Et Et Et Et Et Scheme 2.7 "Preparation of Fluorinated Porphyrins" ,OTos O HN—NH N=N R' CF3 1)NH 2OH 2) TosCI, pyr R' R' Ag zO R' "tosyloxime' 'diaziridine" "diazirine' Scheme 2.8 Synthesis of 3-trifluoromethyl-3-phenyl diazirine 84 Results and Discussion The methodology discussed below illustrates the first example of the utility of the trimethylsilyltrifluoromethyl (TMSCF3) reagent for introducing a trifluoromethyl group at the [3-position of the porphyrin macrocycle. The electron withdrawing groups most commonly employed at the meso positions of the porphyrins are the perhalophenyls and perfluoalkyl groups obtained by total synthesis.109 In addition, the reaction of porphyrin (22) (obtained by treating the related formyl analog with TMSCF3) , with tetrapropyl ammonium perruthenate (TPAP)-N-methyl-morpholine N-oxide (NMO) provides the first example of introducing the trifluoroacetyl group at the P-position of the porphyrin system 23 (Scheme 2.9). Scheme 2.9 Progress towards the synthesis of diazirine-TPP (25) 85 Results and Discussion The synthesis towards 25 got off to a slow start because of the difficulties encountered in finding a suitable metal to coordinate the central nitrogen atoms. Direct formylation attempts on the free base of TPP were unsuccessful. Introduction of divalent central metals is known to activate porphyrin reactivity.110 This is thought to occur as a result of the increased electronegativity and the 'ruffling'"', or deviation from planarity, known to occur in the porphyrin macrocycle. Certain trends have been observed; for example, the following metalloporphyrins' degree of electronegativity follows the order, MgP>ZnP>CuP>NiP>PdP.112 The problem that often follows however, is the harsh conditions required to demetallate the porphyrin after subsequent modifications. To this end, a certain amount of trial and error must be carried out before finding the conditions appropriate for the required reactivity. The first attempts to formylate TPP in the series of reactions in this work were carried out using nickel as the central metal. Table 2.8 illustrates the progress made in finding suitable metallation conditions from the vast number found in the literature."3'"4'115 Most of the problems encountered centered around the major difficulty of getting both the free base porphyrin and the metallic reagent simultaneously into the same solution under reactive conditions. This is due to the fact that good solvents for the porphyrins in their unionized forms are generally poor solvents for the simple metallic ions and vice versa. Hence the yields were low, the rates of reaction slow, and it was very inconvenient to prepare a large amount of material at one time. In addition, work-up and product purification were both troublesome and tedious. The use of DMF as a reaction solvent seemed to represent a compromise to the 86 Results and Discussion problem. While the solubility is low at room temperature, at the reflux point all of the porphyrinic material was soluble. Nickel Salt Reaction Time Solvent Temperature Yield Ni(COCH 3 ) 2 1 week CHC1 3 100°C 3% Ni(acac)2 24 h o-xylene 150°C 10-55% NiCl 2 »6H 2 0 6-24 h DMF 180°C 95% Table 2.8 TPP metallation conditions Subsequent to metallation, the task of formylating nickel TPP remained an arduous one. The Vilsmeier-Haack procedure is well known for the formylation of aromatic rings. In the formylation of nickel TPP however, the HC1 biproduct of the reaction resulted in the demetallation of the porphyrin. This competing reaction resulted in decreased yields of the desired product, and an increase of TPP in the product mixture which proved impossible to remove. Table 2.9 lists the various modifications that were attempted to increase product yield. Initially, it was thought that reactant concentrations might be the cause for poor reactivity and these were adjusted accordingly. When the suspicion of acid formation arose, two strategies were attempted to resolve the problem. First, the reaction mixture was evacuated upon formation of the Vilsmeier complex in an attempt to remove the HC1 formed. The second approach involved the addition of a proton sponge, namely l,8-bis(dimethyl amino) naphthalene107. It was hoped that this strong base would neutralize the acid formed, without interfering with the reaction thanks to the base's weak nucleophilic character. The results varied, and it seems the benefit of the proton sponge was outweighed by the amount of sponge required and the difficulty in isolating the product from the intractable mixture. 87 Results and Discussion NiTPP POCl 3 DMF Proton Sponge Reaction Time Yield 0.149 0.013 0.020 - 48 h 33 % 0.223 13.17 19.84 - 8h 28 % 3.100 31.00 46.50 - 24 h 30% 0.797 95.64 143.46 23.33 14 h 40% 3.172 380.68 571.02 95.19 12 h 20% 0.760 107.28 129.14 50.74 24 h 30 % Table 2.9. Formylation of NiTPP (all concentrations shown are in millimoles) Amidst the frustration of these early formylation reactions, a colleague in the Dolphin group facing similar challenges discovered an alternative published method. The answer was found in a translated article from a Russian publication, by Ponomarev and Maravin. 1 1 6 It so happens that the cobalt complex of tetraphenylporphyrin is considerably more reactive to formylation. Metallation is comparatively facile, and involves the addition of cobalt acetate in methanol to a refluxing solution of the porphyrin. The reaction is monitored by UV/vis spectroscopy and takes less than an hour to complete. In addition to the obvious advantages in synthesis, it was a relief to forgo the handling of the nickel salts, whose toxicity has been documented extensively."7 The method of formylation proposed shortened the reaction time to 30 minutes, decreased the amount of reagents consumed, and lowered the temperature of the process. At the end of the reaction (determined by chromatography) the product is not hydrolyzed, rather the intermediate, a metallated phosphorus complex is isolated. Treatment of the latter with sulfuric acid not only results in the desired aldehyde, but also serves to remove the metal simultaneously. Inspection by T L C did reveal a secondary product, however 88 Results and Discussion not enough of this product was isolated for extensive characterization. The suspicion is that this side product is the result of attack on the formyl group, or perhaps sulfonation of the porphyrin ring. Nonetheless, yields increased to 65% and product isolation was much easier than with the nickel-TPP. It should be noted, however, that even this method of formylation leaves much to be desired. The characteristically low solubility of porphyrins, TPP in particular, requires large volumes of solvent. For example, for 0.744 mmol of CoTPP, 300 mL of dichloroethane, followed by 500 mL of water, and 300 mL of benzene in the sulfuric acid work-up were used. Many would consider this a large scale reaction, and yet less than 400 mgs of product are obtained. In addition, attempts at scaling-up the reaction were unsuccessful. The products listed in the experimental section of this work are all free base porphyrins. Although progress in this series was also made on the metallated NiTPP, namely as far as the oxime, these products have not been included on the basis of redundancy. The superior yield and ease of preparation of the free base formyl-TPP via CoTPP will undoubtedly be the one to explore further. The conversion of formyl-TPP (21) to trifluoromethanol-TPP (22) (Scheme 2.9) proved as elusive as the formylation did initially. Although Pandey 1 0 7 reported a similar reaction on a formyl group in the meso position of octaethylporphyrin in excellent yields (95%), the communication gave no indication of the proprotions of reagents, solvents or 89 Results and Discussion work-up used. A paper by Prakash et. al.118 which reports the nucleophilic trifluoromethylation reaction of carbonyl compounds by Rupert's reagent's"9 (TMSCF3) was referred to, but gives no porphyrinic examples. It recommended the use of an "equimolar excess" of T M S C F 3 for every mole of aldehyde substrate. The reaction did not work on the porphyrin until the ratio of T M S C F 3 had systematically been increased to 15 times that of the formyl-TPP. Throughout the assays, it was thought that an increased amount of tetrabutylammonium fluoride (TBAF) catalyst might also improve the yield of the reaction. This error in logic resulted in many unsuccessful reactions, as well as expensive reagents squandered. In fact, one mechanistic explanation points to a competition for the TMSCF3 reagent between the fluorine generated from T B A F and the reaction site on the porphyrin. What was far more successful in increasing the reaction yield, was maintaining dry reaction conditions to assure that TMSCF3 would not be hydrolyzed. Once this was realized, and all extra precautions were taken to assure dry reaction conditions (including distillation of the solvent-THF) a much better conversion (89%) was achieved. The final hurdle in this reaction was overcome by the discovery that trifluoroacetic acid (TFA) was not an appropriate acid for cleaving the TMS from the intermediate 21a to afford the alcohol (Scheme 2.10). This milder acid was suggested for use with the porphyrin, but it turns out that [0.5M] HC1 is required to cleave the strong oxygen-silicon bond. The HC1 protonated the porphyrin as well, turning the solution a vibrant green colour, but was easily neutralized by washing with saturated sodium acetate solution. 90 Results and Discussion 21 21a 22 Scheme 2.10 Formation of trifluoromethanol-TPP An oxidation similar to that of trifluoromethanol-TPP (22) to trifluoroacetyl-TPP (23) was also referred to by Schiau et. al.120 Ley and co-workers121 first described the use of tetrapropylammonium perruthenate (PfyN) (R.UO4) and N-methylmorpholine N-oxide for the oxidation of alcohols to aldehydes and ketones. Similar to the aforementioned formation of trifluoromethanol, several reaction conditions had to be worked out to apply the chemistry to the porphyrin molecules in question. Again, the amount of reagents needed to be increased (6.4: 1, NMO:alcohol instead of the suggested 3.2:1). It became particularly important to assure the driest possible conditions in this case as well. It is suspected that water in the NMO and solvent (CH2CI2) was responsible for repeatedly poisoning the ruthenium catalyst. Once precautions for a dry reaction were taken, the yield for the reaction increased from 20 to 74%. It was also found that addition of one or two extra equivalents of catalyst, towards the end of the reaction, helped to push it to completion. The remaining 26% can be attributed to a substantial impurity, visible by T L C as a purple band directly under the desired product. Despite efforts to characterize this side-product, MS, NMR and UV-vis spectroscopies only revealed it to be a porphyrin of lower molecular weight than the desired product. 91 Results and Discussion In Scheme 2.7, the conversion of trifiuoroacetophenone to the tosyloxime is shown in one step. The oxime is not purified or characterized in the literature procedure,51 presumably due in part to the ease with which the tosylation is carried out on that intermediate. Unfortunately, the tosylation of the trifluoromethyl oxime-TPP (24) was never achieved in the present work. The oxime (24) is in fact, the last intermediate in the synthesis to be isolated and characterized. One of the challenges in the formation of the oxime was finding a suitable solvent for the reaction, and making the necessary changes in reaction to accommodate this solvent. The literature precedence cited a reaction carried out in ethanol, in which the pH would be continually adjusted to 6 with addition of 4M NaOH. Not only would the porphyrin be insoluble in ethanol, but also any suitable organic solvent would not allow the pH to be easily monitored. Addition of molar equivalents of NaOH was not successful, nor was the use of excess hydroxylamine»HCl. After some thought about the equilibrium involved in the reaction, it was decided to attempt the reaction in a high boiling, dry solvent with molecular sieves. The hope was that the water could be driven off, pushing the reaction to the products side of the equilibrium. After only one hour of reflux in dry benzene with molecular sieves, a light green spot (Rf 2 4 =0.42, 100% CH 2 C1 2 eluent) appeared below that of the trifluoroacetyl-TPP (R f 2 3 = 0.61). Many hours later when T L C revealed near-completion of the reaction, the product was isolated at 60% yield. It is anticipated that upon scale-up of future reactions, the use of a Dean-Stark trap will further improve the yield of this reaction. 92 Results and Discussion It is difficult to postulate about the reasons for the seemingly facile tosylation reaction not working. By this point in the synthesis, despite efforts to scale-up, there was only a minute amount of 24 left and it was time to 'pass the torch' to the graduate student who would assume ownership of this project in the future. Trifluoroacetyl-TPP, the first example of a trifluoroacetyl group in the (3 position, and its intermediates were fully characterized by standard methods. The 'H NMR spectra shown in Figure 2.29 highlights the subtle changes in the aromatic region of the spectra, that support the signal assignments; these are outlined in detail in the experimental section. The aldehyde signal and [3-proton signals at 9.46 and 9.24 ppm respectively, compare favourably with the literature."6 The introduction of the trifluoromethyl group is far enough removed from the P-proton that its shift is hardly affected (9.24 - 9.12 ppm). The aldehyde signal disappears, and the proton signal from the alcohol carbon appears further upfield, at 8.84 ppm. Oxidation to the trifluoroacetyl results in the disappearance of the 81 proton altogether, and the p-proton is shifted back to 9.25 ppm. The 1 9 F spectra shown in Figure 2.30, illustrate how simple fluorine coupling experiments confirmed the oxidation of trifluoromethanol-TPP to trifluoroacetyl-TPP. Thus the fluorine nuclei in the trifluoromethyl moiety are coupled to the single proton on the carbon also bearing the alcohol, producing a doublet (-76.87 ppm). The same trifluoromethyl moiety in the trifluoroacetyl functionality, on the other hand, does not have any neighbouring protons and appears as a singlet (-73.88 ppm). These fluorine nuclei also 93 Results and Discussion Results and Discussion appear to sense the deshielding effect of the carbonyl, which results in their shifting downfield by 3 ppm. -76.40 -73.80 -73.90 -74.00 -74.10 -74.20 -74.30 Figure 2.30 Coupled l 9 F spectra of 22 and 23. Despite literature precedence for each of the reactions applied towards the synthesis of 25, many difficulties arose in adapting the procedures to porphyrin chemistry. As a result of optimizing five of the eight reactions in the formation of the diazirine, scale up is now possible and large amounts of the trifluoroacetyl have been formed. It is anticipated that the last three steps in the reaction are feasible, and that upon completion of the synthesis the interesting photochemical properties of 25 will be explored. 95 Conclusion 3 Conclusions and Suggestions for Further Studies In the first part of this thesis, the synthesis, isolation and characterization of a series of novel pyrazoline porphyrins (10, 11, 12) were described. These azo-type derivatized PP-D M E porphyrins were found to extrude molecular nitrogen on thermal and light activation. Irradiation with UV light and long wavelength light were found to initiate the observed photochemistry, and the mechanism presumably involved the stepwise formation of a 1,3-biradical. A single product for each respective pyrazoline porphyrin was formed on completion of both the thermal and photochemical reactions (15, 16, 17). These novel cyclopropane derivatives were also purified and characterized. The optimized diazomethane reaction was also carried out on a chlorin to yield a novel pyrazoline derivatized MePPP (19). Like the PP-DME pyrazolines, activation with heat, U V light and long wavelength light yielded the cyclopropane product of MePPP (20). The optical characteristics of the pyrazoline MePPP made for an interesting photochemical study. The photoconversion was monitored by UV-vis spectroscopy to reveal that this photochemical reaction can be taken to complete conversion to the respective cyclopropane, without any intervening products. The phenomenon of long wavelength activation in these systems is very fortuitous considering the potential application of crosslinking photosensitizers in PDT. The mechanism of activation was addressed, in particular the ability of the low energy photoactivation to lead to nitrogen extrusion. We believe that the mechanisms of the 96 Conclusion short and long wavelength photochemistry are the same and that this highly exothermic cycloelimination reaction of pyrazolines is possible with the wavelength of light used. That the porphyrin macrocycle has the ability to absorb light at low wavelengths and activate photoactive moieties at its periphery is very encouraging for the design of future potential crosslinking photosensitizers, and is a key discovery of this work. Several attempts to trap the biradical intermediates produced on photoactivation of the pyrazolines were carried out in solution, to investigate the potential efficacy for crosslinking. Unfortunately, the lifetime of these biradical intermediates proved to be much shorter than expected. No adducts were formed with the crosslinking substrate solutions, even when the pyrazoline porphyrins were dissolved in neat substrate solutions. Although these results were disappointing, solution studies alone may not reveal the modified photosensitizers crosslinking abilities in vivo. A variety of in vitro and in vivo testing regimes are available to assess these crosslinking systems. Initial experiments were scheduled to take place at QLT Inc., but could not be carried out this past year due to the company's many commitments. The assays that are foreseen will focus on establishing whether crosslinking improves the cellular retention of the photosensitizer upon light activation. L1210, a tumor cell line, will be incubated in the dark with the MePPP pyrazoline 19 and the unfunctionalized MePPP 18. Incubation will be followed by irradiation with the appropriate long wavelength of light. The cells will then be washed after 1 hr of incubation, followed by the extraction of drug from the cells.61 Comparisons between crosslinked systems, (light activated and dark control), and 97 Conclusion cells incubated with unfunctionalized photosensitizers will provide evidence for selective retention in a biological environment. The last section in this work described the progress made towards the synthesis of a diazirine derivatized porphyrin (25). It is anticipated that this photoactive functionality will yield a longer-lived carbene intermediate upon photolysis, endowing it with greater crosslinking abilities. Each of the reactions carried-out on TPP had to be modified for porphyrin chemistry from the so-called small molecule literature precedents to yield the desired products. The reactions leading up to trifluoroacetyl-TPP (23) provide the first example of introducing a trifluoroacetyl group at the (3 position of the porphyrin. Each of the intermediates (21, 22, 23, 24) were isolated and characterized. Five of the eight reactions planned in the synthesis of (25) were optimized and scaled-up. This will afford the new investigator with a sufficient amount of the trifluoroacetyl intermediate (23) to pursue the remaining reactions effectively. Early in the designs of this project, an azide derivatized porphyrin was also envisioned. The nitrene intermediate has been applied extensively in photoaffinity labeling studies, making the azide functionality a worthy synthetic target. The success in this project elicited the interest of another graduate student who is currently working towards the synthesis of the diazirine, and has also made progress towards a nitrene derivatized porphyrin. 98 Experimental 4 Experimental 4.1 Instrumentation and Materials Elemental Analyses (EA) Microanalyses were carried out in the microanalytical laboratory in the Department of Chemistry, University of British Columbia by Mr. Peter Borda using a Carlo Erba Elemental Analyzer 1106. Analysis was attempted on all pyrazoline and cyclopropyl derivatives of protoporphyrin and methylpyropheophorbide, as well as trifluoroacetyl-TPP. In many cases, the results are unacceptable due to the likely presence of water that could not be removed by heating, because of the thermal lability of the compounds in question. Mass Spectra (MS) Mass spectrometric analyses were carried out by the B. C. Regional Mass Spectrometry Center at the University of British Columbia, Department of Chemistry. Low and high resolution mass spectra were obtained by liquid secondary ion mass spectrometry (LSIMS), and were determined on a KRATOS Concept IIHQ hybrid mass spectrometer. Molecular ions are designated as M + . UV-vis Spectra UV-vis spectra were taken on a Cary 50. Wavelengths for each absorption maximum (^max) are reported in nanometers (nm), and extinction coefficients (e (M_lcm"')) are given in parentheses. 99 Experimental Fluorescence Instrumentation The fluorescence measurements were performed on a SLM-AMINCO AMINCO-Bowman Series 2 Luminescence Spectrometer using a pulsed xenon lamp as the excitation source. Excitation was at 350 nm and 660nm with a 2 nm bandwidth and emission was at 430 nm with 16 nm bandwidth. Fluorescence emissions were corrected for lamp fluctuations using the reference signal from the excitation source. Electrochemical Studies Cyclic Voltammetry was carried out at the University of Western Ontario, with the assistance of Prof. Mark Workentin using an E.G. & G. PAR 283 potentiostat interfaced to a personal computer using PAR 270 electrochemistry software. The electrochemical cell was maintained at 25° C and contained 0.1 mol L - l TEAP in 25mL of methylene chloride purged by argon. At the beginning of every experiment the working electrode, a glassy carbon 3 mm disk, was freshly polished with 1 mm diamond paste and ultrasonically cleaned in ethanol for fifteen minutes. The counter electrode was a platinum plate and a silver wire immersed in a glass tube containing 0.1 M TEAP in the desired solvent with a fine sintered bottom was used as a quasi-reference. Ferrocene was used as an internal redox reference; the potential was calibrated against the saturated calomel electrode (SCE). To compensate for the cells internal resistance the iR compensation was adjusted to at least 95% of the oscillation value. In a typical experiment, 1 mM sample of the appropriate porphyrin was used. 100 Experimental Nuclear Magnetic Resonance Spectrometry (NMR) NMR spectra were recorded either by the author, by Dr. Nick Burlinson, by Marietta Austria or Liane Darge of the University of British Columbia Chemistry Department NMR Service Laboratory. *H NMR Proton nuclear magnetic resonance spectra ('HNMR) were recorded on the following spectrometers: Bruker WH-400 (400 MHz), Bruker AV-400 (400 MHz) and Buker AMX-500 (500 MHz). The positions of the signals are given as chemical shifts (8) in parts per million (ppm) with respect to tetramethylsilane (TMS) at 8 0 ppm; however, the internal reference standard used in each case was the residual proton signal present in the deuterated solvent. Reported chemical shifts are followed in parentheses by the number of protons, the multiplicity of the peak, the coupling constant (J) in Hz, and the atomic assignment. The following abbreviations are used in reference to the multiplicities: s = singlet, d = doublet, t = triplet, q = quartet, dd = doublet of doublets, dt = doublet of triplets, m = multiplet, br = broad. In some cases, in order to ascertain structures, H M Q C (Heteronuclear Multiple Quantum Coherence, 'H- ' 3 C correlation) and NOESY (Nuclear Overhauser Effect Spectroscopy) experiments on the Bruker AMX-500 spectrometer, COSY (Correlated SpectroscopY), NOESY and homonuclear decoupling experiments on the Bruker WH-400 spectrometer were carried out. 101 Experimental 1 3 C NMR The following spectrometers were used to record the carbon nuclear magnetic resonance spectra ( l 3 C NMR): Bruker AM-400 at 100.6 MHz, Bruker AV-300 at 74.6 MHz, and Bruker AMX-500 at 128.5 MHz. All spectra were determined with broad band proton decoupling. The position of the signals are given as chemical shifts (8) in parts per million (ppm) with respect to tetramethylsilane at 8 0 ppm; however, the internal reference standard used in each case was the central transition of the solvent carbon atom. Reported chemical shifts are followed in parentheses by the carbon assignments, which were often made possible by an APT (attached proton test). Crystallographic Analysis (X-RAY) The X-ray crystal structure was determined using single-crystal X-ray analysis, on a Rigaku / ADSC C C D area detector with graphite monochromated Mo-Ka radiation, and was drawn with a locally modified version of the ORTEP program at the 50% probability level. Structures were determined by Dr. Brian Patrick in the Chemistry Department of the University of British Columbia. Chromatography Chromatographic purifications of compounds were carried out using silica gel 60, 70-230 mesh, supplied by E. Merck Co. Thin layer chromatography (TLC) was carried out on pre-coated silica gel plates (Merck 60, 230-400 mesh, with aluminum backing and fluorescent indicator (F254). Preparative thin layer chromatography was prepared on pre-coated 10 x 10 cm 0.5 or 1 mm thick Whatman or Merck silica gel plates. 102 Experimental Preparative thin layer chromatography was also carried out using the Chromatotron®, when necessary, whereby circular plates of Merck 60, 230-400 mesh with fluorescent indicator (F254) were prepared on site. Reaction Conditions Due to the inherent light sensitivity of these compounds, all reactions were performed in a blacked-out fume hood or surrounded by aluminum foil. Reagents and Solvents Unless otherwise specified, reagents were used as supplied by the Aldrich Chemical Company. Solvents were reagent or HPLC grade and purified using standard literature methods when necessary. Deuterated solvents were supplied by Cambridge Isotope Laboratory. 4.2 Photochemical Studies - General Procedures Irradiation Sources Photochemical irradiations were carried out with either a 250W Osram H L X 64655 arc lamp in an Oriel lamp housing, a Rayonet Photochemical Chamber Reactor (Model RPR-100) or a 672 nm Light Emitting Diode. The light output from the Oriel lamp passed through a glass filter: P70-600-S-533G-Corion 103 Experimental Solution State Irradiation Spectral-grade solvents were used for irradiations. For preparative-scale irradiations, the solution of the substrate in the appropriate solvent was placed in a 50 mL septum-sealed tube and deoxygenated with nitrogen or argon for 30 minutes, with stirring, prior to reaction. Efficient stirring was maintained throughout the irradiation period when possible. For irradiations carried out with the Osram lamp and the LED, the substrate was placed 6-10 cm from the source, and supported on a retort stand. In the case of the Rayonet Reactor, the tubes were suspended by copper wire in the chamber and were not stirred during irradiation. Analysis of Photochemical Reactions The photochemical reactions were monitored by thin layer chromatography. To insure exclusion of oxygen, these were purged with nitrogen or argon, a large needle was inserted through the septum and aliquots were removed with a capillary. In some cases UV-vis spectrometry was also performed on these aliquots. 4.3 Preparation of Protoporphyrin Pyrazolines 10,11 and 12 The pyrazolines of protoporphyrin-dimethylester were synthesized according to a modified procedure for the reactions of diazomethane reported by Black. 6 8 A diazomethane apparatus appropriate for production of (1-50 mmol) of diazomethane was used. First, the cold finger was filled with acetone and dry ice. A solution of potassium hydroxide (13.36 g, 0.238 mol) in water (20 mis) was then added to the reaction vessel, 104 Experimental followed by ethanol (10 mis). A 60 ml separation funnel was then placed over the reaction vessel, sealed and supported by a rubber septum that had a whole bored through it. The 100 ml receiver flask containing PP-DME (9) (0.5 g, 0.846 mol) in a minimum amount of methylene chloride and a stir bar was placed under the condensor, and cooled in an ice bath. An ether trap was not used. The separation funnel was then charged with diazald (12.74 g, 0.59 mol) in a minimum amount of diethylether (~ 40 mis). (Often, solutions were sonicated to facilitate dissolution). The reaction vessel was heated to 60° in a water bath, and the apparatus was kept behind a blast shield from this point on. Diazald in ether was slowly added to the reaction vessel via the separation funnel, at a rate attempting to match that of the formation of yellow diazomethane gas apparent on the condensor. Once the addition was complete, the separation funnel was flushed with a minimum amount of ether, to dissolve residual amounts of diazald. The 'diazo-balloon' was then attached to the top of the receiver flask, and placed behind the blast shield or in a canister, left to stir in the dark, overnight. The reaction was followed by thin layer chromatography using only make-shift plastic capillaries to remove aliquots (any sharp edges solicit the risk of detonation of diazomethane). In the event the reaction had not proceeded to completion, the solvent (and diazomethane) were evaporated and the reaction mixture re-exposed to the aforementioned diazomethane procedure. (The advent of the 'diazo-balloon' greatly reduced the need for re-exposure.) After satisfactory conversion, the reaction was quenched by removal of solvent and diazomethane-carrying ether (usually in fume hood, under a gentle stream of nitrogen.) 105 Experimental Separation of the three pyrazoline products was only effective using the chromatotron, starting 100% methylene chloride to introduce the product mixture to the circular silica plate, and increasing the polarity thereafter (0.5% -> 5 % THF in CH 2 C1 2 eluent). The recovered yield of the three products was 78%, in a ratio of 1(10): 1(11):2(12). Recovered starting material accounted for a remaining 20%. EA : Calculated for C37H4oN604 : C, 70.23; H, 6.37; N, 13.28. Found C, 69.85; H, 6.39; N, 12.57 MS LR +LSIMS (matrix: thioglycerol) : Exact mass calculated for C37H4oN604 (M+l): 633.75. Found 633. MS HR +LSIMS (matrix: thioglycerol) : Found C, 37; H, 41; O, 4; N, 6 (M+l) 633.31948, (dev 0.87). XH NMR (400 MHz, CDC13): 5 9.94 (1H, s, H-5), 9.91 (1H, s, H-20), 9.83 (1H, s, H-15), 9.61 (1H, s, H-10), 8.14 (1H, m, JT R A N S= 17.8, JC I S= 11.5, H-31), 6.80 (1H, m, J U M , L-N= 9.7, JL-O, L-P= 2.5, HL-8'), 6.29 (1H, m, JT R A N S= 17.9, J G E M = 1.3, H-32), 6.13 (1H, m, JC I S= 11.5, 106 Experimental J G E M = 1.3, H-32), 5.45 (1H, m, J 0 - P = 17.8, J 0-M= 9.8, J 0 . N , O-L= 2.7, H 0-8 3), 4.74 (1H, m, JP_o= 18.4, Jp.M, P.N= 9.4, JP.L=2.8, Hp-83), 4.28 (2H, t, J = 7.5, H-131), 4.27 (2H, t, J = 7.5, H-171), 3.65 (3H, s, H-135), 3.64 (3H, s, H-175), 3.54 (3H, s, H-121), 3.51 (3H, s, H-18'), 3.48 (3H, s, H-71), 3.47 (3H, s, H-21), 3.21 (2H, t, J = 7.7, H-132), 3.20 (2H, t, J = 7.7, H-172), 2.74 (1H, m, JM-N=22.4, JM-P=9.7, JM-L=3.1, HM-8 2), 2.11 (1H, m, JN-M=22.7, JN-L, N-P=9.7, JN-o=3.2, HN-8 2), -3.72 (2H, s, 7VH) UV-VIS (CH2C12) Xmax (rel. intensity): 406 (1), 502 (0.08), 536 (0.06), 572 (0.05), 626 (0.02) OCH3 H3C0 11 E A : Calculated for C37H4oN604 : C, 70.23; H, 6.37; N, 13.28. Found C, 69.88; H, 6.34; N, 12.58 MS LR +LSIMS (matrix: thioglycerol) : Exact mass calculated for C3 7H4oN604 (M+l): 633.75. Found 633. MS HR +LSIMS (matrix: thioglycerol) : Found C, 37; H, 41; O, 4; N, 6 (M+l) 633.31937, (dev 0.69). 107 Experimental *H NMR (400 MHz, CDCI3): 5 9.95 (1H, s, H-10), 9.79 (1H, s, H-15), 9.69 (1H, s, H -20) , 9.57 (1H, s, H-5), 8.05 (1H, m, J t r a n s= 17.8, J c i s= 11.5, H-81), 6.81 (1H, m, J L . M , L-N= 9.8, J L . 0 , L-P= 2.6, HL-3'), 6.23 (1H, m, J t r a n s= 17.8, J g e m = 1.2, H-82), 6.10 (1H, m, J c i s= 11.5, J g e m = 1-2, H-82), 5.46 (1H, m, J0-P= 17.9, J 0-M= 9.8, J0-N,O-L= 2.6, H 0-3 3), 4.75 (1H, m, JP-o= 17.9, Jp.M, P-N= 9.3, JP.L=2.8, HP-3 3), 4.28 (2H, t, J = 7.8, H-131), 4.24 (2H, t, J = 7.8, H-171), 3.66 (3H, s, H-135), 3.65 (3H, s, H-175), 3.49 (3H, s, H-71), 3.46 (3H, s, H-21) , 3.43 (3H, s, H-121), 3.42 (3H, s, H-181), 3.21 (2H, t, J = 7.8, H-132), 3.18 (2H, t, J = 7.8, H-172), 2.72 (1H, m, JM-N=13.0, HM-3 2), 2.10 (1H, m, JM-N=13.0, HN-3 2), -4.20 (2H, s,/VH) 1 3 C NMR (125.8 MHz, CDC13) (quaternary carbons are unresolved, C = 19): 5 97.62, 96.86, 96.66, 95.89, 129.99, 85.89, 120.57, 77.81, 21.65, 21.62,51.69, 51.67, 11.42, 12.41, 11.83, 11.51, 36.79 (2xC), 27.72. UV-VIS (CH2C12) Xmax (rel. intensity): 404 (1), 502 (0.08), 536 (0.06), 572 (0.05), 626 (0.02) 108 Experimental E A : Calculated for CsgH^NgC^ : C, 67.64; H, 6.27; N, 16.61. Found C, 67.50; H, 6.26; N, 15.60 MS LR +LSIMS (matrix: thioglycerol) : Exact mass calculated for CjgH^NgC^ (M+l): 675.79. Found 675. MS HR +LSIMS (matrix: thioglycerol) : Found C, 38; H, 43; O, 4; N, 8 (M+l) 675.33975, (dev-1.45). *H NMR (400 MHz, CDC13): 8 10.08 (1H, s, H-20), 10.03 (1H, s, H-15), 9.81, 9.79 (1H, s, H-5), 9.73, 9.71 (1H, s, H-10), 6.91 (2H, m, H L -3 \ 81), 5.55, 5.50 (2H, m, J 0-P= 17.9, J 0-M= 10.0, J0-N,O-L= 2.5, H 0 -3 3 , 83), 4.82, 4.78 (2H, m, H P -3 3 , 83), 4.37 (2H, t, J = 7.4, H-13'), 4.36 (2H, t, J = 7.4, H-171), 3.64 (3H, s, H-135), 3.64 (3H, s, H-175), 3.60 (3H, s, H-18'), 3.56, 3.55 (3H, s, H-71), 3.54, 3.53 (3H, s, H-121), 3.50, 3.49 (3H, s, H-21), 3.26 (2H, t, J = 7.7, H-132), 3.25 (2H, t, J = 7.7, H-172), 2.84 (2H, m, H M -3 2 , 82), 2.24, 2.22 (2H, m, H N -3 2 , 82), -3.75 (2H, s, NH) UV-VIS (CH2C12) ? w (rel. intensity): 400 (1), 500 (0.10), 534 (0.07), 570 (0.05), 622 (0.03) 4.3.1 Photoproducts of Protoporphyrin Pyrazolines (15), (16), and (17) Stirred, deoxygenated solutions of each of the pyrazolines of protoporphyrin-dimethylester in benzene were irradiated in the Rayonet reactor (35 A bulbs) and with red-filtered light (Corion 600). The following table outlines the solution concentrations, irradiation times, and isolated product yield: 109 Experimental Rayonet Photochemical Chamber Reactor Cone. (g/mL), mmol Time (h) Product yield (%) A-ring pyrazoline 3.06 x 10'4, 0.017 22.5 82.4 B-ring pyrazoline 5.00 x 10"J, 0.016 21 52.4 Di-adduct pyrazoline 2.29 x 10-4, 0.014 5 50.0 650 nm filtered light Cone. (g/mL), mmol Time (h) Product yield (%) A-ring pyrazoline 8.00 x 10"\ 0.006 14 35.3 B-ring pyrazoline 2.40 x 10"4, 0.019 7 52.3 Di-adduct pyrazoline 2.60 x 10"4, 0.019 38 78.5 Benzene was removed in vacuo, and the residue dissolved in dichloromethane. The crude compounds were chromatographed on silica gel 60, 70-230 mesh, 3%-10% EtAce/CH2Cl2 gradient eluent). The appropriate fractions were pooled and evaporated to give the respective yields of the desired cyclopropyl derivatives. A large proportion of the starting material was recovered in each case, which accounts for the incomplete conversion to products. 110 Experimental MS LR +LSIMS (matrix: thioglycerol) : Exact mass calculated for C37H40N4O4 (M+l): 605.747. Found 605. MS HR +LSIMS (matrix: thioglycerol) : Found C, 37; H, 40; O, 4; N, 4 (M+l) 605.31293, (dev 0.24). *H NMR (400 MHz, CDCI3): 8 10.44 (1H, s, H-5), 10.16 (1H, s, H-20), 10.01 (1H, s, H-15), 10.00 (1H, s, H-10), 8.28 (1H, m, J t r a n s= 17.8, J c i s= 11.5, H-31), 6.35 (1H, m, J t r a n s= 17.7, J g e m = 1.4, H-32), 6.15 (1H, m, J c i s= 11.5, J g e m = 1.4, H-32), 4.40 (2H, t, J = 7.6, H-131) , 4.36 (2H, t, J = 7.6, H-171), 3.71 (3H, s, H-121), 3.68 (3H, s, H-181), 3.65 (3H, s, H-135), 3.65 (3H, s, H-175), 3.62 (3H, s, H-71), 3.59 (3H, s, H-21), 3.27 (2H, t, J = 7.9, H-132) , 3.26 (2H, t, J = 7.8, H-172), 3.05 (1H, m, J d 0 ub= 8.2, J d o u b = 8.2, H-81), 1.67 (2H, m, H-82), 1.46 (2H, m, H-82), -3.76 (2H, s, NH) UV-VIS (CH2C12) lmax (rel. intensity): 402 (1), 502 (0.08), 538 (0.07), 572 (0.04), 626 (0.02) E A : Calculated for C 3 7H 4oN 404 : C, 73.49, H, 6.67; N, 9.26. Found C, 72.02; H, 6.88; N, 7.93 Experimental MS LR +LSIMS (matrix: thioglycerol) : Exact mass calculated for C37H40N4O4 (M+l): 605.747. Found 605. MS HR +LSIMS (matrix: thioglycerol) : Found C, 37; H, 41; O, 4; N, 4 (M+l) 605.31293, (dev 0.24). *H NMR (400 MHz, CDCI3): 5 10.38 (1H, s, H-10), 10.14 (1H, s, H-15), 10.07 (1H, s, H-20), 10.02 (1H, s, H-5), 8.28 (1H, m, J t r a n s= 17.6, J c i s= 11.4, H-81), 6.33 (1H, m, J t r a n s= 17.7, J g e m = 1.4, H-82), 6.14 (1H, m, J c i s= 11.5, J g e m = 1.4, H-82), 4.40 (2H, t, J = 7.7, H-131) , 4.37 (2H, t, J = 7.8, H-171), 3.68 (3H, s, H-121), 3.67 (3H, s, H-181), 3.64 (3H, s, H-171), 3.63 (3H, s, H-135), 3.63 (3H, s, H-175), 3.61 (3H, s, H-21), 3.26 (2H, t, J = 7.7, H-132) , 3.26 (2H, t, J = 7.6, H-172), 3.06 (1H, m, J d o u b = 7.6, J d o u b = 7.6, H-31), 1.66 (2H, m, J(?)=2.0, H-32), 1.45 (2H, m,J(?)=1.7, H-32), -3.75 (2H, s, NH) UV-VIS (CH2C12) X m a x (rel. intensity): 404 (1), 502 (0.07), 538 (0.06), 572 (0.05), 626 (0.03) X-RAY : Please refer to Appendix 1 112 Experimental E A : Calculated for C38H42N4O4 : C, 73.76, H, 6.84; N, 9.05. Found C, 73.66; H, 7.54; N, 7.11 MS LR +LSIMS (matrix: thioglycerol) : Exact mass calculated for C38H42N4O4 (M+l): 619.747. Found 619. MS HR +LSIMS (matrix: thioglycerol) : Found C, 38; H, 4 3 ; O, 4; N, 4 (M+l) 619.32852, (dev 0.15). *H NMR (400 MHz, CDC13): 5 10.41 (1H, s, H-10), 10.39 (1H, s, H-15), 10.04 (1H, s, H-20), 10.02 (1H, s, H-5), 4.41 (2H, t, J = 7.7, H-131), 4.39 (2H, t, J = 7.8, H-171), 3.68 ( 3 H , s, H-121), 3.67 ( 3 H , s, H-181), 3.64 ( 3 H , s, H-171), 3.63 ( 3 H , s, H-135), 3.63 ( 3 H , s, H-175), 3.61 ( 3 H , s, H-21), 3.26 (2H, t, J = 7.7, H-132), 3.27 (2H, t, J = 7.6, H-172), 3.05 (1H, m, H - 3 1 and H-81), 1.66 ( 4 H , m, J ( v i c )=1.9, H - 3 2 and H-82), 1.45 ( 4 H , m, J (?)=1.5, H-3 2 and H-82), -3.74 (2H, s, NH) UV-VIS (CH2CI2) X m a x (rel. intensity): 400 (1), 500 (0.09), 534 (0.07), 568 (0.05), 622 (0.04) 4.4 Preparation of Methylpyropheophorbide Pyrazoline (19) The pyrazoline of methylpyropheophorbide was synthesized according to a modified procedure for the reactions of diazomethane reported by Black In most cases, this reaction was carried out immediately following the addition of diazomethane to protoporphyrin-dimethylester. Thus, enough potassium hydroxide was added to the reaction vessel to accommodate the formation of diazomathane for both reactions. The portion of the aqueous solution required for this reaction was (2.88 g, 0.051 mol) of 113 Experimental potassium hydroxide, followed by addition of ethanol (10 mis). The separation funnel was then charged with diazald (2.75 g, 0.013 mol) in a minimum amount of diethylether. Upon formation, the diazomethane was delivered to a 50 ml receiver flask containing methylpyropheo-phorbide(0.5 g, 0.911 mol) in a minimum amount of methylene chloride. The reaction was then left to stir overnight. The crude compound was chromatographed on silica gel 60, 70-230 mesh, 2%-10% EtAce/CH2Cl2 gradient eluent). The appropriate fractions were pooled and evaporated to yield 0.419 g (78% yield) of the desired pyrazoline. 15% of the starting material did not react, and was recovered. E A : Calculated for C35H38N603 : C, 71.16, H, 6.48; N, 14.23. Found C, 70.72; H, 6.72; N, 13.63 114 Experimental M S L R + L S I M S (matrix: thioglycerol) : Exact mass calculated for C 3 5 H 3 8 N 6 0 3 (M+l): 591.73. Found 591. M S H R + L S I M S (matrix: thioglycerol) : Found C, 35; H, 39; 0, 3; N, 6 (M+l) 591.30823, (dev -0.23). * H N M R (400 MHz, CDC13): 5 9.51 (1H, s, H-10), 8.96 (1H, s, H-20), 8.57 (1H, s, H-5), 6.67 (1H, m, JL-M,L-N= 9.7, JL-O,L-P= 2.2, HL-3'), 5.47 (1H, m, J 0 - P = 17.6, J 0 - M = 10.1, J 0 . N , o-i = 2.5, H 0-3 3), 5.26 (1H, d, J = 19.6, H-151), 5.11 (1H, d, J = 19.6, H-151), 4.76 (1H, m, J P . 0 = 18.2, Jp.M, P-N= 9.1, JP-L= 3.1, Hp-33), 4.48 (1H, m, J D O U B = 7.4, J Q U A R T = 7.1, H-18), 4.29 (1H, m, J D O U B = 8.1, J D O U B = 2.2, J D O U B = 2.2, H-17), 3.68 (2H, t, J = 7.4, H-172), 3.65 (3H, s, H-175), 3.59 (3H, s, H-71), 3.25 (3H, s, H-21), 3.17 (3H, s, H-121), 2.75 (1H, m, JM-N=19.7, JM-P= 7.6, JM-o, M-L= 2.4, HM-3 2), 2.68 (1H, m, H-171), 2.55 (1H, m, H-81), 2.28 (1H, m, H-81), 2.28 (1H, m, H-171), 2.11 (1H, m, JN.M=19.6, J N . U N-o=9.7, HN-3 2), 1.80 (3H, d, J = 7.4, H-181), 1.67 (3H, t, J = 7.7, H-82), -1.77 (2H, s, NH) 1 3 C N M R (125.8 MHz, CDC13) (total C = 35) 210.68, 195.67, 71.28, 160.45, 154.72, 150.91, 148.92, 144.99, 141.05, 137.95, 135.96, 135.33, 134.66, 132.81, 130.62, 128.56, 106.28, 104.08, 96.69, 93.01, 85.32, 78.01, 51.75, 51.61, 50.01, 48.01, 30.93, 29.85, 27.32, 23.10, 19.38, 17.33, 11.94, 11.45, 11.11 U V - V I S (CH2C12) ?imax: 411.0 (107, 100), 506.4 (10, 800), 536.5 (9, 500), 608.0 (8, 070), 663.5 (46, 061) 115 Experimental 4.4.1 Photoproduct of Methylpyropheophorbide Pyrazoline (20) A stirred, deoxygenated solution of the pyrazoline of methylpyropheophorbide in benzene (2.75 x 10~4 g/mL, 0.023 mmol) was irradiated in front of a 672 nm L E D panel for 14 hours. Thin layer chromatography revealed completion of the reaction, and benzene was removed in vacuo.. The residue was dissolved in dichloromethane, and the crude compound was chromatographed on silica gel 60, 70-230 mesh, 3%-10% EtAce/CH 2 Cl 2 gradient eluent). The appropriate fractions were pooled and evaporated to give (0.0130 g, 0.023 mmol, 98.6%) the cyclopropane derivative of methylpyropheophorbide. MS LR +LSIMS (matrix: thioglycerol) : Exact mass calculated for C35H38N403 (M+l): 563.712. Found 563. *H NMR (400 MHz, CDC13): 5 9.55 (1H, s, H-10), 9.46 (1H, s, H-20), 8.41 (1H, s, H-5), 5.22 (1H, d, J = 19.6, H-151), 5.10 (1H, d, J = 19.6, H-151), 4.43 (1H, m, J d o u b = 7.4, J q u a r t = 7.1, H-18), 4.24 (1H, m, J d o u b = 8.5, J d o u b = 2.2, J d o u b= 2.2, H-17), 3.68 (2H, t, J = 7.6, H-116 Experimental 172), 3.64 (3H, s, H-175), 3.58 (3H, s, H-71), 3.35 (3H, s, H-21), 3.24 (3H, s, H-121), 2.80 (1H, t, H-31), 2.67 (1H, m, H-171), 2.52 (1H, m, H-81), 2.30 (1H, m, H-171), 2.27 (1H, m, H-81), 1.77 (3H, d, J = 7.3, H-181), 1.68 (3H, t, J = 7.8, H-82), 1.57 (2H, m, J = 1.9, H-32), 1.30 (2H, m, J = 1.6, H-32), -1.65 (2H, s, M i ) UV-VIS (CH2C12) ^ : 410.5 (112,000), 505 (9,800), 536 (9,200), 603 (8,400), 658 (46,400) 4.5 Progress Towards the Synthesis of Diazirine-TPP (25) Formyl-tetraphenylporphyrin (21) Formyl-tetraphenylporphyrin (21) was synthesized according to a modification of the procedure reported by Ponomarev1. A solution of CoTPP (500 mg, 0.744 mmol) in dichloroethane (300 mL) was added to a pre-formed Vilsmeier complex 117 Experimental (dimethylformamide: 1.9 mL, 1.79 g, 24.52 mmol and phosphorusoxychloride: 2.25 mL, 3.70 g, 24.14 mmol) in a 2L RBF and heated at 60°C for 30 minutes. The solvent was then evaporated in vacuo, and 500 mL of cold water added rapidly to the oily residue. After 10 minutes, the precipitate was removed by water aspirator filtration through a coarse sintered glass filter and left to air dry overnight. The resultant crude salt, a dark green powder, was used in the following steps without purification. The immonium salt, having been returned to the 2L flask, was then dissolved in 10-15 mL of concentrated sulfuric acid and stirred for an hour. 300 mL of benzene and then a maximum amount saturated sodium acetate solution were added and left to reflux for an hour. Upon cooling, the organic layer was separated and the aqueous layer checked for neutrality. In the event the product was not yet neutralized, the separation funnel was charged with additional sodium acetate solution. The organic layer was passed through cotton and evaporated to dryness. The solvent was removed by evaporation and the residue chromatographed on silica gel 60, 70-230 mesh, 50-100% CH2Cl2/hexanes as gradient eluent). Of the four bands apparent, the fastest moving, peach coloured band accounted for a small proportion of unreacted CoTPP (Rf = 0.8). The dark green band (Rf = 0.5) was the desired product, and two lighter green bands (Rf = 0.3, Rf - 0.24) are yet undefined because their presence is so slight. The appropriate fractions were pooled and dried to afford 0.232 g (49%) of the desired purple crystals, formyl-tetraphenyl porphyrin. 118 Experimental MS L R +LSIMS (matrix: thioglycerol) : Exact mass calculated for C 4 5 H 3 0 N 4 O 3 (M+l): 699.42. Found 699; J H N M R (300 MHz, CDCI3): 5 9.46 (1H, s, H-81), 9.24 (1H, s, H-7), 8.92 (4H, m, H-2,3,12,13), 8.82 (2H, s, H-17,18), 8.23 (8H, m, H-5 2, 102, 152, 202), 7.77 (12H, m, H-5 3' 4, 103'4,153'4, 203'4) 1 3 C N M R (75 MHz, CDC13) (quaternary carbons are unresolved, C = 28): 5 189.26, 142.43, 141.76, 141.57, 134.98, 134.63, 134.58, 134.55, 134.55, 133.34, 133.32, 130.79, 130.05, 130.03, 128.98, 128.14, 127.92, 127.92, 127.36, 127.36, 126.84, 126.84, 126.84, 126.81, 122.60, 120.59, 120.28, 120.00. UV-VIS (CH2C12) Xmax (rel. intensity): 430.9 (1), 466, (0.06), 526.0 (0.07), 567.0 (0.03), 606.0 (0.03), 662.0 (0.03) Trifluoromethanol-tetraphenylporphyrin (22) 22 Trifluoromethanol-tetraphenylporphyrin (22) was synthesized according to a modification of the procedure reported by Prakash and Olah." 8 All flasks and syringes 119 Experimental were flame-dried and cooled under argon. To a stirring solution of formyl-tetraphenylporphyrin (21) (25 mgs, 0.039 mmol) in distilled tetrahydrofuran (15 mL) was added TMS-CF3 (1.16 p:L, 0.585 mmol). The mixture was cooled to 0° under argon, and a catalytic amount of T B A F (lOuL , 0.01 mmol) was added. The reaction was monitored by T L C whereby disappearance of the green starting material (Rf = 0.5) and appearance of the reddish, faster moving trifluoromethylated siloxy intermediate (Rf = 0.78) signaled completion of the reaction. Hydrolysis was achieved by addition of 0.5M hydrochloric acid (until the solution takes on a bright green colour). Methylene chloride and a solution of saturated sodium acetate were then added to neutralize the product. The organic layer was separated, the solvent removed in vacuo, and the residue chromatographed on silica gel 60, 70-230 mesh, (50-100% CH2Cl2/hexanes as gradient eluent) to yield 24.7 mgs (89%) of trifuoromethanol-tetraphenylporphyrin. One faint impurity was visible by T L C , a light green spot (Rf = 0.73), but did not impede isolation of the desired product (Rf = 0.34) which moved much slower. MS L R +LSIMS (matrix: thioglycerol) : Exact mass calculated for C 4 6H3|F3N 40 (M+l): 713.74. Found 713. *H N M R (300 MHz, CDC13): 5 9.12 (1H, s, H-7), 8.84 (1H, s, H-81), 8.82-8.72 (6H, m, H-2,3,12,13, 17, 18), 8.20 (8H, m, H-5 2, 102, 152, 202), 7.76 (12H, m, H-5 3' 4, 103'4, 153'4, 203'4), 2.99 (1H, br. s., -OH), -2.68 (2H, s, -NH). 1 9 F N M R (282 MHz, CDC13) (coupled) 5 -76.87 (3F, d, -CHOHCF3) UV-VIS (CH2C12) X m a x (rel. intensity): 420.0 (1), 450 (0.12), 517.1 (0.05), 553.0 (0.02), 596.0 (0.02), 652.0 (0.02) 120 Trifluoroacetyl-tetraphenylporphyrin (23) Experimental 23 Trifuoroacetyl-tetraphenylporphyrin (23) was synthesized according to a description of the use of tetrapropylammoniumperruthenate (Pr4N)(Ru04) and N-methylmorpholine oxide (NMO) for oxidation of alcohols to ketones by Ley and co-workers.121 All flasks and syringes were flame-dried and cooled under argon. To a stirring solution of trifluoromethanol-tetraphenylporphyrin (22) (250 mgs, 0.351 mmol) in distilled methylene chloride (50 mL) was added NMO (0.46 mL, 2.24 mmol). The mixture was then left to stir at room temperature for 10 minutes, under argon, after which time a catalytic amount of TPAP (25 mgs, 0.07 mmol) was added. The reaction was monitored by T L C whereby disappearance of the dark red starting material (Rf = 0.4) and appearance of the dark green, faster moving trifluoroacetyl-tetraphenylporphyrin (Rf = 0.81) signaled completion of the reaction. The mixture was then washed with water (2 x 50 mL), the organic layer separated and dried through cotton, the solvent removed in vacuo, and the residue chromatographed on silica gel 60, 70-230 mesh, (40-60% Experimental CH2Cl2/hexanes as gradient eluent) to yield 199.3 mgs (74%) of trifluoroacetyl-tetraphenylporphyrin. One substantial impurity was visible by T L C , a purplish band (Rf = 0.72), directly under the desired product. E A : Calculated for C46H2 9F3N40 : C, 77.73, H, 4.11; N, 7.88. Found C, 77.97; H, 4.11; N.7.78. MS L R +LSIMS (matrix: thioglycerol) : Exact mass calculated for C 4 6 H 2 9 F 3 N 4 0 (M+l): 711.74. Found 711. ] H N M R (300 MHz, CDC13): 5 9.25 (1H, s, H-7), 9.15 (1H, d, H-12), 9.04 (1H, d, H-3), 9.00 (2H, d, H-2,13), 8.83 (2H, s, H-17,18), 8.32 (8H, m, H-5 2, 102, 152, 202), 7.81 (12H, m, H-5 3 , 4 , 103'4, 153'4, 203'4), -2.43 (2H, s, -NH). 1 3 C N M R (75 MHz, CDC13) (quaternary carbons are unresolved, C = 29): 5 190.06, 180.53 (q, J c -F = 135 Hz), 142.10, 141.707, 141.51, 141.313, 138.15, 137.70, 136.44, 136.44, 134.82, 134.73, 134.58, 129.88, 129.39, 128.93, 128.33, 128.33, 127.94, 127.94, 127.00, 126.93, 126.88, 126.85, 122.32, 121.15, 120.61, 120.42, 118.50. 1 9 F N M R (282 MHz, CDC13) (coupled) 8 -73.88 (3F, s, -COCF 3 ) UV-VIS (CH2C12) ?i m a x (rel. intensity) : 429.0 (1), 562.0 (0.07), 562 (0.02), 606.0 (0.02), 665.0 (0.04). 122 Experimental Trifluoromethyl oxime-tetraphenylporphyrin (24) s4 24 Trifuoromethyl oxime-tetraphenylporphyrin (24) was synthesized according to a modified procedure by Brunner et al.51 All flasks and syringes were flame-dried and cooled under argon. To a stirring solution of trifluoroacetyl-tetraphenylporphyrin (23) (43 mgs, 0.06 mmol) in distilled benzene (15 mL) was added 12 mgs (0.18 mmol) of hydroxylamine hydrochloride (NH 2 OH«HCl). Activated molecular sieves (7A) were also added to the 50 mL RBF. No base was added. The mixture was then left to reflux overnight. The reaction was monitored by T L C whereby disappearance of the dark green starting material (23, Rf = 0.72) and appearance of the lighter green, slower moving trifluoromethyl oxime-tetraphenylporphyrin (24, Rf = 0.46) signaled completion of the reaction. The mixture was then washed with water (2 x 50 mL), the organic layer extracted with CH2CI2 and dried through cotton, and the solvent removed in vacuo. The crude residue was isolated to yield 26 mgs (60%) of trifluoromethyl oxime-123 Experimental tetraphenylporphyrin (24). One minimal impurity was visible by T L C , a light green band (Rf = 0.24), under the desired product, but sufficient amounts could not be isolated for characterization. It should also be noted, that on standing the product (24) reverted back to its precursor (23), presumably due to hydrolysis. MS L R +LSIMS (matrix: thioglycerol) : Exact mass calculated for C 4 6 H3oF 3 N 5 0 (M+l): 725.76. Found 725. *H N M R (300 MHz, CDC13): 8 9.63 (1H, d, H-12), 8.94 (1H, s, H-7), 8.74 (1H, d, H-3), 8.73 (2H, d, H-2,13), 8.64 (2H, s, H-17,18), 8.36 (8H, m, H-5 2, 102, 152, 202), 7.74 (12H, m, H-5 3' 4, 103'4, 153'4, 203'4), -1.70 (2H, s, -NH). UV-VIS (CH2C12) \ m a x (rel. intensity) : 435.9 (1), 539.0 (0.04), 582.1 (0.07), 618.0 (0.02), 679.0 (0.04). 124 Appendix: Crystallographic Analysis of 17 EXPERIMENTAL DETAILS E m p i r i c a l F o r m u l a F o r m u l a Weight C r y s t a l Co lo r , H a b i t C r y s t a l D imens ions C r y s t a l S y s t e m L a t t i c e T y p e L a t t i c e Parameters Space G r o u p Z value F o o o /z(MoKa) Dif f ractometer R a d i a t i o n Detector A p e r t u r e D a t a Images A . C r y s t a l D a t a C37.50H42.5rjN4O4Cl1.50 666.45 red, b lock 0.30 X 0.20 X 0.15 m m tr ic l in ic P r i m i t i v e a = 8.928(2)A b = 12.015(1) A c = 17.212(2) A a = 106.955(7)° 0 = 89.471(3)° 7 = 106.925(2)° V = 1683.8(4) A 3 Pi (#2) 2 1.314 g / c m 3 706.00 2.00 c m - 1 B . Intensity Measurements R i g a k u / A D S C C C D MOKQ (A = 0.71069 A) graphite monochromated 94 m m x 94 m m 464 exposures @ 58.0 seconds 125 <f> osc i l la t ion Range (x=0) u osc i l l a t ion Range (x=90.0) Detector P o s i t i o n Detector S w i n g A n g l e N o . of Ref lect ions Measu red Cor rect ions Appendix: Crystallographic Analysis of 17 0.0 - 190.0° -19.0 - 23.0° 40.32 m m -5.57° 50.3° T o t a l : 12726 Un ique : 5576 ( R j n t = 0.054) Lorentz -po la r i za t ion A b s o r p t i o n (trans, factors: 0.7284 - 1.0000) C . S t ructure So lu t ion and Ref inement S t ruc tu re So lu t ion Ref inement Funct ion M i n i m i z e d Least Squares Weights p - factor A n o m a l o u s D ispers ion N o . Observat ions (I>0.00<r(I)) N o . Var iab les R e f l e c t i o n / P a r a m e t e r R a t i o Res idua ls : R; R w Goodness of F i t Ind icator M a x S h i f t / E r r o r i n F i n a l C y c l e N o . Observat ions (I>3tr(I)) Res iduals (on F , I>3CT(I)): R; RW M a x i m u m peak in F i n a l Diff. M a p M i n i m u m peak i n F i n a l Diff. M a p Di rect M e t h o d s (SIR97) F u l l - m a t r i x least-squares Zw(Fo2 - F c 2 ) 2 w = ^ ( k j = l°UFo) + ^ F o 2 ] - i 0.0000 A l l non -hydrogen atoms 5576 504 11.06 0.103 ; 0.154 1.52 0.00 3026 0.058 ; 0.071 0.57 e"/A3 -0.62 e " / A 3 126 Appendix: Crystallographic Analysis of 17 Table 1. A t o m i c coordinates and B , - , 0 / B e ? a t o m x y C l ( l ) -0.0689(5) 0.5686(3; C l (2 ) 0.1062(7) 0.4078(6; C l (3 ) -0.0555(6) 0.5084(5; C l (4 ) 0.1948(7) 0.5196(6; 0 ( 1 ) 0.4910(5) 0.3519(3; 0 ( 2 ) 0.2743(4) 0.2218(3; 0 ( 3 ) -0.1655(5) 0.0120(3; 0 ( 4 ) -0.2405(3) 0.1773(2] N(21) 0.0958(3) 0.2300(2; N(22) 0.4388(3) 0.3288(2; N(23) 0.4081(3) 0.2691(2] N(24) 0.0657(3) 0.1684(2] C ( l ) -0.0629(4) 0.1800(3] • C(2) -0.1324(4) 0.1729(3] C (3 ) -0.0105(4) 0.2193(3] C (4 ) 0.1318(4) 0.2563(3) C (5 ) 0.2820(4) 0.3100(3] C (6 ) 0.4224(4) 0.3459(3] C (7 ) 0.5746(4) 0.4095(3] C (8 ) 0.6793(4) 0.4314(3) C (9 ) 0.5936(4) 0.3783(3) C(10) 0.6507(4) 0.3768(3] C ( l l ) 0,5657(4) 0.3274(3] C(12) 0.6348(4) 0.3348(3] occupancy z B e 9 occ 0.4369(2) 9.9(1) 1 /2 0.4550(5) 8.5(2) 1 /4 0.5824(2) 12.6(1) 1 /2 0.4847(3) 7.0(2) 1 /4 0.4987(2) 7.37(9) 0.4258(2) 5.54(8) 0.3821(2) 7.8(1) 0.4215(1) 4.64(7) -0.0914(1) 2.11(6) -0.0601(1) 2.07(6) 0.0913(1) 1.97(6) 0.0604(1) 2.08(6) -0.0978(2) 2.23(7) -0.1771(2) 2.27(7) -0.2177(2) 2.34(7) -0.1629(2) 2.07(7) -0.1819(2) 2.20(7) -0.1345(2) 2.09(7) -0.1523(2) 2.01(7) -0.0878(2) 2.20(7) -0.0294(2) 2.07(7) 0.0444(2) 2.13(7) 0.1009(2) 1.94(7) 0.1802(2) 2.07(7) 127 Appendix: Crystallographic Analysis of 17 Tab le 1. A t o m i c coordinates and B j , 0 / B e 9 and occupancy (continued) a t o m X y z C(13) 0.5148(4) 0.2797(3) 0.2175(2) 1.98(6) C(14) 0.3739(4) 0.2378(3) 0.1612(2) 2.06(7) C(15) 0.2254(4) 0.1776(3) 0.1783(2) 2.03(7) C(16) 0.0842(4) 0.1446(2) 0.1322(2) 1.98(6) C(17) -0.0690(4) 0.0839(3) 0.1503(2) 2.23(7) C(18) -0.1762(4) 0.0729(3) 0.0895(2) 2.25(7) C(19) -0.0894(4) 0.1265(3) 0.0322(2) 2.10(7) C(20) -0.1492(4) 0.1333(3) -0.0400(2) 2.24(7) C(21) -0.3009(4) 0.1255(3) -0.2025(2) 3.02(8) C(22) -0.3638(4) 0.0663(3) -0.2773(2) 4.11(9) C(31) -0.0100(5) 0.2282(4) -0.3028(2) 3.05(8) C(71) 0.6046(4) 0.4479(3) -0.2267(2) 2.49(7) C(72) 0.5901(5) 0.3554(3) -0.3083(2) 3.87(9) C(73) 0.7463(5) 0.4369(4) -0.2703(2) 4.9(1) C(81) 0.8507(4) 0.4990(3) -0.0762(2) 2.73(8) C(121) 0.8046(4) 0.3936(4) 0.2090(2) 2.99(8) C(131) 0.5145(4) 0.2658(3) 0.3015(2) 2.43(7) C(132) 0.4402(4) 0.3522(3) 0.3623(2) 3.02(7) C(133) 0.4073(6) 0.3130(3) 0.4375(2) 4.2(1) C(134) 0.2329(7) 0.1723(6) 0.4926(3) 8.3(2) C(171) -0.1033(4) 0.0420(3) 0.2249(2) 2.44(7) C(172) -0.1337(4) 0.1401(3) 0.2960(2) 3.42(8) C(173) -0.1808(4) 0.1007(3) 0.3695(2) 3.17(8) C(174) -0.2875(6) 0.1516(4) 0.4967(2) 5.5(1) 128 Appendix: Crystallographic Analysis of 17 Table 1. Atomic coordinates and BiSO/Beq and occupancy (continued) atom X y z Beq C(181) -0.3493(4) 0.0148(4) 0.0812(2) 3.08(8) C(200) 0.0000 0.5000 0.5000 16.1(6) H(l) 0.356(4) 0.294(3) -0.036(2) 3.3(8) H(2) 0.144(4) 0.205(3) 0.035(2) 3.2(7) H(3) 0.2891 0.3233 -0.2353 2.6226 H(4) 0.7640 0.4140 0.0586 2.4632 H(5) 0.2196 0.1565 0.2295 2.2980 H(6) -0.2638 0.1031 -0.0514 2.5840 H(7) -0.3447 0.0697 -0.1710 3.5190 H(8) -0.3400 0.1960 -0.1858 3.5190 H(9) -0.4774 0.0445 -0.2779 4.8515 H(10) -0.3261 0.1184 -0.3124 4.8515 H(H) -0.3312 -0.0080 -0.2975 4.8515 H(12) -0.114(5) 0.227(3) -0.323(2) 5.0(8) H(13) 0.010(4) 0.152(4) -0.344(2) 5.1(8) H(14) 0.060(5) 0.303(4) -0.311(2) 4.9(9) H(15) 0.5734 0.5195 -0.2267 2.8432 H(16) 0.5603 0.2684 -0.3124 4.6008 H(17) 0.5442 0.3659 -0.3558 4.6008 H(18) 0.8051 0.5013 -0.2928 5.6488 H(19) 0.8212 0.4038 -0.2494 5.6488 H(20) 0.882(3) 0.544(3) -0.116(2) 2.4(6) H(21) 0.918(5) 0.448(4) -0.086(2) 5.0(9) H(22) 0.881(4) 0.548(4) -0.025(2) 4.5(8) 129 Appendix: Crystallographic Analysis of 17 Table 1. Atomic coordinates and B i S 0 / B e , and occupancy (continued) atom X y z H(23) 0.868(5) 0.345(4) 0.183(2) 5.4(9) H(24) 0.846(4) 0.473(3) 0.201(2) 3.8(7) H(25) 0.823(4) 0.403(3) 0.269(2) 4.4(8) H(26) 0.4547 0.1819 0.2978 2.8726 H(27) 0.6234 0.2825 0.3217 2.8726 H(28) 0.5155 0.4352 0.3775 3.5526 H(29) 0.3438 0.3519 0.3367 3.5526 H(30) 0.2222 0.2349 0.5407 10.1717 H(31) 0.1341 0.1047 0.4773 10.1717 H(32) 0.3166 0.1390 0.5049 10.1717 H(33) -0.1969 -0.0297 0.2115 2.9551 H(34) -0.0136 0.0198 0.2408 2.9551 H(35) -0.0366 0.2098 0.3117 4.0382 H(36) -0.2173 0.1673 0.2782 4.0382 H(37) -0.3295 0.2151 0.5299 6.5600 H(38) -0.3697 0.0721 0.4839 6.5600 H(39) -0.1972 0.1485 0.5268 6.5600 H(40) -0.386(4) -0.021(3) 0.124(2) 4.6(8) H(41) -0.415(5) 0.076(4) 0.081(2) 7(1) H(42) -0.387(5) -0.058(4) 0.031(2) 5.4(9) H(43) -0.0465 0.4130 0.4710 20.1154 Beq = -7r2(t/u(aa')2 + Un{Wf + U33{cc*)2 + 2Vl2aa'bb' cos 7 + 2U13aa'cc* cos/3 + 2U2zbb'cct cosa) 130 References References 1. Milgrom, L. R. The Colours of Life, Oxford University Press: New York, 1997, preface. 2. Kuster, W. Z. Physiol. Chem. 1912, 52, 463. 3. Treibs, A. Angew. Chem. 1936, 49, 682. 4. Alafaleq, E. I. Pet. Sci. Technol. 1997, 75, 357. 5. Ali, S. A.; Cesani, F.; Nusynowitz, M . L.; Briscoe, E. G.; Shirtliff, M . E.; Mader, J. T. J. Nuc. Med. 1997, 38, 1999. 6. Reimers, J. R.; Lu, T. X.; Crossley, M . J., Hush, N.S. Nanotechnology 1996, 7, 424. 7. Neys, P. E. F.; Severeyns, A.; Vankelecom, I. F. J.; Ceulemans, E.; Dehaen, W.; Jacobs, P. A. J. Mol. Cat. A 1999,144, 373. 8. Iqbal, R.; Moratti, S. C.; Holmes, A. B.; Yahioglu, G.; Milgrom, L.R.; Cacialli, F.; Morgado, J.; Friend, R.H. J. Mater. Sci-Mater El 2000,11, 97. 9. Keseru, G. M. ; Balogh, G.; Czudor, I.; Karancsi, T.; Feher, A.; Bertok, B. J. Agr. Food Chem 1999, 47, 762. 10. Liu, C. Y.; Bard, A. J. Chem. Materials 1998,70, 840. 11. Sternberg, E. D.; Dolphin, D. Tetrahedron, 1998, 54, 4151. 12. Fischer, H.; Orth, H. Die Chemie des Pyrrols, Academische Verlagsgesellschaft: Leipzig, 1934, Vol I, 327. 13. Fischer, H.; Orth, H. Die Chemie des Pyrrols, Academische Verlagsgesellschaft: Leipzig, 1937, Vol II, Part 1. 14. Fischer, H.; Stern, A. Die Chemie des Pyrrols, Academische Verlagsgesellschaft, Leipzig, 1940, Vol II, Part 2. 15. Commission of the Nomenclature of Biological Chemistry, J. Am. Chem. Soc. 1960, 82,5581. 16. Moss, G. P. Eur. J. Biochem. 1988,178, 277. 17. The Colours of Life, Section 3.1.4. 131 References 18. Smith, K. M . In Porphyrins and Metalloporphyrins, Smith, K. M , Ed.; Elsevier: Amsterdam, 1975, p.399. 19. Porphyrins and Metalloporphyrins, p. 11. 20. Shanmugathasan, S.; Edwards, C ; Boyle, R. Tetrahedron 2000, 56, 1032. 21. Fuhrop, J.-H. In The Porphyrins, D. Dolphin, Ed. Vol. 2, Academic Press: New York, 1978, p. 131. 22. The Colours of Life, p.85 23. Gouterman, M . In The Porphyrins; Dolphin, D., Ed.; Academic Press: New York, 1978, Vol. 2, p. 65. 24. The Colours of life, p.52 25. Baldwin, J. E. Top. Curr. Chem. 1984,121, 181. 26. Morgan, B.; Dolphin, D. Struct. Bond. 1987, 64, 115. 27. Smith, K. M . ; Langry, K. C. J. Org. Chem. 1983, 48, 500. 28. Hunter, C. A. et al, J. Am. Chem. Soc. 1990,112, pp.5773, 5781, 5525. 29. http://www.qlt-pdt.com 30. Dolphin, D. Can. J. Chem. 1994, 72, 1005. 31. Van Lier, J. E. In Photobiological Techniques; Valenzeno, D. P.; Pottier, R. H.; Mathis, P.; Douglas, R. H., Eds. PlenumL: New York, 1991, p. 87. 32. Gilbert, A.; Baggot, J. Essentials of Molecular Photochemistry CRC Press, Boston, 1991, pp.512. 33. Schenck, G. O. Krauch, C. H. Angew. Chem. 1962, 74, 510. 34. Jensen, F.; Greer, A.; Clennan, E. L. J. Am. Chem. Soc. 1998,120, 4439. 35. Fedan, J. S.; Hogaboom, G. K.; O'Donnell, J. P. Biochem. Pharmacol. 1984, 33, 1167. 36. Rao, S.; Krauss, N. E.; Heerding, J. M . ; Swindell, I. R.; Ringel, I.; Orr, G. A.; Horwitz, S. B. J. Biol. Chem. 1994, 269, 3132. 132 References 37. Schuster, D. I.; Probst, W. C ; Ehrlich, G. K.; Singh, G. Photochem. Photobiol. 1989, 49, 785. 38. Morrison, H.; Ed. Bioorganic Photochemistry, Vol. 1, Photochemistry and Nucleic Acids, 1990, Wiley. 39. Fleming, S. A.; Tetrahedron 1995, 51, 12479. 40. Miller, L. J.; Hadac, E. M. ; Powers, S. P. Anal. Biochem. 1994, 220, 434. 41. Lamarche, N., Gaudreau, P.; Massie, B.; Langelier, Y. Anal Biochem. 1994, 220, 315. 42. Brown, R. L.; Gerber, W. V.; Karpen, J. W. Proc. Natl. Acad, Sci. U. S. A. 1993, 90, 5369. 43. Schuster, G. B.; Platz, M . S. Adv. Photochem. 1992,17, 69. 44. Swindell, C. S.; Heerding, J. M. ; Krauss, N. E. J. Med, Chem. 1994, 37, 1446. 45. Wani, M . C.; Taylor, H. L.; Wall, M . E.; Coggon, P.; McPhail, A. T. J. Am. Chem, Soc. 1971,95,2325. 46. Parness, J.; Horwitz, S. B. J. Cell. Biol. 1981, 91, 479. 47. Dorman, G.; Prestwich, G. D. Biochemistry 1994,33,5661. 48. Boyd, N. D.; Cerpa, R.; Kaiser, E. T.; Leeman, S. E.; White, C. F. Biochemistry 1991, 30, 336. 49. Ehret-Sabatier, L.; Kieffer, B.; Goeldner, M . Hirth, C. NATO ASI Ser., Ser. C; Photochem. Probes Biochem. 1989, 272, 107. 50. Smith, R. A. G.; Knowles, J. R. J. Am. Chem. Soc. 1973, 95, 5072. 51. Brunner, J.; Senn, H.; Richards, F. M . J. Biol. Chem. 1980, 255, 3313. 52. Nassal M . J. Am Chem. Soc. 1984,106, 7540. 53. Platz, M . ; Admasu, A. S.; Kwiatkowski, S.; Crocker, P. J.; Imai, N.; Watt, D. S. Bioconjugate Chem. 1991, 2, 337. 54. Hatanaka, Y.; Yoshida, E.; Nakayama, H.; Kanaoka, Y. Bioorg. Chem. 1989,17, 482. 55. Shih, L. B.; Bayley, H. Anal. Biochem. 1985,144, 132. 133 References 56. Bayley, H. Photogenerated Reagents in Biochemistry and Molecular Biology, 1983, Elsevier: Amsterdam. 57. a) Gronemeyer, H.; Govindan, M V. Molecul. Cell Endocr. 1986, 46, l.b) Tometsko, A .M. ; Richards, F. M . , Eds. Ann. N. Y. Acad. Sci. 1980, 346, 1. 58. Bayley, H.; Staros, J. V. in Azides and Nitrenes 1984, Academic Press: New York. 59. Potter, W. R. In Proceedings ofSPIE 1989,1065, 88. 60. Meier, H.; Zeller, K. P. Angew. Chem. Int. Ed. Engl. 1977, 89, 876. 61. Richter, A. M . ; Meadows H.; Ashol, K.J.; Canaan, A. J.; Levy, J.G. SPIE PDT of Cancer II1994, 2325, 189. 62. Engel, P. S.;Nalepa, C. J. Pure Appl. Chem. 1980, 52, 2621. 63. Tschirret-Guth, R. A., Medzihradsky, K. F., Ortiz de Montellano P. R. J. Am. Chem. Soc. 1999, 4731. 64. Durr; Kober Top. Curr. Chem. 1976, 66, 89. 65. Fleet, G. W. J.; Porter, R. R.; Knowles, J. R. Nature 1969, 224, 511. 66. Schrock, A.K.; Schuster, G. B. J. Am. Chem. Soc. 1984,106, 5228. 67. Pascual, A.; Casanova, J.; Samuels, H.H. / . Biol. Chem. 1982, 257, 9640. 68. Black, T. H. Aldrichimica Acta 1983,16, 3. 69. Pizey, J. S. In "Synthetic Reagents"; Halsted: New York, 1974; Vol. II, Chapter 4. 70. Carey, F. A.; Sundgberg, R. J. In "Advanced Organic Chemistry"; Plenum Press: New York, 1990, Third Ed., p.635. 71. Advanced Organic Chemistry , p.636 72. Houk, K. N. J. Am. Chem. Soc. 1973, 95, 4092. 73. Houk, K. N.; Sims, J.; Duke, R. E.; Strozier, R. W.; George, J. K. / . Am. Chem. Soc. 1973, 95, 7287. 74. Barfield, M . ; Sternhell, S. J. Am. Chem. Soc. 1972, 94, 1905. 75. Pinhey, J. T.; Sternhell, S. Tetrahedron. Lett. 1963,275. 134 References 76. Sternhell, S. Rev. Pure Appl. Chem. 1964,14, 15. 77. Barfield, M . ; Spear, R. J.; Sterhell, S. J. Am. Chem. Soc. 1971, 93, 5322. 78. Marshall, J. L.; Erickson, K. C ; Folsom, T. K. J. Org. Chem. 1970, 35, 2038. 79. Meier, H.; Zeller, K. P. Angew. Chem. Int. Ed. Engl. 1977,16, 835. 80. Askani, R. Chem. Ber. 1965, 98, 2551. 81. Wentrup, C. Helv. Chim. Acta. 1972, 55, 1613. 82. Rau, H. Angew. Chem. Int. Ed. Engl. 1973,12, 224. 83. Adam, W.; De Lucchi, O. Angew. Chem. Int. Ed. Engl. 1980,19, 762. 84. Hiberty, P. C ; Jean, V. J. Am. Chem. Soc. 1979,101, 2538. 85. Jones, M . In Organic Chemistry, Norton:New York, 1997, p.716. 86. For reviews, see a) Bonnett, R. Chem. Rev. 1995, 24, 19. b) Sessler, J. L.; Hemmi, G.; Mody, T. K.; Murai, T.; Burrell, A.; Young, S. W. Acc. Chem. Res. 1994,27,43. 87. Redmond, R. W. ; Gamlin, J. N. Photochem. Photobiol. 1999, 70, 391. 88. Mettath, S.; Shibata, M. ; Alderfer, J. L; Senge, M . O.; Smith, K. M. ; Rein, R.; Dougherty, T. J.; Pandey, R. K. J. Org. Chem. 1998, 63, 1646. 89. Newton, R. F.; Wadsworth, A. H. J. Chem. Soc, Perkin Trans. 1 1982, 823. 90. Wiberg, K. B.; O'Donnell, M . J. / . Am. Chem. Soc. 1979,101, 6660. 91. Branchadell, V.; Muray, E.; Oliva, A.; Ortuno, R. M . ; Rodriguez-Garcia, C. J. Phys. Chem. A 1998 102: 10106. 92. Essentials of Molecular Photochemistry, p. 169. 93. Eberson, L. Electron Transfer Reactions in Organic Chemistry, Springer-Verlag: Berlin Germany, 1987. 94. Cowan, D. O.; Baum, A. A. J. Am. Chem. Soc. 1971, 93, 1153. 95. a) Okada, T.; Figita, T.; Kubota, M. ; Masaki, S.; Mataga, N. Chem. Phys. Lett. 1972, 14, 563. b) Leonard, N. J.; McCredie, R. S.; Logue, M . W.; Cundall, R. J. Am. Chem. Soc. 1973, 85, 2320. 135 References 96. Boldyrev, A. I.; Schelyer, P. v. R.; Higgins, D.; Thomson, C ; Kramarenko, S. S. Comput. Chem. 1992,13, 1066. 97. a) Berson, J. A.; Bushby, R. J.; McBride, J. M. ; Tremelling, M . 7. Am. Chem. Soc. 1971, 93, 1544. b) Berson,, J. A.; McDaniel, D. M . ; Corwin, L. R.; Davis, J. H. J. Am. Chem. Soc. 1972, 94, 5508. c) Berson,, J. A.; McDaniel, D. M . ; Corwin, J. Am. Chem. Soc. 1972, 94, 5509. d) Platz, M . S.; Kelsey, D. R.; Berson, J. A.; Turro, N. J.; Mirbach, M . J. Am. Chem. Soc. 1977, 99, 2009. 98. Huisgen, R. Angew. Chem. Int. Ed. Engl. 1963, 2, 565. 99. Neidlein, R.; Salzmann, K. Synthesis 1975, 52. 100. Van Loock, E.; Vandensavel, J. M. ; L'abbe, G.; Smets, G. J. Org. Chem. 1973, 38, 2916. 101. Salaun, J. Topics in Current Chemistry, Vol. 207, 2000. 102. Bonnett, R.; White, R. D.; Winfield, U. J.; Berenbaum, M . C. Biochem. J. 1989, 261, 277. 103. Braichotte, D.; Wagnieres, G.; Philippoz, J-M.; Bays, R.; Ris, H-B.; van den Bergh, H. in Photodynamic Therapy in Biomedical Lasers, Elsevier Science: Amsterdam, 1992, p.461. 104. Chen, F-Q.; Hirano, T.; Hashizume, Y.; Ohmiya, Y.; Ohashi, M. / . Chem. Soc, Chem. Commun. 1994, 2405. 105. Masukawa, T.; Kato, H.; Kakiuchi, T.; Jayasundera, K. P.; Kinoshita, H.; Inomata, K. Chem. Lett. 1998, 455. 106. Boyle, R. W.; Bruckner, C ; Posakony, J.; James, B. R.; Dolphin, D. Org. Synth. 1999, 76, 287-293. 107. Li , G.; Chen, Y.; Missert, J. R.; Rungta, A.; Dougherty, T. J.; Grossman, Z. D.; Pandey, R. K. J. Chem. Soc, Perkin Trans 1, 1999, 1785. 108. Pandey, R. Roswell Park Cancer Institute, personal communication to D. Dolphin. 109. Wijesekera, T. P.; Dolphin, D. in Metalloporphyrins in Catalytic Oxidations, ed. Sheldon, R.; Dekker, M . , New York, 1994, 193. 110. The Colours of Life, p.75. 136 References 111. Senge, M . O.; Gerstung, V.; Ruhlandt-Senge, K.; Runge, S.; Lehmann, I. J. Chem. Soc, Dalton Trans. 1998 , 24, 4187. 112. The Colours of Life, p.75. 113. Fuhrop, J-H.; Laboratory Methods in Porphyrin and Metalloporphyrin Chemistry, Amsterdam: Elsevier, 1975. 114. Porphyrins and Metalloporphyrins, p. 182 115. Adler, A. D.; Longo, F. R.; Kampas, F.; Kim, J. Inorg. Nucl. Chem. 1970, 32, 2443. 116. Ponomarev, G. V.; Maravin, G. B. Chemistry of Heterocyclic Compounds 1982,18, 50. 117. ICNCM. International Committee on Nickel Carcinogenesis in Man., Report of the International Committee on nickel carcinogenesis in man. Scand. J. Work Environ. Health 1990,16,1. 118. Prakash, G. K. S.; Krishnamurthi, R.; Olah, G. A. / . Am. Chem. Soc. 1989, 111, 393. 119. Ruppert, I.; Schlich, K.; Volbach, W. Tetrahedron Lett. 1984, 2195. 120. Shiau, F.; Pandey, R. K.; Ramaprasad, S.; Dougherty, T. J.; Smith, K. M . J. Org. Chem. 1990, 55,2190. 121. Griffiths, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D. / . Chem. Soc, Chem. Commun. 1988, 1625. 137 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061435/manifest

Comment

Related Items