UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Trivalent metal complexes of 3-hydroxy-4-pyrone and 2-acyl-3-hydroxyfurans Lutz, Tammy 1988

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

Item Metadata


831-UBC_1989_A6_7 L87.pdf [ 4.17MB ]
JSON: 831-1.0060357.json
JSON-LD: 831-1.0060357-ld.json
RDF/XML (Pretty): 831-1.0060357-rdf.xml
RDF/JSON: 831-1.0060357-rdf.json
Turtle: 831-1.0060357-turtle.txt
N-Triples: 831-1.0060357-rdf-ntriples.txt
Original Record: 831-1.0060357-source.json
Full Text

Full Text

TRIVALENT METAL COMPLEXES OF 3-HYDROXY-4-PYRONE AND 2-ACYL-3-HYDROXYFURANS by TAMMY LUTZ B S c , University of British Columbia, British Columbia, Canada - 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Chemistry) We accept this thesis as conforming to the required standard The University of British Columbia November 1988 ©Tammy Lutz, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of £ / / / g / W / c / / ? * / The University of British Columbia Vancouver, Canada DE-6 (2/88) ii ABSTRACT Group 13(IIIA) metal complexes (ML3) of 1 -(3-hydroxy-2-furanyl)ethanone (Hima), 3-hydroxy-4H-pyran-4-one (Hpa), and a newly synthesized ligand 2-(N-methylamido)-3-hydroxy-5-methylfuran (Hahm) have been prepared, as well as the Gd complex of Hima. They have been characterized by IR, NMR, and UV spectroscopies, as well as mass spectrometry, and elemental analyses. All were consistent with the proposed tris(ligand)metal(lll) structure. The Al and Ga complexes of isomaltol have been studied by single crystal X-ray diffraction showing two out of the three ligands to be disordered. Solution equilibrium studies (0.15 M NaCl, 25°C) showed that Hpa is a better binder for the Group 13 metal ions than either Hima or Hahm. iii TABLE OF CONTENTS page Abstract ii Table of Contents iii List of Tables v List of Figures vi List of Abbreviations viii Acknowledgements xi Chapter 1. Introduction 1 Chapter 2. Isomaltol Experimental 9 Results and Discussion 2.1. Infrared Spectra 12 2.2. Mass Spectra 15 2.3. 1 H NMR Spectra 16 2.4. 27AINMR 18 2.5. Elemental Analyses 22 2.6. X-ray Crystallographic Analyses 23 2.7. Ultraviolet Spectra 26 2.8. Potentiometric Studies 28 2.9. Synthetic Attempts 31 Conclusion 32 Chapter 3. Hehm and Hahm 34 Experimental 36 A. Results and Discussion: Hehm 3.1a. Infrared Spectrum 39 3.2a. 1 H NMR Spectrum 40 iv 3.3a. 1 3 C NMR 40 3.4a. Mass Spectrum 42 3.5a. Attempts to Prepare Complexes of Hehm 43 3.6a. Hehbm 45 B. Results and Discussion: Hahm 52 3.1b. Infrared Spectra 53 3.2b. 1 H NMR Spectra 56 3.3b. Elemental Analyses 56 3.4b. Mass Spectra 57 3.5b. Ultraviolet Spectra 58 3.6b. Potentiometric Studies 61 Conclusion 63 Chapter 4. Pyromeconic Acid 64 Experimental 65 Results and Discussion 4.1. Infrared Spectra 66 4.2. NMR Spectra 68 4.3. Elemental Analyses 69 4.4. E.I. Mass Spectra 70 4.5. Potentiometric Studies 70 Conclusion 74 Future Work 74 References 75 Appendix 80 V LIST OF TABLES page Table 2.1. Infrared Absorptions of the Metal Complexes of Hima. 14 Table 2.2. Mass Spectral Data for M(ima)3 Complexes. 15 Table 2.3. NMR Properties of Al, Ga, and In Isotopes. 19 Table 2.4. Elemental Analyses of M(ima)3 Complexes. 22 Table 2.5. Logarithms of the Metal-lsomaltol Stability Constants. 29 Table 3.1. Elemental Analysis for Hahm and M(ahm)3. 57 Table 3.2. FAB Mass Spectral Data for the Metal Complexes of Hahm. 58 Table 4.1. Infrared Absorptions of Hpa and M(pa)3. 67 Table 4.2. 1 H NMR Data for Hpa and M(pa) 3. 68 Table 4.3. Elemental Analyses for the Metal Complexes of Hpa. 69 Table 4.4. Logarithms of the Metal-Pyromeconic Acid Overall 70 Stability Constants. Table A.1. Crystallographic Data for Al(ima)3. 80 Table A.2. Final Positional and Isotropic Thermal Parameters for 81 Al( ima) 3 . Table A.3. Calculated Hydrogen Coordinates and Isotropic Thermal 82 Parameters for Al(ima)3. Table A.4. Bond Lengths for Al(ima)3. 83 Table A.5. Bond Angles for Al(ima)3. 84 vi LIST OF FIGURES page Figure 1.1. Speciation diagrams of Al in aqueous conditions. 2 Figure 1.2. Maltol (1) and the pyridinones (2). 5 Figure 2.1. Isomaltol (Hima). ' 7 Figure 2.2. Synthesis of metal complexes of isomaltol. 12 Figure 2.3. Hydrogen bonded chelate structure of Hima. 13 Figure 2.4. Deprotonated isomaltol. 13 Figure 2.5. Isomaltol ligand fragmentation from E.I. MS. 16 Figure 2.6. 80 MHz- 1 H NMR of isomaltol and its metal complexes. 17 Figure 2.7. 2 7 A I NMR of Al(ima)3 in H2O (after the subtraction routine).21 Figure 2.8. Superposition of the two possible binding orientations 24 of A!(ima)3. Figure 2.9. ORTEP diagram of Al(ima) 3. 25 Figure 2.10. Variable pH UV spectra of isomaltol. 27 Figure 2.11. Definition of formation constants. 28 Figure 2.12. Speciation diagrams of M(ima)3. 30 Figure 2.13. Speciation diagram of 0.05 M Al/0.15 M Hima. 31 Figure 2.14. Attempts to form an imine of isomaltol. 32 Figure 3.1. Hehm synthesis. 34 Figure 3.2. 80 MHz- 1 H NMR spectrum of Hehm in CDCI3. 41 Figure 3.3. The chemical shift assignments for the 1 3 C spectrum of 42 Hehm. Figure 3.4. Fragmentation pattern of Hehm. 43 Figure 3.5. Hehbm. 45 Figure 3.6. Major fragmentation peaks seen in the mass spectrum 46 of Hehbm. vii Figure 3.7. SINEPT pulse sequence. 47 Figure 3.8. Population inversion of a 1 H - 1 3 C coupled system. 48 Figure 3.9. The 1 3 C SINEPT spectra of Hehbm. 50 Figure 3.10. Hahm synthesis in ethanol. 52 Figure 3.11. Amide resonance forms. 54 Figure 3.12. Infrared spectra of Hahm and its metal complexes 55 in the 3700 to 2600 cm- 1, and 1800 to 1400 cm- 1 region. Figure 3.13. Variable pH UV spectra of Hahm. 59 Figure 3.14. Hydrogen bonding in Hahm. 61 Figure 3.15. The speciation diagrams of the metal complexes of 62 Hahm. Figure. 4.1. Pyromeconic acid (Hpa). 64 Figure 4.2. The synthesis of the metal complexes of Hpa. 66 Figure 4.2. The E.I. mass spectra of Al(pa)3 and Ga(pa)3. 71 Figure 4.3. Speciation diagrams of Al(pa)3 and Ga(pa)3. 73 Figure A.1. The 1 3 C spectrum of Hehbm. 85 viii LIST OF ABBREVIATIONS Abbreviation Meaning 3 J C H 1 3 C - 1 H coupling through three bonds G n overall formation constant acac 2,4-pentanedione APT attached proton test ax isotopic abundance br broad C D C I 3 deuterated chloroform d doublet d 4 -MeOH deuterated methanol DCI direct chemical ionization deg. degrees DMSO dimethylsulfoxide e molar absorptivity E.I. electron impact EDSA ethyl(dimethylsulfuranylidene)acetate FAB fast atom bombardment Hahm 2-(N-methylamido)-3-hydroxy-5-methylfuran Hdpp . 3-hydroxy-1,2-dimethyl-4-pyridinone Hehbm 2-ethoxycarbonyl-3-hydroxy-4-(1 ',3'-butanedione)-5-methylfuran Hehm 2-ethoxycarbonyl-3-hydroxy-5-methylfuran Hima 1-(3-hydroxy-2-furanyl)ethanone (isomaltol) Hpa 3-hydroxy-4H-pyran-4-one (pyromeconic acid) ix Hz Hertz 1 nuclear spin IR infrared K n : stepwise formation constant -^max wavelength of maximum absorbance M molar m medium intensity band MHz megahertz MS mass spectrometry msec millisecond V vibration (infrared spectroscopy) NMR nuclear magnetic resonance ORTEP Oak Ridge Thermal Ellipsoid Plot ppm parts per million Q nuclear quadrupole Rx relative receptivity o bending deformation s singlet SINEPT selective insensitive nuclei enhanced by polarization T delay time (NMR spectroscopy) T i spin-lattice (longitudinal) relaxation T 2 spin-spin (transverse) relaxation tame 1,1,1 -tris(aminomethyl)ethane tc correlation time UV ultraviolet w weak intensity band w 1 / 2 peak width at half peak height X W x relative peak width A Angstrom {d2v/dz2}2 electric field gradient A delay time during refocussing phase (NMR spectroscopy) x i ACKNOWLEDGEMENTS I would like to thank Dr. Chris Orvig for his support, guidance, and enthusiasm throughout this work. Many thanks to the members of the "Orvig team" past and present, especially the founders: -Bill "our hero" Nelson, Speciation Dave, and Smythe (alias Alexis), for their encouragement and helpful assistance. Thanks also to Fraser for our enlightening Friday afternoon discussions. Special thanks to Dr. Ray Andersen and Dr. Nick Burlinson for their invaluable spectroscopy and NMR knowledge. Mrs. Marietta Austria and Mr. S.O. Chan are also gratefully acknowledged for their expedient service and aid with various NMR experiments. I would like to express my appreciation to Mr. Peter Borda for the prompt elemental analyses, and Dr. Steven Rettig for his excellence in X-ray crystallography. Thanks also to the mass spectrometry laboratory and the UBC technical support staff. Finally, thanks to Clive for helping me maintain my sanity, and to turtle Inalga for the entertaining feeding sessions, and delightful company (she never talks back) on those late nights. xij To all my caring and supportive parents. 1 Chapter 1. Introduction In the past, few workers studied the coordination chemistry of the Group 13(11 IA> metals in aqueous solution because of their extensive hydrolysis. The correlation between aluminum and several dysfunctions such as Alzheimer's disease and dialysis encephalopathy, 1- 2 as well as the use of gallium and indium as nuclei in radiopharmaceuticals, 3 have generated more interest in the aqueous behaviour of the Group 13 metals, and initiated our pursuit in this area. The hydrolysis of the Group 13 metals has been thoroughly examined. Under acidic conditions, in the absence of ligands, these metal ions have a coordination number of six and exist as the tricationic metal hexaquo spec ies . 4 - 5 Depending on pH and concentration, many mono and polynuclear hydrolytic species can be formed. 6 Aluminum and gallium, however, form more high molecular weight polynuclear species than indium. 2 ! I I J I I yr] I I I | I I i i i i 1 l l 10 12 Figure 1.1. Speciation diagrams of Al at (a) 0.1 M and (b) 10"5 M in aqueous conditions7. The dotted lines signify supersaturation with respect to AI(OH)3 solid phase. The numbers (x,y) in the above diagram represent Al x(OH) y hydrolysis species. For example, in Figure 1.1, at pH 7.4 and 10~5 M, aluminum exists primarily as the AI(OH)3 species, yet at 0.1 M concentrations, the predominant species is [Ali3C>4(OH)24]7+. Gallium and indium undergo hydrolysis that is similar to aluminum, but their polynuclear species have not yet been accurately determined.8 Aluminum is an abundant and ubiquitous element.9 With the increased amount of processed foods and convenience products, North Americans ingest between 5 to 50 milligrams of aluminum daily. Aluminum occurs in different forms in such products as processed cheese slices, baking powder, antiseptics, antacids, antiperspirants, deodorants, and is used in the purification of drinking water . 1 0 - 1 3 Recently, considerable evidence has accumulated that suggests a correlation between aluminum and such diseases as Alzheimer's disease and 3 dialysis encepha lopa thy . 1 4 - 1 6 Alzheimer's disease is a progressive dementia. 1 7 It starts with memory loss and disorientation, progresses to loss of intellectual capacities, and results in a state of total dependence. The cause of Alzheimer's disease is not known, but several theories have been postulated, including the aluminum toxicity theory. 1 8 . 1 9 Patients who have Alzheimer's disease have an increased amount of neurofibrillary tangles and neuritic plaques distributed in a patchy manner throughout the brain. 2 0 - 2 1 Autopsies have found an elevated level of aluminum in brain cell nuclei of people suffering from Alzheimer's disease that is proportional to the number of neurofibrillary tangles. The toxicity of aluminum has been indicated by the intercranial injection of an aqueous solution of AICI3 into trained animals resulting in loss of cognitive and memory functions; 2 2 biopsy or autopsy also showed neurofibrillary tangles similar to those found in humans. Why most people can ingest large amounts of aluminum without any deleterious effects and some people cannot, is not understood. It appears that Alzheimer's disease involves a complex mechanism of neurological changes encompassing genetic and environmental factors 2 3 - 2 4 that are not yet clearly understood. There is not enough evidence to link aluminum as a causitive element for acquiring Alzheimer's disease, but it is definitely involved in the pathogenesis of the disease. The in vivo use of gallium and indium as radiopharmaceuticals is also a rapidly expanding area. Both gallium and indium have two radioactive isotopes; 6 7 . 6 8 G a and in . i i 3 m | n . 6 7 G a and n i . i 1 3 m | n are all gamma emitters with energies suitable for easy detection. 6 7 G a and i n . 1 1 3 m | n a | S 0 have suitable haif lives (t-|/2=3.3 days, 2.8 days, and 1.7 hrs. respectively)2 5 so that 4 the amount of radiation exposure to the patient is minimal. ^ G a is primarily a positron emitter which is detected by positron emission tomography (PET). Radiopharmaceuticals are agents containing radioactive nuclei. When the agents are introduced in vivo, they are found to have a particular organ specificity useful for diagnostic or therapeutic purposes. For a compound to be used in vivo, it must be stable at physiological pH, and H2O soluble to some extent. Using Group 13 metals in vivo has one major problem in addition to hydrolysis. A blood plasma protein, transferrin, involved in iron transport can bind two equivalents of a trivalent metal. 2 6 Due to the Group 13 metals having a +3 charge and a size close to that of Fe+ 3 ( G a + 3 in particular2 7), their ions may bind with transferrin when injected in the bloodstream. The formation constant for metal-transferrin complexes are quite large (i.e. log K 1 = 19.53 for gal l ium). 2 8 Therefore, any metal complex used in vivo must have a high formation constant in order to compete with transferrin and maintain its structure, potentially resulting in a unique biodistribution. Gallium-67 citrate, which becomes gallium-transferrin in vivo, is presently used to detect soft tissue tumors 2 9 - 3 0 and abscesses. Although the biodistribution of gallium-transferrin is useful, the presence of transferrin in the blood presents a barrier to the development of new radiopharmaceuticals. Even though radioactive Ga and In complexes are presently used as radiopharmaceuticals, a complex having an 'ideal' biodistribution has not yet been found. In order to develop new radiopharmaceuticals, and to elucidate the role of aluminum in disease, we must first investigate the coordination chemistry of aluminum, gallium, and indium under physiologically relevant conditions. Some studies using gadolinium were also undertaken, and will be discussed where applicable. 5 The first challenge is to find a ligand that will sucessfully bind with the Group 13 metals, and compete with hydrolysis. Past research in our group has shown that bidentate ligand systems containing an cc-hydroxyketone functionality were good ligands for the Group 13 metals, some with high formation constants. In particular, maltol (3-hydroxy-2-methyl-4H-pyran-4-one) (1) was found to form tris(maltolato)metal(lll) complexes that are water soluble, neutral, and of reasonable thermodynamic stability.31 Other related bidentate ligands are the 3-hydroxy-4-pyridinones (2), 3 2 which are synthesized from maltol and a primary amine via a ring opening insertion reaction. 3 3 The formation constants of the tris(pyridinone)metal(lll) complexes were higher than those of maltol, 3 4 and the solubility properties of these ligands and their complexes can be varied by the insertion of different primary amines. However, since the site of insertion of the different amines is far from the binding site, (Figure 1.2) the formation constants were found to be largely unaffected by various alkyl substitutions (R). OL O , C H i O H R I , C H , O H O R= H, alkyl 1 2 Figure 1.2. Maltol (1) and the pyridinones (2). In order to improve the formation constants of the metal complexes, other ligand systems were investigated which would 1) have the potential for 6 substitution at a position closer to the binding site, or 2) have the potential for the formation of a hexadentate ligand. A reasonable candidate was isomaltol, which contains a 13-hydroxyenone moiety (rather than an ct-hydroxy ketone as in maltol and the pyridinones). Initially it was thought that the acyl side chain would be able to undergo condensation with primary,amines to form various Schiff base l igands, 3 5 - 3 6 with the future aim of synthesizing a hexadentate imine ligand (eg. the condensation of 1,1,1 tris(aminomethyl)ethane with three equivalents of isomaltol). Unfortunately, reactions on the side chain were not possible because of preferential ring expansion and the formation of pyridinones. Another furan-based ligand, 2-ethoxycarbonyl-3-hydroxy-5-methylfuran (Hehm), 3 7 similar to isomaltol was then investigated. Hehm has an ester side chain which is more likely to undergo amide formation than a ketone is to undergo Schiff base formation, allowing substitution to occur without ring insertion and opening problems. This ligand was particularly useful as a precursor to other amide ligands. In particular, the methyl amide of Hehm, Hahm, was prepared and studied. 7 Chapter 2. Isomaltol (Hima) Isomaltol, 1-(3-hydroxy-2-furanyl)ethanone (Hima), is a naturally occurring carbohydrate byproduct which has a strong caramel odor, and acts as a flavoring agent in beer 3 8 and roasted coffee 3 9. Both isomaltol and maltol are found in trace amounts in the steam distillate of bread. 4 0 In 1910, isomaltol was thought to have a structure based on a pyrone ring because it had similar properties to maltol 4 1 and it was not until 1961 that a furan structure was proposed for isomaltol. 4 2 This furan structure gives rise to a metal binding site that is a 8-hydroxyenone moiety. O Figure 2.1. Isomaltol (Hima) Isomaltol was studied for two reasons: 1) to determine if the (3-hydroxyenone functionality is a good binder for the Group 13 metals, and 2) to pursue its potential as a precursor to various imine furans and possibly a hexadentate tri-imine. Hima was prepared from lactose in two steps according to J.E. Hodge and E.C. Ne lson . 4 3 The glucose sugar of lactose is dehydrated to form O-galactosylisomaltol, which is subsequently hydrolyzed to isomaltol giving an 8 overall yield of -15%. Hima can also be prepared from maltose 4 4 and is one of the degradation products of sucrose 4 5 , but these processes give lower yields and/or a greater mixture of products. Isomaltol is known to form complexes with sodium 4 6 copper, and iron 4 3 , but no complexes with the Group 13 trivalent metals or gadolinium have been reported. 9 Experimental Instrumentation and Methods All infrared spectra were recorded in the solid form using a KBr pellet on a Perkin-Elmer PE783 spectrophotometer. A polystyrene band at 1601 cm' 1 was used as a reference. The proton NMR spectra (80 MHz) were recorded on a Bruker WP-80, and the 2 7 AI NMR (78.16 MHz) spectra on a Varian XL-300 at 18°. Some 1 H NMR spectra were also recorded on a Varian XL-300 by Marietta Austria of this department. The spectra were referenced to tetramethylsilane (TMS) and AI(H20)6 3 + respectively. Crystallographic data were obtained on an Enraf-Nonius CAD4-F diffractometer for Al(ima)3 by Dr. Steven Rettig of this department, and a Rigaku AFC6R diffractometer for Ga(ima)3 by the crystallographic staff of Molecular Structure Corporation. The UV spectra were recorded on a Perkin-Elmer Coleman 124 (Double Beam) spectrophotometer from 370 to 200 nm. All UV spectra were obtained in 0.15 M NaCl solutions. The mass spectra were recorded on a Kratos MS50 (electron impact ionization, E.I.), an AEI MS 9 (fast atom bombardment, FAB), and a Nermag R 10-10 (direct chemical ionization, DCI) by the mass spectrometry laboratory. The base peak is set at 100% intensity. Potentiometric measurements of the ligands in the absence, and presence, of metal ions were performed by Dr. David Clevette of this laboratory, 10 with an Orion Research EA 920 pH meter equipped with Orion Ross research grade glass and reference electrodes. Mr. Peter Borda of this department performed the C, H, N, and Gd analysis. The elemental analysis of the Group 13 metals were done by the author gravimetrically with oxine. 4 7 Starting materials Lactose (Aldrich) was used as supplied. The AI(N03)3-9H20 (Mallinckrodt), ln(N03)3-5H20 (Alfa) and Gd(N03)3-6H20 (Alfa) metal salts were used as supplied. Ga ingots (Alfa) were dissolved in concentrated hydrochloric acid and standardized. Synthesis of O-aalactosvlisomaltol. The synthesis was performed according to J.E. Hodge and C.E. Nelson 4 3 During the heating of the lactose mixture, the solution was refluxed and 70 ml (instead of 50 ml) of trimethylamine were added. Hvdrolvsis of O-aalactosvlisomaltol. The hydrolysis was performed according to J.E. Hodge and C.E. Nelson 4 3 but required the collection of only 700 ml of distillate from the steam distillation (instead of 1000 ml). The distillate was extracted ten times with chloroform (15 ml each). The chloroform volume was reduced and triturated with petroleum ether to give 2.49 g ( 38% ) of an off-white precipitate. It was then recrystallized from ether. It was found to be soluble in alcohols, acetone, ether, chloroform, benzene, ethyl acetate, and hot water. Synthesis of trisM'somaltolatoteluminumfllh. AKimaV^. To isomaltol ( 1.0 g, 7.9 mmol) in 40 ml H2O was added AI(N03)3-9H20 (0.99 g, 2.6 mmol) in 10 ml 11 of H2O. The pH was adjusted to 6.8 with aqueous 2M sodium hydroxide, and an off-white solid precipitated. Recrystallization from methanol gave 0.89 g (86%) of product. It was soluble in water (0.62 mM), chloroform, methanol, acetonitrile, acetone, slightly soluble in ethanol, and insoluble in diethyl ether. Other characterization data is presented elsewhere. Synthesis of tris(isomaltolato)aalliumnin. Gan'maVfr The preparation was as for Al(ima)3 with isomaltol (1.0 g, 7.6 mmol) and 1.37 M GaCl3 solution (1.92 ml, 2.6 mmol). The yield of the off-white powder was 1.04 g (88%). Recrystallization from methanol/ether gave crystalline rods. It was soluble in methanol, ethanol, chloroform, sparingly soluble in water, and insoluble in diethyl ether. Synthesis of tris(isomaltolato)indium(lll). ln(imaVTt-1/4H?0. The preparation was as for Al(ima)3 with isomaltol (1.38 g, 10.9 mmol) and l n(N03)3-5H20 (1.43 g, 3.7 mmol). The yield of the off-white precipitate was 1.44 g (80%). The crude precipitate was recrystallized from methanol. It was soluble in chloroform, and slightly soluble in methanol and ethanol. Synthesis of tris(isomaltolato)aadoliniumHII). GdfimaV^HgO. The preparation was as for Al(ima)3 with isomaltol (0.78 g, 6.2 mmol) and Gd(N0 3 ) 3 -6H 2 0 (0.86 g, 1.9 mmol). The yield off the off-white product was 0.72 g (68%). The off-white product was recrystallized from methanol. It was soluble in methanol, chloroform, slightly soluble in water and ethanol. 12 Results and Discussion Hima forms neutral tris(isomaltolato)metal(lll) complexes of Al, Ga, In, and Gd with yields of 68-88%. Formation of these metal complexes is obtained by a three to one stoichiometric addition of ligand to trivalent metal in water and adjusting the pH to ~7. All metal complexes were characterized by IR, NMR, MS, and elemental analyses and each was consistent for the proposed tris(isomaltolato)metal(lll) as in Figure 2.2. Figure 2.2. Synthesis of metal complexes of isomaltol, where M = aluminum, gallium, indium, and gadolinium. 2.1. Infrared Spectra Isomaltol exists as a hydrogen bonded conjugate chelate structure. 4 8 This is seen in the infrared spectrum of isomaltol which has a strong broad O-H stretch from 314Q-2670 c m - 1 and a carbonyl stretch occuring from 1620-1550 c m - 1 . Dilution studies in CCU, as well as the weakness and breadth of the O H stretch are evidence for intramolecular hydrogen bonding 4 9 as seen in Figure 2.3. The carbonyl stretch for isomaltol is not separable from the C = C stretch, and contributes to the broadening of this band. 1 3 CH 3 CH 3 o. / O O - H Figure 2.3. Hydrogen bonded chelate structure of Hima Each of the metal complexes of Hima has two weak bands (3140, 3120 cm"1) which are diagnostic of C-H stretching modes in furans. 5 0 Deprotonation of the 8-hydroxyenone moiety of isomaltol enables delocalizaton forming a binding site analogous to acetylacetone (acac). Upon chelation, the O-H stretch disappears, and the broad band in the carbonyl region is resolved into three sharp bands (1615-1508 cm' 1). This is indicative of a lack of hydrogen bonding, thus evidence for the formation of a metal complex. The carbonyl bands for the metal complexes undergo bathochromic shifts relative to Hima due to the donation of electron density from the oxygen to the metal, thus weakening the carbonyl bond. These three bands are assigned CH H Figure 2.4. Deprotonated isomaltol 1 4 as combinations of C=0 and C=C stretches. Assignments of the bands in the infrared spectra of the metal complexes of Hima are shown in Table 2.1. Table 2.1. Infrared Absorptions of the Metal Complexes of Isomaltol (cm-1) Al(ima)3 Ga(ima)3 Gd(ima)3 In(ima)3 assignment 3140 3148 3140 3145 vC-H 1602 1588 1615 1583 1555 1547 1544 1538 vC=0, C=C 1515w 1514m 1508m 1510 1305 1294 1280 1280 vC-0 1162 1167 1164 1166 o-C-Hc 1143 1139 1138 1135 978m 974m 969m 973m oC-Hd 858 856 865 857 479 m 430m - 426w vM-O* a Al l strong (except where labelled m=medium, and w=weak) intensity. bv=vibration; o=bending deformation. c ln plane. dOut of plane. * tentative assignments. 15 2.2 Mass Spectra The mass spectra of M(ima)3 were obtained by electron impact ionization. Each complex has fragment ion peaks, ML-3+, Ml_2+, and ML+, corresponding to the tris, bis, and mono ligand-metal species respectively. In each case, the M L 2 + species gave the base peak. Table 2.2 . Mass Spectral Data for M(ima)3 Complexes. m/z Al Ga* In Gd* ML 3+ 402 444,446 490 530-533,535 M L 2 + 277 319,321 365,363 405-408,410 ML+ 152 194,196 240,238 280-283,285 * Gallium and indium have two isotopes each 6 9 - 7 1 G a , and 1 1 3 - 1 1 5 l n , while gadolinium has five consecutive isotopes 154-158QCJ A S W E | | A S 152,160GCJ. 1 6 Other fragments seen in the mass spectra were the parent ion peak of isomaltol (HL+ at m/z=126), and its fragmentation shown below. 5 1 2.3. 1 H NMR Spectra The proton N M R of Hima in CDCI3 consists of four peaks at 2.41 (s, 3H ), 6.95 (s, 1H ), 6.29 (d, 1H ), 7.29 (d, 1H ) ppm. Due to the paramagnetism of the gadolinium nuclei, N M R was not used as a method for characterization. However, comparison of the other metal complexes with Hima show shifts in all of the protons ranging from chemical shift changes of 0.03 for H a to 0.23 ppm for H c . In all cases, the protons nearest the binding site undergo the greatest shifts upon coordination to the metal. The 1 H N M R spectra of Hima and its metal complexes were obtained in CDCI3, and are shown in Figure 2.6. m/z =111 m/z =83 m/z =55 Figure 2.5. Isomaltol ligand fragmentation from E.I. MS. 17 11! U . U . Hi M - I I I I ' • ' I L . a 1 ppm Figure 2.6. 80 MHz- 1 H NMR of (i) Hima, (ii) Al(ima)3, (iii) Ga(ima)3, and (iv) ln(ima)3 in CDCI3. 18 2.4. 27AINMR The 2 7 AI nucleus is quadrupolar with a nuclear spin of 5/2. The Ti and T2 relaxation times are therefore dominated by nuclear quadrupolar relaxation.52 The local electric field gradients generated by the ligand field symmetry interact with the quadrupolar nucleus, coupling the nucleus to the random Brownian motion of the sample. This provides a very efficient relaxation mechanism, resulting in short relaxation times. The relaxation rates are dependent primarily on the quadrupole moment of the nucleus, and the nuclear spin, according to the equation: 5 3 _L _ J JL 2I+3 /eQ\2 /-d2v\2 Ti - T 2 - 4 0 | 2 ( 2 | - l ) \ h J \ d z 2 j ' T c Where I is the nuclear spin (I > 1), Q is the nuclear quadrupole moment, d 2v/dz 2 is the maximum electric field gradient at the nucleus, and x c is the correlation time for molecular Brownian motion. An ideal nucleus would therefore have a small quadrupole moment and a large nuclear spin so as to have a long enough spin-spin relaxation time to obtain narrow lines in NMR. Of all the Group 13 nuclei, the 2 7 AI nucleus has the most favorable properties for NMR studies. The 2 7 AI nucleus is quadrupolar in nature (I = 5/2), is 100% naturally abundant, highly sensitive (0.206 relative to protons), and has a relatively small quadrupole moment (0.149 x 10" 2 8 m2). See Table 2.3 5 4 19 Table 2.3. NMR Properties of Aluminum, Gallium, and Indium Isotopes. Nucleus, X 27AI 69Ga 7iGa 113m H5 | n Isotopic Abundance % a x 100 60.4 39.6 4.28 95.72 Relative receptivity, R x with respect to 1 H 0.206 0.042 0.056 0.015 0.332 NMR Frequency 1 H at 100 MHz 26.057 24.001 30.496 21.865 21.912 Spin, l x 5/2 3/2 3/2 9/2 9/2 Quadrupole Moment, Q x e x 10" 2 8 m 2 0.149 0.178 0.112 1.14 0.83 (2l+3)/l2(2l-1)=F(l) 0.32 1.33 1.33 0.019 0.019 Q2F(I)/0.0071 relative peak width, W x 1.0 5.94 2.35 13.55 7.18 Relative peak height RXA/Vxwith 2 7AI =100 100 3.4 11.6 0.54 22.4 Table 2.3 shows various properties of the Group 13 isotopes. The 2 7 AI nucleus gives the narrowest intrinsic line width and the greatest relative peak height compared to the other Group 13 nuclei. For these reasons, the 2 7 AI nucleus was studied over the others. The short relaxation times encountered in NMR due to nuclei having a quadrupolar moment may lead to broad NMR signals that provide little useful 2 0 information. The 2 7 A l signal can be as broad as several kilohertz 5 5 and even vanish into the baseline, depending on the local electric field gradients generated by the surrounding ligands. 5 6 If the ligands of an aluminum complex have high symmetry about the quadrupolar aluminum nucleus, the local electric field gradients generated at the nucleus are small, and the relaxation times will be longer than a complex of low symmetry, giving rise to narrow NMR s igna ls . 5 7 - 5 8 Sharp lines are therefore obtained for aluminum complexes that have an Oh or Td symmetry. For this reason, the highly symmetric AI (H20 )6 3 + is used as the zero ppm reference standard. Presence of aluminum in the probe head ceramics results in a broad signal at -62 p p m 5 9 - 6 0 that may interfere with the spectra in samples of low water solubility, <0.01M. This problem can be overcome by acquiring a spectrum of a water blank, and performing a subtraction of the blank from the spectrum of the aluminum complex. 6 1 Phasing problems may arise after subtraction if the matrices of the blank and the aluminum complex solutions are not consistent. The 27AI NMR spectrum of Al(ima)3 was obtained in H2O, using the subtraction routine, which gave a very narrow line width (W-|/2=200 Hz), occurring at a chemical shift of 2.9 ppm (Figure 2.7). This chemical shift is typical for hexacoordinate Al nuclei (which occur from -40 to +20 ppm), 6 2 but is very different from its structural isomer, tris(maltolato)aluminum(lll) which has a W 1/2=900 Hz and has a chemical shift of 39 ppm. 6 3 The lower W1/2 for isomaltol could be attributed to (1) the differences in binding sites, or (2) a fac arrangement of ligands in solution, which has a higher symmetry about the nucleus. The 2 7 AI NMR spectrum of tris(isomaltolato)aluminum(lll) is similar to that of tris(acetylacetonato)aluminum(lll); both have narrow lines near zero ppm 6 4 due to the similarity of their binding sites. 21 I : ' i ' I i i i i | 200 100 " 1 — i — i — i — i — I — I — i — i — i — i — i — i — ! — i — j — i — i — i — i — | — i — i — i — i — j — r -lbc ppm -200 Figure 2.7. 2 7 AI NMR of Al(ima)3 in H2O (after the subtraction routine). 22 2.5. Elemental Analyses Elemental analyses for all complexes were consistent with the proposed tris(isomaltolato)metal(lll) structures. The hygroscopic nature of these complexes required that the samples be dried under vacuum at 85°C for 20 hours before being submitted for analysis. The In and Gd complexes are consistent with elemental analysis as partially hydrated complexes. Table 2.4. Elemental Analyses of M(ima)3 Complexes. C H M C1 8 H1 5 AI0 9 expected % found % 53.74 53.70 3.76 3.73 6.71 6.66 C i 8 H i 5 G a 0 9 expected % found % 48.58 48.31 3.40 3.37 15.67 15.73 'Ci8H15.5lnO9.25 expected % found % 44.11 43.84 3.08 3.39 23.43 23.59 * C i 8 H i 7 G d O i 0 expected % found % 39.27 39.59 3.11 ' 3.00 28.57 28.22 * Analyses consistent as ln(ima)3-1/4H20 and Gd(ima)3-H20. 23 2.6. X-rav Crystallographic Analyses of Al and Ga Complexes of Isomaltol. Both the aluminum and gallium complexes of isomaltol were recrystallized from methanol/diethyl ether, and were studied by X-ray diffraction. The aluminum complex is a monoclinic crystal system, while the gallium complex is triclinic. Fewer reflections were acquired for the Ga(ima)3 complex, so that the bond lengths and other parameters were not as accurately determined as for the Al(ima)3 complex and therefore the Ga(ima)3 complex will not be discussed in detail. Crystallographic data, bond lengths and bond angles for the Al(ima)3 metal complex appears in the Appendix. The structure of the Al(ima)3 complex was solved by Dr. Steven Rettig of the UBC Structural Chemistry Laboratory by direct methods. The positions of the 25 non-hydrogen atoms were determined from an E-map. Both the Al and Ga complexes suffer disorder of two of the three ligands. Due to the extensive derealization of the binding site, the two oxygen atoms appear 'equivalent' and no binding preference is seen in two of the three ligands. The position of all of the atoms in a ligand are the same except one, independent of the way in which the bidentate ligand is coordinated to the metal. This is shown in Figure 2.8, by which superposition of these two possible binding orientations results in only one carbon atom in each ligand occuring in a position that is not already occupied by any other atom. These correspond to C(11') and C(17') in the ORTEP diagram (Figure 2.9). This nonpreferential binding is the cause of the disordered nature of the crystal, and also results in most of the atoms appearing larger than if there was a geometrical binding preference. This also explains why C(11) and C(17) 24 a b Figure 2.8. Superposition of the two possible binding orientations (a and b) in Al(ima)3, results in only two carbons that do not overlap (C(11') andC(17')). See text. are the only atoms that are resolved as having two distinct orientations in the Al(ima)3 complex. Even though the structure is disordered, interesting information can be obtained by examination of the bond lengths of the ordered ligand. The binding site of isomaltol undergoes considerable derealization when coordinated to the aluminum ion. Comparison of Al(ima)3 bond lengths with those in the tris(acetylacetonato)aluminum(lll) 6 5 complex show many similarities. For example, in the ordered ligand, one finds the two carbon-carbon bonds associated with the binding group to be the same within experimental error (C(1)-C(2) = 1.382(6) A, C(2)-C(3) =1.377(6) A). These are also the same as the average C-C distance (1.380(14)) in Al(acac)3. The hydroxyl bond length (C(3)-0(2) = 1.281(5) A) is also equal to the average C - 0 bond length in Al(acac)3, while the carbonyl bond is shorter (C(1)-0(1) = 1.261(5) A). Overall, ima - undergoes almost as much derealization as the acac anion upon coordination to aluminum. 25 c ( i r ) Figure 2.9. ORTEP diagram of Al(ima)3. 2 6 2.7. Ultraviolet spectra of isomaltol The UV spectrum of isomaltol in ethanol has been reported in the literature, showing a broad band at 280 nm (e=16,000) 6 6 but no variable pH studies have been reported. Using a 5.2 x 10 - 5 M solution of isomaltol in 0.15 M NaCl, the initial pH was 5.1 having a Xmax=285 nm (e=14,100) with a shoulder band at a higher wavelength due to K-K* transitions. 6 7 As the pH is raised with 0.1 M NaOH, the shoulder band increases in intensity (>.max=315 nm) at the expense of the 285 nm band, having an isosbestic point at 298 nm. See Figure 2.10. This is indicative of the presence of two absorbing species in solution; presumably the deprotonated and protonated forms of isomaltol. At pH -5.6, both bands have approximately the same intensity, which is in agreement with the potentiometrically determined pKa of isomaltol (pKa= 5.55). 6 8 When the pH is raised to 6.2, only the band at 315 nm (due to the deprotonated form) is present, and another less intense band at 261 nm appears (probably due to an K-K* transition that has a lower molar absorptivity) which was initially masked by the more intense broad band at 285 nm. Further evidence supporting the two species being the protonated and deprotonated forms of isomaltol was the fact that there was a 1:1 ratio between the number of moles of base added and the number of moles of isomaltol present for complete conversion of the 285 nm band to the 315 nm band. This process was also reversible upon addition of acid. These results will be used for comparison when discussing another furan ligand in Chapter 3. 2 7 Figure 2.10. Variable pH UV spectra of isomaltol (0.05 mM) in 0.15 M NaCl. 28 2.8. Potentiometric Studies Since these bidentate ligands form metal complexes that compete with metal hydrolysis, studies were done to determine their formation constants. The studies were done in 0.15 M (isotonic) NaCl. The complexes form consecutive mono, bis, and tris ligand-metal complexes defined by the following equations, with stepwise formation constants K-|, K 2 , and K3, and overall formation constants 61 , R2. and B3.69 M + L J ^ - ML « [ML] 1 [M] [L] ML + L K2 > ML2 B =_!MkL [M] [L]2 ML2 + L K a - ML3 o [ M L 3l [M] [L]3 Figure 2.11. Stepwise and overall formation constants where M is a trivalent metal and L is the deprotonated bidentate ligand. Potentiometric titrations were done to determine the binding affinities of various ligands, as well as a preliminary test of their usefulness for in vivo studies. Table 2.5 lists the overall formation constants for the Group 13 metals with isomaltol. 7 0 The log B3 values for the Group 13 metals follow trends characteristic of their hexaquo acidities, i.e. Ga > In > A l . 7 1 » 3 5 Comparison of Al(ima)3 with the analogous Al(acac)3 complex (log 63= 23 7 2 ) , indicates that isomaltol is a much weaker binder of aluminum, even though their binding sites appear very similar in the solid state. 29 Table 2.5. Logarithms of The Metal-lsomaltol Stability Constants (stepwise= K, overall= B), Determined by Potentiometric Titration at 25 °C. The Standard Deviation Between Successive Runs is Shown in Parentheses. Isomaltol H Al Ga In -log K a 5.55(1) log K i 5.66(5) 6.63(2) 7.08(2) log K2 4.76(5) 5.68(1) 4.06(2) log K3 4.03(5) 4.05(1) 3.66(3) log B2 10.42(1) 12.31(1) 11.14(1) log B 3 14.45(5) 16.36(1) 14.80(2) The low thermodynamic stability of M(ima)3 becomes apparent when the hydrolysis of the metal ion competes with the formation of the complex to a great degree at millimolar concentrations. This is shown in the speciation diagrams of isomaltol and the various Group 13 metals in Figure 2.12 (a) to (c). For all the metal complexes, the ML3 species exists over a small pH range. At neutral pH, metal hydroxide species predominate. However, at higher concentrations (0.05M Al / 0.15M Isomaltol) mimicking synthetic conditions, Hima easily competes with hydrolysis (Figure 2.13). Now the major form of aluminum is the AIL3 species from ~ pH of 4.5 to 8. 30 100 %of total In / ln(OH)3 In(OH)4 100 % of total Ga %of total Al GaL2 /Ga(OH)4 4 5 6 -log [H+] Figure 2.12. Spec iat ion d iagrams of the metal comp lexes of H ima at 1.0 m M M 3 + / 3.0 m M Hima. The dotted lines in the speciat ion d iagrams indicate the pH regions where the aqueous spec ies are supersaturated with respect to the most stable sol id phase. Spec iat ion d iagrams were created using a Macintosh-compat ib le p r o g r a m 7 3 ba sed on the algorithm C O M I C S . 7 4 3 1 1 2 3 4 5 6 7 8 9 -log IH+] Figure 2 13. Speciation diagram of 0.05 M Al /0.15 M Hima. 2.9. Synthetic attempts Attempts to synthesize a bidentate imine ligand by the condensation of isomaltol and methylamine were unsuccessful, resulting in the formation of a pyridinone. The pyridinones are formed by nucleophilic attack of the nitrogen at the C-5 carbon of isomaltol causing the ring to open, followed by dehydration, and ring c losure. 7 5 The imine however, would be formed by nucleophilic attack at the acyl group. o Figure 2.14. Attempts to form an imine resulted in the formation of a methyl pyridinone. A. Loidl and Th. Severin 7 6 reported a pyridinone synthesis resulting from the addition of isomaltol (or maltol) and methyl amine in aqueous acetic acid. Attemps to synthesize the imine of Hima were therefore carried out in benzene, using a Dean-Stark apparatus to remove the benzene/water azeotrope, driving the reaction to completion as shown in Figure 2.14. Unfortunately, the methyl pyridinone (Hdpp) is the thermodynamically favored product, and altering the reaction conditions only decreased the yield of Hdpp to 12%, without evidence for the desired imine. Conclusion Isomaltol forms tris(isomaltolato)metal(lll) complexes in high yields. The crystal structure of Al(ima)3 shows that the binding site of isomaltol undergoes considerable derealization upon coordination, similar to that of Al(acac)3. Preliminary studies with 6 7 Ga(ima)3 intravenously introduced into rabbits, 33 showed that it forms 67Ga-transferrin in vivo.77 Unfortunately, all of the Group 13 metal complexes of isomaltol have low formation constants, such that the ligand will not be considered for further in vivo applications. Hima was not a good ligand for the formation of other hexadentate ligands, due to nucleophilic attack being favored at the C-5 carbon. 34 Chapter 3. 2-Ethoxycarbonyl-3-hydroxy-5-methylfuran (Hehm) and The failure to form imines of isomaltol initiated the search for a ligand that had a bulkier substituent on the C-5 carbon to inhibit ring opening reactions. In the literature an attractive ligand was found, 2-ethoxycarbonyl-3-hydroxy-5-methylfuran (Hehm), which not only had a methyl group on the C-5 carbon, but also had an ester group, allowing for a way to modify the ligand by the formation of various amides. Little work has been done on such highly substituted furans, especially those containing a 6-hydroxyenone moiety, but a short communication by Mukaiyama 3 7 describes the synthesis of furan derivatives by reacting diketene with various ylids in benzene. The resulting furans were characterized by melting point and elemental analyses. Hehm was prepared by the addition of ethyl(dimethylsulfuranylidene) acetate (EDSA) to diketene as follows; 2-(N-methylamldo)-3-hydroxy-5-methylfuran (Hahm) 3A. Hehm O EDSA Hehm Figure 3.1. Hehm synthesis. 35 Hehm was thought of as a possible chelator for the Group 13 metals, but was primarily of interest as a template for the formation of amides with the aim of synthesizing a hexadentate amide ligand. As an initial foray into this area, the methyl amide of Hehm, Hahm, as well as its tris(ahm)metal(lll) complexes were synthesized and studied. 36 Experimental Starting materials Methylsulfide (Aldrich), ethyl bromoacetate (Aldrich), and methylamine (40% in h^O-Fisher) were used as supplied. Diketene (Sigma) was distilled under reduced pressure prior to use. The AI(N03)3-9H20 (Mallinckrodt) was used as supplied. Ga ingots (Alfa) were dissolved in concentrated hydrochloric acid and standardized. Synthesis of Carbethoxvmethvl Dimethvlsulfonium Bromide. This synthesis was performed according to Payne 7 8 . Synthesis of Ethvl (Dimethvlsulfuranvlidene)acetate (EDSA) The preparation of EDSA was carried out according to P a y n e 7 8 . T h e carbethoxymethyl dimethylsulfonium bromide solution was stirred vigorously using an overhead stirrer, and was cooled for 15 minutes with an ice bath. After addition of the saturated K2CO3 and 12.5 M NaOH solution, the ice bath was removed and the mixture was left to stir for an additional 25 minutes. EDSA was stored at 5°C under nitrogen. Synthesis of 2-Ethoxycarbonvl-3-Hydroxy-5-Methylfuran (Hehm). EDSA (12 g, 75 mmol) was stirred in 40 ml of benzene. To this, freshly distilled diketene (4.8 g, 57 mmol) was added. The mixture was left to reflux for 15 hours. Most of the benzene was removed from the solution, cooled and a precipitate resulted. The precipitate was recrystallized from methanol to give 37 6.9 g (75%) of an off-white crystalline product. It was soluble in methanol, ethanol, chloroform, ether, ethyl acetate, benzene and water. Attempted Synthesis of AKehmV^. In 60 ml of H2O and 40 ml of ethanol, 1.36 g (8 mmol) of Hehm was dissolved by raising the pH to ~ 9.5 with 2 M NaOH. To this basic solution, 0.65 g (2.67 mmol) of AICI36H2O were added. The pH was adjusted to 7.3 and an off-white precipitate weighing 0.46 g resulted (35%). It was soluble in methanol, ethanol, acetone, ethyl acetate, and slightly soluble in toluene. Attempted Synthesis of GafehmV*. Using 80 ml of a 60:40 H 2 0 : E t O H mixture, 1.45 g (8.5 mmol) of Hehm was added and the pH was raised to 9 to achieve complete dissolution. To this 6.40 ml (2.8 mmol) of 0.445 M GaC-13 solution in excess HCl was added. The pH was adjusted to 6.4 with 2 M NaOH to give 0.68 g of a pale yellow powder (42%). It was soluble in methanol, ethanol, and chloroform. Synthesis of 2-(N-methvlamide)-3-hvdroxv-5-methvlfuran (Hahm). Using a 500 ml round bottom flask, 4.67 g (27 mmol) of Hehm was dissolved in 110 ml of ethanol. To this, 2.12 g (27 mmol) of methylamine (40% in H2O ) was added. The solution was left to reflux for 19 hours. The volume was then reduced to give 3.3 g (77%) of an off-white powder. The crude Hahm was recrystallized from ethanol. It was found to be soluble in methanol, ethanol, chloroform, water, and benzene. Synthesis of AI(ahn%3H£0. Hahm (0.475 g, 3.1 mmol) was dissolved in 40 ml of H 2 0 by raising the pH to 9 with 2 M NaOH. AICI 3 -6H 2 0 (0.246 g, 1.0 38 mmol) was then added and the pH adjusted to 7.3. A white precipitate formed weighing 0.36 g (72%). It was then recrystallized from methanol. Al(ahm)3 was soluble in methanol, acetone, and water (3.13 mM). Synthesis of GafahmVj. Hahm (0.528 g, 3.4 mmol) was dissolved in 60 ml of H 2 0 by adjusting to pH 9. To this, 2.5 ml (1.1 mmol) of 0.445 M GaCl3 solution in excess HCl was added. Adjustment of the pH to 7.0 with 2 M NaOH resulted in a grayish-white powder weighing 0.45 g (75%). Crude Ga(ahm)3 was recrystallized from pyridine. It was soluble in pyridine, methanol, water (1.55 mmol), and slightly soluble in acetone. 39 Results and Discussion : Hehm The yield of Hehm was improved by 15%, by using a 1.5 to 1.0 ratio of EDSA to diketene rather than a one to one ratio, as inferred by Mukaiyama, and refluxing for 6 hours instead of one hour. However, any impurities in the starting materials or solvent (benzene) resulted in the synthesis of a red oil, shown by thin layer chromatography to consist of at least eight products. Hexanes extraction of the oil gave rise to Hehm and another product (discussed later on p. 45) that was separable from Hehm by fractional crystallization. For this reason, both the benzene and the diketene were distilled before use, and the EDSA was checked for impurities by 1 H NMR. Using pure starting materials resulted in a 75% yield of Hehm, which was then recrystallized from ethanol. The short communication by Muka iyama 3 7 characterizes Hehm by elemental analyses and melting point only. Other characterization methods will be discussed herein. 3.1a. Infrared Spectrum The IR spectrum of Hehm shows a strong OH stretch at 3240 cm-"1 indicative of a hydrogen bonded, rather than a free, OH group. Hehm also has two strong bands occurring at 1672 and 1630 cm* 1 corresponding to the C=0 and C=C stretches respectively. 7 9 The C=0 stretch occurs at lower frequency than saturated esters due to extensive conjugation and hydrogen bonding analogous to that seen with isomaltol. There are three bands in the C -0 stretch region at 1260, 1181, and 1082 cm- 1 , but it is not possible to assign any one 40 band due to ring vibrations and C-H bending modes occurring in the same region. 3.23.1HNMR Spectrum The 1 H NMR of Hehm in CDCI3 shows the presence of an ethyl group linkage having chemical shifts diagnostic of esters. The long range coupling between the C-5 methyl (labelled e in figure 3.2) and the C-4 ring proton (labelled d in figure 3.2) is further evidence supporting the Hehm structure. The C-5 methyl group splits the ring proton info a quartet with a coupling constant of 0.94 Hz, which is so small that it appears as a doublet unless it is expanded. A broad OH peak is also seen further downfield in CDCI3. The proton NMR spectrum of Hehm is shown in Figure 3.2. 3.3a. 13C NMR Another good NMR probe used for the structural determination of many organic compounds is the 1 3 C nucleus. Due to its low natural abundance (1.1%) and lower sensitivity relative to protons, much higher sample concentrations and/or longer aquisition times are required. Coupled and decoupled 1 3 C spectra can be obtained, but another very powerful method for differentiating between CH3, CH2, CH, and quaternary carbons is called the attached proton test (APT). This involves the use of the 1 3 C pulse sequence; 90x°-x-180x°-x, having a broad band proton decoupler on at all times except during the first delay period, x . 8 0 This experiment has also been called 'J modulation' because after the 90 x° pulse is applied, the position O — C H 2 b C H 3 a 1 >~i J ' 1 — i i i j i 6 5 4 3 2 Figure 3.2. 80 MHz ^H NMR spectrum of Hehm in CDCI 3 42 of the net magnetization for each type of carbon is dependent on JCH (which defines x). When x equals the C H 2 and quaternary carbons have a net magnetization along the -y'-axis while the C H 3 and CH carbons have a net magnetization along the y'-axis. The receiver coil detects the sum of the y' components, and during the spectrum editing, the quaternary carbons (which undergo no JCH coupling) are given a positive value. This results in easily distinguishable positive 1 3 C lines for C H 2 , and quaternary carbons, and negative 1 3 C lines for C H 3 , and CH carbons. The decoupled 1 3 C spectrum of Hehm was obtained in CDCI3, displaying eight 1 3 C signals. The APT spectrum showed four quaternary carbons, one methine, one methylene, and two methyl carbons; in agreement with the proton NMR spectrum. The chemical shifts of the carbon atoms were assigned by comparing the 1 3 C spectrum of Hehm (Figure 3.3) with 1 3 C spectra of isomaltol and other literature compounds. 155.9 156.2 59.7 / 13.6 CH2 "SCH3 13.5 Figure 3.3. The chemical shift assignments for the 1 3 C spectrum of Hehm obtained in C D C I 3 (ppm). 43 3.4a. Mass Spectrum The mass spectrum of Hehm was obtained by electron impact ionization. The spectrum of Hehm shows a parent peak at m/z=170, followed by the loss of ethanol to give a peak at m/z=142. The base peak at m/z=124 is due to a fragmentation typical of B-hydroxy esters by means of the M cLafferty rearrangement81 as follows: O m/z = 170 m/z = 124 Figure 3.4. Fragmentation pattern of Hehm. Lower m/z fragmentations are simply due to the sucessive loss of CO, giving rise to the peaks at m/z=98 and 68 respectively. Elemental analyses of Hehm were also consistent with the proposed C8H10O4 Hehm structure. 3,5a, Attempts to Prepare Complexes of Hehm Metal complexes of Hehm were prepared by a 3:1 stoichiometric addition of ligand to metal. Due to the low water solubility of Hehm, these complexes were prepared in a 60:40 water:ethanol solution. The pH was initially raised to 9, and upon slow addition of the metal salt, the pH dropped to - 2. The pH was then raised again to 6.5 with 2 M NaOH resulting in an off-white precipitate. From the NMR, IR and MS characterization methods, Hehm shows evidence of binding with Al and Ga, but isolation of pure tris(ehm)metal 44 complexes was not possible. Hehm reversibly hydrolyzes under the acidic synthetic conditions, such that as the pH was increased from 2 to 7, only low yields of the complexes were isolated. The isolated product also contained ~ 20% Hehm ligand. Any attempts to purify the complexes resulted in reversion of the complex to its starting materials, as monitored by 1 H NMR as the amount of Hehm present increases. The IR of the metal complexes of Hehm show the expected shifts in the carbonyl band, but there is also a shoulder peak at 1670 cm- 1 as seen in the Hehm spectrum. The 1 H NMR spectra of the metal complexes of Hehm show peaks pertaining to the ligand itself, as well as the peaks that would be expected by the formation of the tris(ehm)metal(lll) complexes. The presence of the ligand peaks in the proton NMR is in agreement with the shoulder bands due to the ligand that were observed in the IR spectra of the metal complexes. The DCI (negative ion detection) mass spectrum of the gallium complex of Hehm shows two parent ion peaks corresponding to [ML3-H]- at m/z= 575 and 577 due to the 6 9 G a and 7 1 Ga isotopes. The base peaks at m/z= 547 and 549 correspond to [ML3-CH2CH3]- (for both gallium isotopes), showing that there may also be the formation of a gallium complex from a hydrolyzed Hehm ligand. 45 3.6a. Hehbm If the reaction of EDSA and diketene is carried out using an excess of diketene, rather than an excess of EDSA, the side product that was formed in the synthesis of Hehm, now becomes the predominant product. This product was characterized by IR, NMR, MS, and elemental analyses and is consistent for a compound having the formula C12H14O6. If two equivalents of diketene react instead of one, the extra diketene unit could be attached to the C-4 ring carbon as shown in Figure 3.5. This compound, 2-ethoxycarbonyl-3-hydroxy-4-(1',3*-butanedione)-5-methylfuran (Hehbm) has not been reported in the literature. The mass spectrum of Hehbm showed a parent peak at m/z=254, followed by loss of a methyl group to give a weak peak at m/z=239. Removal of the CH3-CO group from the C-4 side chain is seen by a strong peak at m/z=43; the loss of both of these groups from the parent mass results in a medium intensity peak at m/z=197. Another major peak occurs at m/z=151, due to the loss of ethanol from the m/z=197 peak. The base peak at m/z=124 is may due to the loss of CO from the m/z=151 fragment peak. A possible fragmentation sequence is shown in Figure 3.6. o o O H Figure 3.5. Hehbm 46 O m/z = 124 Figure 3.6. Major fragmentation peaks seen in the mass spectrum of Hehbm. The proton NMR was obtained in deuterated chloroform, consisting of six peaks at 1.43 (t,3H), 2.19 (s,3H), 2.67 (s,3H), 4.41 (q,2H), 6.23 (s,1H), and 8.45 (s,1H) ppm. By comparison of the proton NMR of Hehm with Hehbm, one can assign the C-5 methyl protons as occurring at 2.19 ppm, while the methyl ketone protons resonate at 2.67 ppm. The methine proton on the C-4 side chain occurs slightly further downfield compared to the ring proton on Hehm. There is also no long range coupling between the methine proton and the C-5 methyl protons as seen with Hehm, thus providing evidence that the methine proton is not directly attached to the C-4 ring carbon. Only one peak due to the OH proton on the C-4 side chain is seen at 8.45 ppm, suggesting that strong hydrogen bonding is occurring in the enol form in solution. 47 Further evidence to support this structure is the 1 3 C NMR obtained in CDCI3. An attached proton test (APT) showed twelve carbons present, three of which are methyl, one methylene, one methine, and the other six carbons are quaternary. The extra diketene unit attached to the C-4 ring carbon shows 1 3 C NMR chemical shifts of 179.4, 99.1, 191.4, and 24.7 ppm relative to TMS, analogous to a similar B-diketone compound. 8 2 The 1 3 C spectrum and its assignments are shown in the Appendix. The positions of the various ring substituents were confirmed by using the NMR technique SINEPT (selective insensitive nuclei enhanced by polarization transfer). A SINEPT pulse sequence is shown in Figure 3.7. 8 3 1 H 9 0 ° x - x - 1 8 0 ° y - x - 9 0 ° y - A /2-180° x-N 2 - decouple 1 3 C 180°x - x - 90° x - A/2 -180° x - Nz - acquire(x') Figure 3.7. SINEPT pulse sequence, where x is the delay time between pulses (x = V4J), and A is the delay time for the refocussing phase which is dependent on the spin multiplicity, e.g. doublets require A equal to1/2j for optimal refocussing. Application of a SINEPT pulse sequence results in enhancement of the 1 3 C signals by polarization transfer. Polarization transfer can only occur between nuclei that are either scalar or dipolar coupled to each other. Since the magnetogyric ratio (and therefore the Boltzmann distribution of populations) for protons is approximately four times that for carbon, the proton transitions are four times as energetic. If one considers a 1 H - 1 3 C coupled system, irradiation 48 of one 1 H line (by a 9 0 ° y pulse) causes a selective inversion of populat ion 8 4 (Figure 3.8). +3 Ho H , +3 90° H , H , -5 (a) +T (b) Figure 3.8. (a) A 1 H - 1 3 C coupled system and (b) the population inversion of one 1 H (H2) line. The relative Boltzmann populations are represented numerically in Figure 3.8, such that the difference in the populations is proportional to the signal intensity (eg. +5 - (-3) = +8 for the H i proton transition, whereas the C i transition is +2). After the selective population inversion, the proton polarization is transferred to the carbon nuclei such that now the population differences of the C i and C2 transitions are +10 and -6 respectively. 8 5 This results in one enhanced positive line and one negative line, which is then refocussed and usually decoupled to give one enhanced positive 1 3 C signal. In the SINEPT experiment, soft pulses are applied to a preselected proton resonance so that only the 1 3 C nuclei coupled to those irradiated protons will show enhanced 1 3 C signals. As with APT, the pulse sequence used involves delay times, x and A , that are dependent on J C H - For optimal polarization transfer, x + 2ce = A + ce =1/2j, where ce is the duration of the 9 0 ° y 49 proton pulse (~ 10 msec). 8 3 The 1 3 C spectra from the SINEPT experiments are shown in Figure 3.9. Irradiation of the methine proton on the C-4 side chain resulted in the enhancement of 1 3 C signals at 99.3 ppm (one bond coupling), 191.6 and 179.6 ppm (two bond coupling), and at 111.3 ppm (three bond coupling). No enhancement was seen at the methyl carbon on the C-4 side chain. These results can be explained by a closer examination of the NMR experiment. The pulse sequence requires a known 3 J C H value for three bond coupling, or an approximation if not known, such that enhancement of the 1 3 C signal will be selective to coupled carbons that are three bonds away. However, some enhancement breakthrough can occur at carbons that are coupled through only one or two bonds if the coupling constants through one and/or two bonds are multiples of the chosen 3 J C H value. Therefore, the 1 3 C enhancements at 99.3, 191.6, and 179.6 ppm are all due to breakthrough. The lack of enhancement expected at the methyl carbon on the C-4 side chain and at the C-4 ring carbon could be due to many factors. A common factor that is often seen when trying to enhance carbon signals through three bonds, is that the time involved in the refocussing sequence is longer than the proton relaxation time. This results in the protons relaxing to their normal Boltzmann distibution of populations before the 1 3 C spectrum is acquired. However, the enhancement breakthroughs as well as the expected enhancement, confirms the presence of the predicted side chain, which is bonded to the C-4 ring carbon. Irradiation of the C-5 methyl resulted in enhancement of the carbon signals at 160.9, 15.5 ppm, as well as 191.6, 99.3, and 24.9 ppm. The 160.9 and 15.5 ppm enhancements are due to one and two bond coupling respectively. No three bond coupling was observed. The other three S I N E P I E X P E R I M E N T o ( I F R 6 2 2 P P M ) • « » « i » M i i i p i i i H ' O N i M i W»I«>I'»IUIII<<>IH>»>»^II>»»H<H' i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | ' i i i i | i i i i | i i i i | i i i i | i ' i i | ' 1 1 1 | 1 1 1 1 ' 1 1 1 1 I 1 1 1 ' 1 ' 1 ' 1 I ' 2 0 0 1 8 0 1 6 0 1 4 0 1 2 0 1 0 P 8 0 6 0 4 0 2 0 P P M 0 ( I F A T 1 5 P P M ) . , . , 1 , ' , I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I I I I I I I I I I I I I J I I I I I I I I I I I I I I I I I I ' I I I I I I I I 'I I I I 'I LL 2J00 1 8 0 1 6 0 1 4 0 120 100 8 0 6 0 4 0 ' | ' I ' " I 2 0 P P M Figure 3.9. The 1 3 C SINEPT spectra of Hehbm due to (a) irradiation of the methine proton and (b) irradiation of the C-5 methyl proton. The decoupled 1 3 C spectrum of Hehbm is shown on page 85. All spectrum were obtained in CUCI3. 51 enhancements are actually due to the irradiation of the methyl group on the C-4 side chain, also showing enhancements due to one (24.9 ppm), two (191.6 ppm), and three (99.3 ppm) bond coupling. Simultaneous irradiation of both methyl groups was seen because the irradiation was "not selective enough to distinguish between the two methyl groups. This SINEPT experiment shows that the C-5. methyl is bonded to a carbon having a chemical shift of 160.9 ppm, and therefore must be bonded to the carbon next to the ring oxygen. This experiment also reinforces the structure of the C-4 side chain. Due to the presence of two metal binding sites on Hehbm, and the lower yield, further use of this ligand was not pursued. 52 3B. Results and Discussion: Hahm Hahm (2-(N-methylamido)-3-hydroxy-5-methylfuran) is a new bidentate amide ligand that has not been previously reported in the literature. Hahm was prepared by refluxing Hehm and methylamine (40% in H2O) in ethanol for 20 hours using a 1:1 stoichiometric ratio (77% yield). Reactions of isomaltol with methylamine resulted in pyridinone insertion products. However, the fact that Hehm contains a bulkier methyl group attached to the C-5 carbon on Hehm rather than a proton, favored the formation of an amide, and no pyridinones were detected. Synthesis of the metal complexes of Hahm were carried out in water, but required that the pH be raised to 9 initially due to the lower water solubility of the protonated form of Hahm. The nitrate salt of the desired metal was then added to Hahm solution, giving a 3:1 Hahm to metal ratio. The pH was then raised again to ~7. The aluminum and gallium complexes were obtained in 72 and 75% yields respectively. Hahm and its metal complexes were characterized by elemental analyses, IR, 1 H NMR, MS, and UV spectroscopy. H s Figure 3.10. Hahm synthesis in ethanol. 53 Secondary amides are known to form intermolecular hydrogen bonds between the carbonyl oxygen and the proton on the nitrogen, existing as cis dimers or trans arrangements. 8 6- 8 7 However, W. Geiger 8 8 and D. Welti 8 9 have done studies on salicylanilides (which have a hydroxyl group beta to a secondary amide carbonyl), and found that the formation of the conjugate six membered ring via intramolecular hydrogen bonding, is favored (as shown in Figure 3.10). Hydrogen bonding between the hydroxyl oxygen and the proton on the amide nitrogen would also involve the formation of a six membered ring, but this form was not favored. 3.1b. Infrared Spectra The infrared spectrum of Hahm, in the solid form, shows a sharp band occurring at 3390 cm* 1 , indicative of a free N-H stretch. 9 0 This supports the formation of a conjugate chelate structure, agreeing with a similar salicylanilide studied by D. Welt i . 8 9 The O-H stretching vibration results, due to hydrogen bonding, in multiple bands which appear as a broad band centered around 3100 cm- 1 . Due to solubility limitations in non-protic solvents, solution studies to investigate the hydrogen bonding in Hahm were not possible. Amides have several diagnostic bands in their infrared spectra. 9 1 A band characteristic of all amides is a strongly absorbing band due to the C=0 stretching vibration (amide I band) which occurs at 1662 cm - 1 for Hahm. This band absorbs at a lower wavenumber than normal ketones due to the resonance effect with the ionic form. See Figure 3.11. 54 p H O - H II I | | C—N—R C=N—R + Figure 3.11. Amide resonance forms. There is considerable mixing of the C=0 and the C=C vibrations such that the bands at 1662, and 1610 cm* 1 are probably combination bands, but being primarily due to vC=0 and vC=C vibrations respectively. A broad band at 1565 cm- 1 (amide II band) is a combination band primarily due to a N-H bending deformation, but is also coupled to a C = N stretching vibration. The C-N stretching (amide III band) region occurs between 1305 and 1220 cm- 1, but due to the presence of many other bands in this region, a definite assignment was not possible. Upon coordination to the metal, the sharp band due to the N-H stretch for the ligand, broadened, and a weak and broad band centered at 3400 cm- 1 appeared. This could be due to some hydrogen bonding occurring between the proton on the nitrogen and the ring oxygen. The band due to the carbonyl stretch shifts from 1662 cm- 1 for Hahm to 1639 and 1634 cm- 1 for the aluminum and gallium complexes respectively. See Figure 3.12. In both the aluminum and gallium spectra, there is one very strong band at 1530 cm- 1 , with weaker shoulder bands at higher wavenumbers. It appears that the two bands at 1610 and 1565 cm- 1 seen in the infrared spectrum of Hahm have shifted into one overlapped band at 1530 cm- 1 . By comparison of the Hahm spectrum with the spectra of the metal complexes, tentative assignments were made for the bands due to M-0 stretches at 420 and 400 cm- 1 for the aluminum and gallium complexes respectively. Since these bands are of low intensity and no isotope experiments were carried out, these are only tentative assignments. T T 3 3 0 0 3 0 0 0 2 7 0 0 cm"' l g 0 0 1 16 00 — r 14 0 0 Isn7fnaor?do^?ra ° i f ( a ) H a h m ' { b ) A l ( a h m ) 3 a n d (c) Ga(ahm)3 in the 3700 to 2600 cm-1, a n d 1 8 0 0 t 0 1 4 0 0 c m . , r e g J Q n ^ 56 3.2b. IH NMR Spectra The proton NMR spectrum of Hahm was easily obtained in D2O, d 4 -MeOH, and d 6 -DMSO. In all solvents, the spectra are consistent with the proposed structure. In d 4-MeOH three peaks are seen at 2.27 (d, 3H), 3.47 (s, 3H), and 5.94 (q, 1H) ppm. Small long range coupling (J=0.93 Hz) between the C-4 ring proton and the C-5 methyl group is also observed. The aluminum complex of Hahm was also obtained in d 4 - M e O H producing a spectrum with three peaks at 2.28 (s, 3H), 3.56 (s, 3H), and 6.11 (s, 1H) ppm showing that the environment about the C-5 methyl is not affected upon chelation to the metal. The coupling constant between the ring proton and C-5 methyl is small for the ligand, and unresolvable for the metal complexes. The ring proton and the amide methyl, on the other hand, undergo shifts of 0.07 and 0.09 ppm respectively, since they are closer to the binding site. The Ga(ahm)3 complex however, decomposed in methanol, showing a broad shoulder or additional peaks due to the free ligand. The spectrum was therefore obtained in d5-pyridine, which displayed three peaks at 1.86 (s,3H), 3.03 (s,3H), and 6.38 (s,1H) ppm, consistent with the proposed structure. In an effort to observe the proton on the amide nitrogen, d 6 -DMSO was used, but decomposition occurred in both the aluminum and gallium complexes and they were less soluble in DMSO than in pyridine or methanol. 3.3b. Elemental Analyses The carbon, hydrogen, and nitrogen analyses are all in agreement with the proposed formulas, except that the aluminum complex analyzed as a trihydrate. See Table 3.1. 57 Table 3.1.Elemental Analyses for Hahm and its Aluminum and Gallium Complexes. compound C H N G7H9NO3 expected % found % 54.20 53.94 5.85 5.82 9.03 8.79 * C 2 i H 3 3 A I N 3 0 i 2 expected % found % 46.16 45.94 5.53 5.27 7.72 7.66 C2 iH27GaN 3 0 9 expected % found % 47.39 47.15 4.55 4.70 7.93 7.99 * Analysis consistent as AI(C7Hg03N)3-3H20. 3.4b. Mass Spectra The mass spectrum of Hahm was obtained using electron impact ionization. The base peak was the parent peak at m/z = 155. Loss of a methyl group results in a peak at m/z = 140, while loss of the N - C H 3 portion of the ligand gives rise to.the peak at m/z = 126. The less volatile metal complexes decomposed using electron impact ionization methods, and therefore were obtained by fast atom bombardment ionization. Peaks due to H 2 M L 3 + , M L 2 + , and ML+ were present for both complexes as shown in Table 3.2. For the aluminum complex, the base peak 58 was due to the Ml_2+ fragment, and a weak M2L-5+ peak was observed at m/z=824, typical of neutral chelate complexes. 6 3 The mass spectrum of the gallium complex showed many peaks due to the matrix (thioglycerol), but the most intense metal-ligand fragment was due to the Ml_2+ species. Peaks due to HM!_3+ occurring at m/z=532 and 534, were slightly more intense than the peaks corresponding to the H2MI_3 +, and have appropriate isotopic ratios. Table 3.2. FAB Mass Spectral Data for the Metal Complexes of Hahm. m/z Al Ga H 2 ML 3 + 489 533,535 M L 2 + 335 377,379 ML+ 181 223,225 3.5b, Ultraviolet Spectra In the solid state, Hahm is a stable off-white powder. However, in H2O, Hahm undergoes decomposition which is time-, concentration-, and pH-dependent. Unusual aqueous solution behavior was first observed when 59 i 60 carrying out potentiometric titrations. The solution behavior of Hahm was then monitored by ultraviolet spectroscopy. The UV spectrum of a -0.1 mM solution of Hahm (in 0.15 M NaCl) taken immediately after dissolution (pH = 5.8) was found to have a ^-max=290 nm with a broad shoulder at -330 nm. After 24 hours, the X. m a x=323 nm. UV spectra of Hahm were also taken at various pH's (Figure 3.14). These spectra show the same conversion from the band at 290 nm to the band at 323 nm as the pH is increased, but was observed over a shorter time period. Under certain conditions, the decomposition can be greatly retarded. At high concentrations (~75mM), the solution is quite acidic (pH - 3) and no decomposition was observed after three months. However, the addition of base accelerates decomposition, as monitored by UV as the conversion of the 290 nm band to the 323 nm band. Thin layer chromatography also shows the presence of at least five products for a solution that has a UV spectrum consisting of only the 323 nm band. A more dilute solution of Hahm (2.0 mM) can also withstand decomposition if the solution is acidified to pH -2 .2 before titrating with base. The UV spectra at different pH's now shows a single absorption band at 290 nm until pH 7.0. A shoulder band then appears at 330 nm that increases in intensity until it is - 30% that of the 290 nm band at pH 11.5. It appears that the 330 and 323 nm bands are due to undesirable species, and can be inhibited by acidification. The deprotonation of the hydroxyl group of Hahm should be a reversible process and should be seen in the UV spectra as it was in the isomaltol case. However, if the 330 and 323 nm bands are due to decomposition species, then the broad 290 nm band must be due to both Hahm and deprotonated Hahm. This is explained as a result of deprotonation of the hydroxyl group on Hahm 61 enabling a strong interaction between the proton on the amide nitrogen and the now negatively charged oxygen.89 The expected bathochromic shift due to deprotonation, is counteracted by the strong hydrogen bonding interaction (shown in Figure 3.14), which causes a hypsochromic shift and results in the UV spectrum showing no apparent shift in wavelength. H (a) (b) Figure 3.14. Hydrogen bonding in Hahm, (a) before and (b) after deprotonation of the hydroxyl group. 3.6b. Potentiometric Studies Potentiometric titrations were done to determine the pKas of Hahm and the formation constants of its aluminum and gallium metal complexes. The titrations were done at ~2.0 mM concentrations and the solution of Hahm was acidified prior to titrating with base, stabilizing the ligand over the time span of the titration. For the determination of the formation constants, titrations are done from pH 2 to 4 because there are no interfering decomposition products in this range. The pKa and the logarithm of the overall formation constants (log G3) for the aluminum and gallium complexes are 7.15(1), 19.5(3) and 20.5(3), respectively. Determination of the pKa of Hahm actually showed two inflection points, one corresponding to the deprotonation of the hydroxyl proton 62 (pK=7.15), and one that could be due to the deprotonation of the amide nitrogen (pK=12.4). However, the UV spectrum of Hahm at pH 11.5 showed - 30% of some other interfering species in solution, therefore whether this value corresponds to either or both species is uncertain. The formation constants listed are slightly higher than those for M(ima)3, but the lack of stability in aqueous solution makes this ligand less desirable. This is the reason the stability constants are reported to only one decimal place. 100-%of total Al %of total Ga 40 20 0 100 80 60 40 20 0 \  M / A1L3 \ / AIL A1L2 / \ ' \ ' \ / \ / V Al(OH>4/\ \ Al(OH)3/ \ • / / , \ Ga / G a L 2 \ / Ga(OH)4 \ GaL3 / \ GaL j \ Ga(OH) \ / / \ A / \ / \ Ca(OH)3 1 2 3 4 5 6 7 8 9 -log [H+] Figure 3.15. The speciation diagrams for the metal complexes of Hahm at 1.0 mM M+3/ 3.0 mM Hahm in 0.15 M NaCl. The corresponding speciation diagrams for the aluminum and gallium complexes of Hahm are shown in Figure 3.15. Although the log 63 values are higher than those of isomaltol, at millimolar concentration, the predominent 63 species was still due to undesirable hydrolysis, at pH greater than 7 for gallium and at pH greater than 8.5 for aluminum. The onset of hydrolysis was delayed until a higher pH compared to isomaltol, but still occurred to a significant extent. At higher concentrations of Hahm, the ligand easily competes with hydrolysis, such that the M L 3 species predominates. Conclusion The synthesis of Hehm can be carried out cleanly and in high yield if the starting materials are pure and an excess of EDSA is used. A close analogue of Hehm, Hehbm, was also formed if a 1:1 ratio of EDSA : diketene was used. Hahm was produced in high yield from Hehm and methylamine. Aluminum and gallium complexes of Hahm have been prepared, having higher formation constants than the isomaltol complexes. Unfortunately, the ligand as well as the complexes have undesirable solution behavior. 64 Chapter 4. Pyromeconic Acid (Hpa) Pyromeconic acid, 3-hydroxy-4H-pyran-4-one (Hpa), is a close structural analogue of maltol, with the omission of a methyl group at the C-2 carbon. Since the methyl group on maltol was adjacent to the binding site, the formation constants of the metal complexes of pyromeconic acid were thought to be greatly affected, and therefore this ligand was of interest. Hpa is a naturally occurring compound, and is also one of several products of pyrolysis of cigarette-paper made of flax pulp and wood pulp (9:1 ) 9 2 . Pyromeconic acid is known to form a tris(pyromeconato)iron(lll) complex 9 3, but no complexes of the Group 13 metals are known. Q O Figure. 4.1. Pyromeconic acid (Hpa) 65 Experimental Starting materials Pyromeconic acid (a gift from Professor J . H. Looker of the University of Nebraska) was used as supplied. The same metal salts were used as in Chapter two. Synthesis of tris(pyromeconato)aluminumflll). Alfpata To pyromeconic acid (1.00 g, 8.9 mmol) in 35 ml of H2O, AI(NC>3)3-9H20 (1.10 g, 2.9 mmol) was added. The pH was then adjusted to -5.2 with 2 M sodium hydroxide resulting in a white precipitate. Recrystallization from hot water gave 0.941 g (84%) of product. It was soluble in water (2.8 mM), and insoluble in other solvents. Synthesis of trisfovromeconatotaalliumnih hemihvdrate. GafpaWOShbO. To pyromeconic acid (0.691 g, 6.16 mmol) in 25 ml of water, 1.37 M GaCl3 (1.38 ml, 1.89 mmol) was added. A beige precipitate resulted upon addition of 2 M sodium hydroxide, having a final pH of - 5.2. Recrystallization from chloroform yielded 0.724 g (95%) of product. It was found to be soluble in chloroform, slightly soluble in water (< 1mM), and insoluble in other solvents. 66 Results and Discussion Pyromeconic acid forms tris(pyromeconato)metal(lll) complexes with aluminum and gallium in high yields of 84 and 95% respectively. The synthesis was carried out in H 2 O using a 3:1 ligand to metal ratio, 6 3 and adjusting the pH to 7 as shown in Figure 4.2. The metal complexes were characterized by NMR, IR, UV, MS, and elemental analyses. + M" +3 pH~7 Figure 4.2. The synthesis of the metal complexes of pyromeconic acid. Where M= Al3+, or G a 3 + . 4.1 Infrared Spectra Pyromeconic acid undergoes intramolecular hydrogen bonding as seen in the infrared spectrum by the broad band at 3140 cm* 1 due to the O-H stretch. Pyromeconic acid and its metal complexes show a four band pattern indicative of ct-pyrones 9 4 in the 1655-1455 cm.'1 region. Overlap of the C=0 and C=C stretches is observed due to the extended conjugation of the carbonyl causing a reduction of the absorption frequency relative to unconjugated ketones. Upon complexation, the band due to the carbonyl stretch shifts to a lower wavenumber. See Table 4.1 for assignments. 67 There is considerable mixing of the C=0 and the C=C stretching vibrations, so the bands are listed in the table are according to the predominant vibrations. The M-0 stretching vibrations were assigned by comparison of the bands present in the spectra of the complexes relative to bands in the spectrum of the ligand itself in the M-0 vibration region. Table 4.1. Infrared Absorptions of Hpa and The Al and Ga Complexes of Hpa. Hpa Al(pa) 3 Ga(pa) 3 assignment13 3140 vO-H 3110(m) 3080(m) 3080(m) vC-H(ring) 1655 1560 1550 vC=0 1620 1608 1601 1565 1520 1510 vC=C 1457 1465 1455 1107 1123 1123 oC-Hc 1001 1010 1010 o C - H d 455 - vM-0 • 433(m) aAII strong (except where labelled m=medium, and w=weak) intensity, ^v ibra t ion; o=bending deformation. c ln plane. dOut of plane. 68 4.2. NMR Spectra The proton NMR spectrum of pyromeconic acid shows three signals, with the two signals due to the protons nearest the ring oxygen occurring farthest down field. Upon formation of metal complexes, the protons nearest the binding site undergo the largest chemical shifts, as seen in Table 4.2. Table 4 .2 . 1 H NMR Data 3 for Pyromeconic Acid and its Metal Complexes (ppm). H a(s,1H) H b (d, 1H) H c (d,1H) Jbc H 8.14 6.58 8.09 5.6 Al 8.50 7.00 8.26 5.3 Ga 8.00 6.70 7.90 5.3 a In D2O, except the gallium complex was obtained in CDCI3. 69 The 2 7 AI NMR spectrum of Al(pa)3 was obtained in H 2 O , and shows a fairly narrow signal (Wi/2=410) at 39 ppm. The chemical shift is typical of a-pyrones, but the signal is narrower than usual (e.g. maltol has a W*i/2=900). This is probably due to the lack of the methyl group on the ring, resulting in a more rapid molecular tumbling rate in solution, and thus a narrower line width. 4.3 Elemental Analyses The elemental analyses are consistent with the proposed structures. Tris(pyromeconato)gallium(lll) analyzed as the hemihydrate. See Table 4.3. Table 4.3. Elemental Analyses for The Metal Complexes of Pyromeconic Acid. C H M C 1 5 H 9 A I 0 9 expected % 50.02 2.52 7.49 found % 49.84 2.50 7.38 *C i 5 H 1 0 GaO9 . 5 expected % found % 43.73 43.49 2.45 2.26 16.92 17.11 •Analysis consistent for Ga(pa) 3 0 .5H2O. 70 4.4. E.I. MS Mass spectra of the aluminum and gallium complexes of Hpa were obtained by electron impact ionization. The base peak was the ML-2+ species. Fragmentations due to the tris and mono ligand complexes were also observed as well as the ligand fragmentation cations as shown in Figure 4.2. Other ligand fragment peaks occur at m/z=1l2 due to HL+, C4H402+ at m/z= 84, C 3 H 3 0 2 + at m/z= 71, and C 3 H 3 0+ at m/z= 55. 4.5. Potentiometric Studies The pKa of pyromeconic acid was determined potentiometrically to be 7.71 (deprotonation of the ring hydroxyl). The formation constants for the metal complexes were also determined in 0.15 M NaCl, and are listed in Table 4.4. Table 4.4. Logarithms of the Metal-Pyromeconic Acid Overall Stability Constants. Hpa H Al Ga -log K a 7.71(1) log Bi 7.88(2) 9.8(1) log 82 14.85(1) 18.3(1) log 83 21.15(1) : 25.6(1) 7 1 100 _ 260°C -' 1 11 1 1 11 1IIII111'l 1 1 II1 xlo* 1 l'i i i i i i i i i i i i i i i i i i i i i 11111 I'I 11 1 1 1 ' 1 1 1 1 1 1 1 1 1 1 ! 1 1 II 1 1 1 1 1 • | 111 II 11 - 111 11 1 1 1' 300 400 100 -i 50 Pa I I i ; i l'i i I l'i , i | ' 100 Gapa 200 100 i 50 A'(P»V xlO 0 i i i i i I i i i I'I'I i i i ; M i I i i , i i i i I I I i i i i I I i M i 260 260 G A'(P«), I II I I I I 1 I I I I I I I I I 'l M I I I I I I | I I I I . I ••'! I I I I 360 lOO-i 50H Alpa Pa I I'I'I I I I I I I'I'I I I I I I'I I I I I I I I I I i i I I I "| i i I i I i IVI I I'I'I i i I I i'i1] , I I I I I'I I I I , I I I I I I I I I . I 100 200 Figure 4.2. The E.I. mass spectra of the aluminum and gallium complexes of pyromeconic acid, from m/z=50 to m/z=400. 72 Only a preliminary titration was performed for Ga(pa)3, and an error limit of ± 0.1 was estimated according to the maximum error encountered from previous titrations. The overall stability constants for the metal complexes of pyromeconic acid are the highest of all three ligand systems, but are still slightly lower than maltol. However, conditional formation constants can also be calculated which take into consideration that at a pH less than the ligand pKa, the concentration of deprotonated ligand available for complexation is decreased due to protonated species. The conditional formation constants can be calculated according to the following equation: 9 5 log (3C = log I33 - 3pKa- 3log {[H+] + Kg} where 8 C = the conditional overall stability constant, and Ka/{[H+] + Ka} = the fraction of unbound deprotonated ligand. Since maltol (pKa=8.50) has a higher pKa than pyromeconic acid (pKa=7.71), Hpa is actually a more effective chelator for aluminum and gall ium at physio logical pH. Tris(maltolato)aluminum(lll) (log 63=21.8) 9 6 has a calculated log 6 C value of 18.4, while the log 8 C value for tris(pyromeconato)aluminum(lll) is 19.7 at pH 7.4. The speciation diagrams displayed in Figure 4.3 show that for both metal complexes of pyromeconic acid, the M(pa)3 species is predominant over a large pH range. Hydrolysis only predominates at pH's greater than 9 and 8.5 for the aluminum and gallium complexes respectively. By comparing the Al(pa)3 speciation diagram with the Al(ima)3, a 0.15 M concentration of isomaltol is required to effectively compete with hydrolysis while pyromeconic acid is just as effective of a chelator of aluminum at 3.0 mM concentration. 73 100 -log [H+] Figure 4.3. Speciation diagrams of Al(pa)3 and Ga(pa)3 at 1.0 mM M 3 + and 3.0 mM pyromeconic acid. Unfortunately, the overall stability constants for the M(pa)3 complexes are still not great enough to compete with other ligands, such as transferrin, in vivo. Hpa metal complexes are also much less water soluble than maltol complexes, which limits it usefulness in isotonic solution. Conclusion Aluminum and gallium form stable tris-ligand complexes with pyromeconic acid in high yields. These have higher formation constants than either isomaltol or Hahm. Pyromeconic acid is also a better chelator than maltol for aluminum and gallium at physiological pH, but it will not be used for in vivo studies due to reasons previously discussed. Future Work This work, evaluation of Hima, Hahm and Hpa as ligands and tris(ligand)metal(lll) complexes, shows that while they can be prepared, their formation constants are not high enough for them to be considered for in vivo applications without further modifications. Future work in this area should therefore be dedicated to the formation of hexadentate ligands. Of the ligands studied, Hehm is the only one with potential for hexadentate ligand synthesis, as the reaction of Hehm with tame (1,1,1,-tris(aminomethyl)ethane) could form an interesting hexadentate amide ligand. This would have a much greater formation constant due to the chelate effect. 75 References 1. A.C. Alfrey, G.R. LeGendre, W.D. Kaehny. New Eng. J. Med. 294,184 (1976). 2. D.R. Crapper, S.S. Krishnan, S. Quittkat. Brain 99,67(1976). 3. R.L Hayes, K.R. Huber. Metal Ions Biol. Syst 16, 279 (1983). 4. S.F. Lincoln. Coord. Chem. Rev. 6, 309 (1971). 5. A. Fratiello, D. D. Davis, S. Peak, R.E. Schuster. Inorg. Chem. 10, 1627 (1971). 6. C.F. Baes Jr., R.E. Mesmer. The Hydrolysis of Cations; Robert Krieger Publishing Co: New York. (1986). 7. ibid p. 122. 8. ibid p.322. 9. F.A. Cotton, G. Wilkinson. Advanced Inorganic Chemistry, John Wiley, 4 t h edition: p.329 (1980). 10. J.R.J. Sorensen, I. R. Campbell, L. B. Tepper, P. D. Lingg. Environ. Health Persp 8, 3 (1974). 11. P.O. GanroL Environ. Health Persp. 65, 363 (1986). 12. A. Lione. Food Chem. Tox. 21, 103(1983). 13. A. Lione. Pharmac. Then 29, 255 (1985). 14. S.M. Ott. Inter. J. Artificial Organs. 6,173 (1983). 15. R.J.S. McGonigle, V. Parsons. Nephron 39,1 (1985). 16. D.R. Crapper McLachlan. Neurobiol. Aging 7, 525 (1986). 17. E. Kokmen. Mayo. Clin. Proc. 59, 35 (1984). 18. D.P. Perl. Environ. Health Persp. 63, 149 (1985). 19. D.R. Crapper-McLachlan, U. DeBoni. Neurotoxicology 1, 3 (1980). 20. D.R. Crapper, S.S. Krishnan, A.J. Dalton. Science 180, 511 (1973). 76 21. D.P. Perl, A.R. Brody. Science 208, 297 (1980). 22. I. Klatzo, H. Wisniewski, E. Streicher. J. Neuropathol. Exp. Neurol. 23, 187 (1965). 23. G.G. Glenner. Human Pathology 16, 433 (1985). 24. R. Katzman. New Engl. J. Med. 314, 964 (1986). 25. S. Moerlein, M.J. Welch. Inter. J. Nuc. Med. Biol. 8, 277 (1981). 26. R.B. Martin. Metal Ions Biol. Syst. 24,44 (1988). 27. R.B. Martin. Metal Ions Biol. Syst. 20,21 (1986). 28. W.R. Harris. Biochemistry 25, 803 (1986). 29. R.M. Lambrecht, N. Morcas. Applications of Nuclear and Radiochemistry, Pergamon: New York, p.359 (1982). 30. S.M. Larson, M.S. Milder, G.S. Johnston. Radiopharmaceuticals, The Society of Nuclear Medicine Inc: New York, p.413 (1975). 3.1. M.M. Finnegan, S.J. Rettig, C.Orvig. J. Am. Chem. Soc. 108, 5033 (1986). 32. W.O. Nelson, S.J. Rettig, C.Orvig. J. Am. Chem. Soc. 109, 4121 (1987). 33. G.J. Kontoghiorghes, L. Sheppard. Inorg. Chim. Acta. 136, L11 (1987). 34. W.O. Nelson. Ph.D. Thesis, University of British Columbia (1988). 35. R. N. Prasad, J.P. Tandon. J. Inorg. Nucl. Chem. 36, 1473 (1974). 36. A Sreekantan, C C . Patel. Proc. Indian Acad. Sci. 87A, 455 (1978). 37. H. Takei, M. Higo, K. Saito, T. Mukaiyama. Bull. Chem. Soc. Jap. 41, 1738 (1968). 38. J . Barrett, S.A. Halsey, T.L Peppard. J. Inst. Brew. 89, 356 (1983). 39. R. Tressl, D. Bahre, H. Koeppler, A. Jensen. Z ; Lebensm.-Unters. Forsch. 167,111 (1978). 40. J.E. Hodge, H.A. Moser. Cereal Chem. 38, 221 (1961). 41. A. Backe. Compt. rend. 150, 540 (1910); 151,78 (1910). 42. B.E. Fisher, J.E. Hodge. J. Org. Chem. 29,776 (1964). 43. J.E. Hodge, E.C. Nelson. Cereal Chem. 38, 207 (1961). 77 44. J.C. Goodwin. Carb. Res. 115, 281 (1983). 45. H. Ito. Agric. Biol. Chem. 41, 1307 (1977). 46. J.C. Goodwin. J. Carb. Chem. 4, 255 (1985). 47. W.R. Rieman, J.D. Neuss, B. Naiman. Quantitative Analysis, McGraw-Hill: p.369 (1950.) 48. B.E. Fisher, J.E. Hodge. J. Org. Chem. 29, 778 (1964). 49. A. Novak. Struct. Bond. (Berlin). 18, 177 (1974). 50. C.W. Bird, G.W.H. Cheeseman. Comprehensive Heterocyclic Chemistry Volume 3; A.R. Katritzky, C Rees, Eds.; Pergamon Press: New York, p. 17 (1984). 51. R. Grigg, M.V. Sargent, D.H. Williams. Tetrahedron 21, 3441 (1965). 52. J.W. Akitt. NMR and Chemistry, Chapman and Hall: New York. (1983). 53. J.W. Akitt. Ann. Rep. NMR Spectroscopy. 5A, 465 (1972). 54. J.J. Delpeuch. NMR of Newly Accessible Nuclei Vol 2, P. Laszlo, Ed.; "Chemically and Biochemically Important Elements," Academic Press (1983). 55. R.W. Briggs.and J.F. Hinton. NMR and the Periodic Table, R.K. Harris, B.E. Mann, Eds.; Academic Press: London, p. 279 (1978). 56. J.W. Akitt, W.S. McDonald. J. Mag. Res. 58, 401 (1981). 57. R.W. Briggs, J.F. Hinton. NMR and the Periodic Table, R.K. Harris, B.E. Mann, Eds.; Academic Press: London, p.280 (1978). 58. F.W. Wehrli. Ann. Rep. NMR Spectroscopy. 9,125 (1979). 59 R. Benn, A. Rufinska. AngewChem. Int. Ed. Engl. 25, 867 (1986). 60. J.W. Akitt, B.E. Mann. J. Mag. Res. 44, 584 (1981). 61. R. Benn, A. Rufinska, E. Janssen, H. Lehmkuhl. Organometallics 5, 825 (1986). 62. J .J . Delpeuch. NMR of Newly Accesible Nuclei, Vol 2, ed. P. Laszlo, "Chemically and Biochemically Important Elements," Academic Press: p.167 (1983). 63. M.M. Finnegan, T.G. Lutz, W.O. Nelson, A. Smith, C. Orvig. Inorg. Chem. 26, 2173 (1987). 78 64. H. Haraguchi, S. Fujiwara. J. Phys. Chem, 73, 3467 (1969). 65. P.K. Hon, C.E. Pfluger. J. Coord. Chem. 3, 67 (1973). 66. R.E. Rosenkranz, K. Allner, R. Good, W.v. Philipsborn, C H . Eugster. Helv. Chim. Acta. 46, 1259 (1963). 67. S.F. Mason. Physical Methods in Heterocyclic Chemistry 2, 61 (1963). 68. Performed by Dr. Dave Clevette of the University of British Columbia. 69. A. Albert, E.O. Serjeant, The Determination of Ionization Constants, Chapman and Hall: New York. (1984). 70. T.G. Lutz, D.J. Clevette, S.J. Rettig, C Orvig. Inorg. Chem., in press. 71. A.F. Cotton, G. Wilkinson. Advanced Inorganic Chemistry, John Wiley, 4 t h edition: New York, p. 333 (1980). 72. A.E. Martell, R.M. Smith. Critical Stability Constants., Plenum: New York, 3,246 (1977). 73. D.J. Clevette. to be published. 74. D.C. Perrin, I.G. Sayce. Talanta. 14, 883 (1967). 75. R.L.N. Harris. Aust. J. Chem. 29,1329 (1976). 76. A. Loidl, Th. Severin. Z Lebens. Unter-Forsch.. 161, 119 (1976). 77. D.J. Clevette, D.M. Lyster, W.O. Nelson, T. Rihela, G.A. Webb, C. Orvig. Manuscript in preparation. 78. G.B. Payne. J. Org. Chem. 32, 3351 (1967). 79. L J . Bellamy. Infrared Spectra of Complex Molecules Volume 2, Chapman and Hall, New York p. 210 (1980). 80. D.W. Brown, T.T. Nakashima, D.L Rabenstein. J. Mag. Res. 45, 302 (1981). 81. R. Grigg, M.V. Sargent, D.H. Williams, J.A. Knight. Tetrahedron 21, 3441 (1965). 82. R.J. Andersen, M.J. Le Blanc, F.W. Sum. J. Org. Chem. 45, 1169 (1980). 83. A. Bax. J. Mag, Res. 57, 314 (1984). 84. A.E. Derome. Modern NMR Techniques for Chemistry Research, Pergamon Press: New York. (1987). 7 9 85. R. Benn, H. Gunther. Angew. Chem. Int. Ed. Engl. 22, 350 (1983). 86. I. Suzuki, M. Tsuboi, T. Shimanouchi, S. Mizushima. Spectrochim. Acta. 16, 471 (1960). 87 R.A. Russell, H.W. Thompson. Spectrochim. Acta. 8, 138 (1956). 88. W. Geiger. Spectrochim. Acta. 19,655 (1963). 89. D. Welti. Spectrochim. Acta. 22,281 (1966). 90. L.J. Bellamy. Infrared Spectra of Complex Molecules Volume 2, Methuen: London. (1958). 91. L.J. Bellamy. Infrared Spectra of Complex Molecules Volume 2, Chapman and Hall: New York, p.181 (1980). 92 A. Yamazaki, K. Maeda. Kami Pa Gikyoshi. 40, 383 (1986). 93. B.F. Anderson, D.A. Buckingham, G.B. Robertson, J . Webb. Acta Cryst. C39, 723 (1983). 94. A.R. Katritzky, R.A. Jones. Spectrochim. Acta. 17, 64 (1961); J. Chem. Soc. 2947 (1960). 95. R.B. Martin. Clin. Chem. 32, 1797 (1986). 96. A.E. Martell, R.M. Smith. Critical Stability Constants, Plenum: New York, 3, 256 (1977). 80 Table A.1. Crystallographic Data for Al(ima)3. formula C 1 8 H i 8 A 1 0 e fw 402.3 crystal system monoclinic space gToup C2/c a (A) 29.215(2) 4(A) 8.2211(7) c(A) 15.052(2) 0 (deg) 98.775(7) V(A 3) 3572.9(6) Z 8 Dc (g/cm3) 1.496 F[0QQ) 1644 1.6 crystal dimensions (mm) 0.25 x 0.44 x 0.45 transmission factors — scan type u>-20 scan range (deg in u>) 1.30 -t- 0.35 tan 6 6can speed (deg/min) 2.0-20.1 data collected -l-/i,+*t ±1 26mat (deg) 55 crystal decay negligible unique reflections 4108 reflections with J > 3<r(7) 2091 number of variables 261 R 0.065 0.080 S • 3.50 max A/a (final cycle) 0.05 residual density (e/A3) -0.31 to +0.37 81 Table A.2. Final Positional (fractional x 10 4, Al x 105) and Isotropic Thermal Parameters {U x 10 3 A 2 ) with estimated standard deviations in the parentheses* for Al(ima)3. Atom X y X V e / U Al 62256( 4) 32517(14) 61776( 7) 58 0(1) 64B2( 1) m e t 3) 6306( 2) 69 0(2) 6093( 1) 3132( 3) 4931 ( 2) 69 0(3) 6629( 1) 4045( 4) 60421 2) 69 0(4) 6346 ( 1) 3495( 4) 7437( 2) 73 0(5) 5996( 1) 5400( 3) 6095( 2) 73 0(6) 5654 ( 1) 2296( 4) 629; ( 2) 64 0(7) 6356( 1) -6B6( 4) 41241 3) 94 0(6) 7339( 2) 5606 ( 5) 8116< 3) 137 0(9) 4B03( 2) 5306( 8) 62671 3) 160 CM ) 651 4 ( 1) 61( 5) 5712 < 3) 65 C(2) 6346 ( 1) 4 17 ( 5) 4625 ( 3) 63 C(3) 61 52 ( 1) 1 652 ( 6) 44771 3) 62 CU) 6047( 2) 167V( 6) 35231 3) 61 C(5) 61 64 ( 2) 201 ( 6) 33591 3) 100 C(6) 6733( 2) -1503( 6) 5979< 3) 69 c m 7 1 27 ( 1) 4730( 5) 66271 3) 64 C(B) 7040( 2) 4943( 6) 747BI 3) 71 C(9) 6663 ( 2) 4324 ( 5) 7667< 3) 65 C( 10) 6741 ( 2) 4727{ 7) 6501 1 3) . 95 C ( 11 )1 7094 ( 4) 5645( 11) 69501 5) 113 CM2) 7571 ( 2) 5265( 6) 63631 3) 62 C( 13) 5586( 2) 5866( 6) 61411 3) 74 CM4) 5242( 2) 4710( 6) 6249 3) 60 CM5) 5279( 2) - 30B5( 6) 6301 1 3) 77 C( 16) 4636( 2) 2337(10) 63571 4) 114 CM 7)1 4557( 3) -3648( 19) 63731 6) 138 CMB) 5456( 2) 75B9( 7) 6081 1 3) 104 CMT F 7740( 6) 5752(28) 71141 16) 116 ( CM7'P 4965( 9) 7307(35) 61631 16) 137( • 1 2 Superscripts refer to occupancy factors 0.705 ( ) and 0.295 < ) 82 Table A .3 . Ca lcu la ted Hydrogen Coord inates (fractional x 10 4 ) and Isotropic Therma l Parameters (U x 1 0 3 A 2 ) for Al(ima)3. Atom y X V l , o H(4) 5916 2482 3089 93 H(5) 6122 -258 2757 112 H(6a) 6604 -1935 6495 116 H(6b) 6672 -2267 5475 116 H(6c) 7068 -1346 6144 116 H(10) 6548 4371 9234 108 H ( H ) 1 7196 6163 9516 129 HC12B) 1 7618 4543 6641 107 H( 12b) 1 7548 5301 5727 107 HC12C) 1 7638 6385 6622 107 H( 16) 4759 1188 6377 127 H(17)1 4234 3557 6437 151 H( 18a) 1 5662 8203 6533 136 H(lBb) 1 5482 7994 5480 136 H(lBc) 1 5135 7701 6192 136 Occupancy factor 0.705. 83 Table A.4. Bond Lengths (Angstroms) with Estimated Standard Deviations in The Parentheses for Al(ima)3. Bond Length(A) Bond Length(A) A l • - 0 ( 1 ) 1 . 9 0 5 ( 3 ) 0 ( 9 ) - C O 7 ' ) 1 . 7 5 ( 3 ) A l - 0 ( 2 ) 1 . 6 6 0 ( 3 ) C O ) - C ( 2 ) 1 . 3 8 2 ( 6 ) A l - 0 ( 3 ) 1 . 9 1 9 ( 3 ) C ( 1 ) - C ( 6 ) 1 . 4 6 5 ( 6 ) A l - 0 ( 4 ) 1 . 8 8 6 ( 3 ) C ( 2 ) - C ( 3 ) 1 . 3 7 7 ( 6 ) A l -0(5) 1 . 8 8 7 ( 3 ) C ( 3 ) - C ( 4 ) 1 . 4 2 9 ( 6 ) A l -0(6) 1 . 8 7 6 ( 3 ) C ( 4 ) - C ( 5 ) 1 . 2 8 9 ( 7 ) 0( 1 ) - C O ) 1 . 2 6 1 ( 5 ) C ( 7 ) - C ( 8 ) 1 . 3 5 4 ( 6 ) 0(2) -C(3) 1 . 2 8 1 ( 5 ) C ( 7 ) - C ( 1 2 ) 1 . 4 7 2 ( 6 ) 0(3) - C ( 7 ) 1 . 2 7 1 ( 5 ) C ( 8 ) - C ( 9 ) 1 . 3 7 2 ( 6 ) 0(4) -C(9) 1 . 2 6 3 ( 5 ) C ( 9 ) - C O 0 ) 1 . 4 2 9 ( 6 ) 0(5) - C 0 3 ) 1 . 2 7 1 ( 5 ) C ( l O ) - C O l ) 1 . 2 7 ( 1 ) 0(6) - C O S ) 1 . 2 7 5 ( 5 ) C ( l 2 ) - C O 1' ) 1 . 2 0 ( 2 ) 0 ( 7 ) - C ( 2 ) 1 . 3 9 5 ( 5 ) ' C ( 1 3 ) - C ( U ) 1 . 4 0 9 ( 7 ) 0 ( 7 ) - C ( 5 ) 1 . 4 0 6 ( 6 ) C ( 1 3 ) - C 0 8 ) 1 . 4 6 6 ( 7 ) 0 ( 6 ) - C ( 8 ) 1 . 3 9 1 ( 5 ) C 0 4 J - C 0 5 ) 1 . 3 4 2 ( 7 ) 0 ( 8 ) - C O D 1 . 5 4 2 ( 9 ) C ( 1 5 ) - C ( 1 6 ) 1 . 4 4 6 ( 6 ) 0 ( 8 ) - c ( n ' ) 2 . 0 5 ( 2 ) C 0 6 ) - C O 7 ) 1 . 3 5 ( 1 ) 0 ( 9 ) - C ( 1 4 ) . 1 . 3 8 2 ( 5 ) C 0 8 ) - C 0 7 ' ) 1 . 4 2 ( 3 ) 0 ( 9 ) - C ( 1 7 ) 1 . 5 5 0 ) • 84 Table A.5. Bond Angles (degrees) with Estimated Standard Deviations in The Parentheses for Al(ima)3. Bonds Angle(deg) Bonds Angle(deg) 0(1) - A l -0(2) 94.1 (1) CO )-C(2)-C(3) 127.9(4) 0(1) - A l -0(3) 8 8 . 4 ( 1 ) 0(2 )-C(3)-C(2) 125.6(4) 0(1) - A l -0(4) 89 .1(1) 0(2 )-C(3)-C(4) 126.6(4) 0(1 ) - A l -0(5) 177.2(1) C(2 )-C(3)-C(4) 107.8(4) 0(1) - A l -0(6) 86 .9 (1) C(3 )-C(4)-C(5) 105.6(4) 0(2) - A l -0(3) 88.0 (1) 0(7 )-C(5)-C(4) 114.4(5) 0(2) - A l -0(4) 176.8(1) 0(3 )-C(7)-C(8) 120.4(4) 0(2) - A l -0(5) 88.0(1 ) 0(3 )-C(7)-C( l2) 120.0(4) 0(2) - A l -0(6] 91.2(1) C(8 )-C(7)-C(12) 119.6(5) 0(3) - A l -0(4) 92.0(1) 0(8 )-C(8)-C(7) 122.1(5) 0(3) - A l -0(5) 89.9(1) 0(8 >-C(8)-C(9) 109.5(4) 0(3) - A l -0(6] 175.1(2) C(7 )-C(8)-C(9) 128.4(4) 0(4) -hi -0(5] 88.8(1 ) 0(4 >-C(9)-C(8) 123.6(4) 0(4) - A l -0(6] 89.1 (1 ) 0(4 >-C(9)-CO0) 125.9(5) 0(5) - A l -0(6] 94.9(1) C(8 l-C(9)-CO0) 110.5(5) Al -0(1)-C(1 1 129.3(3) C(9 -C(10)-C(11 ) 106.4(5) Al -0(2)-C(3 123.9(3) 0(8; -COD-COO) 112.0(6) Al -0(3)-C(7 128.3(3) C(7] -C02)-CO 1 ' ) 97 ( D Al -0(4)-C(9 125.2(2) 0(5] -C(13)-C(14) 119.9(4) Al - o ( 5 ) - c ( i : i) 127.4(3) 0(5] -C(13)-C08) 121.7(5) Al -0(6)-Cd5) 124.4(3) C(14)-C(13)-C08) 118.4(5) C(2) -0(7)-C(5 103.3<4) 0(9) -C(14)-C(13) 116.6(6) C(8) -0(8)-C ( l ) 101.3(5) 0 ( 9 ) -C(14)-C(15) 114.7(6) C(8) -0(8)-C(1 i*) 81.3(7) C03)-C04)-C(15) 128.7(4) C(11)-0(8)-C( I T ) 170.6(8) 0(6) -C(15)-C(14) 124.6(4) C ( l4)-0(9)-C( 17) 97.8(6) 0(6) -C(15)-C(16) 124.2(6) Cd4)-0(9)-C( 17') 92(1) C(14 )-C05)-C06) 111.2(5) CO7)-0(9)-C( 17') 170(1) C05)-C(16)-C(17) 102.0(7) 0(1) -C(1)-C(2 119.0(4) 0(9) -C(17)-C(16) 114.2(6) 0 ( 1 ) -C(1)-C(6 119.2(4) C03)-C(18)-C07' ) 95(1) C(2) -C(1)-C(6. 121.8(4) 0(8) - C ( i r ) - C ( l 2 ) 119(2) 0(7) -C(2)-C(1 123.2(4) 0(9) - C07D-C08) 118(2) 0(7) -C(2)-C(3 108 .9(4) Figure A.1. The 1 3 C NMR spectrum of Hehbm in CDCI 3 . 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items