UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Magnetic and medicinal properties of lanthanides Barta, Cheri A. 2007

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

Item Metadata

Download

Media
831-ubc_2007-317341.pdf [ 10.59MB ]
Metadata
JSON: 831-1.0059743.json
JSON-LD: 831-1.0059743-ld.json
RDF/XML (Pretty): 831-1.0059743-rdf.xml
RDF/JSON: 831-1.0059743-rdf.json
Turtle: 831-1.0059743-turtle.txt
N-Triples: 831-1.0059743-rdf-ntriples.txt
Original Record: 831-1.0059743-source.json
Full Text
831-1.0059743-fulltext.txt
Citation
831-1.0059743.ris

Full Text

M A G N E T I C A N D M E D I C I N A L P R O P E R T I E S O F L A N T H A N I D E S by Cheri A . Barta B .S . (Honors), University of Nebraska-Kearney, 2002 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF D O C T O R OF P H I L O S O P H Y in T H E F A C U L T Y OF G R A D U A T E S T U D I E S (Chemistry) T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A August 2007 © Cheri A . Barta, 2007 A B S T R A C T Two distinct projects involving the use of lanthanide metal ions are discussed in this thesis. In the first study, the magnetic interactions in heterometallic d/f complexes were investigated. The most powerful permanent magnet known to date consists of a Nd-Fe core; however, the magnetic exchange between these two metal ions is poorly understood. B y purposeful design, new cationic d/f complexes using a multidentate amine phenol ligand were synthesized that selectively coordinate divalent first row t/-block transition metal ions (TM) and lanthanides (Ln) in close proximity, a feature desirable for magnetic studies. The synthesis, characterization, and magnetic studies of the d/f cluster complexes with the formula [LnTM2(bcn)2] +, where H^bcn = ?ra-A /,A / ' ,A^"-(2-hydroxybenzyl)-l,4,7-triazocyclononane), TM(II) = Zn, Cu and N i , and Ln(III)= L a , N d , Gd , Dy and Y b are I, discussed herein. In addition, the core structure of [TM(Hbcn)], [TM2(bcn)3], [La(bcn)], and [La 3(bcn) 2] have been examined. Select magnetic studies were also completed on a similar trimetallic d/f species, [LnNi2(tam)2]+, where retain = tris-1,1,1 -(((2-hydroxybenzyl)amino) methyl)ethane, and Ln(III) = Gd, D y and Y b , used to compare the magnetic behaviors between geometrically diverse TM(II)-Ln(III). systems. The second study illustrates the usefulness of Ln(III) in metallopharmaceuticals. La(III) ions are known to be functional mimics of Ca(II) ions and have been shown to affect the bone remodeling cycle. Exploiting this disruption to the homeostasis of bone has potential for the treatment of bone density disorders, such as osteoporosis. In an effort to find new orally active agents for these disorders, a series of Ln(III) containing complexes incorporating small, non-toxic, bidentate pyrone and pyridinone ligands have been synthesized and characterized (LnL3, Ln(III) = La , ii Eu, Gd, Tb, Y b , L = 3-oxy-2-methyl-4-pyrone (ma"), 3-oxy-2-ethyl-4-pyrone (ema"), 3-oxy-l,2-dimethyl-4-pyridinone (dpp") and 3-oxy-2-methyl-4(7//)-pyridinone (mpp~)). Preliminary biological analyses undertaken in this thesis include cytotoxicity, cell uptake and bidirectional transport studies in Caco-2 cells and in vitro hydroxyapatite (HA) binding studies analyzed by I C P - M S , fluorescence assays, U V - v i s spectrophotometric assays and IR. The LnL3 species were found to have IC50 values at least 6 times greater than that of cisplatin, > 98 % HA-binding capacity, and permeability coefficients in the moderate range. iii T A B L E OF CONTENTS Abstract i i Table of Contents iv List of Tables v i i i List of Figures x List of Schemes xi i i List of Equations xiv List of Abbreviations xv Acknowledgments xxv CHAPTER 1 Introduction 1.1 Chemistry of Lanthanides 1 1.2 Magnetic Properties of Lanthanides 5 1.3 Medicinal Properties of Lanthanides 15 1.3.1 Medicinal characteristics 15 1.3.2 Diagnostic uses 20 1.3.3 Approved therapeutic agents 23 1.3.4 Anticancer agents 24 1.3.5 Treatment of autoimmune disorders 26 1.4 Thesis Overview 27 1.5 References 29 iv CHAPTER 2 d/f Amine Phenol Complexes: Synthesis and Magnetic Characterization 2.1 Introduction 35 2.2 Experimental 42 2.2.1 Materials 42 2.2.2 Instrumentation 43 2.2.3 [LaZn 2 (tam)2]C10 4 44 2.2.4 H 3 bcn 45 2.2.5 [TM(Hbcn)] complexes 48 2.2.6 [TM 3 (bcn) 2 ] complexes 50 2.2.7 [La(bcn)] 52 2.2.8 [La 3 (bcn) 2 ] 3 + 52 2.2.9 [LnTM 2 (bcn ) 2 ] + complexes 53 2.2.10 X-ray crystallographic analyses 55 2.3 Results and Discussion 56 2.3.1 The [LnTM 2 ( t am) 2 ] + and [LnTM 2 ( t rn) 2 ] + complexes 56 2.3.2 Synthesis of H 3 bcn 67 2.3.3 Synthesis and characterization of [TM(Hbcn) ]»nH 2 0 complexes 70 2.3.4 Synthesis and characterization of [TM 3 (bcn ) 2 ]*«H 2 0 complexes 77 2.3.5 Synthesis and characterization of [La x(bcn)^]' ? + complexes 80 2.3.6 Synthesis and characterization of [LaTM 2(bcn) 2]C104 • « H 2 0 complexes. 82 2.3.7 Magnetic studies of the [LnNi 2 ( t am) 2 ]C104 '«H 2 0 complexes 93 2.3.8 Magnetic studies of the [ L n T M 2 ( b c n ) 2 ] C 1 0 4 » « H 2 0 complexes 100 2.4 Conclusions 118 v 2.5 References , 120 C H A P T E R 3 Lanthanide Containing Compounds for Therapeutic Care in Bone Resorption Disorders 3.1 Introduction 125 3.2 Experimental 131 3.2.1 Materials • 131 3.2.2 Instrumentation 131 3.2.3 Synthesis 132 3.2.4 Caco-2 cell culture 135 3.2.5 Biological studies ; 137 3.3 Results and Discussion 145 3.3.1 Ligand synthesis 145 3.3.2 Metal complexes 146 3.3.3 Cellular toxicity and uptake studies 156 3.3.4 In vitro hydroxyapatite binding studies 163 3.4 Conclusions 168 3.5 References 170 C H A P T E R 4 Conclusions and Future Work 4.1 Magnetic Studies of a^fHeterometallic Complexes 175 4.1.1 Proposed ligands for d/f complexes 175 4.1.2 Proposed ligands for/7/heteronuclear complexes 177 4.2 Bone Density Agents 178 vi 4.2.1 Metal-ligand stability studies 178 4.2.2 Relevant animal models 179 4.2.3 Proposed bifunctional ligands for bone density disorders 180 4.3 Preliminary Experimental for Relevant Bone-seeking B F C 183 4.3.1 Materials 183 4.3.2 Instrumentation 183 4.3.3 Preliminary synthesis 184 4.4 Results and Discussion for Purposed Bone-seeking B F C 189 4.5 References 197 APPENDIX 200 vii LIST OF T A B L E S Table 1.1 Properties of lanthanides 3 Table 1.2 Magnetic properties of L n 3 + ions at room temperature 8 Table 1.3 Potential lanthanide isotopes for use in diagnosis and therapy 19 Table 1.4 M R I contrast agents approved for clinical use in Canada 21 Table 2.1 "H N M R data for Fbtrn proligand and complexes 66 Table 2.2 Analytical data for [LnTM 2 (bcn) 2 ] + complexes 84 Table 2.3 Observed and calculated magnetic susceptibilities for [LnCu 2(bcn) 2]t. •. 105 Table 2.4 Observed and calculated magnetic susceptibilities for [LnNi 2 (bcn) 2 ] + . . . . 114 Table 3.1 Elemental analyses, % yield and +ESI-MS results for all prepared LnL3 complexes 148 Table 3.2 Selected IR stretches of the Ln(III) complexes 150 Table 3.3 ' H N M R data for La(III) complexes 151 Table 3.4 1 3 C N M R data for the La(III) pyrone complexes 153 Table 3.5 Fluorescence data for the Eu(III) and Tb(III) complexes , 155 Table 3.6 Cytotoxicity data for the LnL3 complexes 159 Table 3.7 Percent of Ln(III) bound to H A after the indicated time period 164 Table Al Crystallographic parameters for structures in Chapter 2 200 Table A2 Selected bond lengths (A), intrametallic distances (A) and angles (°) for [LaZn 2 (bcn) 2 ] + . 202 Table A3 Selected bond lengths (A) and angles (°) for H 3 bcn 203 Table A4 Selected bond lengths (A) and angles (°) for {H 3 }[Ni(bcn)] 2 + 204 Table A5 Selected bond lengths (A), intrametallic distances (A) and angles (°) for [Cu 3 (bcn) 2 ] 2 + 204 viii Table A6 Selected bond lengths (A), intrametallic distances (A) and angles (°) for [GdNi 2 (bcn) 2 ] + , 13 205 Table A7 Selected bond lengths (A), intrametallic distances (A) and angles (°) for [YbNi 2 (bcn) 2 ] + , 15 206 Table A8 Selected bond lengths (A), intrametallic distances (A) and angles (°) for [GdZn 2 (bcn) 2 ] + , 3 207 ix LIST OF FIGURES Figure 1.1 Successive splitting of the 4 I ground term of N d 3 + 6 11 Figure 1.2 A proposed explanation for the ferromagnetic interaction between Cu(II)-Gd(III) 11 Figure 1.3 Angular momentum coupling of Cu(II)-Ln(III) 14 Figure 1.4 Ligand precursors of well-known pharmaceuticals. 17 Figure 2.1 Diagrams of selected ^heterometal l ic complexes 37 Figure 2.2 Ligands of interest as well as the proposed binding motif for the d/f heterometallic complexes 40 Figure 2.3 ' H N M R spectra for H 3 tam, [Zn(Htam)] and [LaZn 2 ( tam) 2 ] + 58 Figure 2.4 +ESI-MS spectrum of [LaFe 2(tam) 2] + 59 Figure 2.5 Table of angles for [M(tam)] complexes, selected angles (A , B, C) and bond lengths (between the numbered 1, 2 and 3 sp 3 carbons) used to determine deformations of the tam3" ligand upon coordinating a metal ion 60 Figure 2.6 O R T E P diagram of the [LaZn 2(tam) 2] + cation 63 Figure 2.7 +ESI-MS spectra of [LaTM 2 ( t rn) 2 ] + , T M = Mn(II) and Zn(II) 65 Figure 2.8 O R T E P diagram of the H4bcn + cation 70 Figure 2.9 *H N M R spectra of H 3 bcn and [Zn(Hbcn)] 71 Figure 2.10 Expanded section of the alkane region for the ' H N M R spectrum of [Zn(Hbcn)] 72 Figure 2.11 ' H N M R spectra of [Ni(Hbcn)] and H 3 bcn 74 Figure 2.12 O R T E P diagram of the {H 3 } [Ni(bcn)] 2 + cation 75 Figure 2.13 *H N M R spectra of [Zn 3(bcn) 2] and [Zn(Hbcn)] 78 Figure 2.14 O R T E P diagram of the [Cu 3 (Hbcn) 2 ] 2 + cation 79 Figure 2.15 ' H N M R spectra of [La 3 (bcn) 2 ] 3 + and [La(bcn)] 81 x Figure 2.16 IR spectra of selected [LnTM 2 (bcn)2] + 86 Figure 2.17 +ESI-MS spectra of [LaZn2(bcn)23+ 87 Figure 2.18 ' H N M R spectra of H 3 bcn and [LaZn 2 (bcn) 2 ] + 89 Figure 2.19 O R T E P diagrams of the [GdNi 2 (bcn) 2 ] + cation 90 Figure 2.20 O R T E P diagrams of the [GdZn 2 (bcn) 2 ] + cation 92 Figure 2.21 Temperature dependence of uefffor [LnNi 2 ( tam) 2 ] + at 500 G 96 Figure 2.22 u e ffV5. T plots of L a C u 2 and L a N i 2 100 Figure 2.23 Temperature dependence of |o,eff for L n Z n 2 complexes 103 Figure 2.24 Temperature dependence of XM^for L n C u 2 complexes 104 Figure 2.25 AXM?" for L n C u 2 complexes 107 Figure 2.26 XM VS. T and u . e f fW. Tplot for G d C u 2 110 Figure 2.27 Mvs. / /curve for G d C u 2 112 Figure 2.28 Temperature dependence of XM^for L n N i 2 complexes 113 Figure 2.29 M vs. H curve for G d N i 2 115 Figure 2.30 AXM?" for complexes L n N i 2 complexes 117 Figure 3.1 Structures of the 3-hydroxy-4-pyrones and 3-hydroxy-4-pyridinones 129 Figure 3.2 ' H N M R spectra of Hma and La(ma) 3 152 Figure 3.3 Fluorescence spectra for the Eu(III) and Tb(III) complexes 154 Figure 3.4 Cel l uptake of L n L 3 in Caco-2 cells determined by I C P - M S 161 Figure 3.5 Apparent permeability of La(ma) 3, La(ema) 3, La(dpp) 3 and Yb(ma) 3 162 Figure 3.6 Mean percentage + SD of Ln(III) bound to H A samples 165 Figure 3.7 Percent ligand concentration + SD remaining in supernatant of the Ln(III)-HA binding studies determined by U V - v i s spectroscopy 166 Figure 3.8 IR spectra from H A binding studies 168 xi Figure 4.1 Potential ligands for d/f complexes utilizing the tren backbone 177 Figure 4.2 Potential backbones and functional groups for bone-seeking B F C 178 Figure 4.3 Known bone seeking agents 182 xii LIST OF SCHEMES Scheme 2.1 Synthesis of [LaZn 2 (tam) 2 ] + 57 Scheme 2.2 Synthesis of [LaTM 2 ( t rn) 2 ] + , T M = Mn(II) or Zn(II) 64 Scheme 2.3 Synthesis of Fbbcn 68 Scheme 2.4 Synthesis of the [LnTM 2 (bcn) 2 ] + 83 Scheme 3.1 Deprotonation (and resonance forms) of 3-hydroxy-4-pyridinones 130 Scheme 3.2 Synthesis of Hmpp 145 Scheme 3.3 3-Hydroxy-4-pyrones and 3-hydroxy-4-pyridinones of interest and their Ln(III) complexes 146 Scheme 4.1 Mannich reaction of salicylaldehyde with tacn 176 Scheme 4.2 Formation of the amide/anhydride of Kemp's triacid 190 Scheme 4.3 Synthesis of the triamine 192 Scheme 4.4 Synthesis of 2,4-dibenzyloxybenzoic acid 192 Scheme 4.5 Synthesis of the protected (3), mono-protected (4) and deprotected (5) tri-substituted tren proligand 194 Scheme 4.6 Synthesis of the protected (6), mono-protected (7) and deprotected (8) 195 tri-substituted nitrilotriacetic acid proligand Scheme 4.7 Synthesis of a potential B F C using protected phosphate directing groups 196 xiii LIST OF EQUATIONS Equation 1.1 Spin-only magnetic moment 5 Equation 1.2 Zeeman Hamiltonian for Ln(III) 7 Equation 1.3 Magnetic moment for Ln(III) :. 7 Equation 1.4 Spin Hamiltonian quantifying intramolecular interactions 9 Equation 1.5 Spin-orbit term coupling constant 12 Equation 1.6 Energy levels of the spin-orbit states 12 Equation 2.1 Magnetic susceptibility for TM(II), Xid 94 Equation 2.1 Magnetic susceptibility for Ln(III), %4/ 94 Equation 2.3 Effective magnetic moments, peff, for d/f complexes 95. Equation 2.4 Conversion of u.eff to y^T 95 Equation 2.5 ueff for coupled spins 98 Equation 2.6 Bleaney-Bowers equation 101 Equation 2.7 Empirical approach for determining magnetic interactions in the d/f complexes, A (XM^)- 102 Equation 2.8 Spin Hamiltonian 109 Equation 2.9 Cu(II)-Gd(III) exchange integral, J c u G d , using a modified spin Hamiltonian 109 Equation 2.10 Discrepancy factor for R(XM) 109 Equation 2.11 Discrepancy factor for R((xe.r) 109 Equation 2.12 Magnetic saturation I l l Equation 2.13 Bri l louin function I l l Equation 3.1 Apparent permeability 141 x i v LIST OF ABBREVIATIONS ~ approximate ° Degree(s) A l 0 angstrom, 1 x 10" meter a alpha particles, alpha 8 Bohr magneton, beta (3 + positron particles (3" beta particles /?4f-3d transfer integral for an electron moving from the 3d orbital of Cu(II) to the 4/orbitalofGd(III) /55d-3d transfer integral for an electron moving from the 3d orbital of Cu(II) to the 5d orbital of Gd(III) 5 chemical shift in parts per mil l ion (ppm) A change, clockwise handedness (chelate isomers) y gamma rays, gamma C, spin-orbit coupling constant X wavelength, spin-orbit term coupling constant Xmax wavelength of maximum absorption (UV-visible) X,em emission wavelength Xexc excitation wavelength A counterclockwise handedness (chelate isomers) p micro (10"6) PB Bohr magneton (units) Pcaic Linear absoption coefficient of the material (crystallography) x v u-efr effective magnetic moment 9 correction term for intermolecular magnetic interactions XM molar magnetic susceptibility XMT- temperature dependent magnetic susceptibility A apical, surface area of the insert membrane A c acetyl Anal . analytical A N O V A analysis of variance A T C C American Type Culture Collection atm atmosphere A T P adenosine 5'-triphosphate B basolateral Bn benzyl Bnma 3-benzyloxy-2-methyl-4-pyrone Bnmpp 3-benzyloxy-2-methyl-4(7i7^)-pyridinone B C A bicinchoninic acid protein assay B F C bifunctional chelator B O P T A benzyloxypropionictetraacetate br broad B S A bovine serum albumin °C degrees Celsius C 0 initial concentration Calcd calculated x v i CC49 specific monoclonal antibody C C D charge-coupled devise C D I TV.Af'-carbonyldiimidazole cm centimeters C N coordination number C O S Y correlation spectroscopy ( N M R ) coth hyperbolic cotangent C T C charge transfer configuration 2D two-dimensional d doublet ( N M R ) , day(s) dd doublet of doublets ( N M R ) ddd doublet of doublet of doublets ( N M R ) D c a i c calculated density (crystallography) D C C dicyclohexylcarbodiimide D C U dicyclohexylurea D M F dimethylformamide D M A P 4-dimethylaminopyridine D M S O dimethylsulfoxide D N A deoxyribonucleic acid D03A-butrol l,4,7-tris(carboxymethyl)-10-(l-(hydroxymethyl)-2,3-dihydroxypropyl)-1,4,7,10-tetraazacyclododecane dt doublet of triplets ( N M R ) D T P A diethylenetriaminepentaacetic acid D T P A - B M A diethylenetriaminepentaacetic acid Z)w(methylamide) xvii D T P A - D M E A diethylenetriaminepentaacetic acid ^(methoxyethylamide) dQ/dt flux rate of mass transport across the monolayer D X A dual energy X-ray absorptiometry E energy E A elemental analysis E B S S Eagle's Balanced Salt Solution E C electron capture E D C l-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride E D T A ethylenediaminetetraacetic acid E D T M P ethylendiaminetetramethyphosphonate E M E M Eagle's Min imum Essential Medium ESI electrospray ionization E t O H ethanol eV electron volt fac facial F D A Food and Drug Administration (USA) FTIR Fourier transform infrared fw formula weight g gram g Lande factor gs Zeeman Hamiltonian G gauss G C ground configuration xviii G l gastrointestinal G O F goodness of fit h hour(s) H magnetic field H Hamiltonian H A hydroxyapatite H A P e n A^,A^'-ethylene-6w-(o-hydroxyacetophenomeimine) H 3 bcn tris-N,N',N' '-(2-hydroxybenzyl)-l ,4,7-triazacyclononane H B S S Hank's Balanced Salt Solution Hdpp deferiprone, 3-hydroxy-1,2-dimethyl-4-pyridinone H E D P hydroxyethylidene-l,l-diphosphonic acid H E D T A hydroxyethylenediaminetetraacetic acid Hema ethylmaltol, 3-hydroxy-2-ethyl-4-pyrone H E P E S 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid hfac hexafluoroacetylacetone Hma maltol, 3-hydroxy-2-methyl-4-pyrone Hmpp 3-hydroxy-2-methyl-4(7//)-pyridinone HOBt 1 -hydroxybenzotriazole H P - D 0 3 A 1,4,7- tris(carboxymethyl)-10-(2'-hydroxypropyl)-1,4,7,10-tetraazacyclododecane H S A B Hard-soft acids-bases H 3 tam tris-1,1,1 -(((2-hydroxybenzyl)amino)methyl)ethane H 3 t rn tris(2' -hydroxybenzylaminoethy l)amine H I V Human Immunodeficiency Virus x i x H z hertz (s"1) . IC50 drug concentration at which 50% of cells are viable relative to the control ICP inductively coupled plasma IR infrared J coupling constant ( N M R ) , exchange integral (magnetics), total angular moment quantum number k Boltzmann constant k kilo (10 3) K Ke lv in L liter, generic ligand L total orbital angular momentum quantum number L D H lactate dehydrogenase L n lanthanide(s) logK stability constant (or formation constant) m meta-m meter, mi l l i (10" ) M molar (moles per liter for concentration), mega (10 6) M magnetic response max. maximum M e O H methanol mer meridional M H z megahertz min minutes min. minimum xx mol mole (6.02 x 10 molecules) M R I magnetic resonance imaging M S mass spectrometry MS-325 l-(i?)-(4,4-diphenylcyclohexylphosphorylmethyl)diethylenetriamine pentaacetic acid, trisodium salt M S B mass susceptibility balance M T S 3 -(4,5 -dimethylthiazol-2-yl)-5 -(3 -carboxymethoxyphenyl)-2-(4-sulfophenyl)-2//-tetrazolium bromide M T T 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide m/z mass to charge ratio n nano, number N Avogadro's number N a temperature-independent paramagnetism N A D H nicotinamide adenine dinucleotide N A D P H nicotinamide adenine dinucleotide phosphate Net3 triethylamine N H S A^-hydroxysuccinimide N M R nuclear magnetic resonance NP-40 nonylphenyl-polyethylene glycol o ortho-obsd observed O R T E P Oak Ridge Thermal Ellipsoid Program O V X ovariectomized p para-x x i P probability of variation between conditions (P-value for A N O V A ) P a p p apparent permeability P A - D O T A 1,4,7,10-tetraaza-/V-(l -carboxy-3-(4-nitrophenyl)propyl)-/v r ',/V",/V" tris(acetic acid)cyclododecane PBS phosphate buffer solution Pd/C palladium on carbon P D T photodynamic therapy P E T positron emission tomography pH -log[H +] p K a acid dissociation constant ppm parts per mil l ion ppt parts per trillion q quartet ( N M R ) R discrepancy factor RI residual factor (crystallography) ref. reflections ROS reactive oxygen species rpm rotations per minute R T room temperature s second, singlet ( N M R ) S spin S total spin angular momentum quantum number S spin operator S p a i r coupling of spins from the Ln(III) and Cu(II) ions xxii ST total spin SD standard deviation S F U Simon Fraser University S P E C T single photon emission computed tomography SQUID superconducting quantum interference device t triplet ( N M R ) tt triplet of triplets T temperature T i longitudinal relaxation time t./2 half-life tacn 1,4,7-triaazacyclononane TAG-72 tumor-associated glycoprotein 72 tame 1,1,1 -fra(aminomethyl)ethane T E E R . transepithelial electrical resistance tex texaphyrin T H F tetrahydrofuran T M d-block transition metal T O F time of flight tren ^r/5(2-aminoethyl)amine TsCl /?-toluenesulfonyl chloride LP energy gap between the excited and ground states U B C University of British Columbia U V ultraviolet xxi i i V volume vis visible vs. versus wR2 least-squares residual (crystallography) w/v weight by volume w/w by weight Z atomic number xxiv A C K N O W L E D G M E N T S I would like to thank my family and friends for all their support throughout this journey. Thanks to Kayla , Chedda, Jason, the volleyball crew, the Nicola Lake group, A T M gatherings, and many chemistry colleges that made this process more enjoyable than I could have imagined. A special thank you to my parents who have been tremendous in their constant encouragement and their continuous confidence—they both have been instrumental to my success. To my three brothers, whose tough love has taught me how to persevere (Nick B. , thanks for the help with a few select reactions and for taking me on several B . C . adventures that wi l l be remembered for many years). Thank you Dr. Rossi for always making me smile and for always knowing the perfect thing to say to keep me motivated. Your friendship has meant the world to me. Thanks to my boss, Dr. Chris Orvig for his limitless encouragement, confidence, and for the independence that allowed me to explore many different facets of inorganic chemistry. Dr. Kathie Thompson's constant optimism, expertise, and endless hours of discussions spanning many different subjects are also sincerely appreciated. Thank you Dr. Bob Thompson for all the insightful conversations about magnetism and to Dr. Kishor Wasan and Dr. Kristina Sachs-Barrable for their expertise and for all their hard work. How can I ever thank the Orvig group members enough? It was always a joy to come into the lab and see what adventures we could cook up next. Thanks Meryn, Lauren, Cara, Ne i l , Vishakha, Michael, T im, Dave, Simon, Chuck, Kathy, Karen, Dr. Chen, Paloma and Yasmin for your knowledge, all the fond memories, and especially for your friendships (you all are very excellent). I also would like to acknowledge the support staff: Marrieta Austria ( N M R ) , Ms . Liane Darge ( N M R ) , Zorana Danilovic ( N M R ) , Dr. Nick Burlinson ( N M R ) , Mr . Minaz Lakha (EA) , Dr. Y u n Ling (MS), the xxv mechanical and electronic shops, Brian Ditchburn, and the office staff. Special thanks to Dr. Brian Patrick (X-ray) for all his help and patience when teaching me about the fine arts of crystallography, to Marshall Lapawa (MS) for his endless help in the M S lab and many enlightening chats over beer, Dr. Elena Polishchuk for all her persistence with the uncooperative Caco-2 cells, Candice Martins for teaching me countless aseptic techniques and for being a friend, Burt Mueller for all his time, patience and enthusiasm for ICP-MS. I would also like to acknowledge Viv ian Lai for her know-how and helpful discussions about acid digestion, Dr. Cecil ia Stevens for passing on her ' love' of the S Q U I D , my three summer students, Rabiya Jali l , Mathias Winkler, and Jessica Jia, for teaching me patience and organization skills as well as for all their hard work and for many laughs. Thank you Dana Zendrowski and Jonathan Lau for your constant support. I would also like to acknowledge the W o l f and Maclachlan labs for their intriguing lunch room discussions, chemical expertise, and use of their instruments. xx vi C H A P T E R 1 Introduction 1.1 Chemistry of Lanthanides Lanthanides were first discovered in 1794 by Johann Gadolin who isolated a black mineral near Stockholm, Sweden he believed contained the oxide of a single new element.1 It was not until the mid 1800s that the mineral was determined to contain no fewer than ten different elements, eight being lanthanides (abbreviated collectively as Ln). For a long time, these new elements were relegated to the bottom of the Periodic Table as an appendix. Mendeleev's revolutionary Periodic Table, first published in 1869, only accommodated lanthanum, cerium, didymium (now known to be a mixture of praseodymium and neodymium), and another mixture called erbia. To complicate matters, the pronounced similarities of the properties of L n and the development of new, often misinterpreted spectroscopic analysis permitted misidentification; around 70 'new' elements were erroneously claimed to be the fourteen actual L n metals.3 Promethium, the only naturally occurring radioactive L n , was not discovered until after World War II, while the thirteen other L n were isolated and correctly identified by 1907. Strictly speaking, lanthanides are confined to the fourteen elements following lanthanum in the period table (Ce, Z = 58 to L u , Z = 71; Table 1.1). Together with lanthanum, yttrium and scandium, lanthanides are historically known as the "rare earths" (lanthanum (La, Z = 57), although technically not a L n , has many of the same characteristics 1 of lanthanides and w i l l be considered herein as a L n for the remainder of this thesis). This misleading term was assigned to these metal ions mainly because all members were isolated from the earth as oxides; however, the 4/elements are actually quite abundant. Cerium, the 26 t h most abundant element on earth, has a crustal concentration greater than that of nickel or copper. Even thulium and lutetium, the rarest L n , are more abundant than are silver or platinum metals.' Although L n are found all over the world including the Sierra Nevada Mountains in western U S A , they are primarily mined from the rare earth reserves in China, which contain over 70% of the world's stock of lanthanides.2 A large percentage of the L n are extracted from two minerals, monazite (Ln- and T h P 0 4 ) and bastnaesite (LnFCCb) from which they are meticulously separated by solvent extraction, ion exchange chromatography or fractional crystallization. 4 The valence electrons for neutral lanthanides are distributed amongst the 4f, 5d, and 6s orbitals. These orbitals experience different degrees of stabilization upon removal of electrons with 4/being stabilized the most and 6s the least. 1 ' 5 Thus, electrons are first selectively removed from the 5d and 6s orbitals upon ionization. The most common oxidation state for L n is +3 with an electron configuration of [Xe]4/ 1 . 4 Although uncommon, +2 (Eu, Yb) and +4 (Ce, Tb) oxidation states are also known; 6 however, only the trivalent L n ions wi l l be considered for the remainder of this thesis. The progressive reduction in the Ln(III) radii with increasing atomic number is a trademark of the lanthanide ions, and is known as the lanthanide contraction (15%> contraction between La(III) and Lu(III), ~ 1%> between successive lanthanides).7 This property stems from the /electrons having a poor ability to screen other valence electrons from the nuclear charge. A n increased effective nuclear charge is experienced by all valence 2 electrons upon the addition of one extra proton, balanced by an equal increase in the electronic charge causing each ion down the row to contract slightly compared to its predecessor. A s a result, the charge density, and subsequently the acidity, of the Ln(III) cations increase from left to right through the series.1 T a b l e 1.1. Properties of lanthanides.2' Element Symbol Atomic Number Electronic Configuration Trivalent ionic radius (A) Atomic L n 3 + C N 6 C N 8 Lanthanum L a 57 [Xe]4f°5d 1 6s 2 [Xe] 1.032 1.160 Cerium Ce 58 [Xe]4f'5d 16s 2 [Xe]4f 1 1.01 1.143 Praseodymium Pr 59 [Xe]4f 36s 2 [Xe]4f 2 0.99 1.126 Neodymium N d 60 [Xe]4f 46s 2 [Xe]4f 3 0.98 1.109 Promethium Pm 61 [Xe]4f 56s 2 [Xe]4f 4 0.97 1.093 Samarium Sm 62 [Xe]4f 56s 2 [Xe]4f 5 0.958 1.079 Europium E u 63 [Xe]4f 76s 2 [Xe]4f 5 0.947 1.066 Gadolinium G d 64 [Xe]4f 75d'6s 2 [Xe]4f 7 0.938 1.053 Terbium Tb 65 [Xe]4f 96s 2 [Xe]4f s 0.923 1.040 Dysprosium D y 66 [Xe]4f 1 0 6s 2 [Xe]4f 9 0.912 1.027 Holmium Ho 67 [Xe]4f"6s 2 [Xe]4f 1 0 0.901 1.015 Erbium Er 68 [Xe]4f 1 2 6s 2 [Xe]4f" 0.890 1.004 Thulium T m 69 [Xe]4f 1 3 6s 2 [Xe]4f 1 2 0.880 0.994 Ytterbium Y b 70 [Xe]4f 1 4 6s 2 [Xe]4f 1 3 0.868 0.985 Lutetium L u 71 [Xe]4f 1 4 5d'6s 2 [Xe]4f 1 4 0.861 0.977 ' 3 The bonding interactions of Ln(III) are primarily ionic due to the 4f electrons being shielded by the outer 5s and 5p electrons.2 The internal nature of the 4f electrons cause the Ln(III) to have little stereochemical preference in the first coordination sphere compared to d-block metal ions that exhibit directional bonding. The unshielded d orbitals are further influenced by crystal-field effects increasing the preference for certain types of geometries (e.g. octahedral, square planar, etc.), a phenomenon not observed with L n ions. 7 In addition, the large ionic radii (between 0.86 A and 1.03 A, coordination number = 6) of the Ln(III) allow for higher coordination numbers (CN) compared to <i-block metal ions . 8 ' 9 Due to the lability of the Ln-solvent bonds and the lack of geometric preferences, it is difficult to predict the geometry and C N of Ln(III) complexes. Instead, geometries and C N around the Ln(III) are primarily determined by the electrostatic interactions and interligand steric constraints such as ligand conformation, types of donor atoms, size, number and solvation effects.6 A variety of C N and geometries have been observed for the Ln(III) ions, with C N ranging from 2 - 1 2 and assorted geometries reported from linear to near icosahedral. 1 ' 3 In addition, Ln(III) are considered to be hard Lewis acids and prefer to coordinate hard Lewis bases (O > N > S ) 2 ' 1 0 following Pearson's hard and soft acids and bases ( H S A B ) rules. 1 1 Detailed studies of Ln(III) ions solvated with water or acetonitrile suggests a preferred C N to be 8 - 9, with the C N decreasing with increasing atomic numbers (decreasing ionic radii). 7 The larger trivalent lanthanides adopt a thermodynamically preferred nine-coordinate tricapped trigonal prismatic arrangement both in solution and in the solid state. The smaller Ln(III) exhibit eight-coordinate square antiprismatic arrangements.1 2 Due to their tunable C N , geometries and their binding preferences, Ln(III) are attractive in the field of catalysis 1 3 and useful for diagnostic and therapeutic medicinal purposes. 1 4"' 8 In addition, the shielding of the 4/electrons by the filled 5s and 5p orbitals 4 leave the spectroscopic and magnetic properties of Ln(III) largely unaffected by their 7 8 environment. ' Thus, chelated lanthanide ions have very similar properties to their corresponding free ions making the Ln(III) attractive for several additional applications • 19 20 21 including luminescent probes, superconductors, N M R shift reagents, light emitting 22 23 25 diodes and magnetic materials. " Herein, the magnetic and medicinal properties of Ln(III) wi l l be further investigated. 1.2 Magnetic Properties of Lanthanides With the exception of La(III) and Lu(III), all Ln(III) contain unpaired electrons and thus, are paramagnetic. 2 5 The 4/orbitals, however, are not largely influenced by crystal field effects, unlike <i-block transition metals, leading to orbitally degenerate states. This large unquenched orbital angular momentum of Ln(III) voids many of the traditional equations used to analyze the magnetic properties of transition metals. For example, the spin-only formula shown in Equation 1.1, used to predict the magnetic moment, (j.eff, of t/-block elements does not accurately describe the magnetic properties of Ln(III). Meff = g~Y S(S+l) ( L 1 ) g is the Lande factor and S is the total spin angular momentum quantum number. In order to closely predict the u eff for lanthanides, additional parameters are needed that account for their orbital angular momentum and spin-orbit coupling. With the exception of Gd(III) and Eu(II) which have f electron configurations and orbitally non-degenerate ground states,26 the unquenched orbital angular momentum (defined 5 as term symbol L = total orbital angular moment quantum number) and the spin angular momentum (defined as term symbol S = total spin angular moment quantum number) further interact to produce several new microstates (Figure 1.1). These distinct energy levels, arising from spin-orbit coupling, are defined as J = total angular momentum quantum number, where J=L + S,L + S-\, L + S - 2 , . \ L - S |. Doublet Electronic Spin Orbital Spin-orbit Magnetic configuration coupling coupling coupling field Figure 1.1. The successive splitting of the 4 I ground term of the N d 3 + ion with an/ electron configuration. For most Ln(III), the magnitude of spin-orbit coupling is sufficiently large, making the excited J levels thermally inaccessible. 2 7 Thus, the magnetic behavior is commonly determined by the ground spin-orbit coupled state. The most notable exceptions are Eu(III) 6 and Sm(III) that have relatively low lying excited states ( 7 F 0 with 7 F[ at 350 cm"1 and the latter a ground state at 6 H 5 / 2 with 6 H 7 / 2 at 700 cm" 1). 2 5 The jueff for Ln(III) can be approximated by accounting for orbital angular momentum and spin-orbit coupling using Equation 1.2 and 1.3 at room temperature. 3 S(S+\)-L(L+\) " 2 2 J ( J + ] ) (1.2) J is the total angular momentum quantum number, gy is the Zeeman Hamiltonian of a given J multiplet, S is the total spin angular momentum quantum number, and L is total orbital angular momentum quantum number. The J values are additionally split into multiplets (2J+ 1) from the internal magnetic field generated by the motion of the orbitals. These microstates often have small energy differences comparable to thermal energy (Figure 1.1). A t low temperature, some of the low-lying J multiplets become depopulated giving rise to large deviations from the expected magnetic behavior. Thus, Equations 1.2 and 1.3 are usually only accurate for Ln(III) ions at higher temperatures. Nonetheless, the characteristic large p eff values arising from the paramagnetic behavior of Ln(III) have led to several novel magnetic materials (Table 1.2). 7 , 2 5 In fact, the strongest permanent magnet known, an alloy of neodymium, iron and boron containing 68 atoms (Nd8Fe56B4) with six distinct iron sites, two neodymium sites and one boron, has been used since the 1970s. 2 4 The magnetic moments of the iron and neodymium ions align in the same direction producing a bulk magnetic field fifty times that of steel, and ten times greater than ferrite, an iron oxide material that accounts for 90% of the mass of all magnets produced world wide. Table 1.2. Magnetic properties of L n 3 + ions at room temperature. 2 3 ' 2 7 L n 3 + / Ground State Heff (M-B) XT (cm 3 K mol"1) L a 0 "So 0 0 Ce 1 2F5/2 2.54 0.80 Pr 2 3 H 4 3.58 1.60 N d 3 3.68 1.64 Pm 4 % 2.83 0.90 Sm 5 6 H 5 / 2 0.85 0.09 E u 6 7 F 0 0 0 G d 7 8o a 7/2 7.94 7.88 Tb 8 7 F 6 9.72 11.82 Dy 9 6Hi5/2 10.63 14.17 Ho 10 5l8 10.60 14.07 Er 11 J-15/2 9.59 11.48 T m 12 3 H 6 7.57 7.15 Y b 13 2F7/2 4.53 2.57 L u 14 'So 0.00 0 8 Commonly, the interactions between two paramagnetic centers are effectively quantified using the spin Hamiltonian, H , in Equation 1.4. H = (-7)(5,-52) (1.4) J is the isotropic interaction parameter that takes negative values for antiferromagnetic coupling and positive values for ferromagnetic interactions, and S\ and §2 are the spin operators associated with the two magnetic centers. This equation has been useful in understanding the magnetic interactions between transition metals; however, Equation 1.4 is invalid in the presence of orbital degeneracy, requiring the introduction of several new parameters. To date, the unquenched orbital angular momentum of lanthanides has obviated the development of a simple model rationalizing magnetic interactions or structural magnetic correlations between f- and <i-block metals. In addition, the internal nature of/electrons gives rise to weak magnetic interactions between/centers, making the analysis of experimental data difficult. Suitable strategies for elucidating the magnetic exchange properties of Ln(III)-containing compounds have relied on synthesizing complexes where the /-block metal ion is coupled to d-block transition metal ions (TM) and/or organic radicals. 2 5 It was thought that the interaction with the less shielded s, p, or d electrons would be stronger than that of the/electrons, producing measurable effects at higher temperatures. Most magnetic studies on c///heterometallic complexes have been limited to Cu(II)-Gd(III) coupled systems because Gd(III) has an orbital non-degenerate ground state, simplifying the magnetic analysis. In all but one case, ferromagnetic interactions for the •y o -yr Gd-Cu species were observed. ' In addition, Gd(III) has been found to exhibit a 9 ferromagnetic spin-exchange when coupled with organic radicals , 3 0 ' 3 1 V 0 2 + , 3 2 N i ( I I ) , 1 5 ' 3 3 " 3 6 Fe(II), 3 2 or Co(II). 3 7 Many researchers have hypothesized that this ferromagnetic interaction is an intrinsic property of Gd(III); however, only a few models attempting to explain this interaction have been developed. Gatteschi and co-workers 3 8 ' 3 9 rationalized the ferromagnetic exchange from the Cu(II)-Gd(III) systems as a spin polarization effect. Kahn and co-workers , 4 0 ' 4 1 believed the interaction between the 4f-3d ground-state and the 3dcu-^5dad charge-transfer configuration resulted in the observed parallel spins. Gatteschi attributed the ferromagnetic interactions to a modified Goodenough-Kanamori rule, 4 2 first observed in manganites. It was hypothesized that the /-orbitals would be extremely weakly delocalized towards any ligand moiety surrounding the rare-earth due to the 4/electrons being efficiently shielded by the occupied 5s and 5p orbitals. The lack of orbital overlap between the Ln(III) and the ligand would obviate any electronic exchange through the ligand. Instead, Gatteschi and co-workers suggested the ferromagnetic exchange was mediated by the derealization of the Cu(II) unpaired electrons toward the empty 6s orbitals of Gd(I I I ) . 3 8 ' 3 9 The 3fi?orbital of the rf-block metal and the 6s orbital of the Ln(III) potentially interact more strongly with the ligand than the 4/magnetic orbitals of Gd(III). On the basis of Hund's rule, 4 3 the seven/electrons of the rare earth would be forced to align parallel to the Cu(II) electron. Kahn and co-workers proposed an alternative mechanism suggesting that the ferromagnetic exchange was a product of the interaction between the 4f-3d ground configuration (GC) and the metal-metal charge transfer configuration (CTC) shown in Figure 1.2.4 0'4 1 Two excited states for the Gd-Cu system, in which the total spin Sj = 3 for an antiferromagnetic exchange or 5T = 4 for a ferromagnetic exchange ( S o d = 7/2, Sbu = 1/2), 1 0 arise upon the electron-transfer from the 3d singly occupied orbital centered on Cu(II) to an empty 5d orbital centered on Gd(III). Both spin states are stabilized by the G C - C T C interactions; however, the latter exchange is lower in energy defined by Hund's rule, 4 3 and thus is significantly populated forming a ferromagnetic exchange. + 4 - 4 - 4 - 4 - 4 - 4 -Af 3d charge transfer configuration 5d ground configuration S = 3 S = 4 W 4 - 4 - 4 - 4 - 4 - + 4 - * - 4 - • Af ^ - 3 t f = 0 3d S = 4 J = S = 3 "S^A AU'2 - A 2 S = 3 S = 3 S = A (a) •(b) (c) Figure 1.2. A proposed explanation for the ferromagnetic interaction between Cu(II)-Gd(III): 4 0 (a) ground configuration (GC) and charge transfer configuration (CTC) ; (b) spin states of G C and C T C without interaction; (c) spin states of G C and C T C upon interaction. U ' is the energy gap between the excited and ground states, / ^ ^ i s the transfer integral for an electron moving from the 3d obital of Cu(II) to the 5d orbital of Gd(III). A is the energy separation between the two states generated by the Af5dx configuration of Gd(III). 11 A recent paper from Hirao and co-workers validated both Kahn's and Gatteschi's theories through theoretical and computational calculations, suggesting that both mechanisms are plausible; 4 4 however, the authors hypothesize that the actual exchange is more likely a mixture of these two current theories. Furthermore, Kahn and co-workers suggested that the overall interaction with Af from Ln(III) forms an antiferromagnetic exchange for n < 1 and a ferromagnetic exchange for n > 7 by assuming the spin-spin interactions are unaltered throughout the Ln(III) series (Figure 1.3). 4 0 ' 4 1 This theory arises from the interaction of the 3d electrons with the scalar products of the spin and orbital angular moment vectors of the Ln(III). The relative orientation of the spin and orbital moments is defined in Equation 1.5. X = ±(,I2S (1.5) C, is the single electron spin-orbit coupling constant that is always positive and ranges between 600 and 3000 cm" for Ln(III) ions, X is the spin-orbit term coupling constant and S is the total spin angular momentum quantum number. The spin-orbit term coupling constant can either possess a + value or a - value. The direction of the vector is determined by Equation 1.6 which defines the energy levels of the spin-orbit states J: E = (X/2)[J(J + \)-L(L + l)-S(S+ 1)] (1.6) J is the total angular momentum quantum number, S is the total spin angular momentum quantum number, L is the total orbital angular momentum quantum number, and X is the spin-orbit term coupling constant (defined in Equation 1.5). Because the energy of the J 12 states must always be positive, the spin-orbit term coupling constant must change signs to allow for a non-negative quantity. The X has a + value for electron shells that are less than V2 filled and takes on a negative value of subshells that are more than V2 filled. Only the J states with the lowest energy are significantly populated at RT. Thus, when elucidating the magnetic interactions of Ln(III)-containing species, only the energy of the lowest J state has to be considered. In order to obtain positive energy values for J and knowing that the spin-orbit term coupling constant for the 4/ 1 of Ln(III) with n < 7 is positive and the term coupling constant is negative for n > 7, 2 7 the J values must be calculated following Hund's third rule. Hund's third rule states that the J level with the lowest energy has the smallest value for ions with half-filled subshells. 4 3 In contrast, the lowest energy level of the spin-orbit states for subshells more than half-filled have the highest J value. Therefore, the energy states of interest are defined by J = L - S for Ln(III) with less than 7 / electrons and by J = L + S for Ln(III) with more than 7/electrons, resulting in positive energy values for J using Equation 1.6. The reversal of the moments for the L n configurations is thought to affect the magnetic exchange (Figure 1.3).45 Those Ln(III) with spin-orbit coupling where the L and S vectors are aligned in an antiparallel fashion (J= L-S) w i l l form antiferromagnetic interactions assuming the electron on the TM(II) w i l l align parallel with the spin vector of the Ln(III). Ln(III), where J - L + S exhibit a ferromagnetic exchange where the 3d electron of the TM(II) (that is being transferred to the 5d orbital of the Ln(III)) aligns in parallel with the 5 and L vectors from the 4/electrons in Ln(III). This theory defines the Kahn and co-workers prediction where the 4f from Ln(III) form ferromagnetic interactions when n > 7 and antiferromagnetic exchange where n < 7. 13 Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Ln 3+ C u 2 + Antiferromagnetic Interaction Ln 3 + - C u 2 + Ferromagnetic Interaction Ferromagnetic Interaction Figure 1.3. Angular momentum coupling of Cu(II)-Ln(III). 4 5 The S p a i r is the ferromagnetic coupling of the spin angular momenta from the Ln(III) and Cu(II) ions, S is the spin angular momenta for the 3d electron of the Cu(II) being transferred to the 5d orbital of the Ln(III) (based on Kahn's theory 4 0) and for the Ln(III) (based on Hund's rule 4 3 ) , and L is the orbital angular momentum. Due to the increased interest for new molecular-based magnetic materials, there has been a significant amount of research in synthesizing new (i/heterometallic TM(II)-Ln(III) complexes to better understand the intermetallic interactions; however, designing ligands that 14 c chelate both a d-block and/ : block transition metal in close proximity has been a large challenge in this field. Thus, the magnetic properties of Ln(III) still remains scantily investigated and poorly understood. 1.3 Med ic ina l Properties of Lanthanides 1.3.1 Medicinal characteristics Medicinal properties of lanthanides were first discovered in the mid-nineteenth century when Ln(III) were found to have anti-emetic properties. 4 6 The mode of action for the first therapeutic L n used, cerium oxalate, is still not understood although it is thought to work at the local level in the gastrointestinal tract. Nonetheless, insoluble cerium oxalate was used until the mid-twentieth century before being replaced with antihistamines. In the early twentieth century, Ln(III) salts were used for the treatment of tuberculosis and as anticoagulants;4 7 however, development was hampered due to severe side effects. These fortuitous discoveries have lead to several novel innovations regarding the biological relevance of lanthanides. Currently, three Ln(III)-containing therapeutic drugs have been approved for clinical use including cerium nitrate as a topical cream for the treatment of burn wounds , 4 8 ' 4 9 b 3 S m - E D T M P for palliative treatment of bone pa in , 5 0 ' 3 1 and lanthanum carbonate as a phosphate binder for the treatment of hyperphosphatemia in renal 52 dialysis patients. Several diagnostic Gd(III) magnetic resonance imaging (MRI) contrast agents 1 4 are also clinically used with many more Ln(III) drugs currently being developed. Ln(III) are primarily considered to be non-toxic when orally ingested due to the absence of active or facilitated transport mechanisms for the Ln(III) metal ions. 5 3 Thus, 15 Ln(III) cannot readily cross the cell membranes, specifically the gastrointestinal barrier. Endogenous concentrations of Ln(III) are found in the human body with detectable concentrations of 0.002 - 12.0 ug/L in blood and blood plasma, Ce(III) being the most abundant. N o specific physiological or biochemical functions of these metal ions are known. 5 4 Nonetheless, Ln(III) cations can be toxic when administered intravenously (IC50 = 0.1 mmol/kg for G d 3 + ) 5 5 mainly due to their similarities to Ca(II). Ln(III), known to be functional mimics of Ca(II) ions, share similar ionic radii (0.86 - 1.22 A-vs. 1.14 A, respectively), donor atom preferences (O > N > S), and almost identical coordination numbers in protein binding sites ( C N = 6 - 9) . 5 6 Because of the larger charge density, Ln(III) usually exhibit larger stability constants in protein binding sites than does Ca(II) and thus are more resistant to demetaliation. 5 7 In vitro, Ln(III) can readily replace Ca(II) ions, causing acute cellular toxicity by blocking voltage and receptor operated calcium channels. 4 7 Upon large doses of Ln(III), skeletal, smooth, and cardiac muscle contractions become inhibited causing decreased blood pressure followed by fatal cardiovascular collapse and pulmonary paralysis. 5 8 Ln(III) ions are also known to clear rapidly from blood and be redistributed to tissues, primarily the liver and bone, resulting in chronic toxicity associated with hepatotoxocity and edema. 4 Metal toxicity and pharmacology can be minimized or optimized by the chelation of metal ions with carefully chosen organic ligands. The thermodynamic and hydrolytic stability, as well as the kinetics of complexation/decomplexation, can be modified to provide desired resistance to metabolic demetaliation and ligand exchange with ubiquitous proteins and metal scavengers in vivo.53'59 The oral administration and effective biodistribution, as well as the passivation of the metal ions through cell membranes can also be tuned using 16 appropriate organic ligands. Necessary requirements for orally available and active agents include low molecular weight (< 600), neutral charge at physiological p H (~ 7.4), reasonable lipophilicity, and water so lub i l i ty . 5 9 , 6 0 High synthetic yields and nontoxic metabolic products also need to be considered. A burgeoning research effort is being conducted on developing new ligands that chelate specific metals in biological systems while imparting desired medicinal effects (Figure 1.4). Figure 1.4. Ligand precursors of well-known pharmaceuticals (Gd- and Lu-Tex, Xcytrin and Antrin®, Lutrin®, respectively; Gd(III) complex of MS-325, Vasovist™; [Gd(HP-D 0 3 A ) ( H 2 0 ) ] , ProHance®; 1 5 3 S m - E D T M P , Quadramet®; [Gd(DTPA)(H 2 0)] 2 " , Magnevist®). 17 Upon contrived chelation, the medicinal properties of Ln(III) can also be extended to radiopharmaceuticals. Several radioactive Ln(III) isotopes possess desirable physical characteristics and the necessary availability for medicinal use (Table 1.3).16 The half-lives of several Ln(III) ions are long enough for the radiopharmaceutical to be synthesized, transferred to the patient and elicit a desirable effect in the target organ (optimal i\a = 2- 12 h). The mode of decay of many Ln(III) isotopes also has medicinal relevance. Ionizing radiation in the form of a-particles, P-particles (most commonly) or Auger-electons is desired for therapeutic treatment.61 In addition, y emitters (80 - 200 K e V optimal) can be imaged by y-scintigraphy or single photon emission computed tomography (SPECT) . Alternatively, positron (|3+) emitters are required for positron emission tomography ( P E T ) . 6 2 High thermodynamic stability and kinetic inertness must be achieved to maintain the integrity of the chelating molecule upon radiolysis from the nuclide and to confer immunity to enzymolysis and exchange with other metals that are ubiquitous in the biological environment. 6 3 Uncomplexed radiometals can form insoluble hydroxides or phosphates in the aqueous biological environment. These compounds may then deposit in healthy tissues, such as bone, causing other medical ailments, including secondary cancers. 6 4 Using multidentate ligands that exhibit high stability constants (logK > 20.0) 6 3 allows for efficient drug delivery, maximized residence time of radioactivity at the target site, and minimal radiation damage to normal tissue—all important considerations in this field. Considerable research funds are currently being invested on developing new Ln(III) containing drugs. Many of the medicinal uses for Ln(III) w i l l be discussed in detail, including their diagnostic use for M R I and as spectroscopic probes. The three F D A -approved therapeutic Ln(III) pharmaceuticals w i l l also be highlighted, as well as other Ln(III) containing complexes for the treatment of cancer and autoimmune disorders. 18 Table 1.3. Potential lanthanide isotopes for use in diagnosis and therapy. 1 6 ' Isotope decay mode t l / 2 source 1 4 1 C e P" 32.5 d reactor 1 5 3 G d 7 241.6 d reactor 1 5 3 S m P\y 46.7 h reactor H 9 T b short range a-particles 4.15 h cyclotron . 6 i T b P" 6.91 d reactor 1 5 3 D y p + , E C a 6.3 h cyclotron 1 5 7 D y y, E C a 8 h cyclotron 1 6 5 D y P" 2.3 h cyclotron/reactor 1 6 6 H o P",y 26.8 h reactor 1 6 5 E r y 10.3 h reactor 1 6 9 E r low p", y 9.4 d reactor 1 7 1 E r low P", y 7.5 h reactor 1 6 7 T m y, E C a 9.6 d cyclotron 1 7 0 T m low p", y, E C a 128.6 d reactor . 6 9 Y b y , E C a 32 d reactor/cyclotron , 7 7 L u P",y 6.7 d reactor 1 7 6 m L u y 3.7 h reactor 'EC = Electron Capture radiation 19 1.3.2 Diagnostic uses Ln(III) have mainly been used for diagnosis, specifically as M R I contrast agents.1 4 Images of organs and tissues can be obtained essentially by acquiring a ] H N M R spectrum of appropriate parts of the human body. The contrast agents help detect problems by distinguishing differences between signals from protons in healthy and diseased tissue. 1 2 G d 3 + , the metal ion commonly utilized in M R imaging, has a large paramagnetic moment due to its seven unpaired electrons (4/ ) , and imparts short longitudinal relaxation times (7/.) to bound water molecules. Currently, 30 - 35 % of all M R I studies utilize (Gd(III)-based) contrast agents to optimize clinical imaging capabilities (Table 1.4). 1 4 ' 6 8 Since the development of these agents, it is estimated that over 30 metric tons of Gd(III) has been administered to millions of patients worldwide. 7 2 The first M R I contrast agent, approved for clinical applications in 1988, consisted of the octadentate complex, [Gd(DTPA)(H20)] 2 ", where D T P A = diethylenetriaminepentaacetic acid, marketed as Magnevist® (Figure 1.4).69 However, kinetic lability and low proton relaxivity plagued Magnevist®, requiring modifications in the backbone, as well as in the chelating pendant groups of the ligand. Macrocylic polyaza-based polycarboxylate proligands have been used to form thermodynamically stable products by taking advantage of the macrocyclic effect. 7 0 One such agent is [Gd(HP-D03A) (H 2 0) ] , where H P - D 0 3 A = 1,4,7- tris(carboxymethyl)-10-(2'-hydroxypropyl)-1,4,7,10-tetraazacyclododecane, an F D A approved cyclic Gd(III) contrast agent marketed as ProHance® (Figure 1.4). 20 Table 1.4. M R I contrast agents approved for clinical use in Canada. 7 1 ' Complex Generic name Brand name [Gd(DTPA)(H 2 0) ] 2 - gadopentetate dimeglumin a ' b Magnevist® [ G d ( D T P A - B M A ( H 2 0 ) ] gadodiamide 3 ' 0 Omniscan™ [ G d ( D T P A - B M E A ( H 2 0 ) ] gadoversetamidea'd OptiMARK® [Gd(HP-D03A)(H 2 0) ] gadoteridol 3 ' 6 ProHance® [Gd(D03A-butrol)(H 2 0)] gadobutrol Gadovist® [Gd(BOPTA)(H 2 0) ] 2 " gadobenate dimeglumine 3 ' 8 MulitHance® MS-325 gadophostriamine trisodium h Vasovist™ a F D A approved b DTPA = diethylenetriaminepentaacetic acid C D T P A - B M A = diethylenetriaminepentaacetic acid bis(methylamide) d D T P A - D M E A = diethylenetriaminepentaacetic acid di(methoxyethylamide) e BOPTA = benzyloxypropionictetraacetate f HP-D03A = l,4,7-tris(carboxymethyl)-10-(2'-hydroxypropyl)-l,4,7,10-tetraazacyclododecane gD03A-butrol = l,4,7-tris(carboxymethyl)-10-(l-(hydroxymethyl)-2,3-dihydroxypropyl)-1,4,7,10-tetraazacyclododecane hMS-325 = l-(R)-(4,4-diphenylcyclohexylphosphorylmethyl)diethylenetriamine pentaacetic acid, trisodium salt Current research in M R I contrast agents involves adding directing groups to the chelating portion to allow accumulation of the contrast agents in specific tissues of interest. 7 3 For example, the newest Gd(III) contrast agent approved in Canada, Vasovist™, consists of a modified D T P A ligand with one attached lipophilic diphenylcyclohexyl group that reversibly binds to the human blood protein serum albumin (Figure 1.4).18 This functionalized group allows for imaging of the blood vessels approximately one hour after administration. Research groups have had success imaging biological functions by adding 21 M R I contrast agents that sense biochemical environments through enzyme-induced relaxitivity changes or changes due to the concentration of a particular substance of interest.1 The characteristic absorptions, as well as the narrow emission bands and long life-times of the excited states (ranging from ps - ms) of the luminescence spectra also make Ln(III) attractive for use as spectroscopic probes. 5 7 ' 7 4 Eu(III) and Tb(III) ions, with suitable energy gaps between the lowest emission level and the ground state, are commonly used to identify donor groups and metal binding sites in proteins. 2 1 Ln(III) have specifically been instrumental in probing the binding sites and biological functions of the spectroscopically silent Ca(II) ion in calmodulin, trypsin, and parvalbumin due to the Ln(III) luminescence being sensitive to the surrounding environment. Ln(III) have also been useful for the examination and determination of nucleic acids and as analytical probes, mainly as fluoroimmunoassays, to detect small concentrations of enzymes, hormones, steroids, and various drugs by utilizing appropriate antibodies bound to luminescent Ln(III) i ons . 5 7 ' 7 3 Several Ln(III) radioisotopes are also currently being investigated for diagnostic purposes including 1 4 1 C e , 1 5 3 S m , l 5 3 G d , 1 6 1 T b , 1 5 7 D y , 1 6 7 T m , 1 6 9 Y b and 1 7 7 L u . When attached to microspheres, 1 4 *Ce and 1 5 3 Gdare both attractive y-emitters for the imaging of cerebral blood flow and acute myocardial necrosis, respectively. 1 6 1 6 1 T b - D T P A - o c t r e o t i d e , 7 6 1 6 7 T m -and 1 6 9 Yb-ci t ra te 7 7 have shown promise for intra-operative scanning of tumors and cancer diagnosis. 1 6 7 T m - and 1 6 9Yb-citrate, as well as 1 5 7 D y - H E D T A , ( H E D T A = hydroxyethylenediaminetetraacetic acid), 7 8 1 5 3 S m - , 1 7 7 L u - , 1 5 7 Dy-gluconate, 1 6 and 1 5 3 S m - , 1 6 6Ho-tetraaza macrocycles 7 9 ' 8 0 with a variety of pendant arms selectively accumulate in bone making excellent reagents for skeletal applications. 22 1.3.3 Approved therapeutic agents Cerium nitrate, Flammacerium K , is commonly used to treat burn wounds as a convenient and cost effective alternative to sophisticated skin grafts. 4 7 The initial rationale for using Ln(III) for the treatment of burn wounds was based on the antibacterial effect of several cerium salts including acetate, stearate, chloride, and nitrate, as observed in the late nineteenth century. Combination therapy with silver sulfadiazine, another metal-based pharmaceutical, demonstrated a 50% reduction in the mortality rate of burn victims 4 8 It was first thought that the efficacy was due to the synergistic antimicrobial interactions between the two agents, but research has since suggested otherwise. Currently, it is believed that the topical application of cerium nitrate leading to a firm impermeable biological dressing, not only prevents bacterial infections, but also inhibits the burn-associated immune response allowing for the wound to stay in a clean, healthy state, while awaiting a skin graft 4 9 Lanthanum carbonate, marketed as Fosrenol®, is the latest F D A approved metal-based, drug. Fosrenol® is a phosphate-binding agent for the treatment of hyperphosphatemia common in chronic kidney disease pateints. 4 7 ' 5 2 Without effective treatment, hyperphosphatemia can lead to several severe pathological consequences including hyperparathyroidism, a collection of bone diseases characterized by bone pain, brittle bones, skeletal deformities and fractures, and lethal cardiac and vascular tissue calcification. 8 1 Lanathum carbonate has been shown to effectively bind phosphate with a high affinity and retain the bound phosphate over rigorous gastrointestinal p H ranges. 8 2 The insolubility of LaPC>4 allows for effective removal of phosphate with minimal tissue uptake of the Ln(III). 8 3 85 23 153 S m - E D T M P , the only lanthanide containing radiopharmaceutical currently on the market, under the commercial name Quadramet®, is used to treat cancerous tumours that have metastasized to bone causing inflammation and painful disruption of the periostium (Figure 1.4).61 1 5 3 S m , exhibiting a reasonable half-life of 1.9 d and a maximum tissue penetration of 3.1 mm, 5 1 is strongly chelated by a multidentate polyaminophosphonate ligand (logK = 22.4) 8 6 and has significant uptake by bone. The selective uptake is thought to be mediated by interactions between the phosphonate groups and hydroxyapatite where the E D T M P ligand bridges between the calcium found in the hydroxyapatite and the chelated 153 87 Sm(III). Bone is ablated with appropriate concentrations of p-radiation from decay of 153 86 the proximal Sm resulting in a 70 - 75 % response rate amongst the patients. 1.3.4 Anticancer agents La(III), Ce(III) and Nd(III) complexed with coumarins have demonstrated anti-proliferative activity on various cancer cell lines. 4 7 Tb(III) has also been shown to enhance cytotoxicity of cisplatin; 1 7 however, the most promising anti-cancer agents that have oo progressed into clinical trials are the Gd(III)- and Lu(III)-texaphyrins (Figure 1.4) and 1 7 7 L u - C C 4 9 . 8 9 A redox active Gd(III)-texaphyrin (Gd-tex, Xcytrin®) complex is in phase III clinical trials for treatment of brain metastases from non-small cell lung cancer. 4 7 Texaphyrins, first synthesized by Sessler and co-workers, are expanded pentaaza porphyrins, consisting of a 22 re-conjugated macrocycle with five coordinating nitrogens. 9 0 The large central cavity can easily accommodate large metal ions such as lanthanides, whereas the water solubility of the molecule can easily be modified by the addition of polyethylene glycol end groups. 9 1 The 24 activated Gd-tex, by pulse radiolysis 2-5 h prior to injection of the metallopharmaceutical, initiates a two electron reduction of 0 2 producing reactive oxygen species (ROS), such as two superoxide anions, hydrogen peroxide (H2O2) and hydroxyl radicals (OH«), while simultaneously undergoing a two electron oxidation of ascorbate to dehydroascorbate.9 2 The R O S species cause necrosis of the cell by interacting with different cellular components, namely by inducing oxidative damage of D N A . Lu-texaphyrin (Lu-tex), ideal for photodynamic therapy (PDT), is being investigated as a photo-activated agent for photoangioplasty (a treatment of artherosclerotic plaques in coronary heart disease, Antrin®), breast cancer (Lutrin®) and treatment of age-related macular degeneration.' P D T is a type of therapy where photosensitized molecules are injected into the body, allowed to accumulate in the targeted tissue, and then irradiated with 93 light. Upon excitation of a certain wavelength of light via an optical fiber, the photosensitized molecules become active, producing cytotoxic singlet oxygen in the targeted cells that is thought to damage D N A , resulting in cell death. The treatment is especially effective for localized disorders. Lu-tex is a promising candidate exhibiting several advantages over current P D T agents.9 4 The most attractive property of Lu-tex is its ability to absorb light in the far-infrared region (732 nm), allowing for greater tissue penetration and less competitive absorption from natural chromophores, namely hemoglobin. Several bifunctional chelators (BFC) incorporating Ln(III) have also been developed to treat ailments including cancer. 6 3 ' 9 5 B F C function by adding a chelating portion to the directing group of the pharmaceutical, thereby forming target specific agents that reduce the probability of ablating healthy tissue. One particularly promising biolabeled metalloradiopharmaceutical is 1 7 7 L u - C C 4 9 . The radiometal 1 7 7 L u is linked to P A - D O T A (1,4,7,10-tetraaza-/V-(l -carboxy-3-(4-nitrophenyl)propyl)-7v", N' ',N" '-fra(acetic 25 acid)cyclododecane). The metal decays to produce P" radiation that penetrates up to 0.3 mm, ideal for treating small ovarian tumors. 8 9 In addition, the use of the octadentate polyaminocarboxylate ligand forms a thermodynamically stable and kinetically inert radiometal complex ( logK= 25.4). The amino group on P A - D O T A is easily labeled with biomolecules, specifically a monoclonal antibody CC49 that targets the extracellular glycoprotein T A G - 7 2 , commonly expressed on the outside of ovarian cancer cel ls . 9 6 l 7 7 L u -CC49 is currently undergoing phase III clinical trials and has exhibited a 73 - 75 % effectiveness rate. 9 7 1.3.5 Treatment of autoimmune disorders Ln(III) have also shown promise as inhibitors of undesired autoimmune responses that cause liver toxicity, lung cell damage and rheumatoid arthritis. 4 7 Kupffer cells are resident macrophages in the liver and the lung that release toxic cytokines and free radicals upon activation, which cause organ damage. In the case of liver cirrhosis, the severity of the alcohol induced liver damage is directly correlated with the activity of these cel ls . 9 8 " 1 0 0 Kupffer cells not only k i l l healthy hepatocytes, but also stimulate proliferation and production of collagen in stellate cells, believed to play an important role in the development of fibrosis. It has been demonstrated that GdCfj can deactivate, and even destroy, Kupffer cells blocking early liver injury during chronic exposure to alcohol 4 7 GdCl3 has also been shown to attenuate organ damage from ischemic injury sustained during aortic cross-clamping performed in aortic surgery. 1 0 1 Blood supply to the lower extremities is commonly blocked during the surgical procedure leading to tissue damage that predisposes cardiac patients to lung injury. GdCl3 is believed to relieve cellular damage in 26 the lungs similarly to that in the liver cells where the activity of the Kupffer cells is decreased, leading to depressed levels of inflammatory cytokines. 1 0 1 Rheumatoid arthritis, argued to be one of the most painful and debilitating forms of arthritis, is an autoimmune disorder in which the immune system begins to attack the healthy tissue around the joint cavi ty . 1 0 2 The synovial membrane that lines most joint cavities throughout the body becomes inflamed and begins to thicken forming a tissue mass that initiates damage to the cartilage and fibrous tissue surrounding the joint. This painful autoimmune disorder eventually leads to joint debilitation. Radiosynovectomy is a procedure in which radioactive isotopes, including lanthanide radioisotopes ( 1 6 6 H o , l 6 5 D y , and 1 5 3 Sm) , are placed in the body and are directed to ablate the synovial lining. Currently, radiosynovectomy is being investigated as an alternative type of treatment to that in current use due to its potential for repeated use. 1 0 2 " 1 0 5 The nuclides deliver radiation to the synovial membrane, inducing the formation of free radicals. These highly reactive species begin to damage different cellular components including the D N A . The main concern for these agents is leakage out of the joint cavity causing accumulation and subsequent ablation of healthy tissue throughout the body. 1 0 6 To circumvent these problems, lanthanide radioisotopes have been attached to particles, specifically microspheres and colloids of specific size, that impede leakage. 1.4 Thesis Overview The diverse properties of Ln(III) stressed in this introduction are further explored throughout this thesis. Both the magnetic properties and the therapeutic medicinal properties of Ln(III) wi l l be investigated. 27 The magnetic interactions in lanthanide-containing materials are still not well-understood due to their complicated intrinsic properties. In order to elucidate the magnetic exchange between TM(II) and Ln(III) ions, several new t^heterometallic complexes have been synthesized using multidentate amine phenol ligands. These ligands coordinate divalent first row d-block transition metal ions (TM) and lanthanides (Ln) in close proximity, which is desirable for magnetic studies. The synthesis, characterization and magnetic analysis of these d/f complexes wi l l be presented in Chapter 2. In addition, synthesis of the T M L and the L n L species using the appropriate ligand, L , to probe the core structure surrounding the metal ions wi l l be outlined. Chapter 3 details the use of Ln(III) for bone density disorders. In an effort to find new orally active agents for these disorders, a series of Ln(III) containing complexes incorporating small, non-toxic, bidentate pyrone and pyridinone ligands have been synthesized and characterized. Preliminary biological analyses completed on these complexes wi l l also be discussed in Chapter 3, including cytotoxicity, cell uptake, bidirectional transport studies in Caco-2 cells and in vitro hydroxyapatite (HA) binding studies. Chapter 4 contains final conclusions and suggestions for future work on both the magnetic and medicinal projects summarized in this thesis. 28 1.5 References 1. Cotton, S., Lanthanide and Actinide Chemistry. John Wiley & Sons, Ltd.: Uppingham, U K , 2006. 2. Katlsoyannis, N . ; Scott, P., The f elements. Oxford University Press: New York, 1999. 3. Greenwood, N . N . ; Earnshaw, A . , Chemistry of the Elements. Pergamon Press: New York, 1984. 4. Palasz, A . ; Czekaj, P. Acta Biochim. Pol. 2000, 47, 1107. 5. Evans, C. H . , Biochemistry of the Lanthanides. Plenum Press: N e w York, 1990. 6. Choppin, G . R., Lanthanide Probes in Life, Chemical and Earth Sciences. Elsevier: Amsterdam, 1989. 7. Bunzl i , J.-C. G . ; Piguet, C. Chem. Rev. 2002,102, 1897. 8. Bunzl i , J : -C . G . ; Choppin, G . R., Lanthanide Probes in Life, Chemical and Earth Sciences. Elsevier: New York, 1989. 9. Shannon, R. D . Acta Cryst. 1976, A32, 751. 10. Rizkalla, E . N . ; Choppin, G. R., Handbook on the Physics and Chemistry of Rare Earths. Elsevier Science: Amsterdam, 1991. 11. Sigel, A . ; Sigel, H . , Metal Ions in Biological Systems. Marcel Dekker: New York, 2003; V o l . 40. 12. Pearson, R. G . J. Am. Chem. Soc. 1963, 85, 3533. 13. Mikami , K . ; Terada, M . ; Matsuzawa, H . Angew. Chem. Int. Ed. 2002, 41, 3554. 14. Caravan, P.; El l ison, J. J.; McMurry , T. J.; Lauffer, R. B . Chem. Rev. 1999, 99, 2293. 15. Bayly, S. R.; X u , Z . ; Patrick, B . O.; Rettig, S, J.; Pink, M . ; Thompson, R. C ; Orvig, C. Inorg. Chem. 2003, 42, 1576. 16. Nayak, D. ; Lahir i , S. J. Radioanal. Nucl. Chem. 1999, 242, 423. 17. Kostova, I. Curr. Med. Chem. 2005, 4, 591. 18. Woods, M . ; Kovacs, Z . ; Sherry, A . D . J. Supermolec. Chem. 2002, 2, 1. 19. Thunus, L . ; Lejeune, R. Coord. Chem. Rev. 1999, 184, 125. 29 20. Sleight, A . W. Acc. Chem. Res. 1995, 28, 103. 21. Parker, p . ; Will iams, J. A . G . J. Chem. Soc, Dalton Trans. 1996, 3613. 22. de Bettencourt-Dias, A . Dalton Trans. 2007, 2229. 23. Sakamoto, M . ; Manseki, K . ; Okawa, H . Coord. Chem. Rev. 2001, 219-221, 379. 24. Coey, J. M . D. Endeavour 1995,19, 146. 25. Benelli , C ; Gatteschi, D. Chem. Rev. 2002, 102, 2369. 26. Hughes, I. D. ; Dane, M . ; Ernst, A . ; Hergert, W.; Ltiders, M . ; Poulter, J.; Staunton, J. B . ; Svane, A . ; Szotek, Z. ; Temmerman, W. M . Nature 2007, 446, 650. 27. Earnshaw, A . , Introduction to Magnetochemistry. Academic Press: New York, 1968. 28. Kahn, O. Adv. in lnorg. Chem. 1995, 43, 179. 29. Costes, J.-P.; Dahan, F.; Dupuis, A . ; Laurent, J.-P. Inorg. Chem. 2000, 39, 169. 30. Lescop, C ; Luneau, D. ; Rey, P.; Bussiere, G . ; Reber, C. Inorg. Chem. 2002, 41, 5566. 31. Bencini, A . ; Benelli , C ; Caneshi, A . ; Dei , A . ; Gatteschi, D . Inorg. Chem. 1986, 25, 572. 32. Costes, J.-P.; Dahan, F.; Dupuis, A . ; Laurent, J.-P. Chem. Eur. J. 1998, 4, 1616. 33. Shiga, T.; Ito, N . ; Hidaka, A . ; Okawa, H . ; Kitagawa, S.; Ohba, M . Inorg. Chem. 2007, 46, 3492. 34. Yamaguchi, T.; Sunatsuki, Y . ; Kojima, M . ; Akashi, H . ; Tsuchimoto, M . ; Re, N . ; Osa, S.; Matsumoto, N . Chem. Commun. 2004, 1048. 35. Kahn, M . L . ; Lecante, P.; Verelst, M . ; Mathoniere, C ; Kahn, O. Chem. Mater. 2000,12, 3073. 36. X u , Z . ; Read, P. W. ; Hibbs, D. E . ; Hursthouse, M . B . ; Mal ik , K . M . A . ; Patrick, B . O.; Rettig, S. J.; Seid, M . ; Summers, D. A . ; Pink, M . ; Thompson, R. C ; Orvig, C. Inorg. Chem. 2000 ,39,508. 37. Cu i , Y . ; Chen, G ; Ren, J.; Quian, Y . ; Huang, J. Inorg. Chem. 2000, 39, 4165. 38. Benelli , G . ; Caneschi, A . ; Gatteschi, D . ; Guil lou, O.; Pardi, L . Inorg. Chem. 1990, 29, 1750. 39. Benelli , C ; Caneschi, A . ; Fabretti, A . C ; Gatteschi, D . ; Pardi, L . Inorg. Chem. 1990, 29, 4153. 40. Kahn, O.; Guil lou, O. Research Frontiers in Magnetism; O'Connor, C. J., Ed. ; World Scientific: London, 1993. 30 41. Andruh, M ; Ramade, I.; Codjovi, E . ; Guil lou, O.; Kahn, O.; Trombe, J. C. J. Am. Chem. Soc. 1993,115, 1822. 42. Goodenough, J. B . , Magnetism and the Chemical Bond. Interscience: New York, 1963. 43. Hund, F. Z. Physik 1925, 33, 345. 44. Paulovic, J.; Cimpoesu, F.; Ferbinteanu, M . ; Hirao, K . J. Am. Chem. Soc. 2004,126, 3321. 45. Kahn, M . L . ; Mathoniere, C ; Kahn, O. Inorg. Chem. 1999, 38, 3692. 46. Jakupee, M . A . ; Unfried, P.; Keppler, B . K . Rev. Physiol. Biochem. Pharmacol. 2005,153, 101. 47. Fricker, S. P. Chem. Soc. Rev. 2006, 35, 524. 48. Garner, J. P.; Heppell, P. S.Burns 2005, 31, 539. 49. de Gracia, C. G . y3wmy 2001, 27, 67. 50. Resche, I.; Chatal, J.; Peching, A . ; E l l , P.; Duchesne, G. ; Rubens, R.; Fogelman, I.; Houston, S.; Frauser, A . ; Fischer, M . ; Wilkins, D. Europ. J. Cancer 1997, 33, 1583. 51. Goeckeler, W. ; Edwards, B . ; Volkert, W.; Holmes, R.; Simon, J.; Wilson, D. J. Nucl. Med. 1987, 28, 495. 52. Abrams, M . J.; Bridger, G . J.; Fricker, S. P.; Idzan, S. R. World patent WO 02/00227 A. 2002. 53. Thompson, K . H . ; M c N e i l l , J. H . ; Orvig, C. Chem. Rev. 1999, 99, 2561. 54. Buseth, E . ; Wibetoe, G . ; Martinsen, I. J. Analyt. Atom. Spectrometry 1998, 13, 1039. 55. Pierre, V . C ; Melchior, M . ; Doble, D. M . J.; Raymond, K . M . Inorg. Chem. 2004, 43, 8520. 56. Pidcock, E . ; Moore, G . R. J. Biol. Inorg. Chem. 2001, 6, 479. 57. Elbanowski, M . ; Makowska, B : J. Photochem. Photobiol. A 1996, 99, 85. 58. Coirault, C ; Sauviat, M . P.; Chemla, D. ; Pourny, J. C. ; Lecarpentier, Y . Eur. Respir. J. 1999, 14, 1297. 59. Thompson, K . H . ; Orvig, C. Coord. Chem. Rev. 2001, 219-221, 1033. 60. Lipinski , C . A . ; Lombardo, F.; Dominy, B . W. ; Feeney, P. J. Adv. Drug Delivery Rev. 1997, 23, 4. 61. Volkert, W. A . ; Hoffman, T. J. Chem. Rev. 1999, 99, 2269. 62. Anderson, C. J.; Welch, M . J. Chem. Rev. 1999, 99, 2219. 31 63. Fichna, J.; Janecka, A . Bioconjugate Chem. 2003 , 1 4 , 3. 64. Heeg, M ; Jurisson, S. Acc. Chem. Res. 1 9 9 9 , 32, 1053. 65. Welch, M . J.; Reduanly, C. S., Handbook of Radiopharmaceuticals. John Wiley & Sons: London, 2003. 66. M a , D. ; Ketring, A . R.; Ehrhardt, G . J.; Jia, W. J. Radioanal. Nucl. Chem. 1 9 9 6 , 206, 119. 67. Burgess, J. Chem. Soc. Rev. 1 9 9 6 , 25, 85. 68. Lane, B . S.; Burgess, K . Chem. Rev. 2003,1 0 3 , 2457. 69. Laniado, M . ; Weinmann, H . J.; Schorner, R. F.; Speck, U . Physiol. Chem. Phys. Med. NMR 1 9 8 4 , 16, 157. 70. Guoxin, T.; Yongjun, Z. ; Jingming, X . ; Ping, Z . Inorg. Chem. 2 0 0 3 , 42, 735. 71. Health Canada, http://www.hc-sc.gc.ca. accessed M a y 28, 2007. 72. Caravan, P.; Ell ison, J. J.; McMurry , T. J.; Lauffer, R. B . Chem. Rev. 1 9 9 9 , 99, 2293. 73. Alves, F. C ; Donato, P.; Sherry, A . D. ; Zaheer, A . ; Zhang, S.; Lubag, A . J. M . ; Merritt, M . E. ; Lenkinski, R. E. ; Frangioni, J. V . ; Neves, M . ; Prata, M . I. M . ; Santos, A . C ; Lima, J. J. P. d.; Geraldes, C. R. G . C. Invest. Radiol. 2 0 0 3 , 38, 750. 74. Parker, D . ; Dickins, R. S.; Puschmann, H . ; Crossland, C ; Howard, J. A . K . Chem. Rev. 2 0 0 2 , 102, 1977. 75. Greenbaum, N . L . ; Mundoma, C ; Peterman, D . R. Biochemistry 2 0 0 1 , 40, 1124. 76. Gatteschi, D. ; Sessoli, R. Agnew. Chem., Int. Ed. 2 0 0 3 , 42, 268. 77. Hisida, K . ; Ando, A . J. Nucl. Med. 1973, 1 4 , 615. 78. Subramanian, G . ; McAfee , J. G . ; Blair, R. J.; O'Mara, R. E . ; Greene, M . W.; Lebowitz, E . J. Nucl. Med. 1 9 8 4 , 25, 930. 79. Marques, F.; Guerra, K . P.; Gano, L . ; Costa, J.; Campello, M . P.; Lima, L . M . P.; Delgado, R.; Santos, I. J. Biol. Inorg. Chem. 2 0 0 4 , 9, 859. 80. Margues, F.; Gano, L . ; Campello, M . P.; Lacerda, S.; Santos, I.; L ima, L . M . P.; Costa, J.; Antunes, P.; Delgado, R. J. Inorg. Biochem. 2006,1 0 0 , 270. 81. Albaaj, F.; Hutchison, A . J. Drugs 2 0 0 3 , 63, 577. 82. Pennick, M . ; Dennis, K . ; Damment, S. J. P. J. Clin. Pharmacol. 2 0 0 6 , 46, 738. 83. Damment, S. J. P.; Greaves, P.; Downes, N . J. Am. Soc. Nephrol. 2 0 0 3 , 14, F-P0651. 32 84. Damment, S. J. P.; Beevers, C ; Gatehouse, D. G . Mutagensis 2005, 20, 29. 85. D'Haese, P. C ; Spasovski, G . B . ; Sikole., A . Kidney Int. Suppl. 2003, 85, S73. 86. Goeckeler, W. ; Troutner, D. ; Volkert, W. ; Edwards, B . ; Simon, J.; Wilson, D . Nucl. Med. Biol. 1986,13, 479. 87. Volkert, W. ; Deutsch, E . , Advances in Metals in Medicine. J A I Press: New York, 1993. 88. Sessler, J.; Mil ler , R. Biochem. Pharm. 2000, 59, 733. 89. Meredith, R.; Patridge, E . ; Alvarez, R.; Khazaeli, M . ; Plott, G . ; Russell, C ; Wheeler, R.; L i u , T.; Grizzle, W.; Schlom, J.; LoBuglio, A . J. Nucl. Med. 1996, 37, 1491. 90. Sessler, J.; Hemmi, G . ; Mody, T.; Murai , T.; Burrell, A . ; Young, S. Acc. Chem. Res. 1994, 27, 43. 91. Sessler, J.; Murai , T.; Hemmi, G. Inorg. Chem. 1989, 28, 3390. 92. Magda, D . ; Lepp, C ; Gerasimchuk, N . ; Lee, I.; Sessler, J.; L i n , A . ; Biaglow, J.; Mil ler , R. Int. J. Rad. Oncol. Biol. Phys. 2001, 51, 1025. 93. Kosenich, G . ; Babushkina, T.; Lavi , A . ; Langzam, Y . ; Mal ik , Z . ; Orenstien, A . ; Ehrenberg, B . J. Porphyrins Phthalocyanines 1998, 2, 383. 94. Sessler, J.; Tvermoes, N . ; Guldi , D. ; Mody, T. J. Porphyrins Phthalocyanines 2001, 5, 593. 95. L i u , S.; Edwards, D. S. Bioconjugate Chem. 2001, 72, 7. 96. Chappell, L . ; Rogers, B . ; Khazaeli, M . ; Mayo, M . ; Buchsbaum, D. ; Brechbiel, M . Bioorg. Med. Chem. 1999, 7,2313. 97. Alvarex, R.; Partridge, E . ; Khazaeli, M . ; Plott, G. ; Autin, M . ; Kilgore, L . ; Russel, C ; L i u , T.; Brizzle, W. ; Schlom, J.; LoBugl io , A . ; Meredith, R. Gynec. Oncol. 1997, 65, 94. 98. Koop, D. R.; Klopenstein, B . ; Iimuro, Y . ; Thurman, R. G . J. Pharmacol. Exp. Ther. 1997, 51, 944. 99. Harstad, E . B . ; Klassen, C. D. Toxicol. Appl. Pharmacol. 2002,180, 178. 100. Rivera, C. A . ; Bradford, B . U . ; Hunt, K . J.; Adachi, Y . ; Schrum, L . W. ; Koop, D. R.; Burchardt, E.-R. ; Rippe, R. A . ; Thurman, R. G . Am. J. Physiol. Gastrointest. Liver Physiol. 2001, 257, 200. 101. Okutan, H . ; Savas, C ; Ozguner, I. F.; Yonden, Z . ; Eren, V . C ; Delibas, N . J. Exp. Med. 2004, 203, 267. 102. Song, J.; Suh, C. H . ; Park, Y . B . ; Lee, S. H . ; Yoo , N . C ; Lee, J. D . ; K i m , K . H . ; Lee, S. K . Eur. J. Nucl. Med. 2001, 28, 489. 33 103. Sledge, C. B . ; Zuckerman, J. D. ; Zalutsky, M . R.; Atcher, R. W.; Shortkroff, S.; Lionberger, D. R.; Rose, H . A . ; Hurson, B . J.; P. A . Lankenner, J.; Anderson, R. J.; Bloomer, W. A . J. Am. Rheum. Assoc. 1 9 8 6 , 29, 153. 104. . Louw, W. K . A . ; Dormehl, I. C ; van Rensburg, A . J.; Hugo, N . ; Alberts, A . S.; Forsyth, O. E. ; Beverley, G . ; Sweetlove, M . A . ; Marias, J.; Lotter, M . G . ; van Aswegen, A . Nucl. Med. Biol. 1 9 9 6 , 23, 935. 105. Davis, M . A . ; Chinol, M . J. Nucl. Med 1 9 8 9 , 30, 1047. 106. Unni , P. R.; Chaudhari, P. R.; Venkatesh, M . ; Ramamoorthy, N . ; P i l la i , M . R. A . Nucl. Med. Biol. 2002,29,199. 34 CHAPTER 2 d/f Amine Phenol Complexes: Synthesis and Magnetic Characterization1 2.1 Introduction With the increased interest in new molecular-based magnetic materials, elucidating the magnetic interaction between d/f transition metals has become imperative. 1 ' 2 Many obstacles, including weak (^interactions and large anisotropic effects in lanthanide ions (except La(III), Gd(III) and Lu(III)), have slowed the progress of research conducted on the magnetic exchange between transition metal (TM) and lanthanide (Ln) ions. Thus, few systematic studies on the structure and magnetic properties of d/f systems have been completed. Most magnetic studies on heterometallic d/f complexes have been confined to Cu-Gd systems first studied by Gatteschi in 1985 (Figure 2.1). 3 Gd(III), a 8S7/2 ion, has no first-order orbital momentum in its single-ion ground state. Thus, a simple analysis based on a traditional spin-only Hamiltonian exchange model can quantify the magnetic exchange between Cu(II) and Gd(III) metal ions. The majority of reported Cu-Gd complexes have been synthesized using salen type ligands. These proligands are commonly formed via a 2:1 condensation of 3-* Portions of this chapter can be found in the two submitted manuscripts. Barta, C. A . ; Bayly, S. R.; Read, P. W. ; Patrick, B . O.; Thompson, R. C ; Orvig, C. Inorg. Chem. Submitted August 2007. 35 formylsalicyclic acid or 3-(hydroxy or methoxy)salicylaldehyde and various diamines (Figure 2.1) . 4 - 1 0 Upon reacting these ligands with appropriate stoichiometric amounts of metal ions, isolated dimetallic or trimetallic d/f compounds are formed; these incorporate dissimilar N 2 0 2 and O 4 metal binding s i tes ." ' 1 2 The softer N 2 0 2 pocket is attractive to d-block metals, while the O 4 pocket incorporating two bridging phenolato oxygen donors selectively binds the Ln(III) ion. Other isolated Cu-Gd species with different geometric demands (selected examples shown in Figure 2.1) ~ include Matsumoto and co-workers' cyclic cylindrical C u 2 G d 2 complex utilizing the ligands l-(2-hydroxybenzamide)-2-(2-hydroxy-3-methoxybenzylideneamino)ethane and hexafluoroacetylacetone.1 8 A trinuclear Cu-Gd-Cu single chain magnet has also been formed using 2-hydroxy-iV-benzamide. 1 9 Other groups, using oxamidates or oxamides, have produced GdCus clusters by reacting Cu-JV,/V'-bis(3-aminopropyl)oxamididate with Gd(N03)3 in a 4:1 ratio. 2 0 Pecoraro's group was successful in synthesizing a metallocrown by using picoline hydroxamate to ligate five Cu(II) and one central Ln(III) in a pentagonal-planar arrangement, where the Ln(III) was varied (Figure 2.1). 2 1 The coordination sphere of Cu(II) in these species range from four-coordinate almost planar, to distorted tetrahedral, to five- or six-coordinate species, with a range of intermetallic T M - L n distances (3.198 4 - 3.649 1 6 A). Regardless of the geometry around the Cu(II) ion or the proximity of the d/f metal ions, all but one of the Cu-Gd species exhibited a ferromagnetic exchange. Unfortunately, a crystal of the antiferromagnetic trimetallic p-phenolato-u-oximato Cu-Gd-Cu system synthesized by Costes et al. has not been isolated, 36 and thus the geometric demands of the system possibly effecting the magnetic interaction cannot be elucidated 22,23 Schiff base [ (CuHAPen) 2 Gd(H 2 0) 3 ] + 3 cationic complex, HAPen = iV.jV-ethylenebis-(o-hydroxyacetophenomeiminato)3 Ladder structure o t 'Ln 2 [Cu(L ) ]3» 23 H 2 0 , L = 1,3-propylenebis (oxamato) 2 6 ' 2 8 bridged by picoline hydroxamic acid and N 0 3 " groups2 1 Tetranuclear [CuLCu(hfac) 2] 2 , 1 8 hfac = hexafluoroacetylacetone, L = I -(2-hydroxybenzamido)-2-(2-hydroxy-3-methoxybenzylideneamino)ethane Trinuclear [Cu 2 Ln(L) 2 ] complex, 3 1- 3 2 L = 2,6-di(acetoacetyl) pyridine Antiferromagnetic [(LCu) 2 Ce(N0 3 ) 3 ] , L = u-phenolato-u-oximato l igand 2 2 2 3 Figure 2.1. Selected t?//heterometallic complexes where the atoms in red represent the Ln(III) metal ions and the atoms in blue are the Cu(II) metal ions. The magnetic exchange for many of the C u - L n compounds, where L n ± La(III), Gd(III) or Lu(III), have not been elucidated due to masked interactions from the large first-order orbital contributions of the Ln ( I I I ) . 1 8 ' 2 1 ' 2 4 " 3 0 To combat this problem, Costes et al. synthesized a N i - L n system isostructural to the Cu-Ln species of interest.2 5 Ni(II) was held in a square planar environment giving rise to a low-spin, diamagnetic state. The non-Curie 37 behavior associated with the/electrons of the Ln(III) was elucidated using the N i - L n species. Assuming identical crystal-field effects for the two complexes, the magnetic behavior between the d/f metal ions for the C u - L n species could be unmasked by subtracting the first-order orbital contributions of the Ln(III) determined by the N i - L n magnetic data. A similar method was used by Matsumoto who synthesized N i 2 L n 2 systems in order to determine the magnetic exchange in the cyclic cylindrical C u 2 G d 2 complex. 1 8 Kahn analyzed the magnetic exchange in the Ln 2 Ni3 and Ln 2 Cu3 polymeric ladder-type compounds by using the analogous Ln2Zn3 2 6 ' 2 8 where the Zn(II) is J 1 0 diamagnetic metal ion, and thus has no magnetic contributions (Figure 2.1). Okawa and co-workers elucidated the magnetic coupling of their 2,6-di(acetoacetyl)pyridine T M - L n - T M systems, where T M = Cu(II) and Ni(II), utilizing an equivalent Zn-Ln-Zn species (Figure 2 .1 ) . 3 1 , 3 2 In all the above examples, a ferromagnetic coupling between the Cu(II) and Ln(III), where L n = Gd(III), Tb(III), Dy(III) and an antiferromagnetic exchange for L n = Ce(III), Pr(III), Nd(III), Sm(III) has been determined. Species with Ho(III), Er(III) and Yb(III) have been shown to exhibit both antiferromagnetic and ferromagnetic interactions, attributed to the differences in the bridging ligand structure between the TM(II) and the Ln(III) (phenolato vs. oxamato vs. enolate oxygen atoms). Additional T M - L n species would contribute to conclusions on the magnetic interactions of these species. Costes et al. were the first to report a ferromagnetic exchange between N i - G d , where Ni(II) was held in a distorted octahedral environment and therefore paramagnetic. 3 3 Since then, only a limited number of molecular N i - G d Schiff base complexes have been reported, 3 4 ' in addition to one N i - G d cryptate, one N i 3 L n 2 ladder complex where L n = La(III), Ce(III), Pr(III), Nd(III), Gd(III), Tb(III), Dy(III), Ho(III) and Er(III), 3 7 three N i 2 L n clusters 38 where L n = Ce(III), Pr(III), Nd(III), Gd(III), Tb(III), Dy(III), Ho(III) and E r ( I I i ) , 3 2 ' 3 8 ' 3 9 and one D y - N i - D y 4 0 system have been synthesized with paramagnetic Ni(II). The magnetic properties of these systems mirrored the Cu-Ln species, as they both exhibit a ferromagnetic interaction for L n = Gd(III), Tb(III) and Dy(III) (and possibly Ho(III) and Er(III)) and an antiferromagnetic exchange for L n = Pr(III) and Nd(III) (and possibly Ce(III) and Eu(III)). The limited magnetic studies reported for the published d/f complexes, mostly using Cu-Gd systems with Schiff base ligands of similar geometric demands, have kept the field widely unexplored and thus poorly understood. Therefore, it was deemed desirable to design new ligands with different geometric demands that can afford discrete d/f mixed-metal complexes where the TM(II) and Ln(III) can be varied while maintaining the same core structure. Uti l izing three different amine phenol ligands previously reported by the Orvig and Wieghardt groups 4 1" 4 3 (Figure 2.2), the synthesis of a series of magnetically and structurally unusual [Ln(TM)2L2] + complexes was attempted. Examination of the magnetic properties of these species was undertaken to allow for a comprehensive study on the effect of the varied geometric-, spin- and orbital-contributions on the overall magnetic behavior between d a n d / metals (Figure 2.2). Various contributions of the divalent first-row <i-block metal ion and the trivalent Ln(III), where T M = Co(II) - Zn(II) and L n = La(III) - Lu(III) (excluding Pm(III)), with the three different chelating ligands, Ffjtam (tris-1,1,1 -(((2-hydroxybenzyl)amino)-methyl)ethane),4 1 H 3 t rn (rra(2'-hydroxybenzylaminoethyl) amine), 4 2 and Ffjbcn (tris-N,N',N' '-(2-hydroxybenzyl)-l ,4,7-triazacyclononane) 4 3 were investigated. It was proposed that the three polyamine phenolate ligands would selectively coordinate the TM(II) by the amine nitrogens and deprotonated phenolate oxygens, rendering 39 the resulting anionic [TM(L)]" complex a potentially preorganized O 3 ligand for the coordination of the Ln(III) ion (Figure 2.2). In the case of H 3 t rn , the extra amine donor could preferentially coordinate the T M in an octahedral geometry using the N4O2 pocket, leaving an unbound phenolate oxygen to more easily accommodate the Ln(III) ion. The F^tam and Habcn have a slightly more sterically demanding N3O3 binding pocket. Figure 2 .2. Ligands of interest and proposed binding motifs for the d/f/d heterometallic complexes. Selected trimetallic N i - L n species using the tripodal heptadentate (N4O3) F ^ t m 3 9 and T O -hexadentate (N3O3) F^tam ligands were previously reported by the Orvig group; however, only the N i - L n - N i series were obtained where the Ni(II) was coordinated in a pseudo-octahedral geometry by the four amine donors and two phenolate oxygens of the trn3" ligand. Alternatively, the Ni(II) ion was chelated by the three amines and three phenolate oxygens in 40 the N3O3 pocket of tam3". Two of the fully deprotonated [Ni(L)]" moieties spontaneously self-assembled around the Ln(III) in a bicapped fashion generating the cationic [LnNi2(L)2] + complex. The Ln(III) in the [LnNi2(trn)2]+ species were seven-coordinate via two bridging and one non-bridging phenolate donor from each of the trn3", and one coordinating solvent molecule. The number of coordinating solvent molecules in the [LnNi2(tam) 2] + was varied depending on the size of the Ln(III) ion (the larger Ln(III) accommodated more solvent molecules). In-house magnetic susceptibility studies on the [LnNi 2 (L)2] + complexes (Ln = La(III), Dy(III), and Yb(III) for L = tam3" and L n = La(III), Pr(III), Nd(III), Sm(III), Eu(III), Gd(III), Tb(III), Dy(III), Ho(III), Er(III), Tm(III), Yb(III) and Lu(III) for L = trn3") were completed • using a superconducting quantum interference device (SQUID). The [LnNi 2 ( t rn) 2 ]C104»«H 2 0 cationic complexes exhibited ferromagnetic exchange for L n = Gd(III), Tb(III), Dy(III), Ho(III) and Er(III) (and possibly Tm(III) and Yb(III)) following the trend of other documented N i - L n species. These results also suggest that Kahn's prediction are relevant for TM(II)-Ln(III) species other than Cu(II)-Gd(III) complexes, where a ferromagnetic coupling is observed for Ln(III) ions w i t h / - / 1 3 configurations. 3 7 ' 4 4 Alternatively, antiferromagnetic interactions were observed for the [LnNi2( tam) 2 ]C104 '«H 2 0 species where L n = Dy(III) and Yb(III). It has been proposed that the observation of apparent antiferromagnetic interactions in the tam3" species were due to a combination of diminishing paramagnetic contributions of the Ln(III) ion at low temperatures in addition to saturation effects arising from the large applied magnetic field (10,000 G) that could mask ferromagnetic interactions between the d/f/d metal ions. Thus, in the present study, the [LnNi2(tam)2]C104»«H 20 species were re-synthesized and their magnetic studies were 41 repeated at a lower magnetic field in hopes of elucidating their magnetic interactions. Attempts to synthesize the other [ L n T M 2 ( L ) 2 ] + using divalent T M = Mn(II), Fe(II), Cu(II), Ni(II), Zn(II), a selective cross-section of the Ln(III) (Ln = La(III), Nd(III), Gd(III), Dy(III) and Yb(III)), and both tripodal H 3 tam and H 3 t rn ligands were also undertaken in the current work. A major part of the research described in this thesis chapter is based on the synthesis and characterization of the geometrically more demanding H 3 bcn ligand and its subsequent [TM(Hbcn)], [TM 3 (bcn) 2 ] and [LnTM 2 (bcn) 2 ] + complexes. In addition, an empirical determination of the magnetic interactions between the d/f/d metal ions of the [LnTM 2 (bcn) 2 ]C104»/ i 7H 2 0 species studied here was.completed using the analogous diamagnetic [LnZn 2 (bcn) 2 ]C104»«H 2 0 species, similar to the analysis by Costes, 2 5 K a h n , 2 6 , 3 7 Okawa, ' Matsumoto and co-workers. 2.2 Experimental 2.2.1 Materials Reagents were purchased from commercial sources and used without purification unless otherwise stated. Hydrated Ln(III) perchlorate salts were obtained from Alfa Aesar as 50% (w/w) solutions in water and evaporated to dryness. N M R solvents were obtained from Sigma-Aldrich and used as received. Methanol, ethanol and acetonitrile were dried over activated 4 A molecular sieves prior to use. Water was deionized (Barnstead D8902 and D8904 Cartridges) and distilled (Hytrex II GX50-9-7/8 and GX100-9-7/8) before use. Yields 42 for the analytically pure compounds were calculated based on the respective metal ion starting material. 1,4,7-Triazacyclononane was synthesized using a modified published procedure. 4 5 Caution: Perchlorate salts are potentially explosive and should be handled with extreme care and used only in small quantities. 2.2.2 Instrumentation ' f l N M R spectra were recorded on a Bruker Avance 300 spectrometer (300 M H z ) or a Bruker Avance 400 inverse spectrometer (400 M H z ) at room temperature (RT) and calibrated with listed deuterated solvents. Infrared spectra were recorded as nujol mulls (KBr windows) in the range 4000-500 cm"1 on a Mattson Galaxy Series FTIR-5000 spectrophotometer and were referenced to polystyrene (1601 cm"1). Elemental analyses of C, H , and N were performed by Mr . M . Lakha at the University of British Columbia (UBC) . UV-vis ible spectra were recorded in methanol using a Hewlett-Packard 8543 diode array spectrophotometer in a 1 cm cuvette. Mass spectra were obtained on a Bruker Esquire Ion Trap (electrospray ionization mass spectrometry, ESI -MS) spectrophotometer. Approximately 1 mg of sample was dissolved in 1 m L of methanol or acetonitrile for +ESI-M S analysis. Room temperature magnetic susceptibilities on powdered samples of [TM3(bcn)2]»ftH20 and [TM(Hbcn) ]» /7H 2 0 complexes were measured on a Johnson Matthey M S B - 1 balance standardized with C o C l 2 ' 6 H 2 0 and CuS04*5H 2 0. Variable-temperature and applied field magnetic susceptibility measurements were performed on two different Quantum Design M P M S X L S Q U I D magnetometers over the temperature range 2 - 300 K at 10,000 and 500 G at U B C and Simon Fraser University (SFU). Diamagnetic corrections for 43 the constituent atoms used Pascal's constants4 6 (6.10 x 10" 4 cm 3 mol" 1 , plus 3.89 x 10"5 cm 3 mol" 1 per water). A specially designed sample holder, as described previously, 4 7 was used to minimize the background signal. Magnetic susceptibilities were corrected for the background signal generated by the holder. 2.2.3 [LaZn2(tam)2]Cl04 [LaZn 2(tam)2]C104*5H 20. To a solution of Zn(C10 4 ) 2 -6H 2 0 (46.5 mg, 0.125 mmol) in dry M e O H under A r was added H 3 tam (54.0 mg, 0.125 mmol). The pale yellow solution was stirred for 15 min before hydrated La(C10 4 ) 3 -6H 2 0 (34.0 mg, 0.063 mmol) was added.. The solution became clear upon the addition of the La(III) and was stirred for an additional 30 min at R T before N E t 3 (38.0 mg, 0.38 mmol) diluted in 5 m L of dry M e O H was added dropwise. The solution turned yellow after stirring for 5 min. Colorless crystals deposited upon slow evaporation of the solvent. The solid was filtered, washed with cold M e O H and dried in air. The air-dried products were obtained in yields of 29.7 mg, 21 %. Anal . Calcd. (found) for C 5 2 H 7 o C l ] L a , N 6 O i 5 Z n 2 : C , 47.16 (47.45); H , 5.33 (5.35); N , 6.35 (6.55). ' H N M R (300 M H z , R T , C D 3 O D ) : 5 = 7.03 (d, 6H, o-benzyl H, 3 J = 7.3 Hz) , 7.01 (t, 6H, m-phenol H, 3J = 8.2 Hz), 6.84 (d, 6H, o-phenol H, 3J= 7.9 Hz), 6.59 (t, 6H,/>phenol H, 3J = 7.7 Hz), 4.25 (d, 6H , benzyl CH2,2J= 12.42 Hz), 2.93 (d, 6H, benzyl CH2,2J = 11.7 Hz), 2.59 (d, 12H, CH2, 2J= 13.2 Hz), 2.53 (d, 12H, CH2,3J= 13.2 Hz), 0.61 (s, 6H; CH3). IR (cm"1; K B r disk): 3990 (br, v 0 - H ) , 3259 (s, v N . H ) , 1946 and 1904 (w, v N . H(ring)), 1594 and 1562 (s, 8N-H,)> 1478 and 1453 (s, v c =c), 1279 and 1085 (s, 5C-o), 758 (m, OC-H ) , 1100 (s, vac*), 600 (m, VTM-O), 453 (s, v L n . 0 ) . M S (+ESI, M e O H ) : m/z 1135 [L+H] + . . 44 2.2.4 H3bcn Diethylene-l,4,7-triamine tritosylate. Diethylenetriamine (8.26 g, 0.08 mol) and N a O H (9.6 g, 0.24 mol) were dissolved in 80 m L of water. The solution was added dropwise top-toluenesulfonyl chloride (TsCl , 45.6 g, 0.24 mol) dissolved in 240 m L diethyl ether. The mixture was stirred for 2 h at R T after which time a white precipitate formed at the interface and sank to the bottom of the flask. The product was filtered and washed with water and diethyl ether. The resulting white solid was vacuum dried without further purification to yield 30 g, 70 %. ' H N M R (300 M H z , R T , ( C D 3 ) 2 C O ) : 5 = 7.75 (d, 4H , H9,3J = 8.2 Hz), 7.64 (d, 2 H , //4, V = 8.3 Hz), 7.41 (d, 4H , / / 1 0 , 3 J = 9.2 Hz), 7.39 (d, 2H, H5,3J= 8.01 Hz), 6.56 (t, 2H, N / / , 3J = 5.82 Hz), 3.20 (tt, 4H , H2,3J = 7.53 Hz , 2J = 3.18 Hz), 3.06 (tt, 4H , H\, 3J = 7.53 Hz, 2J = 3.09 Hz), 2.43 (s, 6H , H\2), 2.42 (s, 3H, HI). 1 3 C N M R (75.5 M H z , R T , ( C D 3 ) 2 C O ) : 5 = 144.07 (C3and C8), 138.87 (C6 and C I 1), 130.76 (C9), 130.61 (C4), 128.18 (C10), 127.94 (C5), 50.28 (C2), 43.18 (CI) , 21.48 (C12 and CI). M S (+ESI): m/z 588 [L + N a ] + . Ethyleneglycol bistosylate. Over 30 min, small portions of T s C l (114 g, 0.6 mol) were , slowly added to ethylene glycol (18.6 g, 0.3 mol) dissolved in 500 m L of 0 O 1 i o 0 2 S s o 2 pyridine under A r and cooled to 0 C using an ice bath. After being continuously r i i r i i • l*=y^ stirred at 0 C under A r for 5 h, the solution was poured onto crushed ice (-500 e C H 3 C H 3 mL) and acidified using concentrated HC1 until p H = 1 (-700 m L of HC1). The white precipitate was washed with cold water, recrystallized from hot methanol and dried for 48 h in vacuo to yield 50 g, 50 %. ' H N M R (300 M H z , RT, CD 3 C1): 5 = 7.71 (d, 4H , H3,3J 45 = 6.6 Hz), 7.32 (d, 4 H , 774, 3J = 8.0 Hz), 4.17 (s, 4H, HI), 2.43 (s, 6H , H6). 1 3 C N M R (75.5 M H z ; R T , CDC1 3 ) : 5 = 145.46 (C2), 132.40 (C5), 130.13 (C3), 128.08 (CA)', 66.92 (CI) , 21.81 (C6). M S (+ESI): m/z 393 [L + Na] + . 1,4,7-Triazacyclononane tritosylate. Diethylene-l,4,7-triaminetritosylate (24.1 g, 0.0427 _ 1 so 2 mol) was dissolved in 100 m L of anhydrous D M F dried over 4 A < Ts > _ 3fS NTs NTs T s - A \ J sieves. Small portions of 60% N a H in oil (previously rinsed with 6 C H 3 petroleum ether, 6 g, 0.25 mol) were added and the solution was o heated to 70 C and stirred under A r for 30 min. A solution of ethyleneglycol ditosylate (15.8 g, 0.0427 mol) in dry D M F (200 mL) was added dropwise over a 5 h period. The solution was stirred for 20 h under A r and maintained at 70 °C. A yellow precipitate formed after the solution volume was reduced to -200 m L under reduced pressure and slowly added to -1000 m L ice and water. The solution was stored at 4°C for 5 h and then filtered; the filtrate was washed with water, ethanol and ether, dried in vacuo overnight, and used without further purification to yield 24.1 g, 94 %. ' H N M R (300 M H z , R T , CD 3 C1): 6 = 7.69 (d, 6H, 7/3, 3J = 8.1 Hz), 7.32 (d, 6H, H4,3J = 8.1 Hz), 3.42 (s, 12H, 7/1), 2.43 (s, 9H, 7/6). 1 3 C N M R (75.5 M H z , R T , CDCI3): 5 = 143.88 (CI), 132.57 (C5), 129.86 (Ci), 127.47 (C4), 51.86 (C6), 21.51 (CI) . MS(+ESI ) : m/z 614 [L + Na] + . l,4,7-Triazacyclononane»trihydrochlorate (tacn»3HCl). 1,4,7-Triazacyclonone tritosylate / ^ N ^ A CI (24.1 g, 0.0407 mol) was added to 40 m L of concentrated H2SO4. The 1< + H 2 + > + 3 2 N H 2 brown slurry was heated at 1 6 0 ° C f o r 3 0 min under A r . The solution was cooled to R T before being added dropwise to cold E t O H (200 mL). Approximately 300 m L 46 of diethyl ether was added to the solution to afford a brown precipitate. After storage at 4°C overnight, the brown precipitate was filtered, dissolved in 150 m L of H 2 O and decolorized by boiling the solution for 15 min in the presence of Norit and Celite. After the solution was clarified by filtration, the solvent was removed under reduced pressure leaving a light brown oil . A n off white precipitate immediately formed upon the addition of 25 m L of concentrated HC1, followed by 150 m L of ethanol. The solid was filtered out, washed with ethanol and ether, and recrystallized from H 2 0 / E t O H (1:8) to yield 9.2 g, 95 %. ' H N M R (300 M H z , RT, D 2 0 ) : 5 = 3.43 (s, 12H, 771). 1 3 C N M R (75.5 M H z , RT, D 2 0 / C D 3 O D ) : 6 = 45.78 (CI). 1,4,7-Triazacyclononane (tacn). Tacn»3HCl (9.2 g, 0.0388 mol) was dissolved in 20 m L NH NH g, 2.8 equivalents). Tacn was extracted into chloroform by continuous extraction for 48 h. The chloroform layer was removed and the aqueous layer was extracted with chloroform ( 3 x 3 0 mL) . The combined chloroform layers were dried over sodium sulphate and filtered. The solvent was removed under reduced pressure affording a pale yellow oil to yield 3.6 g, 7 2 % . ' H N M R (300 M H z , R T , D 2 0 ) : 5 = 2.78 (s, 12H, H\). 1 3 C N M R (75.5 M H z , R T , D2O/CD3OD): 8 = 46.92 (CI) . M S (+ESI): m/z 130 [L + H ] + . M S (+ESI, high resolution): 130.1347 [L+H] + . A',7V',A'"-rm(2-hydroxybenzyl)-l,4,7-triazacyclononane (Hsbcn). To a solution of tacn (3.1g, 0.0279 mol) in dry ethanol was added salicylaldehyde (10.22 g, 0.0837 mol) dropwise. The solution was allowed to stir for 2 h under A r . Solvent was removed under reduced pressure leaving an orange oi l . The oil was dissolved in dry dichloromethane and allowed to deionized H 2 O . The p H was adjusted to -13 by addition of N a O H pellets (-4.3 47 HO. stir under Ar . N a B H ( O A c ) 3 (35.48 g, 0.1674 mol) was slowly added to the solution over a period of 3 h. The white cloudy solution was OH allowed to stir for an additional 3 h under A r after which excess NaBH(OAc )3 was quenched by the addition of a saturated solution of sodium bicarbonate. The product was extracted from the aqueous layer with chloroform (3 x 50 mL), dried over M g S C ^ and volatiles were removed in vacuo. The product was recrystallized in M e O H / H 2 0 yielding an off white solid that was obtained in excellent yield with respect to tacn, 10.23 g, 82 %. Anal . Calcd. (found) for C27H33N3O3: C , 72.46 (72.34); H , 7.43 (7.34); N , 9.39 (9.64). ' H N M R (300 M H z , R T , CDC1 3 ) : 6 = 7.19 (t, 3H, H6,3J = 7.8 Hz), 6.93 (d, 3H, H 8 , 3 J = 7.2 Hz), 6.90 (d, 3H, H5,3J = 7.2 Hz) , 6.76 (t, 3H, HI, 3J = 7.5 Hz), 3.77 (s, 6H, HT), 2.43 (s, 12H, H\). *H N M R (300 M H z , R T , C D 3 O D ) : 5 = 7.22 (t, 3H, H6,3J = 8.1 Hz), 7.16 (d, 3H, HS, 3J = 7.2 Hz), 6.87 (t, 3H, HI, 3J = 8.1 Hz), 6.84 (d, 3H, H5,3J= 7.2 Hz), 3.86 (s, 6H , HI), 2.85 (s, 12H, HI). 1 3 C N M R (75.5 M H z , R T , CD3OD): 5 = 164.18 (C4), 133.15 (C6), 131.07 (C8), 123.33 (C3), 120.79 (C7), 117.10 (C5), 60.50 (C2), 51.89 (CI) . MS(+ESI ) : m/z 448 [L + H ] + . 2.2.5 [TM(Hbcn)] complexes [Cu(Hbcn)]»H20. To a solution of H 3 bcn (27.8 mg, 0.062 mmol) in 20 m L methanol was added C u ( C 1 0 4 ) 2 » 6 H 2 0 (22.8 mg, 0.062 mmol). The pale yellow solution became greenish-brown immediately after the addition. The solution was stirred for one hour at RT. The solvent was evaporated to ca. 2 m L . Green crystals formed after slow evaporation of the solvent and were filtered off, washed with cold water and air dried to yield 19.5 mg, 62 %. 48 Anal . Calcd. (found) for C27H33CUN3O4: C , 61.52 (61.80); H , 6.31 (6.34); N , 6.34 (6.31). M S (+ E S - M S ) : m/z 509 [M(HL)+H] + . u e f f = 1.79 u B . [Ni(Hbcn)]»H20. To a solution of I-^bcn (27.9 mg, 0.062 mmol) in 20 m L methanol was added N ^ C I C M ^ ^ ^ O (22.9 mg, 0.062 mmol). The pale yellow solution turned cloudy immediately after the addition. After the solution was stirred for 1 h, the yellow powder was filtered off, washed with cold methanol and air dried to yield 16.9 mg, 54 %. Anal . Calcd. (found) for C ^ r f ^ N i C V : C , 62.09 (62.34); H , 6.37 (6.26); N , 8.05 (8.30). ' H N M R (300 M H z , R T , CD3OD): 5 = 7.14 (m, 6H, o-benzyl H and w-phenol H), 6.83 (m, 6H, o-phenol H andp-phenol H), 3.82 (s, br, 6H , benzyl CH2, v , / 2 = 20 Hz,) , 2.79 (m, 12H, CH2). M S (+ESI): m/z 504 [M(HL)+H] + . u e r T = diamagnetic. H3[Ni(bcn)]2C104. To a solution of H3bcn (27.9 mg, 0.062 mmol) in 20 m L acetonitrile was added Ni(C104)2*6H20 (27.9 mg, 0.062 mmol). The solution turned a golden yellow color immediately after the addition and was stirred for 1 h at RT. The solution was filtered through glass wool and allowed to stand at RT, whereupon purple needle crystals deposited. The crystals were filtered off, washed with cold water and air dried to yield 6.6 mg, 21 %. Anal . Calcd. (found) for C54H 6 3ClN 6 Ni 2 Oio: C , 58.49 (58.14); H , 5.81 (5.89); N , 7.46 (7.67). ' H N M R (300 M H z , R T , CD3OD) is uninformative. M S (+ESI): m/z 505 [M(HL)+H] + . ja e f f = 2.77 UB. [Zn(Hbcn)]»H20. To a solution of Hjbcn (27.9 mg, 0.062 mmol) in methanol was added Zn(C104)2,6H20 (23.1 mg, 0.062 mmol). The pale yellow solution became clear 49 immediately after the addition and was continuously stirred for 1 h at RT . Evaporation of the solvent to ca. 2 m L afforded a white precipitate which was filtered off, washed with cold methanol and air dried to yield 25.9 mg, 82 %. Anal . Calcd. (found) for C 2 7H33N30 4 Z n : C, 61.31 (61.02); H , 6.29 (6.30); N , 7.94 (8.20). ' H N M R (300 M H z , R T CDC1 3 ) : 5 = 7.14 (d, 3H, o-benzyl H, 3J = 7.2 H z ) , 6.98 (t, 3H, m-phenol H, 3J = 7.2 H z ) , 6.78 (t, 3H,/>phenol H, 3J= 7.2 H z ) , 6.14 (d, 3H, o-phenol H, 3J= 7.2 H z ) , 4.23 (d, 3H , benzyl CH2, 2 J = 11.4 H z ) , 3.39 (d, 3H, benzyl C / / 2 , V= 11.4 H z ) , 2.96 (ddd, 3H, CH2, 2J(A,A ') = 15.0 H z , 3J(A,B) = 14.1 H z , 3J(A,B') = 3.9 H z ) , 2.66 (dd, 3H, CH2, 2J(A ;A) = 15.3 H z , 3J(A\B and.4 \B') = 5.4 H z ) , 2.31 (dd, 3H, CH22J(B\B) = 12.5 H z , 3J{B\A and 5 ^ ') = 3.9 H z ) , 1.97 (ddd, 3H, CH2, 2J(B,B') = 12.8 Hz, 3 J(B,A) = 12.3 H z , 3J(B,A ') = 6.0 H z ) . M S (+ESI): m/z 511 [ M ( H L )+H ] + . 2.2.(5 [TMs(bcn)2] complexes [Cu 3 (Hbcn)2](C104)2 'H 2 0. To a solution of H 3 bcn (27.9 mg, 0.062 mmol) in methanol was added Cu(C104) 2 *6H 2 0 (22.9 mg, 0.062 mmol). The pale yellow solution turned greenish-brown immediately following the addition. The solution was refluxed for 1 h and allowed to cool before the addition of a third equivalent of C u ( C 1 0 4 ) 2 ' 6 H 2 0 (11.5 mg, 0.03 mmol). The solution was allowed to stir at R T for 30 min before N E t 3 (18.8 mg, 0.186 mmol) was added dropwise. The solvent was evaporated to ca. 2 m L affording light brown precipitate after the additional solvent was slowly evaporated. The powder was filtered, washed with cold water and air dried to yield 16.1 mg, 24 %. Anal . Calcd. (found) for C54H64CU2CU3N6O15: C, 49.94 (50.28); H , 4.97 (5.08); N 6.47 (6.58). M S (+ESI): m/z 1080 [ M 3 L 2 + H ] + . p eff=2.86 p B -50 [Ni3(bcn)2]»2H20. To a solution of H 3 bcn (27.9 mg, 0.062 mmol) in methanol was added N i ( C 1 0 4 ) 2 » 6 H 2 0 (22.9 mg, 0.062 mmol). The pale yellow solution immediately turned clear following the addition. The solution was refluxed for 1 h and allowed to cool before addition of the third equivalent of Ni(C104) 2 *6H 2 0 (11.4 mg, 0.031 mmol). The solution was allowed to stir at R T for 30 min and NEt3 (18.8 mg, 0.186 mmol) was added dropwise. The solvent was evaporated to ca. 2 m L and a light purple precipitate formed after additional solvent was slowly evaporated. The powder was filtered off, washed with cold methanol and air dried to yield 28.4 mg, 43 %. Anal . Calcd. (found) for C^H^NeNiaOg: C, 58.90 (59.29); H , 5.86 (5.84); N , 7.63 (7.83). M S (+ESI): m/z 1065 [ M 2 L 3 + H ] + . p.eff= 5.54 p B -[Zn3(bcn)2]. To a solution of H^bcn (27.9 mg, 0.062 mmol) in methanol was added Zn(C104) 2 *6H 2 0 (23.1 mg, 0.062 mmol). The pale yellow solution immediately turned clear following the addition. The solution was refluxed for 1 h and allowed to cool before addition of the third equivalent of Zn(C104) 2 *6H 2 0 (11.6 mg, 0.031 mmol). The solution was allowed to stir at R T for 30 min and NEt3 (18.8 mg, 0.186 mmol) was added dropwise. The solvent evaporated to ca. 2 m L afforded white precipitate that was filtered, washed with cold methanol and air dried to yield 37.7 mg, 56 %. Anal . Calcd. (found) for C54H6oN606Zn3: C, 59.76 (59.80); H , 5.57 (5.72); N , 7.74 (7.89). ' H N M R (300 M H z , R T , CDC1 3 ) : 8 = 7.13 (d, 6H, o-benzyl H, V= 6.4 Hz), 6.98 (t, 6H, m-phenol H, 3J = 7.7), 6.78 (t, 6H,/>phenol H,3J = 7.8 Hz), 6.13 (d, 6H, o-phenol H, 3J= 8.2 Hz), 4.23 (d, 6H, benzyl CH2, 2J = 10.8 Hz), 3.39 (d, 6H, benzyl CH2,2J= 12.0 Hz), 3.07 (ddd, 6H, CH2, 2J(A,A ') = 16.2 Hz , 3J(A,B) = 15.0 Hz, 3J(A,B') = 2.1 Hz), 2.65 (dd, 6H, CH2, 2J(A\A) = 15.06 Hz , 3J(A\B andA'.B1) = 3.9 Hz), 2.34 (dd, 6H, CH2,2J(B',B) = 12.3 Hz , 3J(B',A and B',A') = 3.6 Hz) , 1.96 (ddd, 6H, CH2, 2J(B,B') = 13.5 Hz , 3J(B,A) = 12.3 Hz , 3J(B,A ') = 5.4 Hz). M S (+ESI): m/z 1085 [ M 2 L 3 + H ] + . 51 2.2.7 [La(bcn)J [La(bcn)]»H20. To a solution of Hsbcn (27.9 mg, 0.062 mmol) in methanol was added La(C104)3«6H 20 (33.8 mg, 0.062 mmol). The pale yellow solution turned clear upon addition and was refluxed for 1 h before three equivalents of N E t 3 (18.8 mg, 0.186 mmol) diluted in 2 m L of methanol was added dropwise to the reaction mixture. After 30 min, the solution became turbid. The off-white powder was filtered, washed with methanol and water, and dried under vacuum to yield 7.59 mg, 12 %. Anal . Calcd. (found) for C 27H32LaN 304: C, 53.92 (54.00); H , 5.36 (5.58); N , 6.99 (7.16). ' H N M R (300 M H z , RT, C D 3 O D ) : 5 = 7.19 (dd, 3H , o-benzyl H, 3J = 15.4 Hz , 2J = 1.7 Hz), 7.14 (dt, 3H, m-phenol H, 3J = 8.3 Hz , 2J = 1.6 Hz) , 6.85 (d, 3H, o-phenol H, 3J= 8.1 Hz) , 6.81 (t, 3H,/>phenol H, 3J= 7.3 Hz), 3.82 (s, 6H , benzyl CH2), 2.87 (dt, 6H, CH2,3J= 10.5 Hz , 2 J = 3.6 Hz), 2.78 (dt, 6H, CH2,3J= 10.5 Hz , 2J= 3.6 Hz). M S (+ESI): m/z 584 [ M L + H ] + . 2.2.8 [La3(bcn)2]3+ [La3(bcn) 2](C104)3»8H20 To a solution of H3bcn (27.9 mg, 0.062 mmol) in methanol was added La(C104)3»6H 2 0 (33.8 mg, 0.062 mmol). The mixture was refluxed for 8 h. After the solution was cooled to R T , an additional equivalent of La(C104)3»6H20 (16.9 mg, 0.031 mmol) was added to the solution, which has allowed to stir at R T for 30 min. NEt3 (18.8 mg, 0.186 mmol) diluted with 5 m L of methanol was added dropwise to the solution and the solution was allowed to stand at R T , whereupon a white powder deposited. The solid was 52 washed with a small amount of cold methanol and dried in air to yield 13.0 mg, 14 %. Anal . Calcd. (found) for C 5 4 H 7 6 C l 3 L a 3 N 6 0 2 6 : C , 37.10 (37.16); H , 4.38 (4.12); N , 4.81 (4.54). ' H N M R (300 M H z , R T , C D 3 O D ) : 8 = 7.21 (dd, 6H , o-benzyl H, 3J= 15.3 H z , 2J = 1.8 Hz), 7.14 (dt, 6H , m-phenol H, 3J = 7.5 Hz , 2J = 1.5 Hz), 6.85 (dd, 6H , o-phenol H, 3J = 7.2 Hz , 2 J = 1.2 Hz), 6.84 (dt, 6H,/>phenol H, 3J = 8.4 Hz , 2J= 1.2 Hz) , 3.83 (s, 6H , benzyl CH2), 2.80 (s, 12H, CH2). M S (+ESI): m/z 1503 [ M 3 L 2 + 2 C 1 0 4 ] + , 448 [ H 3 L + H ] + . 2.2.9 [LnTM2(bcn)2j+ complexes See Table 2.2 for a list of the characterization data for [LnTM2(bcn)2]C104*«H20. General procedure for [ L n C u 2 ( b c n ) 2 ] C K V r t H 2 0 , Ln(III) = L a , N d , G d , Dy, Y b . To a solution of 0^0104 ) 2 * 6 ^0 (92.6 mg, 0.25 mmol) in methanol was added H 3 bcn (111.9 mg, 0.25 mmol). The clear solution immediately turned brown upon addition of the hydrated metal perchlorate salt. The solution was allowed to reflux while being stirred for 1 h. After the solution was cooled to R T , one half equivalent of hydrated Ln(C104) 3 «6H 2 0 (68.2 - 72.4 mg, 0.125 mmol) was added. The brown solution was stirred for an additional 30 min at R T and N E t 3 (75.6 mg, 0.75 mmol) diluted in 5 m L of methanol was added dropwise. After 5 min of stirring, the solution turned green and was immediately filtered through a cotton plug. A pale green powder deposited upon slow evaporation of the solvent. The precipitate was filtered, washed with cold ethanol and dried in air yielding 40-70 %. *H N M R (300 M H z , RT, CD3OD) is uninformative. 53 General procedure for [LnNi 2 (bcn ) 2 ]C10 4 viH 2 0, Ln(III) = La , Nd, Gd, Dy, Yb. To a solution o f N i ( C 1 0 4 ) 2 ' 6 H 2 0 (91.4 mg, 0.25 mmol) in acetonitrile was added H 3 bcn (111.9 mg, 0.25 mmol). The pale yellow solution immediately formed a blue-green precipitate. The solution was refluxed and allowed to stir for 1 h. Hydrated L n ( C 1 0 4 ) 3 » 6 H 2 0 (68.2 - 72.4 mg, 0.125 mmol) was added after the solution was cooled to R T , affording a pale purple solution. The solution was stirred for an additional 30 min at R T before N E t 3 (75.6 mg, 0.75 mmol) diluted in 5 m L acetonitrile was added dropwise. After stirring for 5 min, the solution turned yellow. Pale purple perchlorate salts deposited upon slow evaporation of the solvent (except in the case of [GdNi 2 (bcn) 2 ] + and [YbNi 2 (bcn) 2 ] + where prismatic purple crystals formed). The solid was filtered, washed with cold acetonitrile and dried in air yielding 45-85 %. ' H N M R (300 M H z , R T , C D 3 O D ) was uninformative. General procedure for [LnZn 2 (bcn) 2 ]CI04»«H20, Ln=La, Nd, Gd, Dy, Yb. To a solution of H 3 bcn (111.9 mg, 0.25 mmol) in methanol was added Z n ( C 1 0 4 ) 2 ' 6 H 2 0 (93.1 mg, 0.25 mmol). The pale yellow solution immediately turned clear upon addition of the hydrated metal salt. The solution was refluxed and allowed to stir for 1 h. The hydrated L n ( C 1 0 4 ) 3 « 6 H 2 0 (68.2 - 72.4 mg, 0.125 mmol) was added to the solution after it had cooled to RT. The clear solution was stirred for 30 min at R T and N E t 3 (75.6 mg, 0.75 mmol) diluted in 5 m L methanol was added dropwise. The solution turned slightly yellow after stirring for an additional five minutes. A n off-white powder deposited upon slow evaporation of the solvent (except in the case of [GdZn 2 (bcn) 2 ] + where prismatic colorless crystal formed) and was filtered, washed with cold ethanol and dried in air yielding 60-85 %. ' H N M R (400 M H z , R T , CDC1 3 ) : 5 = 7.12 (d, 6H, o-benzyl H, 3 J = 6.2 Hz), 6.96 (t, 6H, m-54 phenol H, V = 7.1 Hz) , 6.75 (t, 6H, m-phenol H3J= 7.3 Hz) , 6.12 (d, 6H, o-phenol H,3J = 8.1 Hz), 4.20 (d, 6H, benzyl CH2, 2 J = 11.7 Hz), 3.37 (d, 6H , benzyl CH2,2J= 11.7 Hz), 2.95 (ddd, 6H, CH2,2J(A,A ') = 16.0 Hz , 3J(A,B) = 14 Hz, 3J(A,B') = 3.9 Hz) , 2.64 (dd, 6H, CH2, 2J(A\A) = 15.5 Hz , 3J{A\B andA'.B') = 5.32 Hz), 2.30 (dd, 6H , CH2,2J(B\B) = 12.5 Hz, 3J(B',A and B'.A ') = 4.3 Hz), 1.95 (ddd, 6H, CH2,2J(B,B') = 12.8 Hz , 3J(B,A) = 12.0 Hz , 3J(B,A ') = 5.5 Hz). 1 3 C N M R (75 M H z , RT, C D 3 O D ) : 5 = 167.02 (ring C-OH) , 132.41 {m-Q, 131.16 (o-benzyl Q, 125.05 (benzyl Q, 121.92 (p-phenol Q, 116.83 (o-phenol C), 63.48 (benzyl C H 2 ) , 57.64 ( C H 2 ) . 2.2.10 X-ray crystallographic analyses Dr. B . O. Patrick determined the structures of the perchlorate salts of the H3[Ni(bcn) 2] + , [LaZn 2 (tam) 2 ] + and [GdNi 2 (bcn) 2 ] + cations. A l l other crystal data were collected and solved by the author. Single crystals of the perchlorate salts of [LaZn 2 ( tam) 2 (CH 3 OH)(H 2 0)]C104, H 3 [Ni (bcn) ] 2 C10 4 »3CH 3 CN, [Cu 3(Hbcn) 2](C10 4) 2 , [YbNi 2 (bcn) 2 ]C10 4 »CH 3 CN, [ G d Z n 2 ( b c n ) 2 ( C H 3 C N ) 2 ] C 1 0 4 ' C H 3 C N , [GdNi 2 (bcn ) 2 (CH 3 CN) 2 ]C10 4 «CH 3 CN and [H 4 bcn]C10 4 , isolated by slow evaporation of the solvent, weremounted on a glass fiber and measurements were made on a Rigaku/ADSC C C D area detector with graphite monochromated M o - K a ; radiation (X = 0.71073 A). A l l calculations were performed using the teXsan crystallographic software package of the Molecular Structure Corporation, and S H E L X L - 9 7 . 4 9 The structures were processed and corrected for Lorentz and polarization effects, and absorption using the d*TREK program. 5 0 The structures were solved by direct 51 52 methods and expanded using Fourier techniques. The non-hydrogen atoms were refined 55 anisotropically. Hydrogen atoms were fixed in calculated positions, but not refined. The S Q U E E Z E function in P L A T O N 5 3 was used for the [LaZn 2(tam)2(CH 3OH)(H20)]C104 to correct the raw data such that electron density in the solvent accessible volumes is effectively ignored. O R T E P diagrams from the X-ray structures are shown in Figures 2.6, 2.8, 2.12, 2.14, 2.19 and 2.20. Cystallographic parameters, selected bond lengths, and interatomic distances are outlined in the Appendix. 2.3 Results and Discussion 2.3.1 The [LnTM2(tam)2]+ and [LnTM2(trn)2]+ complexes Considerable time was spent attempting to synthesize the analogous [LnTM2(L)2] + cationic species ( T M = Mn(II), Fe(II), Co(II), Cu(II), Zn(II), L n = La(III), Nd(III), Gd(III), 3 3 4* Dy(III),Yb(III) and L = trn " and tarn "), analogues of the previously made [LnNi2(tam)2] and [LnNi 2 ( t rn) 2 ] + complexes. 3 8 ' 3 9 [LaZn 2(tam)2]C104'5H 20 was the only trimetallic tam3" species successfully isolated regardless of the utilized solvent (analytical grade and anhydrous M e O H , C H 3 C N , E t O H , acetone, nitromethane and ether), temperature (0°C, R T or reflux), base (NaOH, L i O H , N a H , pyridine, methylamine and triethylamine), metal salts (chlorides, nitrates and perchlorates), aerobic or anaerobic conditions, stoichiometry (adjusting amounts of base, TM(II) salts and Ln(III) salts), or reaction time (ranging from 15 min - 2 weeks). [LaZn2(tam)2]C104»5H20 was synthesized under A r at R T using dry M e O H , CIO4" salts and triethylamine (Scheme 2.1). 56 HO - o < NH 1. TM(C10 4 )2 -6H 2 0, M e O H , RT, 15 min [LaZn2(tam)2] i+ H N HN OH 2. 0.5 eq. Ln(C10 4 ) 3 6 H 2 0 , 3 eq. N E t 3 diluted in M e O H , RT, 30 min Scheme 2 .1 . Synthesis of [LaZn 2 (tam) 2 ] + The elemental analysis (EA) confirmed the proposed stoichiometry for hydrated [LaZn 2 ( t am) 2 ]C10 4 »5H 2 0 where Zn(II) is chelated by the N 3 0 3 of tam3" and La(III) is bound by the 6 deprotonated phenols of the two bridging [Zn(tam)]" moieties. Electrospray ionization mass spectrometry in the positive ion mode (+ESI-MS) gave a diagnostic mass spectrum with the expected isotopic patterns corresponding to the parent ion peak [LaZn 2(tam) 2] + . Peaks in the IR region corresponding to the C10 4 " counterion (1110 cm"1) confirm the presence of the cationic [LaZn 2 ( t am) 2 ]C10 4 »nH 2 0 species. The appearance of new peaks at 600 cm"1 and 440 cm" 1, assigned to VTM-O and VLn-o stretching frequencies, respectively, suggest the complexation of the ligand to the Zn(II) and La(III) metal ions. 5 4 " 5 7 In addition, the broad band for the VOH disappears upon complexation indicating complete deprotonation of the phenolic O H groups. New bands observed at 3400 cm"1 are attributed to coordinated and hydrogen bonded water molecules . 5 8 , 5 9 Vibrations at 3362 cm"1 for coordinated secondary amines and the bathochromic shift (AS = 16 and 19 cm"1) of the 5N-H bending vibrations observed at 1594 and 1562 cm"1 evince a normal coordination mode of the amine nitrogens to the divalent Zn(II) in the N 3 0 3 binding pocket of tam3". In addition, the vc-o peaks are shifted towards lower wavenumbers (A5 = 16 cm"1) with respect to the proligand, 57 attributed to the coordination of the Zn(II) and La(III) ions, further suggesting the chelation of La(III) by the phenolato moieties from the two [TM(tam)]" pre-organized ligands. The structure of the [LaZn 2(tam)2]+ complex was further elucidated by the ' H N M R spectrum. The benzylic hydrogens and the backbone ethylenic CH2 groups are split into disterotopic A B doublets in the ' H N M R spectra confirming the rigidity of the molecule upon complexation of the TM(II) and Ln(III) metal ions (Figure 2.3). H4 H6 H5 H7 J J H 3tarn H5 H7 H4 H6 (Htam)] H5 H7 H4H6 WVL [LaZn 2(tam) 2] + H3 H3 1' NH 1 ( 2 H2 H3A H N — \ " P » 7 6 5 HI \ HI H2 ^ H2 HI H3B 1 1 * 1 * 1 i 1 * * 1 1 1 1 * * 1 1 1 * * 1 1 1 1 1 1 1 1 * 1 1 * 1 1 1 1 1 * 1 1 1 ' ' i 1 1 * * < * 1 * 1 * * 1 1 ' ' i 1 1 * 1 * * * 1 1 1 ' 1 1 i 1 1 1 * 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 LO 0.5 (ppm) Figure 2 .3. ' H N M R spectra of H 3 tam, [Zn(Htam)] and [LaZn 2(tam) 2] +(300 M H z , C D 3 O D , RT, * solvent). 58 A large downfield shift was seen for the hydrogen otho to the phenolato oxygen atom (A5 -0 .13) due to the donation of % electrons from the aromatic benzene to the chelate ring. Strain in the organic backbone is detected by the significant up field shift (A8 - 0.33 ppm) of the methyl group despite its considerable distance from the binding moiety of the tam3" ligand. In addition, only one set of hydrogen resonances were observed suggesting symmetry between the two bridging ligands. Other [LnTM 2 ( t am) 2 ] + species, where T M = Fe(II) and Zn(II) gave +ESI-MS spectra where the parent ion peak solved for the cationic complex with 100 % relative intensity, an example of which is shown in Figure 2.4. Small amounts of the Co(II), Cu(II) and Mn(II) species prepared by refluxing in dry acetonitrile and using either CI" or NO3" salts were also, observed by +ESI-MS (relative intensities < 10 %). Unfortunately, the [LnTM 2 ( tam) 2 ] + cationic complexes could not be isolated regardless of numerous attempts by crystallization in a variety of solvents and co-solvents, column chromatography, and addition of bulky anions such as NaBPL}, N a P F 6 and NaBF4. Slow evaporation of solvent only afforded the thermodynamically stable [TM(Htam)] products as confirmed by +ESI-MS. 100 -, [LaFe 2 ( tam) 2 ] + = 1114 1113.7 a 50 A JJU . J J L Figure 2.4. +ESI-MS spectrum of [LaFe 2(tam) 2] + . 59 The limited success in synthesizing the [LnTM2(tam)2] cationic complexes may be attributed to the steric demands of the tripodal tarn3" ligand, which limit the size of the coordinated metal ion. It has previously been determined that the N3O3 cavity of the amine phenol ligand selectively accommodates A l 3 + , which has an ionic radius of 0.53 A. This is shown by the lack of deformations in the organic framework. 4 1 Nonetheless, the tarn3" ligand has been shown to successfully coordinate Ga(III) and In(III) metal ions in distorted octahedral geometries (ionic radius of 0.62 A and 0.80 A, respectively). The organic framework was significantly stressed upon coordination of these metal ions, as revealed by deviations of the bond angles A , B and C and selected atomic distances from ideal tetrahedral values for the sp 3 carbon atoms (109.5°, C - C = 1.54 A and C - N = 1.47 A, Figure 2.5) 4 1 Figure 2.5. Table of angles for [M(tam)] complexes 4 1 (left) and selected angles (A, B , C) and bond lengths (between the numbered 1, 2 and 3 sp 3 carbons) used to determine deformations of the tarn3" ligand upon coordinating a metal ion. Ni(II), with an ionic radius only slightly larger than the ionic radius of In(III) (0.83 A vs. 0.80 A, respectively), has been shown to form complexes with the F^tam proligand, but with additional deformations seen in the previously determined crystal structure of 60 [Ni(H2tam)(CH3CN)]PF 6 '2 .5CH3CN'0.5CH 3 OH. 3 8 The third phenol arm of the H 2tam" was not able to coordinate to the Ni(II) ion due to the strain within the molecule. Instead, the Ni(II) was bound by one phenolato, one phenol, and three amines. Upon sandwiching the Ln(III) ion between the phenolato moieties of the [Ni(tam)]' ligand, the octahedral geometry around the Ni(II) ion has been shown to become additionally distorted. The N - N i - N angles of the [ L a N i 2 ( t a m ) 2 (CH30H ) 1 / 2 ( C H 3 C H 2 O H ) 1 / 2 ( H 2 0 ) ]C104 •0 .5CH3OH»0.5CH 3CH 2OH*4H 2O complex have previously been determined to become compressed compared to the [Ni(H 2 tam)] + (averaged angles of 89.4° vs. 90.3°, respectively). The averaged O - N i - 0 angles also become distorted upon chelation of the Ln(III) (averaged angle of 83.7° vs. 88.4°, respecively). Fe(II) has only a slightly larger ionic radius than Ni(II) (0.85 A vs. 0.83 A, respectively), which allows for the formation of the [LnFe 2 (tam) 2 ] + species in solution as observed by +ESI-MS. However, it can be assumed that [LnFe 2 (tam) 2 ] + has a very strained core and hence the unstable trimetallic complex cannot be isolated. B y using the ligand strain rationale upon chelation of larger TM(II) ions, it is surprising that the H 3 tam can accommodate a Zn(II) ion with a significantly larger ionic radius of 0.89 A; however, significant strain in the molecule is observed in the *H N M R spectrum of [Zn(Htam)] (Figure 2.3). The [Zn(Htam)] complex was used only for analytical comparison (*H N M R and +ESI-MS) and was not characterized by any other means. A relatively easy synthesis of adding H3tam to a solution of Zn(C104) 2 »6H 2 0 in dry M e O H under A r afforded the off-white [Zn(Htam)], which precipitated upon slow evaporation of the solvent. The solid was filtered, washed with cold M e O H , and redissolved for analysis. 61 Two sets of peaks are visible in the ' H N M R spectrum of [Zn(Htam)]. One set of resonances, which integrate for twice as many hydrogens, exhibit sharp peaks significantly shifted downfield from the proligand. The other set of resonances gives broad peaks for the ring and benzylic protons with the same chemical shift as the free ligand. This suggests that one of the phenol arms is not coordinated to the Zn(II) ion (hence the lack of chemical shifts of the second group of resonances in respect to the proligand). A broad peak at 2.76 ppm and a sharp peak at 2.97 ppm assigned to the CH2 backbone protons are both shifted downfield from the proligand confirming the coordination of all three amines to the Zn(II) atom. From the ' H N M R , it appears that the Zn(II) ion is chelated by a N3O2 binding motif, similar to that observed for the [Ni(H2tam)]+ species. Although Zn(II) ions commonly prefer an octahedral geometry, it is possible that the strain in the organic backbone would not allow for the coordination of the third phenol arm to the Zn(II). Only one observable signal shifted downfield for the methyl group implies the formation of one product in solution. Further deformations in the [Zn(Htam)] species were observed upon accommodating a Ln(III) ion, confirmed in the crystal structure of [LaZn 2 (tam) 2 (CH30H)(H 2 0)]C104 (Figure 2.6). The Ln(III) ion is eight-coordinate bicapped by the bridging phenolate oxygen atoms from the two tridentate [Zn(tam)]" units and coordinated by two solvent molecules; one water and one methanol. The capping units are nearly identical to each other with each Zn(II) ion being encapsulated by a fully deprotonated tam3" ligand via three amine (average atomic distances for Z n - N = 2.137 A for both Zn( l ) and Zn(2)) and three phenolato functionalities (atomic distances for Z n - 0 = 2.128 A for Zn( l ) and 2.137 A for Zn(2)) in a distorted octahedral geometry. The Z n - 0 distances are significantly longer than other TM-bridging phenolate oxygens in the literature displaying the harsh steric demands on the l igand. 6 0 " 6 2 62 The metallic core of [ L a Z n 2 ( t a m ) 2 ] + cation is in a V-shape with a Zn(l)-La(l)-Zn(2) angle of 141.89°. Analogous to the previously determined trimetallic Ni(II)-Ln(III)-Ni(II) species, the geometry around the La(III) ion is best described as a square antiprism distorted towards a C 2 v bicapped trigonal p r i s m . 3 8 ' 6 3 The Z n - L a distances are longer than the analogous N i - L a system (average of 3.220 A vs. 3.172 A, respectively) 3 8 due to accommodating the larger Zn(II) ion; however, the averaged La-O(ligand) bond distances of the Zn-La species are slightly shorter than the analogous Ni(II) system (2.46 A vs. 2.53 A). These are both appropriate lengths for L a - 0 bonds. 6 4" 6 6 Figure 2 .6 . O R T E P diagram of the cation in [LaZn 2 ( tam) 2 (CH 3 OH)(H 2 0)]C10 4 . Hydrogen atoms and perchlorate anion are omitted for clarity. Thermal ellipsoids are drawn at 50% probability and ring carbon atoms are omitted. 63 Distortions of the organic framework were detected in the bond angles between the apical carbon and the methylene carbons of the three chelating arms (angle A , Figure 2.5) compared to sp tetrahedral values. Angle A is slightly compressed (107.5°) and angle B is slightly expanded (111°), while significant deformations are seen in angles C and D with values of 114° and 113°, respectively. In order to sandwich smaller lanthanide ions, the already strained organic framework must further lengthen the bonds between the benzyl methylenes and secondary amines, and expand the angles C and D , leading to extremely unstable d/f complexes, hence the lack of these desired species. Similar steric strain is hypothesized for the [LnTM2(trn)2]+ systems. This explains the unsuccessful isolation of the desired trn3" metal complexes regardless of synthetic modifications or isolation techniques. +ESI-MS parent peaks with 100% intensity suggest that [LnTM 2 ( t rn) 2 ] + species, where T M = Mn(II) and Zn(II), were formed in solution shown in Figure 2.7. However, +ESI-MS verified that the species rapidly decompose to the [TM(Htrn)] upon removal of the solvent. . HO HO NH HN. OH 1. Mn(Cl) 2 6 H 2 0 , dry MeOH, reflux under Ar, 2 h 2. 0.5 eq. Ln(Cl) 3 6 H 2 0 , 3 eq. N E t 3 diluted in dry MeOH, reflux under Ar, 2 h 1. Zn(C10 4 ) 2 6 H 2 0 , dry MeOH, reflux under Ar, 2 h 2. 0.5 eq. Ln(C10 4 ) 3 6 H 2 0 , 3 eq. N E t 3 diluted in dry MeOH, reflux under Ar, 4 h [LnMn 2(trn) 2] + [LnZn2(trn)2] S c h e m e 2 .2 . Synthesis of [LaTM 2 ( t rn) 2 ] + , T M = Mn(II) or Zn(II). 64 [LaMn 2 ( t rn ) 2 ] + = 1172 11722 1193.6 [LaZn 2 ( t rn) 2 ] + = 1193 - U • 1 i.L . -I 1 _ . - . . . . . . . . . „_LJ.. « Figure 2 . 7 . +ESI-MS spectra of [LaTM 2 ( t rn) 2 ] + , T M = Mn(II) and Zn(II). The 'FI N M R spectrum of [LaZn 2(trn)] + aided in the elucidation of the compound structure in solution. Further shifting of the proton resonances in the [LaZn 2(trn)] + from the ' H atom resonances in the proligand or the [Zn(Htrn)] complex spectra suggest the coordination of La(III) to the [Zn(trn)]" ligand (Table 2.1). In addition, only one set of resonances is seen for the trimetallic complex, suggesting 3-fold symmetry system. The coordination of the TM(II) and Ln(III) by the phenolato arms is shown by the large downfield shift observed for the hydrogen ortho to the phenolato oxygen atom (A8 - 0 . 1 7 ppm, see H4 in Table 2.1). A l l other aromatic resonances were shifted downfield from the [Zn(Htrn)] species due to the additional La(III) metal ion removing electron density from the ring (A5 - 0.02 ppm). Broadening of the peaks for the backbone and the benzylic protons suggests fluxionality in the [LaZn 2 (trn) 2 ] + species, which may be attributed to steric strain in 65 the ligand and the resulting observed instability of the complex. The non-aromatic signals are also shifted significantly downfield. Table 2.1. ' H N M R data for H 3 t rn proligand and complexes (300 M H z , RT, MeOD) . Hydrogen resonances are listed in ppm. Protons H 3 t r n a [Zn(Htrn)] a ' b [LaZn 2 ( t rn) 2 ] + a ' b ' c H I 2.54 (t) 2.67 (br), 2.81 (br) 2.79 (br) c H2 2.63 (t) 2.71 (br), 2.85 (br) 2.79 (br) c HO^Q H3 3.79 (s) 3.79 (br), 4.31 (br) 4.29 (br) c ( H H4 H5 6.71 (d) 7.00 (t) 6.67 (br), 6.84 (br) 7.0 (br), 7.15 (br) 6.88 (d) NH—1 7.17 (t) L M H6 6.69 (d) 6.67 (t) 6.75 (br) O H 6.77 (br) .. H7 7.08 (t) 7.0 (br), 7.15 (br) 7.20 (d) 2 5 Abbreviations: s = singlet; d = doublet; t = triplet; br = broad. J H H = 8.2-7.0 H z for ring protons; 3 J H H = 5.6 H z for ethylinic protons. b br, vi/2 = 23 Hz . °br, vi/2 = 45 Hz . The ' H N M R spectra for the mononuclear [Zn(Htrn)] is very similar to the spectrum for [Zn(Htam)], where the Zn(II) is coordinated by only two of the phenolate arms (the [Zn(Htrn)] species was prepared in moderate yields by the addition of Zn(C104) 2 -6H 2 0 in dry M e O H to H 3 t rn under A r . The solution was refluxed for 2 h affording an off-white precipitate upon slow evaporation of the solvent. The solid was filtered, washed with cold M e O H , and redissolved for +ESI-MS and ' H N M R analysis). The additional set of 66 resonances seen in ' H N M R spectra for the mononuclear [Zn(Htrn)] is due to the uncomplexed phenol oxygen with similar chemical shifts as the proligand (Table 2.1); however, both sets of resonances assigned to the ethylenic protons, H I and H2, are shifted downfield suggesting the binding of all four amines to the Zn(II) metal. No evidence of the formation of [LaTM2(trn)2J+ was observed in the +ESI-MS for T M = Co(II), Fe(II) or Cu(II). Interestingly, a red-rust precipitate formed immediately upon addition of the Fe(II) metal ion. This species was determined to be the sodium adduct of the [Fe(trn)] by +ESI-MS proposed to be formed by the one-electron oxidation of Fe(II) to Fe(III). A similarly oxidation of iron has been reported for the l,4,7-triazacyclonane-l,4-diacetate Fe(II)/Fe(III) complex. 2.3.2 Synthesis of H^bcn JV-Functionalized 1,4,7-triazacyclononanes (tacn) have been known to form stable complexes with various metal ions since the 1970's, making these ligands attractive for a variety of uses including catalysis, 6 7" 6 9 models for metallobiosites, 7 0 photo-responsive devises, 7 1 M R I contrast agents,7 2 and radiotherapeutic agents.7 3 The coordinating properties of this macrocycle can further be exploited for d/f magnetic studies by attaching phenolate arms to the secondary amines of tacn. The resulting ligand is capable of binding a TM(II) in the N3O3 pocket and a Ln(III) between the phenolates of the resulting [TM(bcn)]" complexes to form compounds analogous to [LnTM2(L) 2] +, L = tam3" or trn3". Unfortunately, polyazamacrocycles are notoriously difficult to synthesize. The most common route is the Richmond-Atkins cyclization using linear polyamine salts, usually 67 protected as sulfonamides, coupled with a tosylated diol in anhydrous D M F A number of other synthetic routes toward the macrocyclic polyamines have been developed, but have not succeeded in minimizing synthetic steps or drastically improving the moderate yields. 7 4 " 7 8 Thus, the Richmond-Atkins route was used to synthesize H 3 bcn with modifications as outlined, resulting in a yield higher than that previously published 4 5 outlined in Scheme 2.3. Scheme 2 .3 . Synthesis of H 3 bcn, where Ts = P - C H 3 C 6 H 4 S O 2 : (a) N a O H , H 20/ether, RT, 70 %; (b) pyridine, 0 ° C , A r , 50 %; (c) anhydrous D M F , 60% N a H , 70°C, A r , 94 %; (d) cone. H2SO4, 160°C, A r ; (e) H 2 0 / N a O H , C H 3 C 1 , RT, 72 %; (f) dry E t O H , salicylaldehyde, N a B H ( O A c ) 3 , R T , A r , 82%. The main differences between the previously published tacn synthesis and the protocol used in this work stemmed from synthesizing the bis-p-toluenesulfonamide salt in situ using 60 % N a H in D M F . Previously, the disodium salt generated by the addition of sodium ethoxide was isolated and redissolved in D M F prior to the coupling reaction with the ditosylated diol. Heating the coupling reactions at slightly lower temperature (70° vs. 100°) 68 for longer time periods (20 h vs. 2 h), increasing the yield. ' Similar harsh conditions using concentrated H2SO4 required to remove the tosyl groups were used, but at higher temperatures (160° vs. 100°) for a shorter time period (30 min vs. 48 hr), resulted in complete detosylation of tacn with minimized decomposition. The phenol arms were successfully attached to the polyamine macrocycle by reductive alkylation instead of utilizing the more common Mannich reaction or nucleophilic substitution involving the appropriate benzyl halide to functionalize the secondary amines of tacn. Sodium triacetoxyborohydride was the only reducing agent found to prepare the desired FFibcn in moderate to high yields, with minimal side-products. H 3 bcn was recrystallized in hot methanol before further use. A crystal of the CIO4" salt of FLtbcn* was isolated and the structure was solved (Figure 2.8). Unfortunately, the proligand was extremely disordered due to the many different low energy conformations of the modified cyclononane tacn backbone. Only isotropic refinements were used for all atoms in the unit. In addition, the placement of the fourth hydrogen could not be located and was assumed to be fluxional around the N and O of the ligand. Nonetheless, a majority of the electron density could be correlated to the structure shown in Figure 2.8 where the tacn backbone is in a twist-boat-chair conformation, using 70 nomenclature proposed from Glaser et al. The bulky phenol moieties along with the disordered tacn backbone are also thought to distort the molecule leading to large C-C bond lengths and angles for some of the connectivities. 69 Figure 2 .8. O R T E P digram of cation in [HUbcnJClO^ Hydrogen atoms and the perchlorate anion are omitted for clarity. Thermal ellipsoids are drawn at 50% and are omitted for ring carbon atoms. 2.3.3 Synthesis and characterization of [TM(Hbcn)]'nH~20 complexes Transition metal complexes were synthesized using one equivalent of T M ( C 1 0 4 ) 2 ' 6 H 2 0 ( T M = Zn(II), Cu(II) or Ni(II)) and H 3 bcn in either methanol or acetonitrile. Elemental analysis was consistent with a stoichiometry of [TM(Hbcn)]»H 2 0, confirming the formation of a neutral encapsulated products, where T M = Zn(II), Cu(II) and Ni(II). For all three of the metal complexes, a parent peak with 100%) relative intensity corresponding to the proton adducts of the [TM(Hbcn)] complexes was observed in the +ESI-MS. A rigid octahedral geometry where the three phenolate oxygen donors and three amine nitrogens are tightly bound to the Zn(II) ion, is postulated from the sharp resonances observed in the ' H N M R data of [Zn(Hbcn)] shown in Figure 2.9. The ring hydrogens, H3-6 70 are split into four distinct signals, all significantly shifted. The largest upfield shift was observed for H3 , the hydrogen ortho to the chelating phenolate oxygen (A8 -0.79 ppm). Splitting of the singlet from the benzylic protons into A B doublets upon complexation additionally confirms the rigidity of the complex. The diastereotopic hydrogens are further distinguished by assuming that'the proton in the pseudoaxial position, H 2 B , is more shielded than is the benzylic proton in the pseudoequatorial environment, H 2 A , due to the proximity of the collinear p-orbital from the neighboring benzene ring. The excess shielding causes the pseudoaxial hydrogen to resonate at lower frequencies (3.39 ppm vs. 4.23 ppm of the pseudoequatorial proton) 80 J H I H O H2 ' / \ V - N , IS 2 O H - 6 4 H , b c n , O H m H5 H3 H4I J w H 2 A H2B .L-Wv. [Zn(Hbcn)] H I A ' H I A j H I B ' H I B i i i r~i i i i i i i i | i i i i i i i i i i i i i r~i i i i i | i i i i [ i i i i j i i i i ] i n i | i i i i | i i i i \ i i i i | r~ 7.5 7.0 6.5 6.0 5.5 5.0 . 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm Figure 2.9. ' H N M R spectra of H 3 bcn and [Zn(Hbcn)] (300 M H z , C D C 1 3 , R T , *solvent). 71 The conformational constraints on the ethylenic backbone imposed by the chelation of Zn(II) renders the ethylenic hydrogens magnetically inequivalent (Figure 2.9 and 2.10). The singlet assigned to the ethylenic backbone ' H atoms in the proligand is split into an A A ' B B ' multiplet in the complex ' H N M R spectra exhibiting first- and second-order couplings upon complexation. 8 1 H I A and H 1 B hydrogens are assigned as the 'axial ' hydrogens; H 1 A ' and H 1 B ' are assigned to the 'equatorial' hydrogens. These assignments are made based on the additional couplings of the HI A and H 1 B protons to the inequivalent 'equatorial' hydrogens resulting in a doublet of doublet of doublets. 8 2 A smaller coupling constant is expected for the protons from the 'equatorial' hydrogens ( 3 / H H < 1 Hz) due to the dihedral angles between the H I A ' - C - C - H I B or H I A ' - C - C - H 1 B ' ( H I B ' - C - C - H I A or H 1 B ' -C - C - H 1 A ' ) being near 90°, resulting in only a doublet of doublets visible at 300 M H z (Figure 2.10). Assignments were confirmed by ' H - ' H C O S Y experiments. H1A —i « • « > 1 • « • • 1— 3.0 2.5 2.0 ppm Figure 2.10. Expanded section of the alkane region for the ' H N M R spectra of [Zn(Hbcn)], assigned based on the twisted conformation of the ethylenic backbone depicted on the right. 72 Interestingly, Ni(II) appears to form two different complexes with ben3", depending on the solvent used. A purple paramagnetic material was obtained when synthesized in acetonitrile, whereas a yellow diamagnetic material precipitated using the analogous reaction in methanol. The purple material is not soluble in alcohols and only sparingly soluble in acetonitrile, whereas the yellow material is soluble in both solvents. Significant differences between the two Ni(II) species were not observed by +ESI-MS and the only difference detected in the fingerprint region of the IR is a prominent band at 1061 cm"1 for the purple material. This peak correlates with a CIO4" counterion suggesting the formation of a cationic complex for the purple Ni(II) species. This is confirmed by E A , where the purple complex is consistent with H3[Ni(bcn)] 2C104. The yellow species has the stoichiometry of [Ni(Hbcn)]-H 2 0. The absence of resonances corresponding to the free ligand and the shifting of the resonances in the 'IT N M R spectrum confirm the coordination of Ni(II) to the polyamine phenol ligand for the yellow diamagnetic species (Figure 2.11). The multiplets in the aromatic region, the broad peaks for the organic framework, and the small chemical shift changes relative to the free ligand (A8 < 0.02 ppm) suggest exchange of the amine nitrogens and phenol oxygens around the metal on the N M R time scale, making elucidation of the coordination sphere around the metal ion difficult. Elevating or lowering the temperature resulted in no significant sharpening, broadening or coalescence of the resonances, implying a fluxional process in solution with a low energy barrier between interconverting states. The broadening of the hydrogen resonances may also be complicated due to the exchange of the coordinated amine nitrogens and phenol oxygens with surrounding solvent molecules. The ' H N M R data/spectrum from the paramagnetic purple Ni(II) species was uninformative. 73 * H4 H6 J H3 H5 H2 H2 HI H I [Ni(Hbcn)] OH 6 5 , 0 H 3 H 3bcn I I i i i i | 1 1 "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 | I I I I | I I I I | I I I I | I I I I | I I 1 I | I 1 I I :0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm Figure 2.11. ! H N M R spectra of [Ni(Hbcn)] and H 3 bcn (300 M H z , C D 3 O D , RT, *solvent). A solid state structure of the purple paramagnetic material H3[Ni(bcn)]2Cl(V 3CH3CN, shown in Figure 2.12, revealed Ni(II) encapsulated, in a distorted octahedral environment by the ben " ligand, via three phenolate O groups and three amine N functionalities. Each of the two independent Ni(II) moieties is located near an inversion center with approximate C3 symmetry about an axis perpendicular to the plane of the ben 3 ligand. The N c i s - N i - N c i s "bite" angle (82.6°) about the Ni(II) center varies from the 90° idealized for an octahedral structure due to the constraints imposed by the five-membered 74 chelate ring of the tacn moiety; however, the N C j S -N i -0 C j S angle is close to 90° (89.1°), suggesting that the six membered ring between the phenolate O and the amine N is large enough to accommodate Ni(II) in an octahedral environment. Figure 2.12. O R T E P diagram of the cation in H3[Ni(bcn)] 2 C104'3CH 3 CN. Hydrogen atoms, solvent molecules and the perchlorate anionic salt are omitted for clarity. Thermal ellipsoids are drawn at 50%. The average N i - 0 and N i - N bond lengths do not differ between the two [Ni(bcn)]" units (Ni ( l ) = 2.044 A and 2.141 A, Ni(II) = 2.031 A and 2.137 A, respectively); however, the N i - N bonds are slightly longer than those observed for [Ni(tacn)2]2+ (2.139 A vs. 2.116 A, respectively), and significantly longer than optimum values for an octahedral NiN6 structure of 2.08 A. The elongation of the N i - N bonds are correlated with the less strongly 75 coordinating tertiary nitrogens and the steric demands on the ligand upon accommodation of the Ni(II) ion, similar to that seen in a N i 4 L species, where L = tetrakis(l ,4,7-triazacyclonon-1-yl). 8 4 Alternatively, the average N i - 0 bond lengths (2.014 A) in the [Ni(bcn)]" species are very similar to other deprotonated phenoxide species. 7 ' 3 9 It is also interesting to note that the N i - N and N i - 0 distances in the axial positions are slighter longer than the comparable bonds in the equatorial positions (2.050 A and 2.136 A vs. 2.021 A and 2.132 A, respectively) attributed to steric strain of the ligand. Three bridging hydrogens equidistant to the six phenolic oxygen atoms were found between the two encapsulated Ni(II) structures. The dimers have intermolecular 0 - 0 contacts (2.705 A) consistent with other 0 - H . . . 0 hydrogen bonds in the literature (near 2.8 A); however, the exact location of the O - H hydrogen atoms could not be located due to the three hydrogens being disordered over the six possible sites. The large thermal ellipsoids on the oxygens, expected for intermolecular hydrogen bonding, suggest that the hydrogen atoms reside between the two halves of the molecule rather than being bound to one half of the complex. One perchlorate and three non-coordinating acetonitrile molecules were also found in the unit cell. Based on the diamagnetic properties, it is postulated that the yellow [Ni(Hbcn)] material is a square planar Ni(II); this hypothesis is strengthened by analyzing the U V - v i s data of the yellow material. One weak band resolved at 586 nm, due to the 1 A j g —> ' A 2 g transition, matches literature values for Ni(II) in the square planar conformation. 4 3 ' 8 5 The purple Ni(II) complex has absorbances characteristic for an octahedral environment, arising at 544 nm and 900 nm from 3 A 2 g ( F ) -> 3 T i g ( F ) and 3 A 2 g ( F ) -»• 3 T 2 g ( F ) transitions, respectively. ' 76 It is also interesting to note that the yellow material can readily be converted to the purple material by dissolution and refluxing in a variety of solvents, including acetonitrile. The purple material, however, could not be converted into the yellow diamagnetic material indicating that the yellow complex is an intermediate. This point is confirmed by allowing a sample of the yellow Ni(II) material to stand in methanol for several weeks. During this time, the solution turns red and purple crystals deposit that solve for the same stoichiometry as the purple paramagnetic Ni(II) complex. A green crystal of [Cu(Hbcn)] was also obtained and its crystal structure was found to be analogous to the previously solved Cu(II) species. 4 3 Three amine nitrogens and two phenolate oxygen atoms from the trisubstituted tacn ligand chelate the Cu(II) ion in a distorted square-based-pyramidal geometry. 2.3.4 Synthesis and characterization of [TMi(bcn)2]*nH20 complexes Addition of one equivalent of T I V ^ C I O ^ ^ F ^ O to two equivalents of deprotonated [TM(bcn)]" formed trimetallic transition metal species with the expected stoichiometry of [ T M 3 ( b c n ) 2 ] » / 7 H 2 0 as confirmed by E A (trimetallic Cu(II) species solved for [Cu3(Hbcn)2](C104)2*H20). Although the product peak was observed with the expected isotope pattern in +ESI-MS, only a 10 % relative intensity was resolved, suggesting that the stability of the neutral [ T M 3 ( b c n ) 2 ] » / 7 H 2 0 complexes was significantly reduced compared to that of the respective [ T M ( H b c n ) ] « « H 2 0 species (the parent peak with 100% relative intensity corresponded to the proton adduct of the [TM(Hbcn)] complexes). The only differences observed in the IR spectra of the monometallic and the trimetallic TM(II) species 77 was the bathochromic shift of the vc-o bands attributed to decreased strength in the phenolate oxygen O - T M dative bond from the oxygens bridging an additional metal. Only one set of resonances is seen in the ! H N M R spectra for [Zn 3(bcn) 2], implying a 3-fold symmetric complex. A n additional small upfield shift of the hydrogen (H3) ortho to the phenolate oxygen (AS -0.01) compared to the [Zn(Hbcn)] spectra (Figure 2.13) was observed. Slight downfield shifts and larger couplings between the protons on the ethylenic bridge are also seen suggesting an increasingly strained environment upon bridging the third Zn(II) metal ion (A5 - 0.11 ppm, A 3 J ~ 1.2 Hz, A 2 J ~ 0.8 H z for H I A ) . H6 J H3 H6 m H3 il ¥A [Zn 3(bcn) 2] H2A H2A [Zn(Hbcn)] f\->. < 2 H2B o i l 6 j V 3 4 H,bcn HI A ' H1B". HI A L T A H1B H2B HI A HI A' HIE' HIE 7fl 65 r i "|" s'! i I n i i i i n r i r r j \ r i i | - r T i i i " " r , , " i i i r t i i ' t I i ~ t r 6.0 SS S.fl 4.5 4.0 3.5 3.0 25 2,0 ppm Figure 2.13. ' H N M R spectra of [Zn 3(bcn) 2] and [Zn(Hbcn)] (300 M H z , C D C 1 3 , R T , * solvent). A n octahedral coordination environment for [TM 3 (bcn) 2 ] is inferred from the U V - v i s spectrum for T M = Ni(II) and H N M R spectra for T M = Zn(II). The U V - v i s spectrum of 78 [Ni3(bcn)2] is nearly superimposible upon that of the octahedral monomeric [Ni(Hbcn)] complex, except for the broadened charge transfer band centered around 303 nm. The crystal structure of [Cu 3 (Hbcn)2](C104) 2 '2CH 3 OH indicates three chelated Cu(II) ions, each 2.96 A apart (Figure 1). The two terminal Cu(II) centers, related by an inversion centre at the bridging Cu(II) ion, are arranged in a distorted square-based-pyramidal environment chelated by three amine nitrogens and two phenolate oxygen atoms. The central Cu(II) is chelated in a square planar geometry by four bridging phenolate donors, all in the same plane with an appropriate average bond distance of 1.94 A. Figure 2.14. O R T E P diagram of the cation in [Cu3(Hbcn) 2 ]C104»2CH 3 OH. Hydrogen atoms, solvent molecules, and perchlorate anion are omitted for clarity. Thermal ellipsoids are drawn at 50% probability. The O-Cu-0 bond angles around the central Cu(II) ion for the two bridging phenolates of the same [Cu(Hbcn)] moiety are appreciably compressed at 77.23°. Alternatively, bond angles between the two capping units are significantly expanded 79 (102.77°), forming a cluster where the alternate bridging O atoms are exactly 180° from each other. The two terminal Cu(II) ions have characteristic basal C u - 0 and C u - N bond lengths of 1.93 A and 2.02 A, respectively. The basal N - C u - N , O-Cu-0 and N - C u - 0 bond angles are representative of a distorted square-planar system. Deviating from 90° and 180° for regular square planar geometry, the N C j S -Cu-N C j s and OCJS-CU-OCJs angles exhibit an averaged value of 80.51° while an averaged value of 163.58° was observed for the trans N - C u - 0 bond. The significantly longer apical C u - N bond distances of 2.27 A, similarly seen in the [Cu(Hbcn)] crystal structure, are also characteristic of square-based-pyramidal Cu(II) complexes. 4 1 A n averaged value of 98.46° was observed for the angles between the apical C u - N bond and the C u N 2 0 2 square-planar base with the angles between the Nbasai-Cu-NapjCai significantly compressed due to the geometric constraints of the ligand (averaged value of 83.33°). The third phenolic pendent arm, not coordinated to the Cu(II) ion, is bent away from the CUN3O2 coordination sphere. 2.3.5 Synthesis and characterization of [Lax(bcn)y]nJr complexes (x = 1, 3; y — 1, 2; n = 0, 3) Methanolic reactions of one equivalent of hydrated La(C104)3»6H20 reacted with the deprotonated ben " ligand afforded a yellow solution from which a white powder was isolated in low yields. Analytical and spectroscopic methods confirmed the proposed stoichiometry of [La(bcn)]»H20 for the isolated precipitate where the La(III) is postulated to be chelated by the three amines and 3 phenolate groups of ben3". The R T ' H N M R spectrum of the complex consisted of four resonances similar to the spectrum seen for the [Ni(Hbcn)] (Figure 2.15). 80 Compared to the proligand ' H N M R resonances, the ring and benzylic proton resonances are slightly shifted upfield (A5 < 0.02 ppm and ~ 0.04 ppm, respectively) from the proligand. The ethylinic protons are split into doublets of triplets, with appropriate coupling constants to suggest that each proton is split by the neighboring vicinal protons in addition to the geminal hydrogen. The absence of additional splitting of the ethylenic hydrogens suggest less twisting of the nine-membered ring upon chelation of the La(III) than for the [Zn(Hbcn)] species. • • ID HI H5 H6 i A K [La3(bcn)2]^  V 2 • B A H3bcn H3 H4 H5 Hf H2 I HI A •. 1 I 1"> 1 ' 'I 1 1 1 1 I 1 1 ' 1 I"1 1 1 ' 1 I 1 1 1 "I 1 ' 1 1 I ' ' 1 1 ' I I "* ' 1 1 ' 1 I I | ' I 'I ] I | I 1 I I ] I I I I I I I I I I I I' V 'I S.O 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 -.1.5 1.0 0.5 ppm Figure 2.15. ! H N M R spectra of [La 3 (bcn) 2 ] 3 + and [La(bcn)] (300 M H z , C D 3 O D , R T , * solvent). If base is not used to deprotonate the phenolates of the H 3 bcn , an off-white powder is isolated that solved for the stoichiometry of H 3[La(bcn) 2] by E A . The absences of CIO4" vibrations in the IR spectrum suggest a neutral species where only three of the six phenolate 81 arms are deprotonated. It is hypothesized that the species consisted of a La(III) ion bridged between two macrocyclic ligands. This species is thought to be extremely unstable due to the large size of the chelate ring and the exposure of the bridged La(III) to potentially coordinating solvent molecules. The instability of H 3 [La(bcn) 2 ] is confirmed by the ' H N M R and +ESI-MS spectra, where only peaks corresponding to the proligand appeared upon dissolving the sparingly soluble product in a variety of N M R solvents. Upon adding an additional equivalent of La(C104)3*6H 20 to the [La(bcn)] moiety, a deposited white powder had the stoichiometry [La 3 (bcn) 2 ](C10 4 )3»H 2 0. The +ESI-MS spectrum confirmed the presence of the trimetallic La(III) species with the expected isotope pattern. In addition, the sharp singlet assigned to the benzylic protons and a broad singlet assigned to the ethylenic protons in the ' H N M R spectra are shifted slightly upfield (A8 -0.03 ppm) compared to the ' H N M R spectra of the [La(bcn)] complex (Figure.2.15). The differences in the spectrum are attributed to the steric strain from the ligand attempting to accommodate three large La(III) ions. The proton resonances assigned to the aromatic ring are also slightly shifted downfield due to the increased donation of electron density from the phenolates to accommodate the extra metal. The broad resonances are attributed to the continuously changing coordination sphere around the metal ions from the exchanging labile ligands and solvent molecules. 2.3.6 Synthesis and characterization of [LnTM2(bcn)2]ClC>4 » n H 2 0 Air-stable complexes of the trinuclear [LnTM 2 (bcn )2 ]C10 4 ' / 7H 2 0 unit were produced using protocols adopted from the synthesis of analogous [LnNi 2 ( tam) 2 ] + and [LnNi 2 ( t rn) 2 ] + 82 cationic species ( T M = Cu(II), Ni(II), Zn(II), L n = La(III), Nd(III), Gd(III), Dy(III), 3 8 39 Yb(III)). ' Spontaneous, self-assembled bicapped structures, where the Ln(III) is bridged between the three deprotonated phenolate oxygens from two [TM(bcn)]" pre-organized ligands, were readily formed. The amount of base added is instrumental to the formation of these complexes; the desired d/f/d complexes could not be isolated i f a slight excess or shortage of the base was used. The complexes precipitated from the respective reaction mixtures affording light purple, off white or green powders for the Ni(II)-, Zn(II)- or Cu(II)-Ln(III) species, respectively (Scheme 2.4). Scheme 2 .4 . Synthesis of [LnTM 2 (bcn) 2 ] + , T M = Cu(II), Ni(II), Zn(II) and L n = La(III), Nd(III), Gd(III), Dy(III), Yb(III). Attempts to recrystallize the complexes (in hot M e O H / H 2 0 or CH3CN/H 2 0) or purification by column chromatography resulted in decomposition of the product and presumed hydrolysis of the Ln(III) ions. Washing the precipitated solid thoroughly with cold solvents to remove unreacted ligand, additional Ln(GC>4)3 and TM(C104) 2 salts, or other side products, was the only way to yield complexes pure by elemental analysis ( if the trimetallic d/f/d species did not crystallize upon standing in solution). The compounds were HO' HO 1 • TM(C10 4) 2 6H 2 0, MeOH or CH 3 CN, reflux, 1 h 2. 0.5 eq. Ln(C104)3 6H 20, 3 eq. NEt 3 diluted in MeOH or CH 3 CN, stir for 30 min [LnTM2(bcn)2]+ 83 characterized by E A , +ESI-MS, IR, and N M R (only of the diamagnetic 1) . A l l results are consistent with the formation of cationic d/f/d species shown in Table 2.2. Table 2.2. Analytical data for [LnTM 2(bcn) 2]C104*/7H 20 complexes. % C % H % N Complex found Calcd. found Calcd. found Calcd. +ESI-MS [LaZn 2 (bcn) 2 ]C10 4 ' 4H 2 0 48.90 48.76 4.82 5.15 6.81 6.32 1157 [ N d Z n 2 ( b c n ) 2 ] C l C v 3 H 2 0 49.08 49.23 5.11 5.05 6.52 6.38 1163 [ G d Z n 2 ( b c n ) 2 ] C l C v H 2 0 50.06 50.10 5.20 4.83 6.68 6.49 1176 [ D y Z n 2 ( b c n ) 2 ] C l C v H 2 0 50.79 50.60 5.12 4.72 6.96 6.56 1181 [YbZn 2 (bcn ) 2 ]C10 4 'H 2 0 50.58 50.19 4.87 4.68 6.45 6.50 1192 [LaCu 2 (bcn) 2 ]C10 4 ' 2H 2 0 50.65 50.26 5.06 5.00 7.06 6.51 1155 [NdCu 2 (bcn) 2 ]C10 4 ' 2H 2 0 50.04 50.05 5.12 4.98 6.72 6.48 1160 [GdCu 2 (bcn) 2 ]C10 4 ' 3H 2 0 48.19 48.10 5.20 5.01 6.18 6.33 1173 [DyCu 2 (bcn ) 2 ]C10 4 'H 2 0 48.15 48.26 4.87 5.08 6.43 6.22 1179 [YbCu 2 (bcn ) 2 ]C10 4 ' 2H 2 0 49.11 49.29 5.05 4.87 6.61 6.34 1188 [LaNi 2 (bcn ) 2 ]C10 4 »3H 2 0 49.71 49.93 5.25 5.12 6.51 6.47 1145 [NdNi 2 (bcn) 2 ]C10 4 ' 2H 2 0 49.14 49.28 4.79 4.75 6.18 6.39 1150 [ G d N i 2 ( b c n ) 2 ] C l ( V H 2 0 50.41 50.62 4.91 4.88 6.89 6.56 1163 [ D y N i 2 ( b c n ) 2 ] C 1 0 4 ' H 2 0 50.04 49.84 5.02 4.72 6.32 6.46 1169 [YbNi 2 (bcn) 2 ]C10 4 50.76 50.71 4.79 4.73 6.71 6.57 1179 84 The elemental analyses for all the lanthanide complexes agreed with the empirical formula [LnTM 2 (bcn)2]C104»rcH 2 0. Hydrated species have been observed with previously synthesized Ln(III)-containing systems where coordinated solvent molecules are associated with the metal in order to fulfill the high coordination number preference of the Ln(III) ions. Regardless of the TM(II) or the Ln(III) used, the IR spectra for the [ LnTM2 ( b c n ) 2 ] C l C v « H 2 0 complexes were nearly superimposable in the 400-3500 cm"1 range, suggesting the formation of isostructural complexes in the solid state. The small < 10 cm"1 variations in the IR peak frequencies between Ln(III) ions with the same ligand are attributed to mass differences amongst the metal ions (Figure 2.16). Vibrations for coordinated reaction solvents, specifically methanol (VCH and VOH stretches between 2800-2700 and 3700-3600 cm" 1, respectively) and acetonitrile (v 2 for C N , v 3 for H C N , and v 4 for C C from 2280-2320 cm"1) seen in the solid state crystal structures are not observed in the IR spectra. 8 6 Instead, the coordinating methanol and acetonitrile are replaced with ubiquitous water suggested by the new broad band at -3450 cm"1 and a narrow band at 600 cm"1 attributed to the O - H stretching vibrations of coordinated water molecules . 5 8 ' 5 9 These peaks appear at shorter wavenumbers than those observed for free water. The exchange of labile coordinating reaction solvents for water upon standing in air is also confirmed by EA. The broad band at -3150 cm"1 is correlated to hydrogen bonded O H groups either between the hydrogen atoms of the coordinated water and/or the oxygen atoms of the ligand chelate, as previously observed in other hydrated Ln(III) complexes. The broad band for the VOH group would be hidden by the coordinated water; however, the absence of VOH vibrations at 1400 cm"1 seen for the [ LnTM2 ( b c n )2 ] C 1 0 4 * « H 2 0 species suggest that all phenol groups are deprotonated. The sharp absorption peaks near 1280 cm"1 and 1100 cm"1 for the vc-o 85 vibration and 1450 cm" for VC-N stretching bands are shifted 20-50 cm" to lower energy with respect to the proligands, evincing normal coordination of the metal ion via the amines and deprotonated phenol oxygens. The vc-o is slightly more bathochromically shifted due to the phenol oxygen bridging two metals compared to the amine nitrogen coordinated to only one metal. In addition, the appearance of new peaks at -622 cm' 1 and between 400-500 cm"' for VTM-O and Vi„.o stretching frequencies, respectively, confirm the complexation of ben3" to the metal ions. 5 4 " 5 7 The bands at 3000-2860 c m - 1 and 1550 c m - 1 for the C H and aromatic ring C=C stretching vibrations did not significantly shift upon coordination of the metal ions. In addition, the presence of v 3 of CIO4" at 1068 cm"1 for all the complexes suggests the formation of cationic complexes. [LaCu 2 (bcn ) 2 ] a o 4 [NdCu 2(bcn) 2]ao<, [GdCu 2 ( b cn ) 2 ]C I0 4 [DvCu 2(bcn)2]CI04 [ Y b a i 2 (bcn ) 2 ]C I0 4 [ L a N i 2 ( b c n ) 2 ] Q 0 4 [YbNi 2(bcn)2]CIO„ [ L a Z n 2 ( b c n y c i 0 4 [ Y b Z n 2 ( b c n ) 2 ] Q 0 4 3500 3000 2500 2000 waven umbers 150a 1000 500 Figure 2.16. IR spectra of selected [LnTM 2 (bcn ) 2 ]C10 4 »«H 2 0 . 86 Electrospray ionization mass spectrometry in the positive ion mode was the primary diagnostic tool for the detection of the cationic d/f/d clusters, giving diagnostic mass spectra with the expected isotopic patterns. The mass spectra are devoid of any other significant peaks besides the parent ion peak for the cationic complex, [LnTM 2 (bcn ) 2 ] + (Figure 2.17). The isotope patterns were also in good agreement with the calculated patterns. [La2n 2(bcn) 2]+= 1159 1159.: 1159.219 1159.2 1157.3 1158.2 I Calculated Experimental Figure 2.17. +ESI spectra of [LaZn 2 (bcn) 2 ] + . The figure on the bottom right is an expanded portion of the +ESI-MS spectrum displaying the isotope pattern closely matching the theoretical isotope pattern displayed on the bottom left. 87 The two TM(II) in the [ L n T M 2 ( b c n ) 2 ] + cationic complexes appear to be in comparable coordination environments in solution, as reflected by the absence of splitting of the absorbance peaks in the U V - V i s . The absorptions at 900 nm and 544 nm for the octahedral Ni(II) are associated with 3 A 2 g ( F ) —• 3 T 2 g ( F ) and 3 A 2 g ( F ) —> 3 T ] g ( F ) transitions, * 38 39 3 3 respectively. ' The A 2 g ( F ) —> T] g(P) transition is obscured by a strong charge-transfer band at 303 nm. A n - n* transition of the K band of the benzene ring appears at 249 nm. 3 6 The spectrum of the d/f/d species is very similar to that of the octahedral [Ni(Hbcn)] system, suggesting that the incorporation of the Ln(III) does not significantly distort the coordination environment around the Ni(II) ion. The coordination geometry around the Cu(II) center appears to change upon coordination to the Ln(III), as monitored by U V - v i s . The K band of the ligand is shifted towards longer wavelengths and a large ligand-to-metal charge transfer band is blue shifted from 429 nm to 414 nm upon accommodation to the Ln(III). The d-d transition is significantly red shifted from 523 nm, observed in other distorted square planar d9 Cu(II) 30 ions, to -670 nm corresponding to an octahedral geometry. The nearly identical weak, broad vi transition at -670 nm from the 2 E g —> 2 T 2 g in the visible spectra for Cu(II) suggests o o that the two Cu(II) centers are indistinguishable. 'ff N M R spectra were instrumental in elucidating the solution structure of the [ L a Z n 2 ( b c n ) 2 ] + complexes. The relative shifts in the resonances conclusively indicate binding of the ligand to the metal ions (Figure 2.18). Interestingly, the ' H N M R spectrum of [ L a Z n 2 ( b c n ) 2 ] + is nearly superimposable with that of [Zn(Hbcn)], suggesting that the T M ( H L ) structure is maintained upon coordination to the La(III) ion. The ring proton resonances are shifted slightly upfield upon accommodating the Ln(III) ion (A8 - 0.02 ppm) 88 compared to those in the ' H N M R spectra for [Zn(Hbcn)]. The ethylenic and benzylic proton resonances, exhibiting the same splitting seen for the [Zn(Hbcn)] and [Zn3(bcn)2] species, are also slightly shifted upfield from the Zn(II) complexes (AS > 0.03 ppm) possibly due to the smaller donation of electron density to La(III). A decreased orbital overlap of the />orbitals on the bridging phenolates with the shielded /-orbitals on the Ln(III) ion is common with f-block metal ions. The absence of additional splitting of the resonances further suggests that the 3-fold symmetry is maintained. In addition, only one set of resonances in the 1 3 C N M R spectrum was observed; one peak for each of the carbons in the ring, one resonance for the benzylic carbon and one peak for the chemically and magnetically equivalent carbons in the ethylenic bridge. no H3 H6 |H4|| 1 HI H2 ff V _ ^ J K 2 Oil r on 5 ^ 3 4 H3bcn 115 IB H6 « 4 J J L H2B H2A [LaZr^Cbcn^f H1A' H1B* H1B 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 i | i i i i | i i i i | 1 i 1 1 1 i 1 1 1 1 i 1 2.0 1.5 0.5 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 ppm 15 3.0 Figure 2.18. ' H N M R spectra of H 3 bcn and [LaZn 2 (bcn) 2 ] + (300 M H z , C D C 1 3 , R T , * solvent). 89 A n O R T E P diagram of [GdNi 2(bcn) 2(CH3CN)2]C104'CH 3CN, representative of the [LnTM2(bcn)2]+ cationic species, is shown in Figure 2.19 along with an expanded view of the coordination sphere. The crystal structures, where L n = G d (III) or Y b (III) revealed a V -shaped metallic core for the [GdNi 2 (bcn) 2 ] + cation with Ni ( l ) -Ln( l ) -Ni (2 ) angle of 142.4° and a nearly linear metallic core of 176.4° for Yb(III). NS Figure 2.19. O R T E P diagrams of the cation in [GdNi2(bcn)2(CH3CN)2]C104«CH3CN and of the expanded view of the Gd(III) coordination sphere. Hydrogen atoms, solvent molecules and the perchlorate anion are omitted for clarity. Thermal ellipsoids are drawn at 50% probability. 90 The Gd(III) ion is eight-coordinate, bicapped by the two tridentate [Ni(bcn)]" units, plus two coordinating acetonitrile molecules. The geometry around the Gd(III) ion is described as square antiprism distorted to C^v bicapped trigonal prism, as determined by a detailed analys is . 3 8 , 6 3 The smaller Yb(III) ion is 6-coordinated by six bridging phenolate oxygens. Uniquely, these d/f/d complexes do not contain coordinated counterions. The intrametallic bonds for N i - Y b are slightly shorter than the N i - G d bonds (2.895 A vs. 3.119 A), with the N i - N i interatomic distance > 5.787 A. A l l L n - 0 bond lengths uniformly reflect the periodic decrease in the ionic radius of the Ln(III) ion (2.367 A for Gd(III) and 2.177 A for Yb(III)). The compressed angles for O-Y b - 0 (average of 77.6°) compared to 84.6° for the Gd(III) complex are also a function of the smaller ionic size of Yb(III). Regardless of the Ln(III) ion, the capping units are almost identical with each Ni(II) chelated in a distorted octahedral geometry via the three amine and three phenolato functions from the fully deprotonated ben3" ligand, similar to the coordination mode seen for the [Ni(Hbcn)]. The N i - N and N i - 0 bond distances of the [GdNi2(bcn)2J+ and [YbNi2(bcn)2]+ cations are slightly longer than the [Ni(Hbcn)] complex (< 0.057 A), but very similar to other literature d/f complexes. [GdZn2(bcn)2(CH 3CN)2]C104'CH 3CN (Figure 2.20) is analogous to the [GdNi 2 (bcn )2 (CH 3 CN )2 ]C10 4 'CH 3 CN species where the Gd(III) is 8-coordinate, bound by two deprotonated [Zn(bcn)]" moieties and by two solvent molecules. For the Zn(II)-Gd(III) system, two acetonitrile molecules occupy the remaining coordination sites analogous to the N i - G d - N i system. The Zn(l)-Gd-Zn(2) angle is slightly smaller (141.68°) than observed for the N i - L n - N i complexes. However, the T M - O , T M - N and G d - 0 bond lengths are nearly identical regardless of the TM(II). The averaged Zn-Gd distance is slightly longer (3.125 A) 91 than the N i - G d distance of 3.119 A attributed to the larger ionic radii of the Zn(II) ion. The interatomic Zn-Zn distance is 5.905 A, slightly longer than the N i - N i distances in the N i - L n -N i complexes. The similarity in the structures of the N i - L n - N i and the Zn-Ln-Zn species allows for the direct comparison of magnetic properties between the different hydrated [LnTM2(bcn)2]ClCv«H20 complexes. Figure 2 . 2 0 . O R T E P diagrams of the cation in [ G d Z n 2 ( b c n ) 2 ( C H 3 C N ) 2 ] C 1 0 4 ' C H 3 C N and of the expanded view of the Gd(III) coordination sphere. Hydrogen atoms, solvent molecules and the perchlorate anion are omitted for clarity. Thermal ellipsoids are drawn at 50% probability. 92 2.3.7 Magnetic studies of the [LnNi2(tam)2]Cl04%nH20 complexes Unfortunately, a theoretical model that reliably accounts for the orbital angular momentum effects of both Ln(III) and TM(II) does not exist; however, a ferromagnetic interaction for six N i - G d complexes in the literature have been determined by a qualitative analysis of the variable temperature magnetic susceptibilities, where L n = Gd(III), Tb(III), Dy(III) (and possibly Ho(III) and Er(III)) . 3 1 ' 3 3 " 3 7 The magnetic data for the [LnNi2(trn)2]C104»rtH20 systems were consistent with the literature examples where a ferromagnetic exchange for L n = Gd(III), Tb(III), Dy(III), Ho(III) and Er(III) was determined by analyzing the shape of the temperature dependent XMT curve; 3 9 however, an antiferromagnetic interaction was determined for the [LnNi2(tam)2]C104*«H20 species, i n where L n = Dy(III) or Yb(III). It was further observed that the antiferromagnetic exchange between the Ni(II) and the Ln(III) ion increased with decreasing size of the Ln(III). Interestingly, a ferromagnetic exchange for a similar trimetallic N i - G d - N i complex using the Schiff base analogue of the F^tam ligand was determined by Matsumoto and co-workers. 3 4 Structural differences between the N i - G d Schiff base complex by Matsumoto 3 4 and the reduced N i - L n Schiff base species previously analyzed in the Orvig lab 3 8 were found to be trivial, making the differences in the magnetic exchange even more puzzling. In addition, antiferromagnetic interactions between Ni(II) and Ln(III) with n > 7 configuration, where n is the number of electron of the Ln(III), have never previously been observed. The unprecedented interaction for the [LnNi2(tam)2]C104»rcH20 species was attributed to magnetic saturation that could mask any existing weak ferromagnetic exchange. The observed apparent antiferromagnetic exchange could also be due to incorrectly 93 accounting for the Ln(III) contributions to the overall susceptibility at low temperature. In order to rule out magnetic saturation effects, the [ L n N i 2 ( t a m ) 2 J C 1 0 4 » « H 2 0 complexes were synthesized using published procedures 3 8 and their magnetic studies were repeated at a lower magnetic field (500 G vs. 10,000 G). The free-ion approximation, where only one level 2 S + 1 L j is thermally populated and second-order contributions are ignored, is a good approximation of the magnetic susceptibility, XM, at room temperature (RT) for many rare earth compounds. Thus, the experimental data w i l l be compared to the calculated values derived from XM = 2x3d + X4fi where X3d and x*f are defined using Equations 2.1 and 2.2. X3d = ( N ^ / 3 A T ) [ 5 ( 5 + l ) ] (2 .1) X4{=(Ngj2/32/3kT)[J(J+\)] (2 .2) N is Avogadro's number, /3 is the Bohr magneton, k is the Boltzmann constant, g is the Lande factor, gj = 3/2 + [5(5+1) - L(Z+1)](2J(J+1))"', L is the total orbital angular moment quantum number, 5 is the total spin angular moment quantum number, and J is the total angular momentum quantum number. Effective magnetic moments, also used in many of the analyses were calculated by Equation 2 .3 . 94 Meff= [Min2 + 2 / / T M 2 ] (2 .3) MLn = gj[J(J + 1)] , JUVM = g[S(S + \)]u\ g is the Lande factor, g J = 3/2 + [S(S+l) -L(L+\)](2/(7+1))"', L is the total orbital angular moment quantum number, S is the total spin angular moment quantum number, and J is the total angular momentum quantum number. The /jeff and XM values can be related to each other using Equation 2 .4 . XMT=(jjeJTl2.%2%f (2 .4) Hejf is calculated using Equation 2.3 and XMT is the temperature dependent magnetic susceptibility per mole of trimetallic complex. It was determined in the present work that the magnetic moments for the [LnN i2 ( t a m )2 ] C 1 0 4 » « H 2 0 species collected at a lower magnetic field (500 G vs. 10,000 G) followed the free-ion approximation for three isolated metals at RT . The p eff for the [DyNi 2 ( tam)2 ]C104 '3H 2 0and [YbNi 2 ( t am) 2 ]C10 4 ' 3H 2 0 (11.09 and 5.61 p B , respectively) closely matched the calculated value (11.37 and 6.05 p B calculated from Equation 2 .3, respectively) expected for two magnetically isolated Ni(II) ions (SNJ = 1) and one Dy(III) ion or Yb(III) ion (J = 15/2, g 5 = 4 /3 and J= 112, g j = 8/7 , respectively, Figure 2.21) at RT. These values also compared reasonably well with the previous data (11.9 and 6.28 p B , respectively) 3 8 collected at 300 K at a magnetic field of 10,000 G . 95 12 11 -10 -9 8 A A ^ A A A A A A A A A A A A O O O O O O 0 0 0 0 0 0 0 0 6 ooo oo o o o o o o o o o <> 50 100 150 200 T / K 250 300 Figure 2.21. Temperature dependence of peff for [ L n N i 2 ( t a m ) 2 ] C 1 0 4 ' « H 2 0 a t 500 G: O , Gd(III); A , Dy(III); O , Yb(III). Similar to what was previously observed, the magnetic moment curves collected at 500 G for both the Dy(III) and Yb(III) species decreased at low temperatures suggesting an antiferromagnetic exchange. The magnetic moment for [DyNi 2 (tam) 2 ]Cl(V3H20 remained relatively constant until 100 K where the moment begins to drop reaching a minimum of 9.75 P B at 2 K . The peff at low temperatures is lower than the expected value of 11.37 U B calculated for the three metal ions. The value at 2 K of [DyNi 2(tam) 2]C104»3H 20 is also notably larger than the value observed at 10,000 G (7.76 pe) possibly due to magnetic saturation effects masking the interaction between the metal ions at the higher field. In addition, the decrease in the upvalues could also be attributed to zero-field splitting of the Ni(II) ions. 3 8 Recently, Gatteschi and co-workers reported a Dy(III)-Ni(II)-Dy(III) system which displayed decreasing XMT1 values at low temperatures and at a relatively small magnetic field 96 of 100 G. Upon removal of the orbital contributions of the Ln(III) ions by using the y^T data of the analogous Dy-Fe-Dy system as the baseline, an increase in the xiviT'data at low temperatures was observed. This increase in the y^T data after removing the first-order orbital contributions from Dy(III) was attributed to a ferromagnetic exchange that had been previously masked by the depopulation of the Dy(III) sub-levels. Dy(III) has a large orbital angular momentum with a 16-fold degeneracy of the 6H| 5/2 ground state. The energy gap between the highest and lowest degenerate level is only on the order of a few hundred wavenumbers. Thus, the highest levels are progressively depopulated as the temperature is lowered leaving only the lowest J sublevel populated at 2 K . This thermal depopulation leads to an inherent decrease of the y^T causing potential deviations from the predicted magnetic behavior (offsetting any possible ferromagnetic interactions between the metal ions), also hypothesized to be present in [LnNi2(tam)2]C104*«H20. In contrast to the results for [DyN i2 ( t a m )2 ] C 1 0 4 « 3 H 2 0 , the moment for [YbN i2 ( t a m )2 ] C 1 0 4 » 3 H 2 0 decreased gradually as the temperature was lowered to 6 K . After 6 K , the moment dropped drastically, reaching a minimum of 4.77 U B at 2 K . This value correlates well to the u eff measured at the higher field (5.06 U B at 2 K , 10,000 G). Regardless of which u eff values is used, both values measured at 2 K are lower than that expected for non-interacting metal ions (6.02 u B calculated from Equation 2.3). This suggests that a small antiferromagentic exchange may exist; however, it is d i f f i c u l t 2 5 , 3 2 , 3 9 to determine the magnetic interaction without removing the intrinsic orbital contributions from the Ln(III). The magnetic moment of [GdN i2 ( t a m ) ] C 1 0 4 » 3 H 2 0 was also measured at R T and closely matched the calculated data for two isolated Ni(II) ions and one uncoupled Gd(III) ion (8.71 p B VS. 8.89 U B , respectively, Figure 2.21) at RT. A decrease in the u eff data was also 97 seen upon lowering the temperature where a value of 6.78 O B was reached at 2 K . This value significantly deviates from the calculated peff (Equation 2.5) for a ferromagnetic coupling of the spins (peff = 11.96 \Xcn = g[STiST+l)]m (2.5) ST = 2S N i + SLn (SKi = 1, SGd = 7/2 for [GdNi 2 ( t am) 2 ]C10 4 »3H 2 0) and g is the Lande factor (g = 2). Every other N i - G d system reported to date has exhibited a ferromagnetic exchange between the Ni(II) and Gd(III) closely following the expected spin exchange values calculated using Equation 2.5. Although the experimental peff values for [GdNi 2 ( t am) 2 ]C10 4 »3H 2 0 are lower than expected for a ferromagnetic spin-exchange, the p e t f data is significantly higher than the calculated value for an antiferromagnetic coupled spin interaction (spin-only peff value of 3.87 U B , S T = 3/2) making it difficult to determine the appropriate magnetic exchange. To our knowledge, only one other remotely analogous trimetallic N i - L n - N i complex (where Ln(III) varies) has been synthesized and magnetically characterized. This system displayed ferromagnetic interactions where n > 7 for the Af Ln(III) metal ions. 3 2 Due to limited literature precedence, it is difficult to hypothesize why the [LnNi 2 ( tam) 2 ] + systems, L n = Gd(III), Dy(III) and Yb(III), exhibit an opposite response than previously reported. When comparing structural data between the other ferromagnetic N i - G d systems and [LnNi 2 ( tam) 2 ] + , the intermetallic N i - G d distances were not significantly different (-3.18 vs. 3.17-3.21, respectively). 3 4 ' 3 5 The L n - 0 bond lengths are also approximately the same 98 length for the other published phenolate N i - G d species. ' ' In addition, the angle between the N i - L n - N i is similar to the reduced Schiff base ferromagnetic [LnTM 2 ( t rn ) 2 ]C10 4 » t tH 2 0 systems, suggesting that the overlap between the metal and oxygen orbitals would be approximately the same. The only structurally significant difference between the [ L n N i 2 ( t a m ) 2 ] C 1 0 4 ' « H 2 0 and the previously reported N i - G d complexes is the intrametallic distance between the Ni(II) ion. The N i - N i intrametallic distances for the [LnNi 2 ( tam) 2 ]C104»«H 2 0 complex are only ~ 5.4 A, whereas the N i - N i distances for the [LnNi 2 ( t rn ) 2 ]C10 4 »«H 2 0 and Okawa's N i 2 G d systems are, reported to be > 7 A . 3 2 ' 3 9 Thus, the lack of evidence for ferromagnetic exchange observed for the [LnNi 2 ( tam) 2 ]C10 4 «/7H 2 0 complexes, where L n = Gd(III), Dy(III), and Yb(III), could also be due to an antiferromagnetic exchange between the two terminal Ni(II) species masking any weak ferromagnetic interaction between the Ln(III) and the terminal Ni(II) ions. Regardless, the unparalleled d/f/d magnetic behavior observed in the [LnTM 2 ( tam) 2 ]C104' t tH 2 0 systems is difficult to explain. The decrease in the u eff values suggest a possible antiferromagnetic exchange for the L n = Gd(III), Dy(III) and Yb(III) complexes; however, the decrease in the peff values could also be due to first-order orbital contributions from the Ln(III) complicated by zero-field splitting effects of the Ni(II) ions. Nonetheless, there has been no evidence for a ferromagnetic exchange in the previous and present magnetic studies of the [LnTM 2 ( t am) 2 ]C10 4 »r tH 2 0 complexes. 3 8 In order to accurately elucidate the presence or nature of the magnetic exchange, however, either a model of d/f magnetic interactions needs to be designed that accurately accounts for the orbital angular effects of the TM(II) and the Ln(III) or more species with different geometric demands need to be synthesized and magnetically characterized. 99 2.3.8 Magnetic studies of the [LnTM2(bcn)2]Cl04*nH20 complexes The magnetic susceptibilities of the [LnTM2 (bcn)2 ]C104*«H20 species ( [ LnTM2 ( b c n ) 2 ] C 1 0 4 » n H 2 0 complexes wi l l be abbreviated as L n T M 2 for the remainder of this thesis) were collected on powdered samples using the S Q U I D magnetometer at U B C over a temperature range of 300 - 2 K and at a magnetic field of either 500 or 10,000 G . The Peff plots for L a C u 2 and LaNi2 corresponds well to the expected curves for isolated Cu(II) and Ni(II) ions (Figure 2.22). The experimental effective magnetic moment of 2.76 up at 300 K is attributed to the two Cu(II) ions with ueff per Cu(II) = 1.95 up and g = 2.26. The LaNi2 complex exhibits a peff value of 4.48 up at R T , corresponding to two Ni(II) ions with ueff per Ni(II) = 3.17 up and g = 2.24. Below 10 K , the u e tr.curve for L a C u 2 and L a N i 2 decreases rapidly due to zero-field splitting of the octahedral Ni(II) transition metal ion or weak intra-and/or intermolecular antiferromagnetic interactions. Nonetheless, the u e rr values determined for the two non-interacting Cu(II) and Ni(II) ions are in the range of previously reported literature values. 89 Figure 2 .22 . ueff vs. T plots of L a C u 2 (O) and L a N i 2 (A) . 100 The Bleaney-Bowers equation was used to determine magnetic interactions between the two terminal Cu(II) metal ions of L a C u 2 , complicating the magnetic data for the other [LnCu2 (bcn )2 ]C104»«H20 species. Because La(III) is diamagnetic, the paramagnetic behavior would be exclusively due to the Cu(II) ions. The exchange integral between the two Cu(II) ions, J, can be determined by fitting the expermental XM or XMT plots to the Bleaney-Bowers equation, Equation 2.6, with J and g as variables. X=2Ng2/3\kT(3-2J/kr)y{ + Na (2.6) N is Avogadro's number, g is the Lande factor,./? is the Bohr magneton, k is the Boltmann constant, T is the temperature, N a is the temperature-independent paramagnetism and J is the exchange integral. N o significant interaction between the two Cu(II) metal ions was observed by fitting the experimental data to the theoretical curve. The variable-temperature magnetic susceptibility for the [LnZn2 (bcn )2 ]C104»r tH 2 0 complexes from 2-300 K was measured and corrected for diamagnetism using Pascal's constants.4 6 The absence of a magnetic exchange between the d/f/d metals for the [LnZn 2 (bcn) 2 ]C10 4 »A?H20 systems make these species an attractive baseline for the analogous paramagnetic Cu(II) and Ni(II) systems, similar to the empirical approach developed by Costes et al.,25 and used by K a h n , 2 6 ' 3 7 O k a w a , 3 1 ' 3 2 Matsumoto and co-workers. 1 8 Costes et al. removed the unquenched orbital angular momentum contributions from the Ln(III) by calculating the difference, A ( X M T ) , between the y^T values for the C u - L n species and the 101 structurally homologous N i - L n complexes, where the diamagnetic Ni(II) was in a square planar geometry. The magnetic exchange between the d/f/d metals for the complexes herein were elucidated using a similar empirical approach defined by Equation 2.7. A ( X M 7 ) = ( X M ^ L n T M - , - ( X M ^ L n Z n , = 2 ( X M ? ) T M + •^TMLn(T) (2.7) T M = Cu(II) or Ni(II), L n = La(III), Nd(III), Gd(III), Dy(III), Yb(III) and JJMUT) is related to the nature of the overall exchange interaction between TM(II) and Ln(III), such that positive or negative values indicate a ferromagnetic or antiferromagnetic interaction, respectively. Similar crystal field effects are assumed to be operative in both the Cu(II), Ni(II) and Zn(II) complexes due to similarities in their core structures. Intermolecular Cu(II)-Cu(II) or Ni(II)-Ni(II) magnetic interactions are assumed to be negligible for the d/f/d systems. The R T magnetic moments for the [LnZ^O^cnhjClO^rcFbO, where L n = Nd(III), Gd(III), Dy(III) and Yb(III) were determined to be u e f f = 3.66, 7.74, 10.64, and 3.95 p B , respectively, shown in Figure 2.23. These values closely correspond to the calculated R T values (3.68, 7.94, 10.63, 4.53 u B using Equation 2.3) where the magnetic contribution is solely originating from the isolated Ln(III) ion. The effective magnetic moment decreases slightly with lowered temperatures from 300-10 K . Below 10 K , the effective magnetic moment decreases significantly reaching a minimum at 2 K , an intrinsic characteristic of Ln(III) attributed to the depopulation of the spin-orbit sublevels similar to that seen in other <i//"heterometallic spec ies . 2 5 , 3 1 ' 3 8 > 3 9 102 12 10 o o o o o o < W A A A A A A A ^$£AA ee e e e — i — 50 O O O O 9 100 150 T / K 200 A 250 0 8 8 8 300 Figure 2.23. Temperature dependence of ueff for L n Z n 2 complexes at 10,000 G : O , N d Z n 2 ; A , G d Z n 2 ; O , D y Z n 2 ; • , Y b Z n 2 . The temperature dependent magnetic susceptibility plots of the trinuclear [LnCu 2 (bcn) 2 ]C104»«H 2 0 complexes are shown in Figure 2 . 2 4 . A l l measurements were taken at either 10,000 G or 500 G . Unfortunately, regression values at high temperatures (>100 K ) are only adequate for data taken at the higher magnetic field (10,000 G , R > 0.99). Thus, all XM l v a l u e s reported at R T wi l l be those measured at a magnetic field of 10,000 G. In order to minimize effects from magnetic saturation, X M ^ values reported herein at < 100 K were those obtained using the magnetic field of 500 G . 103 A t 300 K , the %M^of the N d C u 2 , G d C u 2 , D y C u 2 and Y b C u 2 complexes were determined to be 2.58, 8.80, 15.18, and 3.02 c m 3 K mol" 1, respectively, closely corresponding to the calculated values (2.40, 8.64, 14.93, and 3.33 c m 3 K mol" 1 using Equation 2.3 and 2.4) expected for three uncoupled metal ions (S c u =1/2; J= 9/2, g} = 8/11 for Nd(III); J= 7/2, g} = 2 for Gd(III); J = 15/2, g} = 4/3 for Dy(III); J= 112, g} = 8/7 for Yb(III); Table 2.3). The data for N d C u 2 marginally decreases until 20 K where the value more rapidly decreases reaching a minimum XMT value of 0.875 cm 3 K mol"'at 2 K (Figure 2.24). This value is significantly lower than the calculated value of 2.40 cm 3 K mol" 1 expected for the two Sc u = 1/2 and J= 9/2, gj =8/11 for the Nd(III) ion suggesting antiferromagnetic coupling between two Cu(II) and a Nd(III). 16 14 12 \ _ 10 o E o > 6 4 4 ^ - ^ o o o o o o o o A A A \ A A A • i r i 5 3 ^ _ I 0 o o o o o o o • • • • • • o o 20 —I— 40 T / K 60 O 80 O A A A A • • • • O O O O • A • • 100 Figure 2.24. Temperature dependence of X M T for L n C u 2 complexes at 500 G : • , L a C u 2 ; O , N d C u 2 ; A , G d C u 2 ; O , D y C u 2 ; • , Y b C u 2 . 104 A n opposite response was observed for G d C i i 2 and D y C u 2 , where the j^T values increased dramatically after 50 K and 22 K , respectively (Figure 2.24). A maximum value of 3 1 13.94 cm" ' K mol"' for G d C u 2 was reached at 2 K , roughly matching the calculated value originating from the spin coupling between two Cu(II) ions and a Gd(III) ion (12.37 cm 3 K mol" 1, Equation 2.5). Compound D y C u 2 exhibits maximum y^T- 14.94 c m 3 K mol" 1 at 3 K , matching the calculated value of 14.93 c m 3 K mol" 1 using Sc u =1/2 for each Cu(II) ion and J = 15/2, gj = 4/3 for the D y (III) ion. Table 2.3 Observed and calculated magnetic susceptibilities for [ L n Q ^ b c n h J C l O ^ t t F b O . Ln(III) Ground state gl XM^ c a l c d XiviT'obsd ( R T ) XM^ o b s d (2K) A ( X M 7 ) B sign of La 'So — 0.75 0.97 0.74 — — N d %/2 8/11 2.40 2.58 0.875 0.30 -Gd 8 S7 / 2 2 8.64 8.80 13.94 5.98 + Dy 6 H l 5 / 2 4/3 14.93 15.18 14.84 6.19 + Y b 2S7/2 8/7 3.33 3.02 1.94 1.25 + *yjv\T calculated using Equations 2.3 and 2.4. bA(xM7) values were obtained using Equation 2.7 at 2 K . csign of JTMLH(T) was determined after removing both Ln(III) and TM(I I ) contributions. The similar XMT values for the experimental and calculated data of C u 2 D y suggest the absence of any magnetic exchange between the metal ions; however, the nearly identical values are probably coincidental, arising from the considerable orbital angular contributions of Dy(III) (L = 5) that may offset any observed ferromagnetic spin exchange between the 105 metal ions. Gatteschi and co-workers similarly observed X M ^ values that were lower than expected for a ferromagnetic exchange in their Dy2Cu3 systems. 4 0 They attributed the lower X M ^ t o the depopulation of the J sublevels at low temperatures, intrinsic to Dy(III), that masked the ferromagnetic exchange between Dy(III) and Cu(II) in their complex. The key in elucidating the underlying magnetic exchange of these compounds can be resolved by removing the Ln(III) first-order orbital contributions. A n empirical approach based on the comparison of [ LnCu2 ( b c n ) 2 ] C 1 0 4 » « H 2 0 with the analogous Zn(II) complexes can be used to remove the Ln(III) contributions and, subsequently, determine the magnetic interaction between the d/f/d metal ions (Figure 2.25). Uti l izing Equation 2 .7 , A ( X M ? ) values for G d C u 2 and D y C u 2 increase with decreasing temperatures, indicating a ferromagnetic interaction between the adjacent Cu(II) and Ln(III) ions. Furthermore, maximum A (%M7 ) values of 5.98 cm 3 K mol" 1 (2 K ) and 6.19 cm 3 K mol" 1 (3 K ) for G d C u 2 and D y C u 2 , respectively, are significantly higher than the magnetic contribution of 0.75 cm K mol" for two isolated Cu(II) ions. Okawa and co-workers T 1 determined a A ( X M ^ ) value of ~ 6 cm K mol" for their C u 2 D y species and a value slightly above 4 for their C u 2 G d complexes using an analogous empirical approach, closely matching the magnetic exchange seen for our [LnCu2 (bcn )2 ]C104»«H20 species. Although it appears that the exchange integral, J-vuLn, can easily be extracted from the experimental data using Equation 2 .7 , this value is not a good quantitative determination of exchange parameters in the ground and excited crystal-field states. Actually, this J value is only loosely related to the nature of the overall exchange interaction. Nonetheless, the sign of the value (either negative or positive) can be used to assess the type of magnetic interaction between metal ions. 106 o E CO E o K 4 H 2 i O J A gji i M 1 1 1 1 1 i n f t ^jjjffCOXDOO OO no • • o • 8 A A • • o o • o o • o o • o o 20 40 60 T / K • o o 80 • O o 100 Figure 2.25. A ^ i o r L n C u 2 complexes: O , N d C u 2 ; A , G d C u 2 ; O, D y C u 2 ; • , Y b C u 2 . Furthermore, the Cu(II) contributions to the [ L n C u 2 ( b c n ) 2 ] C l C V / 7 H 2 0 systems can be removed, to minimize any second-order effects of the Cu(II) ion that may complicate the analysis of the magnetic exchange. B y using the analogous [ L a C u 2 ( b c n ) 2 ] C l ( V / 7 H 2 0 magnetic data, the Cu(II) contribution can be subtracted out of Equation 2.7, thereby leaving only the temperature dependent JjMLn(T) value. A negative value suggests an antiferromagnetic exchange between metal ions, and a positive value suggests a ferromagnetic exchange. Positive J values were determined for G d C u 2 and D y C u 2 (5.14 and 3 1 5.35 cm K mol" , respectively), further suggesting a ferromagnetic interaction between Cu(II), and Gd(III) or Dy(III) metal ions. After subtracting the Nd(III) first-order orbital contributions, N d C u 2 exhibited a * i 1 minimum A(XM7) value of 0.30 cm K mol" , below the expected magnetic contribution of 0.75 c m 3 K mol" 1 for two isolated Cu(II) ions, assuming g = 2. The decreasing A(XM7) with 107 decreasing temperature and the sustained negative JTMLII(T) value of -0.507 c m 3 K mol" 1 following the removal of the orbital angular momentum effects from Cu(II) and Ln(III) confirms an antiferromagnetic interaction between the Cu(II) and Nd(III) ions; however, the slight decrease in XMT values and the small J value suggests, at best, a very weak magnetic interaction between Cu(II) and Nd(III). The %M7" value for Y b C u 2 decreases slightly after 20 K and abruptly after 8 K reaching a minimum value of 1.94 cm 3 K mol" 1 at 2 K , lower than the calculated value of 3 1 3.33 cm K mol" . The dramatic decrease in the y^T values below 8 K may be due to the unquenched angular momentum from Yb(III), masking possible ferromagnetic interactions between the d/f/d metals. Using the same empirical approach as above, the A ( X M T ) plots for Y b C u 2 reached a subtle maximum of 1.69 c m 3 K mol" 1 at 12 K . The value slightly decreases after 12 K , 3 1 reaching a minimum value of 1.25 cm K mol" at 2 K . Nonetheless, the A ( X M 7 ) value at 2 K is still larger than the magnetic contributions of two Cu(II) ions. If the Cu(II) contributions are removed, the empirical JrMLn (T) value is determined to be 0.812 cm' " ' K m o l " 1 at 12 K or 0.445 cm"1 K mol" 1 at 2 K . The lack of a dramatic decrease in the A ( X M T ) values and the positive J T M L n ( T ) value suggests the existence of a weak ferromagnetic exchange previously masked by the first-order orbital contributions of the Ln(III) ion; however, the interaction between the Cu(II) and the Yb(III) cannot be confidently confirmed because of the weak magnetic exchange. To quantitatively determine the spin-exchange integral, the [GdCu2(bcn)2]C104*3H20 complex was analyzed using a spin-only expression described by the spin Hamiltonian in Equation 2.8. 108 H - -2/cuGd(Scul ' SGd + Scu2 * SGd) (2.8) «/cuGd is the exchange integral between the adjacent Cu(II)-Gd(III) ions with the two terminal Cu(II) ions assumed to be identical. A positive /integral suggests a ferromagnetic interaction, whereas a negative value is indicative of an antiferromagnetic exchange. The magnetic data of G d C u 2 can be interpreted without considering orbital angular momentum contributions and isotropic effects due to the ground state of Gd(III) (8S7/2,, J= 7/2, L = 0, and S = 7/2) having no first-order orbital angular momentum. The experimental data were fit using Equation 2.9, assuming the Cu(II) and Gd(III) share the same Lande factor. TV is Avogadro's number, j3 is the Bohr magneton, k is the Boltzmann constant, N a is the temperature-independent paramagnetism, g is the Lande factor, 9 is a correction term for intermolecular magnetic interactions, and XM denotes the molecular susceptibility per trinuclear complex. The agreement factors, R(XM?) and R(ueff) defined in Equation 2.10 and 2.11, respectively, were also used to evaluate how well the experimental data fit the theoretical values. N g y (165 ( l d / / f c T ) + 8 4 ( 7 ^ + 8 4 ( 9 J / C T ) + 35) + N a 4kf T-9) ( 5 ( 1 6 ^ T ) + 4UJIa) + 4{9J!kT) + 3) (2.9) R ( X M 7 ) - POOi^obsd ~ XM7"calcd) /^XM^obsd) ] (2.10) R(P-eff) = [£(u0bsd _ Scaled) /S(Pobsd) ] (2.11) 109 The theoretical expression was fit to the experimental y^T vs. T and u eff vs. T data using nonlinear least-squares fitting from 300 - 2 K . The experimental data best fit the theoretical curve using J = 3.67 cm" 1, g = 2.07, and 0 = -0.20 K , and N a = 12 x 10"5 cm 3 mol with an agreement factor of R (XM7 ) = 9.68 x 10"3and R(p eff) = 1.32 x 10' 2 (calculated using Equations 2.10 and 2.11, Figure 2.26). The negative value of G suggests a slight antiferromagnetic interaction between the terminal Cu(II) ions. The positive J value corresponds to a ferromagnetic exchange between the Cu(II)-Gd(III) ions, previously observed in other trimetallic Cu-Gd-Cu systems. 0 50 100 150 200 250 300 T(K) Figure 2.26. jj^Tvs. T and p eff vs. T plots for G d C u 2 at 500 G . The solid lines represent the theoretical curves calculated using Equation 2.9. 110 Okawa and co-workers used the same modified Van Vleck equation (Equation 2 .9) to elucidate the C u - G d interaction for their trimetallic C u 2 G d system with 2,6-di(acetoacetyl)-pyridine as the ligand. 3 1 They determined J= 1.44 cm"1, where the Cu(II) is sandwiched between /?-diketonate sites (Cu-Cu intrametallic distance = 7.298 A) and the Gd(III) ion is chelated by the two central 2,6-diacylpyridine sites (Gd-Cu bond length = 3.649 A). Gatteschi and co-workers used a different equation to model the magnetic susceptibility of a trimetallic C u 2 G d tetradentate Schiff base system; 5 however, their model did not accommodate all possible energy levels of the Gd(III) leading to large Cu(II)-Cu(II) antiferromagnetic interaction parameters (Cu-Cu intrametallic distance = 5.41 A, antiferromagnetic exchange J= - 4.2 cm"1). The large Cu(II)-Cu(II) exchange interaction may have skewed the Cu-Gd J value reported as 7.38 cm"' and has been discounted in other papers (Gd-Cu distance of 3.31 A). Disregarding Gatteschi's result, our [GdCu 2 (bcn ) 2 ]C10 4 »nH 2 0 complex has the largest magnetic exchange yet reported for trimetalic Cu-Gd-Cu systems. The magnetic response, M, at varying fields, H, was used to confirm the ferromagnetic interaction in [GdCu 2 (bcn) 2 ]C104«3H 2 0, where all three metal ions are aligned in parallel with a ground state of ST = 9/2 (Figure 2.27). Uti l iz ing the Bri l louin function, which describes the magnetization of an ideal paramagnet, the spin state of the material can be verified by comparing the ideal curves of different spins with the experimental M vs. H curve. 9 0 The Bril louin function is given in Equations 2 .12 and 2 . 1 3 . M = NgSBJy) (2 .12) (2 .13) y = gupSH(kT)"1, N is Avogadro's number, /up is a Bohr magneton, S is the spin, g is the Lande factor, k is the Boltzmann constant and T is the temperature. A s expected, the experimental Mvs. / / curve closely follows the Bril louin function for 5 = 9/2 suggesting the ferromagnetic ground state of S T = 9/2, where Sod = 7/2 and Sc u = 1/2 for each Cu(II) ion, is significantly populated at 10 K . 50000 i n H T ' / G K " 1 Figure 2.27. Mvs. Hcurve for GdCu2 at 10 K . The solid line indicates the theoretical curve from the Bri l louin faction with ST = 9/2 (g - 2). Overall, the [ L n C ^ b c n h J C l O ^ f k O displayed ferromagnetic interactions when L n = Gd(III), Dy(III) and possibly Yb(III), and an antiferromagnetic interaction for L n = Nd(III) (Table 2.3). Although only a handful o f Cu-Ln-Cu trimetallic d/f/d species have been published (3 Cu-Gd-Cu systems using tetradentate Cu(II) Schiff base l igands , 5 ' 2 3 ' 6 2 and one Cu-Ln-Cu species, L n = La(III), Ce(III), Pr(III), Nd,(III) Sm(III), Eu(III), Gd(III), Tb(III), 112 Dy(III), Ho(III), Er(III), Tm(III), and Yb(III) utilizing a 2,6-di(acetoacetyl)pyridine ligand 3 1), all trimetallic complexes including the [LnCu2 (bcn )2 ]C104»«H20 species studied here have followed Kahn's principles. Further conclusions of how the geometric demands influence the exchange integrals between the metals, however, cannot be made due to the limited number of literature examples. The JJMT value for the [ LnN i2 ( b c n ) 2 ] C 1 0 4 « / 7 H 2 0 species also followed the isolated-ion approximation of three isolated metal ions at R T (Figure 2.28). The y^T for N d N i 2 was determined to be 3.70 cm K mol" at RT, closely matching the calculated value (3.64 cm K mol" 1 calculated using Equations 2.3 and 2.4) expected for two magnetically isolated Ni(II) ions (S N i = 1) and one Nd(II) ion (J= 9/2, g} = 8/11, Table 2.4). A s the temperature decreased, the y^T values reached a minimum of 1.02 c m 3 K mol" 1 at 2 K , well below the 3 1 calculated value of 3.64 cm K mol" suggesting an antiferromagnetic interaction where the spins of the two Ni(II) ions and the Nd(III) are aligned in an antiparallel formation. 20 18 16 14 o 12 E CO E 10 o •— H 8 2 6 4 2 0 <x>o <> o o o o ^ujMtffJfiftfto o d D o a — o o o o o o o o 2 o o 20 40 60 T / K O • 6 80 • 6 O <r 100 Figure 2.28. Temperature dependence of xviT'for L n N i 2 complexes at 500 G : A , L a N i 2 ; O , N d N i 2 ; A , G d N i 2 ; O , D y N i 2 ; • , Y b N i 2 . 113 The XM2" values at R T for G d N i 2 , D y N i 2 and Y b N i 2 were determined to be 10.80, 17.48 and 5.05 c m 3 K mol" 1 , respectively, closely matching the calculated values (9.88, 16.17 and 4.57 cm K mol" , respectively, using Equations 2.3 and 2.4 and Table 2.4) for three magnetically isolated metal ions (Figure 2.28). A t lower temperatures, the interactions between the Ni(II)-Ln(III) metals of G d N i 2 and D y N i 2 is difficult to ascertain due to the absence of a dramatic increase or decrease in the plots previously observed for the [LnCu 2 (bcn) 2 ]C104»t tH 2 0 systems. The XM^data for G d N i 2 rise subtly beginning at 40 K ' reaching a maximum of 17.55 c m 3 K mol" 1 at 13 K and decrease slightly to reach a minimum 3 1 value of 16.06 cm K mol" at 3 K , significantly greater than the calculated value for the three metal ions of 9.88 cm K m o l " 1 . This increase in the XM T values at lowered temperatures suggests a ferromagnetic exchange between the two Ni(II) ions and the Gd(III) metal ion. Table 2.4 Observed and calculated magnetic susceptibilities for [LnNi 2 (bcn) 2 ]C104»«H 2 0. Ln(III) Ground state gi (calcd) XM o^bsd (RT) XM o^bsd (2K) XM r^nax. or min. ( T K ) A ( X M 7 ) b sign of L a 'So . . . 2.85 2.00 2.55 2.55 (2) ~ -N d 4Ia/2 8/11 3.64 3.70 1.02 1.02 (2) 0.44 -Gd ^7/2 2 9.88 10.80 16.06 17.55 (13) 8.86 + Dy 6 H i 5 / 2 4/3 16.17 17.48 14.73 14.85 (3) 6.08 + Y b 2S7/2 8/7 4.57 5.05 4.26 4.26 (2) 3.56 . + ay^[T calculated using Equations 2.3 and 2.4. bA(xM^) values were obtained using Equation 2.7 at 2 K . csign of JjuLn{T) was detemined after removing both Ln(III) and TM(II) contributions. 114 Furthermore, the Bri l louin function (Equations 2.12 and 2.13) can be used to confirm the ferromagnetic interactions between the Ni(II)-Gd(III)-Ni(II) species (Figure 2.29). A s expected, the M vs. H curve follows the Bri l louin function for a Sj = 11/2 (SNJ = 1 and Sad= 7/2), suggesting that the ground state (ST = 11/2) arising from a ferromagnetic interaction is significantly populated at 10 K . 60000 -\ H T 1 / G K " 1 Figure 2.29. Mvs. / / curve for GdNi2at 10 K . The solid line indicates the theoretical curve for Si = 11/2 (g = 2). The curve for D y N i 2 decreases slightly until 30 K where y^T begins to abruptly decrease reaching a minimum value of 14.55 c m 3 K mol" 1 at 5 K and a maximum value of 14.85 cm 3 K mol" 1 at 3 K (Table 2.4). A maximum y^T value of 4.26 c m 3 K mol" 1 at 2 K is 115 determined for Y b N i 2 . The y^T values for D y N i 2 and Y b N i 2 (14.85 c m 3 K mol" 1 at 3 K ; 4.26 3 1 3 1 cm K m o l " at 2 K , respectively) are below the calculated values of 16.17 cm K m o l " and 4.57 cm K mol" , respectively, for two Ni(II) ions and the Dy(III) or the Yb(III) ions. Because of the lowered experimental values compared to the calculated values, it appears that an antiferromagnetic interaction is dominant; however, the decreased XMT values could also be a product of the first-order orbital contributions from the Ln(III). The magnetic data wi l l be better resolved by removing the Ln(III) contributions using Equation 2.7. The differences between the y^T plots of [LnNi 2 (bcn ) 2 ]C104 '«H 2 0 and the analogous L n Z n 2 species are shown in Figure 2.30. N d N i 2 shows a slight decrease to a minimum 3 1 A ( X M T ) value of 0.441 cm K mol" , significantly smaller than the contribution from two isolated Ni(II) ions at 2 K (yjMT = 2 cm K mol" ). The calculated JjuhniT) value, after removing the Ni(II) contribution using [LaNi 2 (bcn) 2 ]C104*«H 2 0, is negative with a value of -2.72 cm 3 K mol" 1 , further suggesting an antiferromagnetic exchange. Compounds G d N i 2 , D y N i 2 and Y b N i 2 have A ^ T ) values of 8.86, 6.08 and 3.56 cm 3 K mol" 1 at 2 K , respectively. These values are all larger than XMT= 2 c m 3 K mol" 1 calculated for two Ni(II) ions. The positive JTMLII(7) values, after removing Ni(II) contributions from the X M ^ p l o t using [LaNi 2 (bcn) 2 ]C104»«H 2 0, were determined to be 5.70, 2.93 and 0.406 cm 3 K mol" 1, respectively, further suggest ferromagnetic exchange interactions between the Ni(II) and Gd(III), Dy(III) and Yb(III) ions. 116 12 10 'o 8 E CO E o 6 ^ 4 H 2 H O o o o o o o o o • D • • • o O o o o o • o 8 8 o o o o 20 40 60 T / K 80 100 Figure 2.30. A ^M ^ f o r L n N i 2 complexes: O , N d N i 2 ; A , G d N i 2 ; O , D y N i 2 ; • , Y b N i 2 . Only four other trimetallic N i 2 L n systems have been reported; 3 2 two of the systems were previously synthesized in our group ' and discussed more in depth earlier in this chapter, and only the N i - G d - N i interaction of one of the other trimetallic N i 2 L n complexes were analyzed. 3 4 Thus, it is difficult to correlate structural geometry with magnetic exchange. Nonetheless, the magnetic data for the [LnNi 2 (bcn) 2 ]C104»«H 2 0 complexes matched the previous works of Okawa and co-workers' N i 2 L n system 3 2 and the reduced Schiff base d/f/d complexes of Bayly et al. 3 9 (ignoring the unprecedented results of the [LnNi 2 ( tam) 2 ]C104»«H 2 0 system) where a ferromagnetic exchange was observed when L n = Gd(III), Dy(III) and Yb(III) and an antiferromagnetic interaction when L n = Nd(III). It also appears that Kahn's principles are valid for Ni(II)-Ln(III) systems, where a ferromagnetic exchange is observed with Af Ln(III) metal ions where n > 7 and an antiferromagnetic 117 interaction where n < 7. In addition, the differences in the geometric demands of the N i - L n species does not appear to affect the magnetic exchange, namely the shorter N i - L n and N i - N i intermetallic distances for [ LnN i2 ( b c n )2 ] C 1 0 4 * « H 2 0 compared to the N i - L n - N i reduced Schiff base complexes 3 4 synthesized by Bayly et al. and the 2,6-di(acetoacetyl)pyridine systems made by Okawa and co-workers (-3.119 A vs. >3.53 A for N i - L n distances; 5.98 A vs. > 7.21 A for N i - N i distances, respectively) 3 2. 2.4 Conclusion The synthesis of a series of [ L n T M 2 ( L ) 2 ] + cationic complexes, where L n = La(III), Nd(III), Gd(III), Dy(III), Yb(III), T M = Mn(II), Fe(II), Co(II), Ni(II), Cu(II), Zn(II), and L = 3 3 3 tarn trn ", and ben " were undertaken. Due to the stringent steric requirements placed on the ligand moeties upon chelating three metal ions, isolation of the [ L n T M 2 ( L ) 2 ] + complexes was only modestly successful. The only tripodal tarn3" species isolated was [LaZn2 ( tam)2 ]C104»5H20, isostructural with the [ LaN i2 ( t a m ) 2 ] C 1 0 4 ' 3 H 2 0 . Magnetic studies were also completed on the [LnNi 2(tam)2]C104*«H20 systems at lower fields than previously reported (500 G vs. 10,000 G). The data reported in this work coincided with the previously published results where an unusual antiferromagnetic exchange was observed between the Ni(II) and Gd(III), Dy(III) and Yb(III) metal ions. Several [TM(Hbcn)], [TM 3 (bcn) 2 ] , [La(bcn)], [La 3(bcn) 2] 3", and [LnTM 2 (bcn) 2 ] + cationic complexes were synthesized and characterized, L n = La(III), Nd(III), Gd(III), Dy(III), Yb(III) and T M = Cu(II), Ni(II), Zn(II). The magnetic interactions of the [ LnTM2 ( b c n ) 2 ] C 1 0 4 « « H 2 0 complexes were determined by subtracting the first-order orbital 118 contributions of the Ln(III) contributions and analyzing the resulting %M rvalues using an empirical approach similar to the data analysis of Costes and co-workers. 2 5 Despite different geometric demands, the d/f/d complexes displayed similar magnetic properties to other literature d- and /-block species, where an antiferromagnetic exchange was observed for the lighter Ln(III) (Ln = Nd(III)), while the heavier Ln(III) ions (Ln = Gd(III), Dy(III), Yb(III)) exhibited ferromagnetic behavior. These conclusions are also in-line with the predictions of Kahn et al.?1'44 where a ferromagnetic exchange is exhibited by Af Ln(III) ions when n > 7. We remain hopeful that a more quantitative analysis w i l l soon be available to more accurately elucidate the magnetic exchange between TM(II) and Ln(III) ions as more d/f complexes are synthesized and characterized. 119 2.4 References 1. Warnsdorfer, W. ; Aliaga-Alcalde, N . ; Hendrickson, D . N . ; Christou, G . Nature 2002, 416, 406. 2. Gatteschi, D . ; Sessoli, R. Agnew. Chem., Int. Ed. 2003, 42, 268. 3. Bencini, A . ; Benelli , C ; Caneschi, A . ; Carlin, R. L . ; Dei , A . ; Gatteschi, D . J. J. Am. Chem. Soc. 1985, 107, 8128. 4. Ramade, I.; Kahn, O.; Jeannin, Y . ; Robert, F. Inorg. Chem. 1997, 36, 930. 5. Bencini, A . ; Benelli , C ; Caneshi, A . ; Dei , A . ; Gatteschi, D . Inorg. Chem. 1986, 25, 572. 6. Costes, J.-P.; Novitchi, G. ; Shova, S.; Dahan, F.; Donnadieu, B . ; Tuchagues, J.-P. Inorg. Chem. 2004, 43, 7792. 7. Costes, J.-P.; Dahan, F.; Dupuis, A . ; Laurent, J.-P. Inorg. Chem. 1997, 36, 3429. 8. Costes, J.-P.; Dahan, F.; Dupuis, A . ; Laurent, J.-P. Inorg. Chem. 1996, 35, 2400. 9. Andruh, M . ; Ramade, I.; Codjovi, E . ; Guillou, O.; Kahn, O.; Trombe, J. C. J. Am. Chem. Soc. 1993, 115, 1822. 10. Costes, J.-P.; Dahan, F.; Dupuis, A . Inorg. Chem. 2000, 39, 165. 11.. Sakamoto, M . ; Manseki, K . ; Okawa, H . Coord. Chem. Rev. 2001, 219-221, 379. 12. Benelli, C ; Gatteschi, D . Chem. Rev. 2002, 102, 2369. 13. Benelli , G . ; Caneschi, A . ; Gatteschi, D. ; Guillou, O.; Pardi, L . Inorg. Chem. 1990, 29, 1750. 14. Gao, S.; Yang, Y . ; X u , G . J. Alloys Compds. 1995, 225, 234. 15. Winpenny, R. E . P. Chem. Soc. Rev. 1998, 27, 447. 16. Shiga, T.; Ohba, M . ; Okawa, H . Inorg. Chem. Commun. 2003, 6, 15. 17. Ghoerghe, R.; Andruh, M . ; Costes, J.-P.; Donnadieu, B . Chem. Commun. 2003, 2778. 18. Kido , T.; Ikuta, Y . ; Suantsuki, Y . ; Ogawa, Y . ; Matsumoto, N . Inorg. Chem. 2003, 42, 398. 19. Costes, J.-P.; Clemente-Juan, J. M . ; Dahan, F.; M i l o n , J. Inorg. Chem. 2004, 43, 8200. 120 20. Sanz, J. L . ; Ruiz, R.; Gleizes, A . ; Lloret, F.; Faus, J.; Julve, M . ; Borras-Almenar, J. J.; Journaux, Y . Inorg. Chem. 1996, 35, 7384. 21. Stemmler, A . J.; Kampt, J. W.; Ki rk , M . L . ; Atasi, B . H . ; Pecoraro, V . L . Inorg. Chem. 1999, 38, 2807. 22. Costes, J.-P.; Dahan, F. ; Dupuis, A . ; Laurent, J.-P. Inorg. Chem. 2000, 39, 169. 23. Costes, J.-P.; Dahan, F. ; Dupuis, A . Inorg. Chem. 2000, 39, 5994. 24. He, F.; Tong, M . - L . ; Chen, X . - M . Inorg. Chem. 2005, 44, 559. 25. Costes, J.-P.; Dahan, F.; Dupuis, A . ; Laurent, J.-P. Chem. Eur. J. 1998, 4, 1616. 26. Kahn, M . L . ; Mathoniere, C ; Kahn, 0 . Inorg. Chem. 1999, 38, 3692. 27. Costes, J.-P.; Auchel, M . ; Dahan, F.; Peyrou, V . ; Shova, S.; Wernsdorfer, W. Inorg. Chem. 2005. 28. Osa, S.; K ido , T.; Matsumoto, N . ; Re, N . ; Pochaba, A . ; Mrozinski , J. Inorg. Chem. 2003, 126, 420. 29. Guil lou, O.; Oushoorn, R. L . ; Kahn, O.; Baubekeur, K . ; Batail, P. Angew. Chem. Int. Ed. Engl. 1992, 31, 626. 30. L i , Y . - T . ; Yan, C . -W. ; Zhang, J. Cryst. Growth Des. 2002, 3, 481. 31. Shiga, T.; Ohba, M . ; Okawa, H . Inorg. Chem. 2004, 43, 4435. 32. Shiga, T.; Ito, N . ; Hidaka, A . ; Okawa, H . ; Kitagawa, S.; Ohba, M . Inorg. Chem. 2007, 46, 3492. 33. Costes, J.-P.; Dahan, F.; Dupuis, A . ; Laurent, J.-P. Inorg. Chem. 1997, 36, 4284. 34. Yamaguchi, T.; Sunatsuki, Y . ; Kojima, M . ; Akashi, H . ; Tsuchimoto, M . ; Re, N . ; Osa, S.; Matsumoto, N . Chem. Commun. 2004, 1048. 35. Yue, Q.; Yang, J.; L i , G . - H . ; L i , G . -D. ; X u , W. ; Chen, J.-S.; Wang, S.-N. Inorg. Chem. 2005, 44, 5241. 36. Chen, Q . -Y . ; Luo, Q. -H. ; Zheng, L . - M . ; Wang, Z . - L . ; Chen, J.-T. Inorg. Chem. 2002, ¥ 7 , 6 0 5 . 37. Kahn, M . L . ; Lecante, P.; Verelst, M . ; Mathoniere, C ; Kahn, O. Chem. Mater. 2000, 12, 3073. 38. X u , Z . ; Read, P. W. ; Hibbs, D. E . ; Hursthouse, M . B . ; Mal ik , K . M . A . ; Patrick, B . O.; Rettig, S. J.; Seid, M . ; Summers, D . A . ; Pink, M . ; Thompson, R. C ; Orvig, C. Inorg. Chem. 2000, 39, 508. 121 39. Bayly, S. R.; X u , Z . ; Patrick, B . O.; Rettig, S. J.; Pink, M ; Thompson, R. C ; Orvig, C. Inorg. Chem. 2 0 0 3 , 42, 1576. 40. Pointillart, F.; Bernot, F.; Sessoli, R.; Gatteschi, D . Chem. Eur. J. 2 0 0 7 , 1 3 , 1602. 41. L i u , S.; Wong, E . ; Karunaratne, V . ; Rettig, S. J.; Orvig, C. Inorg. Chem. 1 9 9 3 , 32, 1756. 42. L i u , S.; Belmini , L . R.; Rettig, S. J.; Thompson, R. C ; Orvig, C. J. Am. Chem. Soc. 1 9 9 2 , 774, 6081. 43. Auerbach, U . ; Eckert, U . ; Wieghardt, K . ; Nuber, B . ; Weiss, J. Inorg. Chem. 1 9 9 0 , 29, 938. 44. Kahn, O.; Guil lou, O. Research Frontiers in Magnetism; O'Connor, C. J., Ed. ; World Scientific: London, 1993. 45. Richman, J. E . ; Atkins, T. J. J. J. Am. Chem. Soc. 1 9 7 4 , 96, 2268. 46. Pascal, P., Chimie Generale. Masson et Cie: Paris, 1949. 47. Ehlert, M . K . ; Rettig, S. J.; Storr, A . ; Thompson, R. C ; Trotter, J. J. Can. J. Chem 1 9 9 1 , 69, 432. 48. teXsan: Crystal Structure Analysis Package; Molecular Structure Corporation: The Woodlands, T X , 1985 & 1992. 49. Sheldrick, G . M . SHELXL-97: University of Gottingen, Germany, 1997. 50. D * T R E K Area Detector Software. Version 7.11, Molecular Structure Corporation: 2001. 51. SIR97; Altomare, A . ; Burla, M . C ; Cammali, G . ; Cascarano, M . ; Giacovazzo, C ; Guagliardi, A . ; Moliterni, A . G . G. ; Palidori, G. ; Spanga, A . J. J. Appl. Cryst 1 9 9 9 , 32, 115. 52. Beurskens, P. T.; Admiraal, G. ; Beurskens, G . ; Bosman, W . P.; de Gelder, R.; Israel, R.; Smits, J. M . M . The DIRDIF-94 program system; Technical Report; Crystallography Laboratory, University of Nijmegen: Nijmegen, The Netherlands, 1994. 53. P L A T O N ; A Multipurpose Crystallographic Tool, Utrecht University: Utrecht, The Netherlands, 2001. 54. Comba, P.; Hambley, T. W. ; Hitchman, M . A . ; Stratemeier, H . Inorg. Chem. 1 9 9 5 , 34, 3903. 55. Alyea, E . C ; Malek, A . ; Vougioukas, A . E. Can. J. Chem. 1 9 8 1 , 60, 667. 122 56. Brown, L . M . ; Mazdiyasni, K . S. Inorg. Chem. 1 9 7 0 , 9, 2783. 57. Guil lon, H . ; Hubert-Pfalzgraf, L . G. ; Vaissermann, J. Eur. J. Inorg. Chem. 2 0 0 0 , 1243. 58. Guoxin, T.; Yongjun, Z . ; Jingming, X . ; Ping, Z . Inorg. Chem. 2 0 0 3 , 42, 735. 5 9 . Kameta, N . ; Imura, H . ; Ohashi, K . ; Aoyama, T. Polyhedron 2 0 0 2 , 21, 805. 60. Wong, W . - K . ; Liang, H . ; Wong, W . - Y . ; Cai , Z . ; Li , K . - F . ; Cheah, K.-W. New J. Chem. 2 0 0 2 , 26, 275. 61. Connor, J. A . ; Charlton, J.; Cupertino, D . C ; Lienke, A. ; McPartl in, M.; Scowen, I. J . ; Tasker, P. A . J. Chem. Soc., Dalton Trans. 1 9 9 6 , 2835. 6 2 . Akine, S.; Taniguchi, T.; Nabeshima, T. Angew. Chem. Int. Ed. 2 0 0 2 , 41, 4670. 6 3 . Muetterties, E . L . ; Guggenberger, L . J. J. Am. Chem. Soc. 1 9 7 4 , 96, 1748. 64. Aveci l la , F. ; Bastida, R.; Bias, A . d.; Fenton, D . E . ; Macias, A.; Rodriguez, A . ; Rodriguez-Bias, T.; Garcia-Granda, S.; Corzo-Suarez, R. J. Chem. Soc, Dalton Trans. 1 9 9 7 , 409. 6 5 . Setyawati, I. A . ; L i u , S.; Rettig, S. J.; Orvig, C. Inorg. Chem. 2 0 0 0 , 39, 496. 66. Lisowski , J.; Starynowicz, P. Inorg. Chem. 1 9 9 9 , 55, 1351. 67. Vance, D. H . ; Czarnik, A . W. J. Chem. Soc. 1 9 9 3 , 115, 12165. 68. Young, M. J.; Chin, J. J. Am. Chem. Soc. 1 9 9 5 , 1 1 7 , 10577. 69 . Lane, B . S.; Burgess, K . Chem. Rev. 2 0 0 3 , 1 0 3 , 2457. 7 0 . Gott, A . L . ; McGowan , P. C ; Podesta, T. J.; Tate, C. W . Inorg. Chim. Acta 2 0 0 4 , 37, 689. 71 . Murphy, B . P.; Quinti, L . ; Kel ly , D . G ; Martin, W. J.; Perotti, A . ; Hursthouse, M. B . ; Gelbrich, T. Inorg. Chem. Commun. 2 0 0 2 , 5, 577. 7 2 . Caravan, P.; El l ison, J. J.; McMurry , T. J.; Lauffer, R. B . Chem. Rev. 1 9 9 9 , 99, 2293. 7 3 . Moore, D. A . ; Fanwick, P. E . ; Welch, M . J. Inorg. Chem. 1 9 8 9 , 28, 1504. 74. Hoyle, R. C ; Richman, J. E . ; Dantas, G . A.; Lightbourne, M . F.; Shinneman, L . S. J. Org. Chem. 2 0 0 1 , 66, 2722. 7 5 . Parker, L . L . ; Gowans, N . D. ; Jones, S. W. ; Robins, D . J. Tetrahedron 2 0 0 3 , 59, 10165. 123 76. Briellmann, M . ; Kaderli , S.; Meyer, C. J.; Zuberbuhler, A . D . Helv. Chim. Acta 1987, 70,680. 77. Chavez, F.; Sherry, A . D . J. Org. Chem. 1989, 54, 2990. 78. Vriesema, B . K . ; Buter, J.; Kellogg, R. M . J. Org. Chem. 1984, 49, 110. 79. Glaser, R.; Shiftan, D. ; Levi-Roso, G . ; Ergaz, I.; Geresh, S. J. Org. Chem. 2002, 67, 5486. 80. Fraser, R. R.; Gurudata; Reyes-Zamora, C ; Swingle, R. B . Can. J. Chem. 1968, 46, . 1595. 81. Cole, E . ; Copley, R. C. B . ; Howard, J. A . K . ; Parker, D . ; Ferguson, G . ; Gallagher, J. F.; Kaitner, B . ; Harrison, A . ; Royle, L . J. Chem. Soc. Dalton Trans. 1994, 1619. 82. Beattle, J. K . ; Elsbernd, H . S. J. Am. Chem. Soc. 1970, 92, 1946. 83. Zompa, L . J . ; Margulis, T. N . Inorg. Chim. Acta 1978, 28, L157. 84. Graham, B . ; Grannas, M . J.; Hearn, M . T. W. ; Kepert, C. M . ; Spiccia, L . ; Skelton, B . W.; White, A . H . Inorg. Chem. 2000, 39, 1092. 85. Lever, A . B . P., Inorganic Electronic Spectroscopy. Elsevier: N e w York, 1984. 86. Bunzl i , J .-C. G . ; Mil ic ic-Tang, A . Inorg. Chim. Acta 1996, 252, 221. 87. Yang, L . ; X u , Y . ; Gao, X . ; Zhang, S.; Wu , J. Carbohydr. Res. 2004, 339, 1679. 88. Rafat, F.; Siddiqi, K . S.; Lutfullah. Synth. React. Inorg. Met.-Org. Chem. 2004, 34, 763. 89. Earnshaw, A . , Introduction to Magnetochemistry. Academic Press: New York, 1968. 90. Darby, M . I. Brit. J. Appl. Phys. 1967,18, 1415. 124 C H A P T E R 3 Lanthanide Containing Compounds for Therapeutic Care in Bone Resorption Disorders* 3.1 Introduction Bone density disorders, including osteoporosis, affect 1 in 4 women and 1 in 8 men over age 50 in North America. 1 A s the population ages, these diseases are incurring substantial annual health care costs, escalating into billions of dollars. Osteoporosis is characterized by low bone mineral density that leads to enhanced bone fragility and a consequent risk of low-impact bone fractures. 2 ' 3 The low bone mineral density is a result of an imbalance between bone resorption and bone formation. Normally, formation and absorption of bone is a tightly regulated cycle wherein the bone matrix is manufactured by osteoblast cells and removed by osteoclast cells. Either increased activity of osteoclasts or decreased bone formation by osteoblasts leads to microarchitectural deterioration of bone tissue. Many contributing factors are known to influence the development of the disease such as smoking, physical inactivity, positive family history, age, type 1 diabetes, rheumatoid arthritis, Cushing's syndrome and certain drugs including glucocorticoids, anticoagulants, and chemotherapy;2 however, most cases of osteoporosis result from inadequate calcium uptake. f Portions of this chapter can be found in the published manuscript Barta, C. A . ; Sachs-Barrable, K . ; Jia, J.; Thompson, K . H . ; Wasan, K . M . ; Orvig, C. Dalton Trans. 2007 (in press). 125 Simple therapeutic recommendations for the prevention of osteoporosis include life-style changes such as introduction of weight-bearing exercise or increased dietary supplementation of calcium and vitamin D . 2 ' 4 ' 5 Other common treatments consist of antiresorptive therapies, e.g., hormone replacement therapy, calcitonin and bisphosphonates. The latter preferentially inhibit bone resorption by decreasing osteoclast activity. Unfortunately, patient compliance is usually low due to adverse gastrointestinal side effects and, in the case of calcitonin, high costs . 2 ' 4 ' 5 A new class of osteoporosis drugs, including oral strontium ranelate, stimulates osteoblast proliferation while also inhibiting osteoclast activity; 6 ' 7 however, uncertainty regarding the potential toxicity of chronic strontium accumulation in bone may limit the utility of this product.' L o w doses of Ln(III) have been shown to act similarly to strontium ranelate.9"11 It is known that Ln(III) are functional mimics of Ca(II) ions, sharing similar ionic radii, donor atom preferences, and almost identical coordination numbers in protein binding sites. In vivo, Ln(III) has been found to exchange with Ca(II) in bone, 9 modifying the bone remodeling cycle; osteoblast proliferation is stimulated 1 0 ' 1 1 and osteoclast differentiation is inhibited, impeding bone resorption." Based on this evidence, lanthanum carbonate (La2 (C03 )3 ) has been proposed as a potential preventative measure for post-menopausal osteoporosis." La2 (C03 )3 is currently being used to treat hyperphosphatemia under the trade name of Fosrenol™, an A n o r M E D discovery. Unfortunately, the extremely low bioavailability (<0.0007 %) of Fosrenol™ requires that high doses of elemental La(III) be administered to control phosphate levels. The high doses of La(III) lead to adverse gastrointestinal (Gl) tract side effects, with consequent poor patient compl iance . 9 One of the main concerns of using Ln(III) for the treatment of osteoporosis is metal toxicity. Metal toxicity has previously been reported for many different transition metals that 126 readily exchange with Ca(II) in vivo including Pb(II), Cd(II), Zn(II), Cr(II), Mg(II), Cu(II) and Fe(II). Aluminium hydroxide, an alternative to L a 2 ( C 0 3 ) 3 for the treatment of hyperphosphatemia, has also been shown to accumulate in the body, especially in bone, resulting in low turnover of bone that can progress into osteomalacia or osteoporosis. It has been shown, however, that the absorption of aluminum is significantly higher than that of Ln(III), with values of 100 ng/ml compared to 0.6 ng/ml, respectively. 1 3 Considering the average bone calcium concentration of 120 mg/g of wet weight and assuming heterogeneous distribution of lanthanum throughout the bone, the molar bone lanthanum/calcium ratio is only ~ 2 x 10" 5. 1 4 Thus, lanthanum could be expected to replace one in 50,000 calcium ions whereas for aluminum the ratio would be one in 16; hence, lanthanum would have a decreased probability of disturbing bone growth and structure compared to aluminum. Since then, human subjects have been used to complete extensive research on the biological effects of L a 2 ( C 0 3 ) 3 , specifically on the histomorphology of bone. Despite accumulation in bone, liver, and the gastrointestinal tract with prolonged use , 1 5 ' 1 6 no adverse consequences of tissue-deposited lanthanum have been observed (no carcinogenicity, 1 7 no genotoxicity, 1 8 nor any change in liver function 1 9). The highest concentration of La(III) observed in the bone of treated dialysis patients (maximum dose of L a 2 ( C 0 3 ) 3 = 3000 mg/day of elemental lanthanum) was 10 ug/g wet weight after 4 years of treatment.1 9 Slow release of the deposited lanthanum was still detected 2 years after treatment, with no association of toxicity. In addition, bone biopsies taken 1 - 2 years following a 12-month administration of La2(CO"3)3 actually showed an increase in osteoblast activity followed by an increase in bone formation rates.1 9 The most common side effect reported was gastrointestinal discomfort attributed to the high doses of insoluble La2 (C03 )3 . 127 Recently, a few groups using rat models have shown that Ln(III), specifically La2(COa)3, has led to bone mineralization defects, increased osteoid area, and decreased bone formation rates. 2 0 ' 2 1 Subsequent experimental studies, however, have suggested that these effects of La2 (C03 )3 on bone may be due to phosphate depletion, not direct toxicity from the Ln(III) salts on bone metabo l i sm. 1 4 ' 2 2 ' 2 3 Upon phosphate supplementation in uremic rats, the bone histomorphology parameters did not show signs of impaired bone mineralization upon chronic treatment with L a 2 ( C 0 3 ) 3 2 4 Adjustments to the ligand structure around Ln(III) ions have the potential to increase the oral bioavailability of Ln(III) for the treatment of bone density disorders while decreasing unwanted side effects (i.e. G l tract irritation). 3-Hydroxy-4-pyrones and 3-hydroxy-4-pyridinones, related classes of small, bidentate, O, O proligands, exhibit high bioavailability and favorable toxicity profiles. 2 5 Consisting of a 6-membered O or N heterocycles, these proligands form thermodynamically stable five-membered chelate rings using the hard hydroxyl O atom and the neighboring ketone functionality. Upon deprotonation within the physiological pH range, both 3-hydroxy-4-pyrones and 3-hydroxy-4-pyridinones form neutral, stable (bis or tris) complexes with trivalent cations including Fe , A l , Ga , In , and Ln(III), " as well as the actinides 3 0 and several divalent ions including Z n 2 + , R u 2 + , [ V O ] 2 + , and [ M o 0 2 ] 2 + . 2 5 Stability is attributed to the substantial charge derealization within the six-membered ring A handful of these metal complexes and their corresponding hydroxypyrone and hydroxypyridinone proligands have been extensively investigated for myriad potential biological applications including treatment of iron overload disorders, 3 1 ' 3 2 aluminum tox ic i ty , 3 3 ' 3 4 oral cancer treatment, radioisotope agents for use in S P E C T imaging, " restoration of iron balance in anemia patients, 3 9" 4 1 insulin enhancement,4 2 and Gd(III) contrast agents for M R I . 2 5 ' 4 3 128 Specific proligands of interest include maltol (Hma, 3-hydroxy-2-methyl-4-pyrone), its ethyl analog ethylmaltol (Hema, 3-hydroxy-2-ethyl-4-pyrone), deferiprone (Hdpp, 3-hydroxy-1,2-dimethyl-4-pyridinone), and 3-hydroxy-2-methyl-4(7#)-pyridinone (Hmpp) (Figure 3.1). 4 4 Maltol and ethylmaltol are approved food additives used to impart desirable malty taste and odor to breads, cakes, beer and other beverages. Deferiprone, sold in some countries as Ferriprox®, an oral iron chelator, has shown promise in initial clinical trials for the binding and removal of excess iron in the treatment of iron overload disorders. Hmpp, the only non-commercially available proligand used in this study, is synthesized in moderate yields by ammonolysis of a benzyl-protected maltol followed by acid deprotection. 4 5 OH R R = C H 3 , X = 0 R = C 2 H 5 X = 0 R = CH3 X = N H R = CH3 X = N C H 3 3-hydroxy-2-methyl-4-pyrone (Hma) 3-hydroxy-2-ethyl-4-pyrone (Hema) 3-hydroxy-2-methyl-4( 1 /f)-pyridinone (Hmpp) 3-hydroxy-1,2-dimethyl-4-pyridinone (Hdpp) Figure 3 . 1 . Structures of the 3-hydroxy-4-pyrones and 3-hydroxy-4-pyridinones used in this project. Hydroxypyridinones are known to be significantly stronger chelators than are their hydroxopyrone analogs due to extensive derealization of the lone pair from the ring nitrogen atom, rendering the carbonyl functionality more basic, as depicted in Scheme 3.1. Thus, the 129 stability of the desired complexes can also be tuned by switching the ligand or the metal to obtain desired properties. 3 2 Scheme 3 .1 . Deprotonation (and resonance forms) of 3-hydroxy-4-pyridinones. Previously, Ln(III) complexes using the same pyrone and pyridinone ligands have been investigated, including Ln(ma) 3 , where L n = Pr(III), Nd(III), Sm(III), Gd(III), Dy(III), Yb(III), Gd(ema)3 and Gd(dpp) 3 ; 4 6 " 4 8 however, these complexes have neither been fully characterized, nor have their potential utilities as bone density agents been investigated. Herein, the synthesis and complete characterization of the neutral Ln(III) ^rw(bidentate ligand) complexes using ma", ema", mpp", and dpp" as the ligands are presented. La(III) j Gd(III) and Yb(III) ions were investigated, with selection of specific lanthanides based on their known medicinal applications, and for size comparison. The luminescent properties of the Eu(III) and Tb(III) complexes were also studied, and the suitability of these compounds as therapeutic agents for the treatment of bone resorption disorders was assessed in cytotoxicity studies, comparative Afunctional transport studies in human colon carcinoma cells with intestinal cell-like properties (Caco-2 cells), 4 8 " 5 1 and hydroxyapatite (HA) binding assays. 130 3.2 Exper imental 3.2.1 Materials N a O H , 3-hydroxy-l,2-methyl-4-pyridinone (Hdpp), hydrated lanthanum carbonate and hydrated lanthanide nitrate salts were obtained from Sigma-Aldrich, Fisher Scientific or A l fa Aesar and used without further purification. A l l solvents were reagent grade from Fisher Scientific. Maltol (Hma) and ethylmaltol (Hema) were obtained from Pfizer or Cultor Food Science. Water was deionized (Barnstead D8902 and D8904 Cartridges) and distilled (Hytrex II GX50-9-7/8 and GX100-9-7/8) before use. Water for H A studies and I C P - M S analysis was collected from an Elgastat Prima 2 reverse osmosis and ultra pure water system (High Wycombe, England). N M R solvents were purchased from Aldrich and used as received. Yields for the analytically pure compounds were calculated based on the respective metal ion starting material. A complete list of compounds prepared and analytical data is presented in Table 3.1. 3.2.2 Instrumentation ' H N M R spectra (300 M H z ) and l 3 C N M R spectra (75 M H z ) were recorded on a Bruker Varian XL-300 spectrometer and referenced internally to residual solvent protons. Infrared spectra were recorded as nujol mulls ( K B r windows) in the range 4000-500 cm"1 on a Mattson Galaxy Series FTIR-5000 spectrophotometer and were referenced to polystyrene (1601 cm' 1 ). Analyses of C, H , and N were performed by M r . M . Lakha at the University of British Columbia. UV-vis ible spectra were recorded in water or 0.1 % saline solution using a Hewlett-Packard 8543 diode 131 array spectrophotometer and a 1 cm cuvette. Mass spectra were obtained on a Bruker Esquire Ion Trap (electrospray ionization mass spectrometry, ESI -MS) spectrophotometer. Fluorescence spectra were obtained in 0.1 % saline solution or methanol on a PTI QuantaMaster fluorimeter using a 1 cm quartz cuvette. A high resolution ICP-MS (Thermo Finnigan E L E M E N T 2 ) was used for determination of metal ion concentration, calibrated against Delta Scientific Ltd. Standards. 3.2.3 Synthesis 3-BenzyIoxy-2-methyl-4-pyrone (Bnma). This intermediate was prepared similarly to the previously published procedure with some modifications. 4 5 3-Hydroxy-2-methyl-4-pyrone (5 g, 40 mmol) was dissolved in 200 m L of methanol and a solution of sodium hydroxide (3 g, 75 mmol in 30 m L of water) was added. After addition of benzyl chloride (7 mL, 61 mmol), the mixture was refluxed overnight and cooled in an ice bath. M e O H was removed from the yellow solution by rotary evaporation, 20 m L of additional water was added.and the product was extracted into methylene chloride (2 x 20 mL). The water layer was discarded and the CH2CI2 extracts were washed with 5 % N a O H ( 3 x 1 5 mL) , then water (1 x 20 mL) . The CH2CI2 layer was dried over MgS04 , filtered, and the solvent was removed by rotary evaporation affording a viscous yellow oi l . The oi l was dissolved in hot ethanol and maintained at' -5°C for 48 h, after which time white cubic, crystals formed. These were filtered out, washed with cold E t O H ( 3 x 5 mL) and vacuum dried (6.8 g, 31.5 mmol, 79 % yield). ' H N M R (300 M H z , R T , CDC1 3 ) 5 = 2.04 (s, 3H, CH3), 5.10 (s, 2H , B n - C / / 2 ) , 6.33 (d, 1H, Hb, 3 J M = 5.1 H z ), 7.29 (m, 5H, B n C6H5), 7.54 (d, l H , / / A , 3 J a , b = 5 . 6 ) . 132 3-Benzyloxy-2-methyl-4(i//)-pyridinone (Bnmpp). Bnma (6.8 g, 31.5 mmol) in 80.5 m L of concentrated N H 4 O H and 30 m L of ethanol was stirred in a round bottom flask at room temperature for 3 days. After each day an additional 10 m L of ethanol was added. E t O H and the excess ammonia were removed by rotary evaporation and the residue washed with 10 m L of acetone. The 3-benzyloxy-2-methyl-4(177)-pyridinone, Bnmpp, product was crystallized from ethanol as colorless crystals, filtered, and dried under vacuum overnight (4.3 g, 20.1 mmol, 64 % yield). ' H N M R (300 M H z , R T , C D 3 O D ) 5 = 2.05 (s, 3H, C i / 3 ) , 5.01 (s, 2H , B n - C / / 2 ) , 6.34 (d, 1H, Hb, 3Jb,a = 5.8 H z ), 7.31 (m, 5H, B n C6H5), 7.86 (d, 1H, Ha, \ b = 5.6 Hz). 3-Hydroxy-2-methyl-4(l/f)-pyridinone (Hmpp). Bnmpp (4.3 g, 20.1 mmol) was debenzylated by treatment with 10 m L of 45 % w/v hydrogen bromide-acetic acid solution for 30 min on a steam bath. Excess solvent was removed by rotary evaporation affording an off white percipitate. The precipitate was filtered out, redissolved in 10 m L of water and brought to p H 7 -8 with the addition of 6 M N a O H . The solvent was removed by rotary evaporation. 3-Hydroxy-2-methyl-4(7//)-pyridone (Hmpp) crystallized from ethanol as light pink crystals after 24 h at 4°C (1.1 g, 8.64 mmol, 43 %). ' H N M R (300 M H z , RT, C D 3 O D ) 8 = 2.17 (s, 3H, CH3), 6.09 (d, 1H, Hb, 3Jb,a=6.9 H z ), 7.39 (d, 1H, Ha, V a ; b= 6.9 Hz), 8.31 (s, 1H, OH). M S (+ESI-MS, MeOH) : m/z 126 [L+H] + General procedure to prepare Ln(ma)3»nH20, Ln(III) = La, Eu, Gd, Tb, Yb. L n ( N 0 3 ) 3 ' 6 H 2 0 (0.433-0.467 g, 1.0 mmol) was added to a solution of Hma (0.378 g, 3.0 mmol) in 10 m L of deionized, distilled water and the solution was heated and stirred until the ligand 133 completely dissolved (-10 min). The pH of the clear solution was raised slowly over 10 min with 1 M N a O H (Ln(III) = La , p H 8.03; Eu, p H 8.00; Gd, p H 8.26; Tb, p H 8.5; Y b , p H 8.62) and the resulting reaction mixture was continuously stirred for 16 h. A white, or slightly off-white precipitate formed (except in the case of Eu(ma)3 for which a yellow solid formed) and was collected by vacuum filtration, washed with 2 x 5 m L of cold water and 2 x 5 m L of cold M e O H . A l l compounds were dried in vacuo to yield 36-59 % of product. General procedure to prepare Ln(ema) 3»nH 20, Ln(III) = La, Eu, Gd, T b , Y b . L n ( N 0 3 ) 3 ' 6 H 2 0 (0.433-0.467 g, 1.0 mmol) was added to a solution of Hema (0.420 g, 3.0 mmol) in 10 m L of ethanol and the solution was heated and stirred until the ligand completely dissolved (-5 min). Triethylamine (0.303 g, 3 mmol) was added slowly over 10 min to the clear solution. The resulting reaction mixture was continuously stirred for 18-24 h. A white or slightly off-white precipitate formed (except in the case of Eu(ema) 3 for which a yellow solid formed) and was collected via vacuum filtration, washed with 2 x 5 m L of cold water and 2 x 5 m L of cold M e O H . A l l compounds were dried in vacuo to yield 53-85 % of product. General procedure to prepare Ln(dpp)3*nH20, Ln(III) = La, Eu, Gd, T b , Y b . L n ( N 0 3 ) 3 » 6 H 2 0 (0.433-0.467 g, 1.0 mmol) was added to a solution of Hdpp (0.417 g, 3.0 mmol) in 10 m L of deionized, distilled water and the solution was heated and stirred until the ligand completely dissolved (-30 min). The p H of the clear solution was raised slowly over 10 min with 1 M N a O H (Ln(III) = La , p H 10.02; Eu, pH 10.20; Gd, p H 10.42; Tb, p H 11.00; Y b , p H 11.13) and the resulting reaction mixture was continuously stirred for 16 h. A white, or slightly off-white precipitate formed (except in the case of Eu(dpp) 3 for which a yellow solid formed) 134 and was collected by vacuum filtration, washed with 2 x 5 m L of cold water and 2 x 5 m L of cold M e O H . A l l compounds were dried in vacuo to yield 40-76 % of product. General procedure to prepare Ln(mpp) 3»nH20, Ln(III) = La, Eu, Gd, Tb, Yb. L n ( N 0 3 ) 3 » 6 H 2 0 (0.144-0.154 g, 0.33 mmol) was added to a solution of Hmpp (0.125 g, 1.0 mmol) in 10 m L of deionized, distilled water and the solution was heated and stirred until the ligand completely dissolved (-30 min). The p H of the clear solution was raised slowly over 10 min with 1 M N a O H (Ln(III) = La , p H 8.10; Eu, p H 8.20; Gd, p H 8.20; Tb, p H 8.30; Y b , p H 8.30) and the resulting reaction mixture was continuously stirred for 16 h. A white or slightly off-white precipitate formed (except in the case of Eu(mpp) 3 for which a yellow solid formed) and was collected by vacuum filtration, washed with 2 x 5 m L of cold water and 2 x 5 m L of cold M e O H . A l l compounds were dried in vacuo to yield 38-64 % of product. 3.2.4 Caco-2 cell culture Caco-2 cells were purchased from American Type Culture Collection ( A T C C , Rockville, M D and Manassas, V A , U S A ) . Cel l culture media were purchased from Gibco B R L (Grand Island, N Y , U S A ) . Sterile Steritop™ 0.22 pm Express membrane bottle top filters were purchased from Mill ipore (Bedford, M A , U S A ) . Sterile 50 ml centrifuge tubes, disposable 10 and 25 m L stereological pipettes were purchased from Starstedt (Montreal, PQ , Canada). Culture flasks, Transwell and polycarbonate plates were obtained from Corning-Costar (Cambridge, M A , U S A ) . Triton X-100 was purchased from Sigma. CellTiter 96® Aqueous One Solution Assay and CytoTox 96 were from Promega (Madison, WI). 135 Caco-2 cells for M T T studies were cultured in 75 c m 2 Falcon tissue culture flasks in Eagle's Min imum Essential Medium ( E M E M ) supplemented with Eagle's balanced salt solution (EBSS) and 2 m M glutamine modified with 1.0 m M sodium pyruvate, 0.1 m M nonessential amino acids, 1.5 g/L sodium bicarbonate and supplemented with 20 % fetal bovine serum. The cells were incubated at 37°C in a humidified atmosphere of 5 % C 0 2 : 9 5 % O2. Medium was changed every third day and the cell stock was routinely subcultured by trypsinization (0.25 % trypsin with 0.53 m M E D T A ) after reaching 80 % confluency. Caco-2 cells for M T S , L D H , cell uptake and transport studies were cultured in E M E M supplemented with 10 % fetal bovine serum (FBS), 292 ug/mL glutamine, 0.1 m M non-essential amino acids, 100 ug/mL penicillin and 100 ug/mL streptomycin at 37°C in humidified air containing 5 % C02:95 % O2 in either T75 flasks, 6-well, 12-well, 96-well or Transwell plates depending on the type of experiment. Media were changed every other day. Cells used for experiments reached 80 to 90 % confluency (in the case of transfer studies when the transepithelial electrical resistance (TEER) values were above 400 Q / c m 2 , measured using Mi l l i ce l l -ERS from Millipore). Data from cell permeability, cytotoxicity and hydroxyapatite binding (vide infra) studies were analyzed by A N O V A and Newman-Keuls post-analysis, where appropriate, and expressed as means + SD. Studies were compared against La2(C03)3 unless otherwise stated. Statistical significance was accepted at P < 0.05. 136 3.2.5 Biological studies M T T assay. A modified procedure was used for the M T T assay ( M T T = 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide). 5 2 Caco-2 cells were seeded onto 96-well plates at a density of 1 x 10 4 cells per well. On the day of the experiment, the culture medium was exchanged for treatment solution of 1 ug/mL, 10 ug/mL, 50 ug/mL, 100 ug/mL, 200 ug/mL and 500 ug/mL (equivalent to 2 - 1092 uM) of Hma, Hema, Hdpp, Hmpp, Ln(ma) 3, Ln(ema) 3, Ln(mpp) 3 , and Ln(dpp) 3 , where Ln(III) = La , Gd and Y b (dissolved in 1 % DMSO/media solution; highest concentration of D M S O 100 ug/mL). After an incubation of 72 h, a 50 u L aliquot of M T T (2.5 mg/mL in PBS) was added to each well of the plate. The plate was further incubated for 3 h at 37°C during which time purple formazan crystals formed at the bottom of the wells. The solution of each well was carefully aspirated leaving the formazan crystals which were dissolved in D M S O (150 uL). The absorbance of each well was measured at 570 nm using a microplate reader (Spectra Max 190, Molecular Devices). Cel l viability was calculated relative to the 100 % control ( D M S O plus media). The IC50 was determined by plotting the percent cell viability vs. drug concentration (logarithmic scale) and finding the concentration at which 50 % of the cells were viable relative to the control. Values are presented as means + SD, based on > 6 independent trials. M T S assay. Caco-2 cells were seeded onto 96-well plates at a density of 40,000 cells/cm . Media was changed every other day. Cells were kept at 37°C in a humidified atmosphere of 5 % C02:95 % O2 until 90 % confluency was reached (after about 2 weeks). On the day of the experiment, the cells were washed 1 to 2 times with P B S and media. The culture medium was 137 exchanged for treatment solutions of 1 ug/ml, 10 ug/mL, 50 ug/mL, 100 ug/mL, 200 ug/mL and 500 ug/mL (equivalent to 2 - 1092 uM) of Hma, Hema, Hdpp Ln(ma)3, Ln(ema)3, Ln(dpp)3, where Ln(III) = L a , G d and Y b (dissolved in 1 % DMSO/media solution), and 1 % Triton X-100 (as positive control for cytotoxicity) in Hanks' Balanced Salt Solution (HBSS) without phenol containing 10 m M H E P E S , p H 7.4. After an incubation period of 24 h, 20 u L of CellTiter® AQueous One Solution Reagent from the CellTiter 96® A Q u e o u s One Solution Cel l Proliferation Assay kit (Promega, Maddison, WI) was added to each well . The plate was wrapped in aluminum foil and incubated for 1 - 4 h at 37°C in a humidified, 5 % C02:95 % 0 2 atmosphere. Absorbance was measured at room temperature at 492 nm with a Multiscan Ascent Multi-plate reader from Labsystems. The percent cell viability was calculated relative to the 100 % control (media plus cells). The IC50 was determined by plotting the percent cell viability vs. drug concentration (logarithmic scale) and finding the concentration at which 50 % of the cells were viable relative to the control. The estimated standard deviations are based on at least 3 trials. L D H assay. A 50 u L aliquot of the treatment solution was removed from the 96-well plate used for the M T S assay after the 24 h incubation period and transferred into a new 96-well plate. Using the protocol for the Cytotox96® Non-Radioactive Cytotoxicity Assay (Promega, Madison, WI), 50 u L of the L D H Substrate M i x was added to each well . After the plate was covered with foil and incubated for 30 min at room temperature, 50 pl of Stop™ solution was added to each well. The absorbance was read at 492 nm with a Multiscan Ascent Multi-plate reader from Labsystems within 1 h of the stop solution addition. Cytotoxicity was calculated relative to 100 % cytotoxity obtained from the Triton X-100 group. The IC50 was determined by plotting the percent cell viability vs. drug concentration (logarithmic scale) and finding the concentration 138 at which 50 % of the cells were viable relative to the control. The estimated standard deviations are based on at least 3 independent trials. Protein Concentration. Total protein content of lysed cells was measured by the bicinchoninic acid protein assay ( B C A ™ Protein Assay Kit) from Pierce Biotechnology, Inc. (Rockford, IL, U S A ) . After performing the cell uptake and bifunctional transport studies, the assay media was aspirated, cells were washed 3 times with PBS and were lysed. A B C A standard curve was constructed in the range of 25 ug/mL to 2 mg/mL. Exactly 10 u L of both standards and cell samples (in triplicates) were aliquoted into a 96-well microtiter plate and 200 uL of reagent (Mix A : B = 50:1) was added to each well . Absorbance was measured after 1 h at room temperature at 540 nm with a Multiscan Ascent Multi-plate reader from Labsystems. Protein concentration of each sample was determined against the standard curve. Caco-2 cell uptake studies. Caco-2 cells, passage number 25 to 30, were seeded at 5000 cells/cm into 6-well plates and incubated in 37°C in a humidified atmosphere of 5 % C02:95 % O2. The medium was changed every other day. After reaching 80 % confluency, the cells were washed with P B S . The medium was removed and replaced with either fresh medium, or medium plus compound (La(ma)3, Gd(ma)3, Yb(ma)3, La(ema)3, La(dpp)3, L a 2 ( C 0 3 ) 3 ) at concentrations of 100 ug/mL (equivalent to 218 - 179 uM) . The plate was incubated at 37°C in a humidified atmosphere of 5 % C02:95 % O2 for 4 h. Treatment solutions were aspirated and cells were washed three times with ice-cold P B S to remove any unbound LnL3. Cells were lysed with 1 m L of lysis buffer (50 m M H E P E S , 150 m M N a C l , 2 m M E D T A , 0.5 % Na-deoxycholate, 1 % N P -40) plus protease inhibitor cocktail (1:1000 dilution). The lysed cell solution was transferred 139 into separate 1.5 m L Eppendorf tubes. These wells were rinsed with another 0.3 m L of lysis buffer; the combined cell/media samples were kept for later analysis by I C P - M S . The protein content was determined by a Pierce B C A assay using 10 u L of the lysed cells (vortexed thoroughly and pelleted by centrifugation at 12,000 rpm for 3 min). The concentrations are reported as % uptake of u M Ln(III)/mg protein ± SD. Group values were compared by A N O V A with a Newman-Keuls post analysis. A P value < 0.05 was considered to be statistically significant. Bifunctional transport assay. Caco-2 cells, passage number 25, were seeded at 5000 cells/cm 2 in polycarbonate membrane 12-well Transwell plates (Corning, Costar Corp.) and incubated at 37°C in a humidified atmosphere of 5 % C02:95 % 0 2 . The media was changed every other day. Uptake studies were conducted when the T E E R readings reached 400 Q/cm 2 . Prior to permeability studies, the cells were washed with PBS and the culture media was replaced with fresh media (control) or media plus compound (concentration of 100 ug/mL, equivalent to 218 -179 u M , for all tested compounds including La(ma)3, Gd(ma)3, Yb(ma)3, La(ema)3, La(dpp) 3 and La 2 (C03)3) on the apical or basolateral side. The integrity of the Caco-2 cell monolayer was monitored by measuring the T E E R values at each time point. The amount of permeated solute was determined by removing 100 u L aliquots from either the basolateral or apical side, opposite to the side on which the compound was added, at 10, 30, 60, 120, 180 and 240 min and replaced with fresh media. The plate was incubated at 37°C in a humidified atmosphere of 5 % C 0 2 : 9 5 % 0 2 in between sample removal periods. A l l media was aspirated after 4 h, vacuum dried, and acid digested for further analysis. The cells were washed 3 times with P B S , the membrane was removed and lysed in 1 m L of 0.1 N N a O H . Exactly 10 u L of the lysed cells (vortexed 140 thoroughly and pelleted by centrifugation at 12,000 rpm for 3 min) were analyzed for protein content using the B C A assay. The amount of Ln(III) ions accumulated in the cells, in addition to the concentration of Ln(III) ions in the apical or basolateral chambers were evaluated by ICP-M S . The mean percentage of compound transported + SD was determined by dividing the Ln(III) concentration found in either the basolateral or apical chamber by the total concentration of Ln(III) found in the study (concentration of Ln(III) in the apical and basolateral chambers, and in the cell membrane). The apparent permeability ( P a p p , cm s"1) coefficients from the bidirectional transport of the Ln(III) complexes were calculated using Equation 3 .1 . Papp=(dQ/dt)/(A.C0) (3 .1) dQ/dt (nmol s"1) is the flux rate of mass transport across the monolayers, A is the surface area of the insert membrane (1.13 cm 2 ) , and C 0 is the initial concentration (uM) of the compound in the donor chamber. The flux rates were determined by plotting the concentration of Ln(III) ions found using I C P - M S as a function of time, with the slope calculated via linear regression. Hydroxyapati te in vitro b inding study. Samples containing exactly 20 mg of dried hydroxyapatite ( H A , Sigma, St. Louis, M O , U S A ) were suspended in 2.00 m L of physiological 0.9 % saline solution, pH 7.4 in 13 x 100 mm test tubes and incubated overnight at 37°C to allow for equilibration of the samples. The metal complexes were dissolved in D M S O , with a final concentration of 5 mg/mL. Appropriate volumes of the D M S O solutions were added in order to 141 achieve 1 u M concentration for all compounds tested (maximum concentration of D M S O in each sample < 0.0005 %) and the mixture was shaken gently. After incubation (5, 15 min, 3 and 24 h; 6 vials for each time point), the solution was centrifuged at 3,200 rpm for 5 min. Supernatants were carefully removed and placed in 10 m L scintillation vials. Each pellet was washed two additional times with 1.00 m L of saline solution to remove any unbound metal ion from the pellet. The washings were collected after centrifugation and the supernanant was filtered through a 0.22 pm Mil l ipore filter. Exactly 1.00 m L of the supernatant samples were vacuum centrifuged overnight (excluding Eu(ma) 3 and Tb(ma) 3 samples). The dried samples were resuspended in 2 m L of concentrated Seastar nitric acid for digestion (see procedure below) and diluted accordingly for I C P - M S analysis: The remaining supernatant solution of appropriate dilutions was used for the determination of the free ligand content by U V - v i s . The Eu(ma) 3 and Tb(ma) 3 samples were used after filtering and diluted appropriately for fluorescence studies. The H A samples were dissolved in 2.00 m L of concentrated Seastar nitric acid, acid digested and diluted accordingly for I C P - M S analysis. The total amount of Ln(III) for each HA-binding experiment was determined by adding the Ln(III) concentration found in the supernatant with the concentration found in the H A samples. Percent HA-binding + SD was determined by dividing the Ln(III) concentration from the H A sample by the total Ln(III) concentration found (HA and supernatant) for each sample. Group values were compared by A N O V A with Newman-Keuls post analysis. A P value < 0.05 was considered to be statistically significant. Ln(III) analysis by I C P - M S . High resolution I C P - M S Thermo Finnigan Element2 was used to analyze the Ln(III) content of the cell lysates, solutions from both the apical and basolateral transport studies, and solutions and solid H A samples from the H A binding experiments. High 142 purity standards ( l n 2 0 3 , L a 2 0 3 , G d 2 0 3 , and Y b 2 0 3 1000 ug/mL in 2 % H N 0 3 ) w e r e purchased from Delta Scientific Laboratory Products L T D (Spring City, P A ) . A l l materials used for the digestions of samples and for I C P - M S including glassware, Teflon boiling chips, sample holders, pipettes, and pipette tips were soaked in 2 % Extran solution for 24 h, rinsed with deionized, distilled water, soaked overnight in 1 % Seastar nitric acid, rinsed with deionized distilled water, and finally left to dry in a dust-free environment prior to use. A l l samples were vacuum centrifuged to complete dryness and digested in acid using the following procedure before analysis. The redissolved digests were diluted accordingly with 1 % nitric acid and their lanthanide contents analyzed by monitoring m/z 139, 158, and 174 for 1 3 9 L a , 1 5 8 G d , l 7 4 Y b , respectively. Samples were analyzed in the 'peak jump' mode and all signals corrected according to the internal U 5 I n standard. Interference corrections for 1 5 8 G d were made by monitoring Dy(III) at m/z 161. A c i d digestion for I C P - M S analysis. Dried samples (vacuum centrifugation) were dissolved in 2.00 m L of Seastar concentrated nitric acid. Samples were transferred to test tubes with two or three Teflon boiling chips and placed in a block heater (Standard Heatblock, V W R Scientific Products). The temperature was slowly raised to 105°C over 1 h and then maintained at 105°C for 24 h. Approximately 2.00 m L of 30 % hydrogen peroxide (Fisher Scientific) was added to each sample before heating at 140°C for 24 h. The temperature was increased to 150°C and samples were evaporated to dryness. Subsequently, samples were redissolved in 3.00 m L of 10 % nitric acid and stored in 4.00 m L high density polyethylene bottles (Wheaton, Mi lv i l l e , NJ) at 4°C. 143 Ligand analysis by U V - V i s assay. A calibration curve consisting of appropriate concentrations of dissolved ligand in 0.9 % physiological saline solution was constructed for Hma, Hema and Hdpp at ^ m a x = 273, 279 and 275 nm, respectively. The undigested supernatants from La(ma)3, Eu(ma)3, Gd(ma)3, Tb(ma) 3 , Yb(ma) 3 La(ema)3, and La(dpp)3-HA binding studies for all four time points were diluted 100 fold with 0.9 % physiological saline solution. The concentration of the ligand in solution was calculated by comparing the absorbences collected from the supernatant solutions to the previously constructed calibration curves. High concentration of the ligand in solution indicates disassociation of the LnL3 complex upon H A binding. Percent ligand in solution was determined by dividing the ligand concentration found in the supernatant by the total ligand concentration added. Eu(ma)3 and Tb(ma)3 analysis by fluorescence assay. A calibration curve over a range of concentrations was constructed by integrating the area under the most intense fluorescence peak (^exc = 315 and 319 nm, ~kem = 641 and 642 nm for Tb(ma)3 and Eu(ma)3, respectively) and plotted against the known concentration. Concentrations of the complexes in solution were determined by comparing their integrated area under the most intense fluorescence peak against the calibration curve. Percent Ln(III) bound was determined by subtracting the amount of Tb(III) or Eu(III) found in the supernatant from the total amount of Ln(III) added to the sample. This value was then divided by the total ligand concentration. Xylenol Orange Assay. Approximately 20 mg of Hma, Hema, Hdpp, H A , LnL3, L n ( N 0 3 ) 3 , and solid samples from the metal-HA binding studies as well as the supernatant from the same experiment were dissolved in minimal amounts of 6 M HC1. The p H was adjusted to 5.0 and 144 diluted appropriately with an aqueous 20 % hexamethylenetetramine buffer. Presence of free Ln(III) ions were confirmed by the immediate appearance of a red color in the solution upon the addition of xylenol orange sodium salt. 3.3 Results and Discussion 3.3.1 Ligand synthesis Hmpp, the only ligand used herein not commercially available, was readily synthesized using a modified literature procedure (Scheme 3.2). 4 5 The hydroxyl group of maltol was benzyl protected using a Will iamson ether synthesis with benzyl chloride to produce Bnma. Direct ammonolysis without employing a blocking group was unsuccessful attributed to the deprotonation of the O H groups in the basic ammonia solution causing increased electon-density at C2 , the position most susceptible to nucleophilic attack by primary amines. 5 4 Continuous stirring of the protected maltol in concentrated ammonia for 3 days, however, produced the corresponding pyridinone, Bnmpp. Debenzylation by acid hydrolysis with 45 % w/v HBr/acetic acid resulted in the desired pyridinone, Hmpp, in moderate yields. H H Scheme 3.2. Synthesis of Hmpp: (a) benzyl chloride, M e O H / H 2 0 , N a O H , 79 %; (b) N H 4 O H , E tOH, 64 %; (c) 45 % w/v HBr/acetic acid, 43 %. 145 3.3.2 Metal complexes The neutral LnL3, where L = ma", ema", mpp", or dpp", complexes were synthesized using simple procedures (Scheme 3 .3) . 4 6 " 4 8 ' 5 5 - 5 8 Metal nitrate salts were added to a solution of the ligand precursor in either water or ethanol in a 1:3 ratio. The proligand was deprotonated by careful adjustment of the p H to ~7 - 11 using aqueous 1.0 M N a O H , or equivalent moles of triethylamine. The precipitated products were collected by vacuum filtration, washed with cold water, ethanol and/or methanol, and dried overnight in vacuo. Due to the tendency of Ln(III) ions to hydrolyze easily with increasing pH, extreme care was taken when adjusting the pH of the solution. The complexes precipitated from the respective reaction mixtures to give white or off-white air stable powders in moderate to high yields, with the exception of the Eu(III) complexes which precipitated as yellow powders. o R = C H 3 , X = O Ln(ma) 3 R = C 2 H 5 X = 0 Ln(ema) 3 R = C H 3 X = N H Ln(mpp) 3 R = C H 3 X = N C H 3 Ln(dpp) 3 Scheme 3 .3 . 3-Hydroxy-4-pyrones and 3-hydroxy-4-pyridinones of interest and their Ln(III) complexes (Ln(III) = La , Eu , Gd, Tb, Yb) 146 Unfortunately, attempts to recrystallize the complexes (in hot M e O H or EtOH) resulted in decomposition of the product and hydrolysis of the Ln(III) ions. Column chromatography was useless for purification of the compounds due to their low solubilities; however, washing the precipitated solid thoroughly with cold solvents to remove unreacted ligand, additional L n ( N 0 3 ) 3 salts or other side products yielded pure complexes, as characterized by elemental analysis (EA, Table 3.1). The compounds were thoroughly characterized by E A , positive ion electrospray ionization-mass spectrometry (+ESI-MS), IR, and N M R (LaL 3 ) . IR, E A and +ESI-MS results are all consistent with the formation of neutral L n L 3 species. The elemental analyses for all the lanthanide complexes confirmed the empirical formula LnL 3 »nH20. Hydrated species have been observed with the previously synthesized Ln(ma) 3 and Gd(ema) 3 systems, where at least one water molecule was associated with the metal compounds (Table 3.1) 4 6 ' 4 7 In addition, the N - H functionality of the mpp" ligand is a potential hydrogen bond donor that may increase hydration numbers for its complexes. Electrospray ionization mass spectrometry in the positive ion mode gave diagnostic mass spectra with the expected isotopic patterns. The most intense peak corresponded to the sodium adduct of the parent ion peak [ N a L n L 3 ] + (Table 3.1). Peaks corresponding to [ L n L 2 ] + species generated by the gas-phase loss of one ligand were observed in addition to peaks with the formula [ L n 2 L 5 ] + , as commonly reported for other trivalent metal-/ra(0,0 bidentate ligand) analogs. 3 7 ' 5 9 147 Table 3.1 Elemental analyses, % yield and +ESI-MS results for all prepared L n L 3 complexes Complex % C found (calcd) % H found (calcd) % N found (calcd) % yield [ N a L n L 3 ] + La(ma) 3 42.04 (42.40) 2.94 (3.20) 4 0 % 537 E u ( m a ) 3 » H 2 0 39.65 (39.57) 3.14 (3.32) 6 2 % 551 G d ( m a ) 3 » H 2 0 39.27 (39.28) 3.11 (3.11) 5 9 % 556 T b ( m a ) 3 » H 2 0 39.15 (38.87) 3.10 (3.04) 5 4 % 557 Yb(ma) 3 39.43 (39.43) 2.76 (3.10) 3 6 % 571 La(ema) 3 45.34 (45.74) 3.80 (4.14) 6 8 % 579 Eu(ema) 3 »2 .5H 2 0 41.02 (41.05) 3.94 (4.2.7) 6 4 % 593 Gd(ema) 3 43.89 (44.18) 3.68 (3.97) 53 % 598 T b ( e m a ) 3 « H 2 0 42.44 (42.61) 3.90 (3.91) 5 6 % 599 Yb(ema) 3 42.72 (42.32) 3.58 (3.89) 85 % 614 L a ( m p p ) 3 » 2 H 2 0 36.33 (36.06) 4.94 (4.71) 6.05 (6.12) 5 2 % 534 E u ( m p p ) 3 » 2 H 2 0 38.58 (38.63) 3.96 (3.68) 7.50 (7.87) 4 6 % 547 Gd(mpp) 3 40.82 (40.49) 3.43 (3.82) 7.93 (8.06) 3 8 % 553 T b ( m p p ) 3 » 2 H 2 0 38.11(37.80) • 3.91 (3.71) 7.41 (7.62) 5 6 % 554 Yb(mpp) 3 42.93 (42.78) 4.12 (3.89) 7.15 (6.92) 6 4 % 568 La(dpp) 3 45.58 (45.95) 4.37 (4.74) 7.59 (7.82) 40 % 576 Eu(dpp ) 3 »H 2 0 43.16 (43.18) 4.48 (4.82) 7.19(7.18) 3 8 % 590 Gd(dpp) 3 »H 2 0 42.77 (42.76) 4.44 (4.67) 7.13 (7.04) 7 6 % 595 Tb(dpp) 3 »2H 2 0 41.39 (41.67) 4.63 (4.88) .6.90 (7.03) 5 4 % 596 Yb(dpp) 3 «3H 2 0 39.32 (39.25) 4.71 (4.54) 6.55 (6.82) 68 % 610 148 The IR spectra for all the complexes were dominated by peaks corresponding to the ligand (Table 3.2) 4 6 ' 6 0 ' 6 1 In addition, the IR spectra for L n L 3 were nearly superimposable within each ligand set (<10 cm"1 variations between Ln(III) ions with the same ligand, attributed to mass differences amongst the metal ions), indicating that each series of the Ln(ma) 3 , Ln(ema) 3, Ln(mpp) 3 , and Ln(dpp) 3 complexes, regardless of the Ln(III) ion, are isostructural in the solid state. The vc=o stretching bands were shifted 40-50 cm"1 to lower energy with respect to the proligand, evincing normal coordination of the metal via the ketonic oxygen and the oxygen of the deprotonated phenol group. The broad band for V 0 H disappeared upon complexation indicating complete deprotonation of the phenol O H group. New bands at 3400 and 3200 cm"1 appeared for the pyrone and pyridinone complexes, respectively, due to lattice and coordinated water molecules. Resonance of the coordinated aromatic bidentate ligands resulted in a hyperchromic shift observed for vc-o due to the increasing double bond character of the C - 0 linkage. In addition, the vc=c vibrational modes were bathochromically shifted indicating a decrease in the double bond character of the aromatic ring due to donation of n electrons to the chelate ring. Similarly, this trend is also seen in Ln(ONO) systems. 6 2 The absence of VNO3 (-1382 cm"1) in the IR spectra indicated the absence of any counterions in the complexes, while the appearance of new peaks between 400-500 cm"1 for VM-O stretching frequencies confirmed the complexation of ligands to their respective metal i ons . 6 3 ' 6 4 The literature crystal structure of Pr(ma) 3 revealed three bidentate deprotonated coordinated ligands and two bound water molecules arranged in an irregular geometry around the Pr(III) ion, perfectly matching the analytical data obtained in this study. In addition, the C-O bond distances of the donor oxygens were not significantly different, suggesting that the 149 negative charge is delocalized upon coordination, as. would be expected in the reported complexes. Table 3.2 Selected IR stretches (cm"1) of the Ln(III) complexes. Hma La(ma)3 Gd(ma)3 Yb(ma) 3 Hema La(ema)3 Gd(ema)3 Yb(ema)3 V O H ( H 2 0 ) 3435 3439 3452 3381 3420 3454 VOH (b,s) 3250 3250 VRing 1555 1513 1514 1516 1555 1505 1515 1516 1485 1470 1470 1471 1451 1450 1452 1456 vc=o. 1650 1597 1599 1604 1650 1599 1605 1608 1615 1571 1574 1580 1614 1590 1592 1595 vco 1294 1332 1333 1334 1317 1333 1333 1334 vc-o-c 1260 1283 1286 1291 1266 1287 1287 1292 . VCH3 2927 2973 2974 2976 2982 2975 2976 2977 V M - 0 427 436 440 425 427 441 Hmpp La(mpp)3 Gd(mpp)3 Yb(mpp) 3 Hdpp La(dpp) 3 Gd(dpp) 3 Yb(dpp) 3 VOH ( H 2 0 ) 3265 3262 3250 -3424 3414 3357 VOH (b,s) . 3270 3150 VRing 1540 1503 1520 1528 1530 1504 1505 1510 1500 1481 1501 1512 1515 1485 1501 1502 1420 1395 1402 1403 1400 1456 1456 1459 vc=o 1645 1598 1599 1601 1630 1591 1597 1598 1565 1538 1544 1548 VCH3 2950 2954 2955 2955 2940 2947 2955 2955 vco 1335 1361 1362 1364 1335 1363 1364 1366 V C N 1250 1294 1295 1299 V C - N - C 476 531 531 535 476 481 486 487 V M - 0 432 435 445 436 442 446 The N M R spectra of LaL3 were also useful in elucidating the structure of the metal complexes. A l l diamagnetic L a L 3 were studied in solution by *H N M R spectroscopy (Table 3.3 and Figure 3.2). Only La(ma)3 and La(ema)3 were analyzed by 1 3 C N M R spectroscopy due to 150 limited solubility of the pyridinone complexes (Table 3.4). The absence of ligand peaks and the relative shifts of the resonances in both the ! H and 1 3 C N M R signals conclusively indicated binding of the ligand to the La(III) ion. Both resonances for the ring hydrogens, Ha and Hb, for the La(III) pyrone and pyridinone metal complexes shifted upfield. Similar upfield shifts were reported in the ' H N M R spectra for the corresponding Ln(III) complexes using pyridinethiones and pyronethiones as ligands. 5 6 Table 3 .3 . ' H N M R data for La(III) complexes (300 M H z , R T , d6 - D M S O , X = O for Hma, Hema, La(ma) 3 and La(ema)3, N for Hmpp, La(mpp) 3 , Hdpp, and La(dpp) 3). The ! H chemical shifts are listed in ppm and the 3J coupling constants are listed in parentheses (Hz). Compound R a #a(d) Hb(d) OH Hma 2.23 8.00 (5.4) 6.33 (5.4) 8.80 La(ma) 3 2.20 7.93 (5.0) 6.23 (3.1) Hema 2.61 (7.2) 1.09 (7.3) 8.01 (4.6) 6.33 (5.4) 8.79 La(ema) 3 2.57(7.5) 1.00 (7.3) 7.90 (5.0) 6.16(4.6) Hmpp 2.17 7.39 (6.9) 6.09 (6.9) 8.31 La(mpp) 3 b 2.15 7.26 (br, vj/2 = 35 Hz) 6.07 (br, vi/2 = 35 Hz) Hdpp 2.25 3.62° 7.54 (7.31) 6.07 (6.85) 8.31 La(dpp) 3 b 2.26 3.63° 7.47 (br, vi/2 = 60 Hz) 6.03 (br, vi/2 - 60 Hz) Abbreviations: s = singlet; d = doublet; t = triplet; q = quartet; br = broad aR=CH3 (s) for Hma, Hmpp, Hdpp, La(ma) 3, La(mpp) 3, La(dpp) 3; R = CH2 (q) CH3 (t) for Hema and La(ema) 3. b The coupling constants for the aromatic protons could not be calculated due to broad peaks instead of defined doublets seen for the proligands. c 8 of the - N C r 7 3 moiety of Hdpp, La(dpp) 3 151 A larger upfield chemical shift was observed for Hb (AS -0.10 - 0.17 ppm) than for Ha (AS -0.07-0.11 ppm) on the La(III) pyrone ligands; whereas a larger chemical shift for Ha (AS -0.13 - 0.07 ppm) compared to Hb (AS - 0.02-0.04 ppm) was seen in the ' H N M R spectra of La(III) pyridinone ligands. This difference in the shifts of the La(III) pyrone and pyridinone ligands can be attributed to the increased electron density in the pyridinone ligands donated from the ring N atom. The nitrogen atom stabilizes the pyridinone complex by adding electron density to the ring current of the ligand from its lone pair of electrons. The chemical shifts of the methyl and ethyl groups on the ligands remain almost unchanged. In addition, there is only one set of broad peaks assigned to the ring hydrogens on the ligands, suggesting that their chemical shifts are either averaged values from the rapidly interconverting optical (A and A) and geometric (fac and mer) isomers at R T , or due to the presence of water molecules in the coordination sphere.5 9 1 1 1 1 1 1 1 1 * 1 1 1 — 1 1 1 1 1 1 1 1 1 1 — 1 1 1 1 1 1 1 1 1 1 1 — M I — 1 1 1 1 1 1 1 1 1 1 1 — 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 J.S 1.0 (ppm) Figure 3 .2 . ! H N M R spectra (300 M H z , R T , d6 - D M S O , * solvent) of Hma (top) and La(ma) 3. 152 13 In the C N M R spectra of La(ma)3 and La(ema)3, the most significant changes in chemical shifts were observed for the C=0 (AS ~ 7.54 - 8.10 ppm) and C - O - L a (A5 ~ 4.92 -14.87 ppm), indicative of La(III) ions being bound to the ligand through the (9,0 moiety. Both these signals were shifted downfield due to donation of electron density from the binding moieties to the metal ion. Peaks representing the other carbons in the ring, C I , C4, and C5, shifted upfield, with the biggest change in the chemical shift seen for the C4 and C5 carbons (A8 13 ~ 2.08 - 9.28 ppm). The C N M R signals from the methyl group of La(ma)3 or from the ethyl group of La(ema)3 did not shift significantly (A8 ~ 0.26 - 0.55 ppm). The broadening of the carbon signals also suggested the presence of isomers at RT. Table 3.4. 1 3 C N M R data for the La(III) pyrone complexes (75 M H z , R T , d6 - D M S O ) . 1 3 C chemical shifts are listed in ppm. 7 Compound CI C2 C3 C4 C5 C 6 a CI Hma 149.97 149.23 172.56 113.56 154.62 13.94 La(ma)3 149.27 154.15 180.10 111.16 152.54 14.30 Hema 154.81 153.39 172.90 113.52 142.39 21.09 10.88 La(ema) 3 153.46 168.26 181.00 110.86 152.07 20.54 11.12 d C6 is CH3 group for Hma and La(ma)3 The unique electronic properties of lanthanide cations, particularly Tb(III) and Eu(III), (long luminescent lifetimes, sharp emission bands, and large energy gap between absorbance and 153 emission band of complexes) render these metal centers attractive for luminescence studies. ' Unfortunately, the extinction coefficients for lanthanide ions tend to be small. This can be easily corrected, however, by utilizing sensitized luminescence with radiationless energy transfer from an aryl chromophore to the metal i o n . 6 7 ' 5 8 By using this 'antenna effect', the Ln(III) ion luminesces with increased intensity in the visible region. Eu(nia) 3 Tb(ma) 3 Eu(ema) 3 Tb(ema) 3 Eu(dpp) 3 Tb(dpp) 3 Eu(mpp) 3 Tb(mpp) 3 500 550 600 650 700 750 800 wavelength (nm) Figure 3 .3 . Fluorescence spectra for the Eu(III) and Tb(III) complexes. The pyrone and pyridinone ligands can photosensitize Ln(III) luminescence due to their U V absorption and appreciable molar absorptivity coefficients. 6 2 Luminescent excited states of the Tb(III) and Eu(III) complexes were generated via excitation of the pyrone and pyridinone moieties using U V wavelengths ranging from 315-324 nm. Energy was transferred to the 5 D 4 or 5Do excited states of the Tb(III) or Eu(III), respectively (Figure 3.3). Their fluorescence was red-shifted ca. 300 nm from the maximum absorbance through possible emission transition states 7 F , 154 j = 6-0 with the most intense peaks resulting from 5D 4—> 7F 5 and 5Do—>7F2 transitions, respectively. 6 9 ' 7 0 Most of the other transition states were not observed as they are weak or outside the fluorimeter detection wavelengths (Table 3.5). The emission bands of the complexes were also red-shifted slightly from the spectrum of the E u ( N 0 3 ) 3 and T b ( N 0 3 ) 3 starting materials, reflecting the substitution of original water and nitrate ligands with the coordinating O atoms of the pyrones and pyridinones. Ligand emission was observed, suggesting incomplete ligand-to-lanthanide energy transfer. The intensity ratio for the hypersensitive 5Do—>7F2 transition and the magnetic dipole transition 5Do—>7F] is approximately 25, indicating the absence of imposed symmetry on the complex observed in the N M R data (most likely the ligand is fluxional around the metal ion). Complexes with 7F2/ 7F] intensity ratio lower than 0.7 have centrosymmetric coordination spheres, whereas an intensity ratio higher than 8 is indicative of low symmetry environments. 7 1 ' 7 2 Table 3.5. Fluorescence data for the Eu(III) and Tb(III) complexes (k in nm). The strongest emission for the Eu(III) and Tb(III) complexes are from 5 D o - * 7 F 2 and 5D 4—> 7F 5 i respectively. A l l other emission peaks seen in the fluorescence spectra are listed. Complex A, e Xc 5 Do^ 7 F2 5Do->7Fo 5 D 0 - 7 F , Complex 5 D 4 ^ 7 F 5 5 D 4 ^ 7 F 6 5 D 4 - * 7 F 4 Eu(ma) 3 319 642 618 704 Tb(ma) 3 315 641 545 705 Eu(ema) 3 320 639 617 703 Tb(ema) 3 317 637 545 701 Eu(mpp) 3 322 647 713 Tb(mpp) 3 324 651 721 Eu(dpp) 3 294 616 682 Tb(dpp) 3 309 620 684 155 3.3.3 Cellular toxicity and uptake studies One of the limitations of Ln(III) ions is their inability to cross cell membranes, particularly intestinal barriers, as exemplified by the F D A approved L a 2 ( C 0 3 ) 3 . 1 2 ' 1 5 Masking the Ln(III) ion with ligands designed to form small, lipophilic, low molecular weight compounds, is expected to increase the bioavailibity of Ln(III) ions, allowing lower doses to be administered resulting in minimal unwanted side effects. A Caco-2 cell assay was used as a model for the absorption and secretion of molecules crossing the intestinal membrane. Caco-2 cells, originally derived from a human colorectal carcinoma line, differentiate at confluency into highly functional epithelial barriers with morphological and biochemical similarities to the small-intestinal columnar epi thel ium. 4 9 " 5 1 ' 6 0 The confluent monolayers exhibit brush border membranes containing hydrolytic enzymes, basolateral membrane ATPases and homone receptors. These cells have been used to evaluate the relative bioavailability and transport mechanisms of diverse drugs in vitro.13'77 When grown as a monolayer on a semipermeable membrane, the passage of drugs across the Caco-2 cell barrier can be monitored by analyzing the concentration of the desired compounds transported to either the basolateral (B, bottom) side through the mimicked intestinal environment (i.e., absorption), or to the apical (A, top) side through secretory transport. In addition, the concentrations of compounds present inside the cells can be measured by digesting the cell membrane to release the cytosol contents. Inductively coupled plasma mass spectrometry, I C P - M S , was used to measure the bifunctional transport of the L n L 3 across the Caco-2 cell barrier, by quantifying the metal ion concentrations of the apical and basolateral fractions, as well as in the Caco-2 cells. Regarded as 156 one of the best analytical tools for the quantitative determination of lanthanides, ICP-MS can 78 detect Ln(III) ion concentrations < 1 ppt. Unfortunately, this study does not allow us to determine the composition of the compound being transported in the biological media, or in the cell; however, one can qualitatively determine i f complexation increases the. bioavailability of Ln(III) ions when presented to an intestinal barrier, compared to the benchmark compound, L a 2 ( C 0 3 ) 3 . The B C A assay, a colorimetric assay used to determine quantitatively the total amount of protein in solution, was also used in conjunction with the cell uptake and cell transport studies. Upon addition of cupric sulfate, the C u 2 + ions are reduced to C u 1 + by peptide bonds in the proteins. Two bidentate bicinchoninic acid ligands act to chelate the reduced cupric sulfate in solution forming a purple-colored product that strongly absorbs light. B y comparing the absorbance of the samples with protein solutions of known concentrations, the amount of protein can be quantified. This assay allowed for the standardization of the cell uptake and cell transport studies by dividing the amount of Ln(III) absorbed or transported by the amount of protein in the cells, correlating to the concentration of Caco-2 cells in the sample. M T T (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) and M T S (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium bromide) assays were used to evaluate the cytotoxicity of Ln(ma) 3 , Ln(ema) 3, and Ln(dpp) 3 complexes using Caco-2 cells; only M T T assays were completed on the Ln(mpp) 3 species. These colorimetric assays, commonly used for the initial stages of drug screening, determine cell viability in vitro by measuring the integrity of the mitochondrial membrane after the addition of selected test compounds. The N A D P H or N A D H produced by mitochondrial dehydrogenase enzymes reduces the tetrazolium salts to colored formazan crystals that can be measured / 157 spectrophotometrically. The quantity of formazan crystals (measured at 490 nm for the M T S assay and 570 nm in the case of the M T T assay) is directly proportional to the number of viable cells in the culture. A low IC50 implies cytotoxity or nonproliferation of the cells at low drug concentrations. The main difference between the two assays stems from the tetrazolium salts used. The M T T assay forms insoluble formazan crystals upon reduction and therefore requires an extra step of dissolving the formazan product in D M S O before being analyzed. Each compound was tested at least 3 times for the M T S assay and at least 6 times for the M T T assay. The incubation periods for the cells plus the tested complexes were 24 h in the case of the M T S assay and 72 h for the M T T assay. Despite incubating the Caco-2 cell line with the metal complexes for a longer time period, the IC50 values for the M T T assay were all slightly higher (Table 3.6); indicating lower toxicity of test compounds. These variations between the M T S and M T T assay could be attributed to several factors including use of different media, different cell passage numbers, and different incubation times. Overall, the metal complexes were considerably more toxic than the proligands, (> 3 times more toxic), but less toxic than cisplatin, a known chemotherapeutic of IC50 = 8.67 + 0.26 u M in Caco-2 cel ls . 7 9 A trend of the toxicity of the metal complex is difficult to ascertain since all the IC50 values were in the same range, independent of the particular bidentate ligand with the lowest I C 5 0 for the M T T assay being 70.5 + 4.4 u M for Yb(dpp)3 and the highest being 161.7 + 2.1 u M for La(dpp) 3. 158 Table 3.6. Cytotoxicity data (by M T T and M T S assay) for the L n L 3 complexes, n > 3 (SD). IC 5o (uM) Complex M T T M T S Hma >800 543.0 (1.4) La(ma) 3 155.8 (8.6) 55.2 (6.3) Gd(ma) 3 118.4 (9.2) 78.1 (8.3) Yb(ma) 3 107.9(4.5) 68 (13) Hema >714 646 (21) La(ema) 3 139(18) 83 (14) Gd(ema) 3 107.8(4.4) 82(10) Yb(ema) 3 144 (25) 146(26) Hmpp 451.7(4.6) 841 (25) La(mpp) 3 142.3 (6.2) 76(18) Gd(mpp) 3 136.1 (1.2) 107(15) Yb(mpp) 3 136.5 (4.7) 53.1 (12.3) Hdpp 719 — La(dpp) 3 161.7(2.1) ~ Gd(dpp) 3 121(16) ~ Yb(dpp) 3 70.5 (4.4) — 159 Another type of assay measuring the cytotoxity of the compounds was performed on the same samples used for the M T S assay. The L D H assay quantitatively measures the cell toxicity, whereas M T T and M T S assays measure the cell viability. The lactate dehydrogenase (LDH) is released by damaged/nonviable cells, and converts the tetrazolium salt into a red formazan product quantified by its characteristic absorbance at 492 nm. Ideally, the cytotoxity values measured from the L D H assay and the M T S assay should be comparable; however, the IC 5o values from the L D H assay for all complexes tested, including both the proligand and corresponding metal complexes, were above 500 p M . Theses values, specifically for the metal complexes are significantly greater than the IC50 observed with the M T S and M T T assays. Research has shown that the L D H activity can be inhibited by Ln(III) ions with concentration > 5 u M . This may explain the minimal formation of formazan in the assay leading to unreliable IC50 OA Q I values. u ' 0 1 Thus, only the M T T and M T S I C 5 0 values are used to assess the toxicity of the Ln(III) complexes. A l l compounds, including La2(C03)3, exhibited similar cell uptake independent of the particular ligand (0.26 - 0.65 % uptake), with the exception of La(dpp)3 that showed cell uptake of 9.1 + 2.3 % (P < 0.001, Figure 3.4). Hdpp is the most lipophilic of the ligands used and this lipophilicity may increase cell passage, an assertion supported by the work herein. Masking of the Ln(III) metal with the dpp" ligand resulted in a considerable increase in cellular uptake of elemental Ln(III) compared to that of La2(C03)3. 160 o • — 12% i 10% 8% i 6% S 4% 2% 0% r ~ ~ i La(ma)3 La(ema)3 La(dpp)3 La2(C03)3 Gd(ma)3 Yb(ma)3 L n L 3 complexes Figure 3.4. Cel l uptake of L11L3 in Caco-2 cells determined by I C P - M S . Values are presented as a mean percentage of Ln(III) added/g protein/well + SD, n = 6. La(dpp)3 exhibits significantly higher cell uptake (P < 0.001 between La(dpp)3 and the other LnL3 complexes). N o reduction in the transepithelial electrical resistance (TEER) was observed during the bifunctional uptake studies indicating that the integrity of the monolayer of the Caco-2 cells was not disturbed at the concentrations of LnL3 and time points tested. Slight differences in apparent permeability ( P a p p , xlO" cm s"1) coefficients from A-to-B vs. B- to-A transport were found between compounds (Figure 3.5). Yb(ma)3 exhibited a lower P a p p value for both A-to-B and B -to-A compared to the other LnL3 species. In contrast, La(dpp)3, presumably the most stable of the complexes due to the pyridinones being stronger chelators, showed a higher A-to-B uptake; however, both the A- to-B and B-to-A P a p p value for La(dpp)3 was not significantly different than the values for the other L a L 3 complexes tested. The apparent permeability of the free proligand was previously assayed by other researchers; when chelated to the Ln(III), the P a p p (xlO" 6 cm s"1) 161 values of ligand/metal complexes were significantly lowered compared to the free proligand. For example, Hddp demonstrated P a p p values of 22.3 ± 1.2xl0" 6 cm s _ 1and 18.5 + 0.8x10"6 cm s"1 for the A-to-B and B- to-A transport, respectively 8 2 whereas La(dpp)3 exhibited P a p p values of 3.5 + 2.8 x lO" 6 cm s"1 and 1.82 + 0.73xl0" 6 cm s"1 for the A-to-B and B- to-A transport, respectively. In any case, the L n L 3 species tested exhibit similar partition coefficients to other small, bioavailable drugs such as dopamine, a sympathetic nervous system agent (9.33 + 3.5xl0" 6 cm s" '), mannitol, used to reduce intracranial pressure (0.38 + 0.12xl0" 6 cm s"1), acebutalol hydrochloride, a cardiovascular agent (0.51 + 0.02xl0" 6 cm s"1), and Zidovudine, an antiretroviral drug for the treatment of H I V (6.93 ± 0.17xl0" 6 cm s"1), all reported by other investigators. 7 4 In addition, the L n L 3 complexes have significantly greater bioavailability (~ 1.6 % for La(dpp) 3) than does the benchmark L a 2 ( C 0 3 ) 3 (0.0007 %), based on prelimary results. 10 7 La(ma)3 La(ema)3 La(dpp)3 Yb(ma) 3 Complexes Figure 3.5. Apparent permeability (P a p p ) of La(ma) 3, La(ema) 3, La(dpp) 3 and Yb(ma) 3 , across a Caco-2 membrane in vitro. Both data sets, basolateral to apical (B —» A ) and apical to basolateral (A —» B) transport, were measured by I C P - M S analysis of respective solutions after 4 h of exposure to L n L 3 , + SD, n = 3. 162 3.3.4 In vitro hydroxyapatite binding studies La(ma)3, Tb(ma) 3, Gd(ma) 3 , Eu(ma) 3, Yb(ma) 3 , La(ema) 3, La(dpp) 3 , and L a 2 ( C 0 3 ) 3 tested at concentrations 1 u M , well below their respective IC50 values, showed high H A binding with a quantitative uptake of > 98 % of the Ln(III) onto H A after 24 h. Min imal variability was observed between lanthanides (Figure 3.7). In addition, > 95 % Ln(III) ions were adsorbed onto the H A in 5 min or less. La(dpp) 3, Gd(ma) 3 and Yb(ma) 3 had slightly slower binding kinetics leading to a slower uptake of La(dpp) 3 (95.8 + 1.7 % at 3 h) compared to the other L a L 3 complexes tested (99.1 + 0.4 % for La(ma) 3 and 99.81 ± 0.01 % for La(ema) 3 at 3 h). In addition, La(dpp) 3, Gd(ma) 3 and Yb(ma) 3 may be more thermodynamically stable compared to the other Ln(III)-pyrone and -pyridinone complexes tested. Yb(III) and Gd(III) ions are stronger Lewis acids than is the slightly larger La(III) ion. Thus, these ions bind more tightly to the hard O, O donors of the pyrone and pyridinone ligands, further competing with the H A binding moieties. Furthermore, pyridinones are known to be stronger metal binders than are their pyrone analogs due to the ring N donating more electron density than its electronegative O counterpart resulting in higher stability constants. 2 5 , 8 3 The maximum concentration of La(III) observed in bone of dialysis patients after 4 y of treatment with the benchmark compound L a 2 ( C 0 3 ) 3 (known to accumulate on bone in vivo with maximum doses of 3000 mg/day of elemental lanthanum) is 10 ug/g wet we igh t . 1 6 ' 1 9 ' 2 2 Considering the average bone calcium concentration of 120 mg/g, lanthanum/calcium ratio would be 2 x 10"5—i.e. lanthanum ions would replace one out of every 50,000 calcium ions. 1 4 In our studies, the maximum concentration of Ln(III) determined on H A was ~5 mg/g wet weight, 500 times more than observed in vivo; however, the concentrations of elemental lanthanum 163 tested were considerably higher at -70 pg/ml compared to the 0.6 ng/ml of La(III) determined to be bioavailable. Nonetheless, the Ln(III)-pyrone and -pyridinone complexes tested showed > 98 % binding to H A , the same as was observed for L a 2 (C03 ) 3 . Table 3.7. Percent of Ln(III) bound to H A after the indicated time period, n >_6 (SD). % bound to H A Complex 5 min 15 min 3 h 24 h La(ma) 3 99.07 (0.43) 99.05 (0.39) 99.08 (0.64) 99.47 (0.66) La(ema)3 99.76 (0.10) 99.81 (0.01) 99.69 (0.12) 99.42 (0.53) La(dpp) 3 95.8 (1.7) 97.83 (0.36) 96.44 (0.68) 98.5 (1.5) Gd(ma) 3 95.95 (0.55) 99.28 (0.54) 99.31 (0.50) 99.45 (0.33) Yb(ma) 3 98.12(0.19) 98.19(0.34) 97.64 (0.60) 99.21 (0.32) L a 2 ( C 0 3 ) 3 99.36 (0.51) 99.28 (0.77) 99.47 (0.41) 99.64 (0.45) The fluorescence of Tb(ma)3 and Eu(ma)3 was also used to determine quantitatively the amount of the Ln(III) ions unbound to H A . It was assumed that any Ln(III) in the samples was chelated by the pyrone ligands due to the high concentration of proligand in solution. Thus, the fluorescence of the supernatant was measured at the A , m a x of the Tb(ma)3 and Eu(ma)3 complexes, and compared to a pre-constructed calibration curve of relevant LnL3 concentrations. The fluorescence results confirmed the I C P - M S data showing > 98 % L n - H A binding after 24 h (P > 0.05 compared to the L n L 3 species analyzed by I C P - M S , Figure 3.6). 164 • 5 min • 15 min • 3 h r s • 24hrs La(ma)3 La(ema):. La(dpp); La 2(CCb); Gd(nia).; Yb(mab Eu(ma)i Tb(ma)3 L n L ? complexes Figure 3.6. Mean percentage + SD of Ln(III) bound to H A samples, n = 6. Tb(III) and Eu(III) concentrations were determined by measuring the fluorescence of the Tb(ma)3 or Eu(ma)3 species in the supernatant, respectively, and subtracting the amount found from the total amount of LnL3 added. A l l other Ln(III) were determined by I C P - M S analysis (no significant difference between values were found, P > 0.05 at 24 h). A colorimetric dye, xylenol orange (sodium salt) indicative of free Ln(III) ions, was also used to confirm the I C P - M S results. 8 4" 8 6 In acidic buffered solutions, xylenol orange reddens in the presence of free Ln(III) ions. In contrast, appearance of a yellow/orange colour indicates the absence of unbound Ln(III) ions. Solutions of proligand and H A were used as negative controls; solutions of Ln(N03)3 served as the positive controls. As expected, the proligands, H A and the LnL3 solutions turned weakly yellowish/orange suggesting the absence of unbound Ln(III) in solution.; however, the Ln(N03)3 solution turned a dark red/purple. Solid samples from the metal-HA binding studies as well as the supernatant from the same experiment were tested for the presence of free Ln(III) ions. The addition of acid was presumed to digest the hydroxyapatite of the solid sample thereby releasing any bound Ln(III) ions into solution. In concordance with the I C P - M S data, the supernatant turned a slightly darker 165 orange/yellow than did the controls, suggesting a low concentration of free Ln(III). The digested L n - H A samples turned dark red, indicating the presence of Ln(III). These results confirm the ICP-MS data indicating a large proportion of the Ln(III) ions are present on the H A matrix. To further confirm these results, xylenol orange was added to an undigested L n - H A sample in the buffer solution. The solution remained yellow/orange, but the particles turned a dark red/purple suggesting the presence of Ln(III) ions on the surface of the H A solid samples. Spectrophotometric determination (UV-vis assay) of ligand concentration in supernatants from La(ma) 3, Eu(ma)3, Gd(ma) 3 , Tb(ma)3, Yb(ma) 3 , La(ema)3, and La(dpp) 3 -HA binding studies indicated that at the 24 h time point, > 97 % of the ligand remained in solution independent of the metal complex tested (Figure 3.7). High concentrations of the ligand in solution indicated dissociation of the LnL3 complex suggesting that the bidentate ligands on the Ln(III) ions were almost completely replaced by the phosphates and carbonates of the H A matrix. 100% i n 80% l . I J • 5 nan • 15 min • 3hr • 24 hr La(ma)i La(ema)5 La(dpp); Gd(ma)5 Yb(ma)3 Eu(ma); Tb(ma)! Complexes Figure 3.7. Percent ligand concentration + SD remaining in the supernatant of the Ln(III)-HA binding studies determined by U V - v i s spectroscopy (no significant difference between samples were determined, P > 0.05 at 24 h), n = 6. 166 IR spectra of the solid H A samples exposed to the Ln(III) complexes were compared to spectra of pure H A , the proligand, and the HA-bound L n L 3 complexes (Figure 3.8). Regardless of the metal ion, the complex or the exposure time of the L n L 3 species to the solid H A , the IR spectra were nearly superimposable upon that of pure H A suggesting that the H A composition had not been modified upon complexation of the Ln(III) salt. 8 7" 9 0 Although not prominent, the hydroxyl stretching (-OH) and bending bands were observed at 3569 and 630 cm"1 for the H A -L n samples. Broad bands at 3450 and 1640 cm"1 were attributed to adsorbed water. The v 3 and vi PO4 3 " vibrations were assigned to broad bands at 1043 and 969 cm"1 respectively. Bands at 602 and 567 cm"1 were from the v 4 vibrations. The band observed at 877 cm"1 corresponds to 2 1 HPO4 " vibrations. The v 3 bands of carbonate ions observed at 1460 and 1425 cm" are generally representative of substitution of phosphate groups for carbonate species. A new peak observed at -474 cm' 1 corresponds vm-o vibrations showing that the Ln(III) is bound to the H A , coordinating to either the phosphate or the carbonate oxygens of the hydroxyapatite. In addition, vibrations from the proligand in the L n - H A samples were not observed, consistently suggesting that the Ln(III) dissociates from the pyrone/pyridinone ligands upon complexation to the H A . The lack of IR shifts arising from the pure H A vs. the Ln(III)-HA is unsurprising due to the similarities of the Ca(II) and Ln(III) ions. Despite the slightly larger unit-cell volume found for Ln(III)-HA compared to Ca(II)-HA, Ln(III) are essentially functional mimics of Ca(II) ions sharing similar ionic radii, donor atom preferences, and almost identical coordination numbers in protein binding sites. 9 ' 9 1 167 3700 3400 3100 2800 2500 2200 1900 1601 1300 1000 700 400 Wavenumbers (cm"1) Figure 3.8. IR spectra from H A binding studies. 3.4 Conclusions Twenty L n L 3 were successfully synthesized using Hma, Hema, Hmpp or Hdpp as proligands. La(III), Gd(III), and Yb(III) complexes were chosen for further testing in order to evaluate size effects and assess their medicinal applications; Tb(III) and Eu(III) complexes were of interest due to their characteristic fluorescence properties. Toxicity, bifunctional transport across a model of an intestinal barrier, and binding coefficients to H A were evaluated for several of the L n L 3 species to assess their potential therapeutic use for bone resorption disorders. B y comparison with Fosrenol™ an FDA-approved lanthanum-containing phosphate binder, the investigated L n L 3 compounds showed similar H A binding (> 98 %) and had greater absorption across an intestinal barrier, compared to lanthanum carbonate. The L n L 3 compounds were 168 significantly less toxic than cisplatin and did not appear to disturb the H A structure upon binding. These novel L n L 3 complexes thus show potential as orally available bone resorption inhibition drugs, having minimal side effects that may also serve as potential preventative supplements for osteoporosis. La(dpp) 3 was identified as the lead compound of this work with high H A binding (98.5 + 1.5 %) and a moderate P a p p value. This compound wi l l be further tested in appropriate animal models of bone disorders for both bioavailability and anti-resorptive activity. 169 3.5 References 1. Government of Canada; http://www.biobasics.gc.ca/english/View.asp?x=770. (accessed Jan. 11,2007). 2. Mauck, K . F.; Clarke, B . L . Mayo Clin. Proc. 2006, 81, 662. 3. Dempster, D. W. ; Lindsay, R. Lancet 1993, 341, 797. 4. McDermott, M . T.; Zapalowski, C ; Mil ler , P. D. , Osteoporosis. Hanley & Belfus: Toronto, O N , 2004. 5. Fordham, J. N . , Osteoporosis. Churchill Livingstone: Toronto, 2004. 6. Karsdal, M . A . ; Christiansen, C. Future Rheum. 2006,1, 683. 7. Marie, P. J. Bone 2006, 38, S10. 8. Verberckmoes, S. C ; Broe, M . E . D. ; D'Haese, P. C. Kidney Int. 2003, 64, 534. 9. Pidcock, E . ; Moore, G . R. J. Biol. Inorg. Chem. 2001, 6, 479. 10. Webster, T. J.; Massa-Schlueter, E . A . ; Smith, J. L . ; Slamovich, E . B . Biomaterials 2004, 25,2111. 11. Nigel , D. ; Joseph, W. ; Michael, D . World Patent WO 02/00227 A2 2002. 12. Fricker, S. P. Chem. Soc. Rev. 2006, 35, 524. 13. D'Haese, P. C ; Couttenye, M . M . ; Lamberts, L . V . Clin. Chem 1999, 45, 1548. 14. Behets, G . J.; Verberckmoes, G. J.; Oste, L.;Bervoets, A . R.; Salome, M . ; Cox, A . G. ; Denton, J.; de Broe, M . R.; D'Haese, P .C. Kidney Int. 2005, 67, 1830. 15. Damment, S. J. P.; G i l l , M . J. Am. Soc. Nephrol. 2003,14, F-P0653. 16. D'Haese, P. C ; Spasovski, G . B . ; Sikole, A . Kidney Int. Suppl. 2003, 85, S73. 17. Damment, S. J. P.; P. Greaves, P.; Downes, N . J. Am. Soc. Nephrol. 2003,14, F-P0651. . 18. Damment, S. J. P.; Beavers, C ; Gatehouse, D. G. Mutagensis 2005, 20, 29. 170 19. Spasovski, G . B . ; Sikole, A . S.; Gelev, J.; Masin-Spasovska, T.; Freemont, I.; Webster, M . ; G i l l , C.;Jones, M . E . ; de Broe,; P .C. D'Haese. Nephrol. Dial. Transplant 2Mb, 21, 2217. 20. Vanholder, R.; Cornells, R.; Dhondt, A . ; Lameire, N . N. Nephrol. Dial. Transplant 2002, 52,2 . 21. Cui , Y . ; Chen, G . ; Ren, J.; Quian, Y . ; Huang, J. Inorg. Chem. 2000, 39, 4165. 22. Behets, G . J.; Dams, G. ; Vercauteren, S. R.; Damment, S. J.; Bouil lon, R.; de Broe, M . E. ; D;Haese, P. C. J. Am. Soc. Nephrol. 2004, 15, 2219. 23. Bervoets, A . R. J.; Oste, L . ; Behets, G . J.; Dams, G . ; Blust, R.; Marynissen, R.; Geryl, H . ; de Broe, M . E . ; D'Haese, P. C. Bone 2006, 38, 803. 24. Canavese, C ; Mereu, C ; Nordio, M . ; Sabbioni, E . ; Aime , S. Curr. Med. Chem. 2005, 12, 1631. 25. Thompson, K . H . ; Barta, C. A . ; Orvig, C. Chem. Soc. Rev. 2006, 35, 545. 26. Santos, M . A . ; G i l , M ; Gano, L . ; Chaves, S. J. Biol. Inorg. Chem. 2005, 10, 564. 27. Marques, F.; Guerra, K . P.; Gano, L . ; Costa, J.; Campello, M . P.; Lima, L . M . P.; Delgado, R.; Santos, I. J. Biol. Inorg. Chem. 2004, 9, 859. 28. Parker, L . L . ; Gowans, N . D . ; Jones, S. W. ; Robins, D . J. Tetrahedron 2003, 59, 10165. 29. Parker, L . L . ; Gowans, N . D. ; Jones, S. W.; Robins, D . J. Tetrahedron 2003, 59, 10165. 30. Gordon, A . E . V . ; X u , J.; Raymond, K . N . ; Durbin, P. W . Chem. Rev. 2003,103, 4207. 31. Hider, R. C ; Zhou, T. Ann. N. Y. Acad. Sci. 2005,1054, 141. 32. L i u , Z . D . ; Hider, R. C. Med. Res. Rev. 2002, 22, 26. 33. Yokel , R. A . Coord. Chem. Rev. 2002, 228, 97. 34. Santos, M . A . Coord. Chem. Rev. 2002, 228, 187. 35. Titan Pharmaceuticals Inc.; http://www.titanpharm.com/pt-gallium.html. (accessed Jan 2006). 36. Clevette, D. J.; Orvig, C. Polyhedron 1990, 9, 151. 37. Zang, Z . ; Lyster, D . M . ; Webb, G . A . ; Orvig, C. Nucl. Med. Biol. 1992,19, 327. 171 38. Green, D . E . ; Ferreira, C. L . ; Stick, R. V . ; Patrick, B . O.; Adam, M . J.; Orvig, C. Bioconjugate Chem. 2005,16, 1597. 39. Barrand, M . A . ; Callingham, B . A . ; Hider, R. C. J. Pharm. Pharmacol. 1987, 39, 203. 40. Barrand, M . A . ; Hider, R. C ; Callingham, B . A . J. Pharm. Pharmacol. 1990, 42, 279. 41. Maxton, D. G . ; Thompson, R. P. H . ; Hider, R. C. J. Nutr. 1994, 71, 203. 42. Thompson, K . H . ; M c N e i l l , J. H . ; Orvig, C. Chem. Rev., 1999, 99, 2561. 43. Pierre, V . C ; Melchior, M . ; Doble, D . M . J.; Raymond, K . M . Inorg. Chem. 2004, 43, 8520. 44. Cavalieri, L . F. Chem. Rev. 1947, 4, 525. 45. Cole, E . ; Copley, R. C. B . ; Howard, J. A . K . ; Parker, D . ; Ferguson, G . ; Gallagher, J. F.; Kaitner, B . ; Harrison, A . ; Royle, L . J. Chem. Soc, Dalton Trans. 1994, 1619. 46. Dutt, N . K . ; Sarma, U . U . M . J. Inorg. Nucl. Chem. 1975, 57, 1801. 47. Ahmed, S.; Burgess, J.; Hurman, B . ; Parsons, S. A . Polyhedron 1994,13, 23. 48. Fregona, D. ; Faraglia, G . ; Graziani, R.; Casellato, U . ; Sitran, S. Gazz. Chim. Ital. 1994, 124, 153. 49. Fogh, J.; Fogh, J. M . ; Orfeo, T. J. Natl. Cancer Inst. 1977, 59, 221. 50. Garcia-Casal, M . N . ; Leets, I.; Layrisse, M . J. Nutr. 2000,130, 5. 51. Vanchon, J. F.; Beaulieu, P. Gastroenterology 1992,103, 414. 52. Mosmann, T. J. Immunol. Methods 1983, 65, 55. 53. Walker, J. M . , The Protein Protocols Handbook. 2nd ed.; Humana Press: New Jersey, 2002. 54. Nelson, W. O. Ph. D . Thesis. University of British Columbia, Vancouver, 1988. 55. Zhang, Z . ; Rettig, S. J.; Orvig, C. Inorg. Chem. 1991, 30, 509. 56. Monga, V . ; Patrick, B . O.; Orvig, C. Inorg. Chem. 2004, 44, 2666. 57. Finnegan, M . M . ; Lutz, T. G ; Nelson, W. O.; Smith, A . ; Orvig, C. Inorg. Chem. 1987, 2d, 2171. 172 58. Nelson, W . O.; Karpishin, T. B . ; Rettig, S. J.;Orvig, C. Can. J. Chem 1987, 66, 123. 59. Nelson, W . O.; Karpishin, T. B . ; Rettig, S. J.;Orvig, C. Inorg. Chem. 1988, 27, 1045. 60. Faraglia, G ; Fregona, D. ; Sitran, S. Main Group Met. Chem. 1994,17, 649. 61. Clevette, D. J.; Lyster, D . M . ; Nelson, W. O.; Rihela, T.; Webb, G . A. ;Orv ig , C. Inorg. Chem. 1990, 29, 667. 62. Tedeshi, C ; Azema, J.; Gornitzka, H . ; Tisnes, P.; Picard, C. Dalton Trans. 2003, 1738. 63. Alyea, E . C ; Malek, A . ; Vougioukas, A . E. Can. J. Chem 1981, 60, 667. 64. Brown, L . M . ; Mazdiuasni, K . S. Inorg. Chem. 1970, 9, 2783. 65. Sabbatini, N . ; Guardigli, M . Coor. Chem. Rev. 1993,123, 201. 66. Choppin, G . R.; Peterman, D . R. Coord. Chem. Rev. 1998,174, 283. 67. Weissman, S. I. J. Chem. Phys. 1942,10, 214. 68. Yang, X . - P . ; Kang, B. -S . ; Wong, W . - K . ; Su, C . - Y . ; L i u , H . -Q. Inorg. Chem. 2003, 42, 169. 69. K i m , W . D. ; Kiefer, G . E . ; Maton, F.; M c M i l l a n , K . ; Muller , R. N . ; Sherry, A . D. Inorg. Chem. 1995,34,2233. 70. Tei, L . ; Blake, A . J.; George, M . W.; Weinstein, J. A . ; Wilson, C. ; Schroder, M . Dalton Trans. 2003, 1693. 71. Kirby, A . F.; Foster, D. ; Richardson, F.S. Chem. Phys. Lett 1983, 95, 507. 72. Kirby, A . F.; Richardson, F.S. J. Phys. Chem. 1983, 87, 2544. 73. N g , A . W. K . ; Lukic , T.; Pritchard, H . P.; Wasan, K . M . Cardiovasc. Drug Rev. 2003, 21, 151. 74. Yazdanian, M . ; Glynn, S. L . ; Wright, J. L . ; Hawi , A . Pharmacol. Res. 1998,15, 1490. 75. Okhee, H . ; Failla, M . L . ; H i l l , D . ; Morris, E . R.; J. C. Smith, J. J. Nutr. 1994,124, 580. 76. Makhey, V . D . ; Guo, A . ; Norris, D . A . ; Hu , P.; Yan , J.; Sinko, P. J. Pharmacol. Res. 1998,15, 1160. 77. Risovic, V . ; Sachs-Barrable, K . ; Boyd, M . ; Wasan, K . M . Drug Dev. Ind. Pharm. 2004, 30, 767. 173 78. Ochsenkuhn-Petropoulou, M . ; Ochsenkuhn, K . ; Luck, J. Spectrochim. Acta Part B 1991, 46,51. 79. van Niekerk, E . ; O'Sullivan, J. F.; Joone, G. K . ; van Rensburg, C. E . J. Invest. New Drugs 2001,19, 211. 80. Xunqing, W. ; Zhenhua, X . ; Xiaoxia , G . Chinese Sci. Bull. 1998, 43, 1625. 81. Wei-guo, L . ; Zhen-wei, L . ; Da-qing, Z . ; Jia-zuan, N . Wuji Huaxue Xuebao 2001, 17, 843. 82. Huang, X . - P . ; Spino, M . ; Thiessen, J. J. Pharmacol. Res. 2006, 23, 280. 83. Santos, M . A . ; Gama, S.; Gano, L . ; Cantinho, G . ; Farkas, E . Coord. Chem. Rev. 2002, 228, 187. 84. Alves, F. C. ; Donato, P.; Sherry, A . D. ; Zaheer, A . ; Zhang, S.; Lubag, A . J. M . ; Merritt, M . E . ; Lenkinski, R. E . ; Frangioni, J. V . ; Neves, M . ; Prata, M . I. M . ; Santos, A . C.;de . Lima, J. J. P.;Geraldes, C. F. G . C. Invest. Radiol. 2003, 38, 750. 85. Y u , L . ; Chen, D . ; L i , J.; Wang, P. G . J. Org. Chem. 1997, 62, 3575. 86. Rudovsky, J.; Botta, M . ; Hermann, P.; Koridze, A . ; Aime, S. Dalton Trans. 2006, 2323. 87. Wakamura, M . ; Kandori, K . ; Ishikawa, T. Colloids Surf., A: Physiochem. Eng. Aspects 2000,164, 297. 88. Lebugle, A . ; Pelle, F.; Charvillat, C ; Rousselot, I.; Chane-Ching, J. Y . Chem. Commun. 2006, 606. 89. Choi , D . ; Kumta, P. J. Am. Ceram. Soc. 2006, 89, 444. 90. Chang, M . C ; Tanaka, J. Biomaterials 2002, 23, 4811. 91. Fleet, M . C ; L i u , X . J. Solid State Chem. 2000,149, 391. 174 C H A P T E R 4 Conclusions and Future Work 4.1 Magnetic Studies of rf^Heterometallic Complexes 4.1.1 Proposed ligands for d/f complexes A series of trimetallic d/f complexes utilizing H 3 bcn and H 3 tam, with various TM(II) and Ln(III) ions, were synthesized and characterized in this thesis (Chapter 2). The studies completed on the [LnNi 2 ( tam)2] + species confirmed the antiferromagnetic nature of the d/f species reported in earlier work. 2 In addition, the magnetic exchange observed in the [LnTM 2 (bcn )2] + species, where T M = Ni(II) and Cu(II), followed Kahn's principles 3 where Ln(III) metal ions with > If electrons exhibit ferromagnetic interactions and Ln(III) with < 7 /electrons display an antiferromagnetic exchange. Only minimal conclusions could be made, however, with regards to how the different geometric demands affected the magnetic interactions between the two TM(II) and the single Ln(III) metal ions due to the limited number of d/f species that have been reported. Thus, additional ligands that afford a variety of d/f complexes would be of interest to better understand the structural influences on the magnetic interactions between TM(II) and Ln(III) metal ions. Tacn (1,4,7-triazacyclononane), tren (tris(2-aminoethyl)amine) and tame (1,1,1-tris(aminomethyl)ethane) are polyamino frameworks used as backbones for many existing ligands, including the proligands H 3 bcn, H 3 t rn and H 3 tam. Furthermore, these three 175 frameworks can be easily functionalized with different end groups allowing for various coordination geometries around the d/f metals. For example, relevant ligands using the same design reported by Murphy et al.4 (Scheme 4.1) might be effective for coordinating d/f metal ions in a different conformation than previously reported in the literature. B y utilizing the Mannich reaction to attach para-substituted salicylaldehydes to the nitrogens of the tacn framework, Murphy et al.4 produced a N3C>6 multidentate ligand. Potentially, dimetallic T M -L n species could be produced where the TM(II) is chelated by the 3 nitrogens from tacn (possibly 4 nitrogens in the case of tren) and by the three phenol groups (possibly only 2 phenol groups for tren). In this case, the Ln(III) might be bridged by the three phenol groups and additionally ligated by the three aldehyde oxygen moieties. This chemistry could also be extended to the tame and tren backbones resulting in additional ^heterometal l ic complexes. Scheme 4.1. Mannich reaction of salicylaldehyde with tacn: (a) M e O H , R T , N 2 ; (b) toluene/N2, reflux. 4 In addition, several other Schiff-base and reduced Schiff base d/f complexes could be synthesized, thereby increasing the literature examples of d- and / -block heterometallic species (Figure 4.1). End groups such as 2,3-dihydroxysalicylaldehyde or 2-hydroxy-3-176 methyoxysalicylaldehyde could easily be attached to the tacn, tame or tren backbones, similar to a ligand design of Costes et al1 and Bismondo and co-workers. 5 B y coupling 2-hydroxy-3-methoxysalicyladehyde to tren, Costes et al. formed a 4/74/ complex utilizing the N4O6 cavity 1 (Figure 4.1, b) that could also be potentially used for (i/heterometallic systems. Figure 4.1. Potential ligands for d/f or f/f complexes utilizing the tren backbone. 4.1.2. Proposed ligands for f/f heteronuclear complexes Attempts at synthesizing the trinuclear Ln(III) species with the rTjtrn, F^tam and Ffjbcn ligands were not exhaustively investigated. These species are worth further consideration due to their interesting magnetic properties. For example, the Gd(III)-Gd(III)-Gd(III) species could potentially exhibit an extremely large magnetic moment i f all 21 electrons align in a parallel fashion (7 from each Gd(III) ion, Af). In addition, the heterometallic Ln(III) species may serve as interesting N M R shift reagents or useful M R I contrast agents. 177 Multidentate ligands containing amide bonds have potential to form additional flf heterometallic complexes. For instance, a multidentate, pre-organized ligand could be synthesized utilizing a procedure similar to Ryu and co-workers, 3 and Raymond and co-6 7 workers. ' B y addition of 3 equivalents of activated 2,3-dibenzyloxybenzoic acid to a triamine backbone, 8 a N4O9 chelating ligand was formed. The first lanthanide has the potential to be chelated by both the amide carbonyl oxygen and the o-hydroxyl group of the benzoic acid. The second lanthanide could be chelated by the catechol O atoms, where the 0-hydroxyl groups bridge the two Ln(III) atoms, potentially mediating any spin exchange. 4.2 Bone Density Agents 4.2.1 Metal-ligand stability studies The cell toxicity and the bioavailability of the 3-hydroxy-4-pyrone and 3-hydroxy-4-pyridinone Ln(III)-containing complexes discussed in Chapter 3 show promising results as orally available agents. La(dpp)3, which had the highest P a p p value, the largest cell uptake, and appropriate hydroxyapatite binding was identified as the lead compound for the continued development in the treatment and amelioration of bone density disorders. The next logical step is to predict the integrity of the complex after exposure to harsh gastric conditions, p H < 1.5. The step-wise stability constants for 1:1 and 1:2 metal to ligand species, as well as the overall stability constant for the 1:3 species can be determined by aqueous potentiometric titrations in the range of 3 < pH < l l . 9 If required, spectrophotometric titrations can also be utilized to determine appropriate constants at 178 extremely low or high p H values, where the p H at < 3 or > 11 approaches ionic strength. The dominant species at p H ~ 7 wi l l be resolved by calculating speciation diagrams. From these studies, one can determine i f the stability of the lead compound should be altered (by use of multidentate vs. bidentate chelators, different chelating groups, etc.) so that the complex remains intact in biological media, while still maintaing the ability to exchange Ln(III) with Ca(II) at the desired bone tissue. 4.2.2 Relevant animal models A few relevant animal models commonly used to test bone density agents, including normal, osteopenia, osteoporotic, and established osteoporotic rat models, should be used to determine the effectiveness of La(dpp) 3 as a bone density agent. The most common rat model, and arguably the most relevant for our studies, is the ovariectomized rat model ( O V X ) in the osteopenia category. This animal model exhibits similar characteristics to human postmenopausal osteoporosis, characterized by decreasing bone density (g of mineral per area or volume of bone) and bone quality (changes in architecture, turnover, microfractures, and mineralization) due to low hormone levels. 1 0 " 1 3 In addition, considerable literature precedence documenting the histomorphometric changes in bone and the effect of diet for this model is available. 1 4 ' 1 6 Preliminary discussions with our collaborator, Prof. Kishor Wasan, and his research assistant, Dr. S. J. Thornton, in the Faculty of Pharmaceutical Sciences at the University of British Columbia ( U B C ) , has led to the plan that ovariectomized Wessler rats at least 3 months old wi l l be used in these studies. The slightly aged rats allow for the effects of the 179 ovariectomy to manifest on their skeleton, while also addressing the concern that the rat model should be based on skeletally mature animals to more closely model human postmenopausal bone loss. The effects of the Ln(III) containing compound wi l l be assessed by introducing La(dpp)3 to the food at varying concentrations for each group of rats. Common non-invasive methods of monitoring bone density via dual-energy X-ray absorptiometry ( D X A ) w i l l be primarily used throughout the study. M R I may also be utilized to analyze the post-treatment effects. After a to be defined amount of time (between 3-6 months), the animals w i l l be sacrificed. The bone mineralization and the osteoclast vs. osteoblast ratio wi l l be determined, while their histology wi l l be compared to the untreated rats (no addition of La(dpp) 3 to the food of the untreated rats, although the diet remains constant for all rats). Toxicology of the La(dpp) 3 complex wi l l also be assessed by monitoring the food and water intake, as well as the renal and liver functions of the O V X rats. Unfortunately, the high costs of the animal model ($75 - 100/OVX rat) and the required extended care (up to 6 months) wi l l limit the number of preliminary studies conducted. 4.2.3 Proposed bifunctional ligands for bone density disorders Modifications of the ligand structure may be needed once the speciation diagram at a variety of p H values have been analyzed and the behavior of the complex in vivo has been determined. There is an upsurge of interest regarding the synthesis of new drugs that selectively accumulate in particular compartments of the body, thereby minimizing unwanted side effects. The design of bifunctionalized chelators (BFC) , in which a directing group 180 targeting a specific biomolecule or tissue is added to the chelating portion of the molecule is particularly attractive to our group. Considering the preferences of Ln(III) metal ions, a purposeful ligand design incorporating several hard oxygen donor atoms, large cavity size, and 5-6 membered chelate rings is ideal for the chelate. Tripodal polydentate amide phenol ligands are especially attractive due to the three-dimensional cavities imposed by the preorganized frameworks. The attachment of a bone targeting functional group to the ligand backbone could be completed before or after chelation of the Ln(III). Several different backbones such as tren, tame, 1 7 Kemp's triacid (cis,cis-l,3,5-trimethyl-l,3,5-tricarboxylic acid), and nitrilotriacetic acid coupled to either 4-aminoresorcinol or 2,4-dihydroxybenzoic acid could afford multidentate 6-coordinate ligands useful for chelating Ln(III) ions via multiple hard amide carbonyl and phenol donors (Figure 4.2). Furthermore, the />hydroxyl groups can be functionalized by adding various directing moieties that selectively accumulate on bone. HO HO C 0 2 H C 0 2 H a = H,N N H , - N H , N H , NH HN HO Kemp's triacid tame tren tacn nitrilotriacetic acid C K J D H H,N 4-aminoresorcinol 2,4-dihydroxybenzoic acid Figure 4.2. Potential backbones (a) and functional groups (b) for bone-seeking B F C . 181 Several target moieties could be investigated to increase the selectivity of the complexes in vivo. Research groups have shown that ligands containing phosphate arms tend to localize on the surface of hydroxyapatite and are retained in bone (Figure 4.3). 1 8 Research has also shown that proximal nitrogen atoms increase the hydroxyapatite binding and skeletal uptake of phosphate-containing compounds. 1 8 Bisphosphonates containing a P-C-P backbone, where two phosphonic acid groups are bonded to the same carbon, are known to be osteoclast-mediated bone resorption inhibitors. 1 9" 2 1 In addition, glutamic acid and aspartic acid peptides are well-documented bone-targeting compounds with a stong affinity for bone (Figure 4 .3 ) . 2 2 ' 2 3 O H 3 C II O V P ° 3 H 2 A / \ A X ,o OH U 3 2 N H 2 N H 2 OH phosphate H E D P glutamic acid aspartic acid Figure 4 .3 . Known bone seeking agents. Thus, by combining one of the aforementioned directing groups with an appropriate chelator, a B F C could be developed that is both an attractive ligand for Ln(III) and can selectively accumulate on bone. A few preliminary reactions have been completed using polyamino or polycarboxylic acid backbones, including tren, tame, nitrilotriacetic acid, Kemp's triacid, with a variety of pendant groups. The synthesis and findings wi l l be briefly discussed herein. 182 4.3 Pre l iminary Exper imental for Relevent Bone-seeking B F C 4.3.1 Materials Reagents were purchased from commercial sources (Aldrich, Fisher, A l f a Aesar) and used without purification unless otherwise stated. The Dowex 550A OH" ion-exchange resin was purchased from Sigma-Aldrich and the dibenzylphosphoryl chloride in benzene 10 % w/v was purchased from Toronto Research Chemicals (North York, ON) . N M R solvents were obtained from Sigma-Aldrich and used as received. Methanol, ethanol, acetonitrile and benzene were dried over activated 4 A molecular sieves prior to use. Anhydrous D M F was purchased from Aldr ich and T H F was dried over Na. Water was deionized (Barnstead D8902 and D8904 Cartridges) and distilled (Hytrex IIGX50-9-7/8 and GX100-9-7/8) before use. 4.3.2 Instrumentation *H N M R spectra were recorded on a Bruker Avance 300 spectrometer (300 M H z ) or a Bruker Avance 400 inverse spectrometer (400 M H z ) at room temperature (RT) and calibrated with deuterated solvents. Elemental analyses of C, H , and N were preformed by Mr . M . Lakha at the University of British Columbia. Mass spectra were obtained on a Bruker Esquire Ion Trap (electrospray ionization mass spectrometry, ESI -MS) spectrophotometer. Approximately 1 mg of sample was dissolved in 1 m L of methanol or acetonitrile for +ESI-MS analysis. 183 4.3.3 Preliminary Synthesis 2,4-Dibenzyloxybenzoic acid (1). 2,4-Dihydroxybenzoic acid (0.5 g, 2.6 mmol), benzyl H O ^ O ^ bromide (3.34 g, 19.5 mmol) and K 2 C 0 3 « 1 . 5 H 2 0 (8.06 g, 48.8 mmol) in ' \ ^ Y acetone were refluxed under N 2 overnight. Excess salt was filtered and the solvent was removed in vacuo to afford a pale yellow oi l . The benzyl attached to the carboxylic acid was hydrolyzed by dissolving the resulting oil in 20 m L of M e O H and 30 m L aqueous solution of 5 M N a O H . The mixture was refluxed for 3 h and the solvent was removed under reduced pressure. The aqueous layer was extracted with 2 x 1 5 m L hexanes and acidified with concentrated HC1 until p H 2. The product was re-extracted into chloroform ( 3 x 1 5 mL) and formed a white solid upon removal of the solvent to yield 0.60 g, 54 %. ] H N M R (300 M H z , R T , C D 3 O D ) : 8 = 7.74 (d, I H , H3, 3 J = 8.7 Hz), 7.41 (m, 2H, HBn v , / 2 = 3 Hz), 7.28 (m, 8H, HBn, vm = 19.5 hz)^ 6.66 (d, I H , H\,5J= 2.4 Hz), 6.58 (dd, I H , HI,3J= 8.7 Hz, 2 J = 2.4 Hz), 5.13 (s, 2H , H5), 5.02 (s, 2H, HA). M S (+ESI): m/z 357 [L + N a f . 7>/5(2,4-dimethoxybenzamide)-c/5-l,3,5-methylcyclohexane (2). TV-hydroxysuccinamide 1 C H NH 3 (131 mg, 1.14 mmol) and 7Y,7V-dicyclohexylcarbodiimide H 3 C A - T — V - C H 3 D _ f| 1 c o ° c o ~ 6 * m m ° l ) w e r e added under N 2 to Kemp's R R 0 M E 5 t r iacid(100mg, 0.387 nmol) in dry D M F (30 mL) and reluxed for 20 h. The solution immediately turned a dark red/brown upon addition of 4-methoxyresorcinol (131 mg, 1.14 mmol). After refluxing for 18 h under N 2 , the solvent was removed in vacuo. The resulting brown residue was extracted in C H 2 C 1 2 ( 3 x 1 5 mL), dried 184 over Na2S04 and filtered. The product was chromatographed on silica gel using a gradient of 5 % M e O H / C H 2 C l 2 up to 50 % M e O H / C H 2 C l 2 , yielding 23 mg, 10 %. ' H N M R (300 M H z , RT, C D 3 O D ) : 5 = 7.94 (s, br, 3H, 777, v 1 / 2 = 12 Hz), 6.81 (s, br, 3H, 774, v 1 / 2 = 10.5 Hz), 6.77 (s, br, 3H, H6, v 1 / 2 = 19.5 Hz), 3.01 (d, 3H, 771 e q , 2 7 = 2.1 Hz) , 2.97 (s, 9H, HI), 2.83 (s, 9H, H5), 2.79 (s, 9H, H3), 2.57 (d, 3H, 771 a x, 2J = 2.4 Hz). M S (+ESI): m/z 590 [L + Na] + . rm(2,4-dimethoxybenzylamide)-/m(ethyl)amine (3). D C C (2.99 g , 13.4 mmol) and/V-10 hydroxysuccinimide (1.38 g, 12 mmol) was added to 2,4-°\\ 1 1 2 /34\_/7~OMe dimethoxybenzoic acid (2.19 g, 12 mmol) under N 2 in 125 m L of / - N H ^ — J N~~\ / — \ v }~0Me dry T H F . The mixture was refluxed for 3 h, after which time the ^ N H MeO V=o solution was allowed to cool to R T and was filtered to remove the <f y— OMe y=/ D C U by-product. Tren (0.584 g, 4 mmol) was added dropwise to MeO the solution and the mixture was refluxed overnight. Excess T H F was removed under reduced pressure affording a light yellow oil that was chromatographed on a silica column using a gradient of 5 % M e O H / C H 2 C l 2 up to 20 % M e O H / C H 2 C l 2 , yielding 1.21 g, 47 %. ' H N M R (300 M H z , R T , CD3OD): 5 = 7.78 (d, 3H, H9,3J= 8.8 Hz) , 6.45 (dd, 3H, 7/8, V= 8.8 Hz, 2J= 2.3 Hz), 6.20 (d, 3H , H6,2J= 2.0 Hz), 3.71 (s, 9H, 7711), 3.64 (s, 9H, 7710), 3.47 (t, 6H, 772, V= 6.1 Hz), 2.77 (t, 6H, 771, V= 6.1 Hz). 1 3 C N M R (75.5 M H z , R T , C D 3 O D ) : 5 = 165.4 (C3), 146.2 (C9), 134.1 (CI), 129.5 (Ci), 114.9 (C6), 106.9 (C5), 99.26 (CA), 56.6 ( C l l ) , 56.13 (C10), 54.09 (CI), 39.0 (CI) . M S (+ESI): m/z 639 [L + H ] + . rm(4-hydroxy-2-methoxybenzylamide)-^m(ethyl)amine (4). A 0.5 M solution of NaSCH2CH3 was made by adding ethanethiol (931 mg, 1.1 m L , 15 mmol) to an ice-cooled, 185 stirred solution of N a H in 60 % oil (400 mg, 10 mmol) in 20 m L dry 2 ^ 4 ^ _ 3 ^ O H D M F under N 2 . The solution was warmed to R T and allowed to stir for N—y y~(\ // 0 H an additional 15 min. Approximately 12 m L (6 mmol) of the solution was added to 3 (1 g, 1.57 mol). The.solution was heated at 115°C for 30 min under N 2 . Excess solvent was removed in vacuo and the mixture was allowed to dry completely under reduced pressure. The product was recrystalized in M e O H yielding 400 mg, 43 %. *H (300 M H z , R T , C D 3 O D ) : 5 = 7.38 (d, 3H, 779,3J = 8.9 Hz), 6.17 (d, 3H, H6,2J= 2.5 Hz), 6.05 (dd, 3H, 778, 3 J= 8.9 Hz , 2J= 2.6 Hz), 3.64 (s, 9H, #10), 3.40 (t, 6H, 772,3J = 5.7 Hz), 2.70 (t, 6H , 771, 3J = 6.2 Hz). M S (+ESI): m/z 597 [L + H ] + . rm(2,4-hydroxybenzylamide)-^m(ethyl)amine (5). Compound 4 (100 mg, 0.157 mmol) was dissolved in 40 m L of dry C H 2 C 1 2 / D M F and cooled to -78°C in dry ice/acetone bath. BJ3r3 was added dropwise (125.4 mg, 0.046 mL, N-N-H ^ i / - 0 H 0.500 mmol) and the yellow solution was stirred for 30 min at 0°C. A HO NH HO ^ - ( 0 white precipitate formed after additional stirring for 3 h. The <^  V O H ' precipitate was filtered and washed with ether and small amounts of cold water. After drying under vacuum, the solution was added to hot E t O H and refluxed for 30 min. The precipitate was filtered and dried to yield 76 mg, 87 % *H N M R (300 M H z , RT, C D 3 O D ) : 5 = 7.35 (d, 3H, H9,3J = 8.7 Hz), 6.11 (s, 3H, H6), 6.05 (dd, 3H, #8, 3 J = 11.7 Hz , V = 1.5 Hz), 3.40 (t, 6H, HI, 3J= 6.2 Hz), 2.87 (t, 6H, 771 , 3 J = 6.2 Hz). M S (+ESI): m/z 555 [L + H ] + . 186 rm(2,4-dimethoxybenzamide)-»'m(methyl)amine (6). Under N 2 , Oxalyl chloride (18 mL, P P 211 mmol) and one drop of D M F were added by syringe to MeO H N - T / — \ s=x ' * J J to 5 \ / 8 ) )— nitrilotriacetic acid (2 g, 10.5 mmol), previously dissolved in 6 / = l 0=\ MeO MeO NH 1 0 / = ( 30 m L of dry benzene (dried over actived 4 A sieves). The Meo solution was stirred for 5 h under N 2 ; and the excess solvent and oxalyl chloride were removed in vacuo affording a yellow oi l which was redissoved in 20 m L of dry benzene. Black residue immediately formed upon addition of 2,4-dimethoxyaniline (4.18 g, 31.39 mmol). A n additional 20 m L of dry benzene was added and the solution was stirred overnight. Excess solvent was removed under reduced pressure. The resulting black residue was dissolved in C H 2 C 1 2 and washed with 3 x 0.1 M HC1 solution and 1 x brine solution. Organic fractions were condensed in vacuo and columned on silica gel using a gradient of 7 % acetonitr i le/CH 2 Cl 2 up to 20 % M e O H / C H 2 C l 2 yielding 1.46 g, 23 %. ' H N M R (300 M H z , R T , C D 3 O D ) : § = 7.67 (d, 3H, 778, 3J= 8.6 Hz), 6.45 (d, 3H, 775, 2J = 2.4 Hz), 6.39 (dd, 3H, HI, V = 9.3 H z , 2 J = 2.7 Hz), 3.68 (s, 9H , 7710), 3.59 (s, 9H, 779), 3.54 (s, 6H, 771). , 3 C N M R (75.5 M H z , RT, C D 3 O D ) : 5 = 166.9 (C2), 156.8 (C6), 149.9 (CS), 121.4 (CI), 120.4 (CA), 103.6 (CS), 98.5 (C3), 59.5 (CI) , 55.5 (C10), 55.4 (C9). M S (+ESI): m/z 619 [L + N a ] + . 7'm(4-hydroxy-2-methoxyamide)-/m(methyl)amine (7). A 0.5 M solution of p o N a S C H 2 C H 3 was made by adding ethanethiol (931 mg, 1.1 MeO HN—< / — \ / = \ 9 H \ V N X H N - \> O H 6 / = i 0 = \ MeO HO NH 5 \ _ _ / 8 7 /— m L , 15.0 mmol) to an ice-cooled, stirred solution of N a H in 60 /=< % oi l (400 mg, 16 mmol) and 20 m L dry D M F under N 2 . The y^0Me HO solution was warmed to R T and allowed to stir for an 187 additional 15 min. Approximately 8 m L (4 mmol) of the solution was added to 6 (0.70 g, 1.17 mmol). The solution was heated at 115°C for 1 h under N 2 . Excess solvent was removed in vacuo. The mixture was extracted into CHCI3 and washed with 2 x NaHC03 and 1 x brine. Organic fractions were condensed in vacuo and chromatographed on silica gel eluting with 10 % M e O H / C H 2 C l 2 to yield 400 mg, 62 %. ' H N M R (300 M H z , RT, CD3OD): 6 = 7.38 (d, 3H, 7/8, 3J = 8.7 Hz), 6.37 (d, 3H, H5,2J= 2.7 Hz), 6.32 (dd, 3H, H13J = 8.8 H z , 2 J = 3.0 Hz), 3.66 (s, 9H, H9), 3.64 (s, 6H, 771). M S (+ESI): m/z 577 [L + H ] + . rm (2 ,4-hydroxyamide ) - fm(methyl)amine (8). Compound 7 (100 mg, 0.173 mmol) was 0 0 dissolved in 40 m L of dry C H 2 C 1 2 and cooled to -78°C in dry HO h n ~ 2 \ _ /—K / = \ _ st~~§a 1 N) m y j ° " ice/acetone bath. B B r 3 (130.42 mg, 0.048 m L , 0.520 mmol) 6/=/ 0=\ HO K ' H was added dropwise and the solution was stirred for 30 min at OH 0°C. The mixture was further stirred at R T for 3 h. Methanol was carefully added to the reaction mixture to quench the excess BBr3, and the solvent was removed in vacuo. The product was washed down a small plug of silica gel using 20 % M e O H / C H 2 C l 2 t o yield 63 mg, 68 %. ' H N M R (300 M H z , R T , C D 3 O D ) : 5 = 6.95 (d, 3H, m , 3J = 8.1 Hz), 6.32 (s, 3H , H5), 6.20 (dd, 3H, H13J= 6.2 H z , 2 J = 1.8 Hz) , 3.51 (s, 6H, HI). M S (+ESI): m/z 535 [L + H ] + . Jm(2 ,4-dimethoxybenzylamide)-rm(2-methyl)ethane (9). D C C (2.77 g , 0.013 mol) and AMiydroxysuccinimide (1:38 g, 0.012 mol) were added to 2,4-dimethoxybenzoic acid (2.19 g, 0.012 mol) under N 2 in 125 m L of dry T H F and refluxed for 3 h. After the solution cooled to RT, the mixture was filtered to remove the dicyclohexylurea ( D C U ) by-product. Tame (0.40 188 MeO OMe g, 0.003 mol) was added dropwise to the solution and the MeO, mixture was refluxed overnight. Excess T H F was removed OMe N H under vacuum and the light yellow oi l was chromatographed on silica gel using a gradient of 5 % M e O H / C H 2 C l 2 up to 10 % M e O H / C H 2 C l 2 yielding 0.52 g, 29 %. ' H N M R (300 OMe M H z , R T , C D 3 O D ) : 5 = 7.67 (d, 3H , 775, 37 = 8.4 Hz), 6.51 (d, 3H, 773, 2J= 2.3), 6.47 (dd, 3H, 774, 3 J = 8.4 H z , 2 J = 2.5 Hz), 3.79 (s, 9H , 777), 3.76 (s, 9H, 776), 1.20 (s, 6H , 772), 0.82 (s, 3H, 771). M S (+ESI): m/z 609 [L + H ] + . 4.4 Results and Discussion for Proposed Bone-seeking B F C A few preliminary syntheses have been attempted to produce B F C that effectively deliver Ln(III) to bone tissue. One of the promising B F C chelators was synthesized by the addition of 4-aminoresorcinal to Kemp's triacid (2) . Kemp's triacid is a cyclohexane derivative in which 1,3,5-cw methyl groups force the 1,3,5-cw carboxyl groups into an all-axial conformation. 2 4 Kemp's triacid and derivatives has been extensively used for enzyme 25 * 26 27 28 models (specifically non-heme diiron enzymes ), molecular recognition, ' fluorescent probes, selective metallic ion separation reagents, therapeutic chelating agents, and potential antitumor agents.3 2 Upon coupling the activated Kemp's triacid with /V'-hydroxysuccinimide and D C C in D M F , the triamide ligand is formed with six hard oxygen donors in close proximity which could be used to f i l l the intrinsically large coordination sphere of the Ln(III). Unfortunately, only small quantities of the triamide were detected by N M R , M S , and T L C . In addition, the 189 highly oxophilic molecule could not be successfully purified after several recrystallizations, liquid-liquid extractions and column chromatography (both silica and alumina). Attempts to increase the yield of the tri-substituted product by forming a slightly more reactive starting material via the addition of thionyl chloride or oxalyl chloride 3 3 to produce the triacid chloride of Kemp's triacid only resulted in minimal products (MS (+ESI) m/z 602 [L + Na] + ) . The major product formed in the latter reaction was the amide/anhydride where one of the carboxylic acid functionalities attacked the neighboring acid chloride to form an anhydride, while the other carboxylic acid was functionalized with the 4-aminoresorcinol (MS (+ESI): m/z 332 [L+Na] + , Scheme 4.2). In addition, small amounts of product correlating to the formation of esters, formed by the unprotected phenol groups of the 4-aminoresorcinol reacting with the activated carboxylic acids on the triacid, were observed in the +ESI-MS spectrum. OR OR Scheme 4.2. Formation of the amide/anhydride of Kemp's triacid: (a) oxalyl chloride, D M F , N 2 ; (b) C H 2 C 1 2 , N 2 . R groups = H , C H 3 or C H 2 C 6 H 5 . Due to the low yields and the drawback of the potential seven membered chelate ring formed between the amide carbonyl and the phenol groups of the ligand, a modified route was sought to synthesize the desired bone-seeking B F C . It has been previously shown that 190 amines on the 1,3,5-position of the 1,3,5-trimethylcylohexane w i l l align in the axial position due to the smaller size of the amino groups relative to the methyl groups. Furthermore, hydrogen bonding helps stabilize the amine groups in the axial position, forming a pre-organized ligand that could potentially be functionalized with various carboxylic acids such as 2,4-dihydroxybenzoic acid. This multidentate species would form 6-membered chelate rings upon coordinating a Ln(III) ion (via the carbonyl oxygen and phenol groups). A modified procedure of A z o v and co-workers was used to synthesize the triamine of Kemp's triacid in moderate yields (Scheme 4.3). Upon addition of thionyl chloride, the anhydride was spontaneously produced by the nucleophilic attack at the carboxylic acid by the highly reactive neighboring acid chloride. Dilutions and lowered temperatures did not prevent the formation of the anhydride. However, the anhydride could easily be converted into two acid chlorides by the addition of 1,1-dichloromethyl methyl ether and Z n C ^ , which acted as the catalyst. The trichloride was transformed to the triazide using N a N 3 and a phase-transfer catalyst (tetrabutylammonium bromide). 3 4 The triisocyanate was synthesized using a stereospecific tris-Curtius rearrangement of the triazide in refluxing toluene and further hydrolyzed under harsh conditions to afford the triamine hydrochloride salt. Upon treatment with an ion-exchange resin, the free triamine was isolated. Longer reaction times were needed for the tris-Curtius rearrangement of the triazide and hydrolysis of the triisocyanate to form the triamine in yields similar to those reported by A z o v and co-workers. A l l characteristic data ( E A , M S , ' H and 1 3 C N M R ) matched the previously published data. 2 5 191 Scheme 4.3. Synthesis of the triamine: (a) S O C l 2 , ether, R T ; (b) C H C l 2 O C H , Z n C l 2 , reflux; (c) N a N 3 , B i ^ N B r , C H 2 C 1 2 / H 2 0 , 0°C; (d) toluene, reflux; (e) 37 % HC1, T H F , reflux, Dowex 550A OH", C H 3 O H . 2 5 The synthesized triamine was further reacted with a protected 2,4-dihydroxybenzoic acid to circumvent any unwanted side-products generated from the unprotected hydroxyl groups. The protected 2,4-dibenzyloxybenzoic acid (1) was synthesized using a modified literature procedure (Scheme 4.4). 3 0 The bulky benzyl protection groups were incorporated to prevent the addition of two benzoic acid moieties to the same amine of the modified Kemp's triacid. Unfortunately, all three of the O H groups, including the carboxylic acid O H group, were benzylated upon the addition of benzyl bromide; however, the hydrolysis of the protected carboxylic acid proceeded at quantitative yields by refluxing the fully benzylated product in a 5 M N a O H : M e O H solution. OH OBn OBn Scheme 4.4. Synthesis of 2,4-dibenzyloxybenzoic acid (1): (a) K 2 C 0 3 , acetone, N 2 , reflux; (b) 5 M N a O H , M e O H , reflux. 192 Unfortunately, attempts to use either the acid chloride of 2,4-dibenzyloxybenzoic acid or an activated carboxylic acid, prepared from TV-hydroxysuccinimide or HOBt , only resulted in the addition of two 2,4-dibenzyloxybenzoic acid functionalities. The absence of the tri-substituted ligand was thought to be a function of the steric bulk caused by the large benzyl protecting groups. Smaller methyl protecting groups (2,4-dimethoxybenzoic acid) were also used to functionalize the triamine utilizing /V-hydroxysuccinimide and D C C coupling reagents. Even though the yield was low, the desired tri-substituted product (2) was isolated after purification by column chromotagraphy. Upon optimization of this synthesis, the tri-functionalized triamine complex may prove to be an effective ligand for Ln(III). Other multidentate chelators could be synthesized upon the addition of 2,4-dihydroxybenzoic acid to polyamino backbones such as tren, tame and tacn. In addition, polycarboxylic acid backbones such as nitrilotriacetic acid could also be used to synthesize new ligands for Ln(III) by forming amide bonds with 2,4-dihydroxyanaline. These potential ligands could form a pre-organized, multidentate species where the p-hydroxyl group can be further functionalized with specific directing groups. Preliminary attempts to couple with 2,4-dibenzyloxybenzoic acid to tren using N-hydroxysuccinimide and D C C in T H F produced minimal yields of the tri-substitued tren (MS (+ESI): m/z 1095 [L + Na] + ) . Greater yields of the tri-substituted product (3) was isolated using the methyl protected benzoic acid (2,4-dimethoxybenzoic acid) and the same coupling agents (7V-hydroxysuccinimide and D C C , Scheme 4.5). D C U by-products were removed by column chromatography. Other attempts to form the desired tri-substituted product either resulted in no observable products or minimal yields ( D C C , H O B T and D M A P ; thionyl chloride; 7V,/V-diisopropylethylamine, E D C and D M F ; C D I and T H F ) . 193 Scheme 4 . 5 . Synthesis of the protected (3), mono-protected ( 4 ) and deprotected ( 5 ) tri-substituted tren ligand: (a) TV-hydroxysuccinimide, D C C , T H F ; (b) N a C H 3 C H 2 S , D M F , N 2 , reflux; (c) B B r 3 , C H 2 C 1 2 , N 2 , -78°C RT . The /?-methoxy group was selectively hydrolyzed ( 4 ) in good yields by sodium 36 thioethoxide. In addition, the o-methoxy group ( 5 ) was selectively removed by B B r 3 to afford the completely deprotected produce. 3 7 ' 3 8 It is interesting to note that the/>methoxy group on the protected tri-substituted tren could not be demethylated by B B r 3 , even with additional equivalents of B B r 3 , elevated temperatures, and longer reaction periods. In addition, the demethylation of the />methoxy with sodium thioethoxide would not proceed i f the o-methoxy was previously demethylated. The necessary order of reaction, where the p-methoxy group must be demethylated first to effectively deprotect the entire molecule, is actually an advantage for further functionalization of the end hydroxyl group. The p-hydroxyl group can be modified with various directing groups without functionalizing the protected o-hydroxyl group. After the p-hydroxyl group is functionalized, it is assumed that the o-methoxy group could be deprotonated using B B r 3 prior to chelation of the Ln(III). 194 Similar preliminary reactions were carried out with the nitrilotriacetic acid framework (Scheme 4.6). The tri-substituted complex (6) was synthesized upon addition of the 2,4-dimethoxyanaline to the triacid chloride of nitrilotriacetic acid (formed by addition of oxyl chloride). Other attempts using a variety of coupling reagents only provided minimal yields of the tri-substituted product (JV-hydroxysuccinimide and D C C in D M F or T H F ; D C C , H O B T , D M A P in D M F ; thionyl chloride). A s observed with the tri-substituted tren species, the /7-methoxy group was selectively removed by sodium ethoxide (7), while the o-mefhoxy group was demethylated using BBr3 (8). Scheme 4.6. Synthesis of the protected (6), mono-protected (7) and deprotected (8) tri-substituted nitrilotriacetic acid ligand: (a) oxalyl chloride, benzene, N 2 ; (b) NaCH3CH 2S, D M F , N 2 , reflux; (c) B B r 3 , C H 2 C 1 2 , N 2 , -78°C -> RT . 195 A n additional test reaction to functionalize the end />hydroxyl group with a phosphate was attempted for the tri-substituted nitrilotriacetic acid produced (8, Scheme 4.7). Addition of dibenzyl chlorophosphate to 8, where the jc-hydroxyl groups were deprotonated by N a H , afforded the protected B F C . A l l the /?-hydroxyl groups were functionalized with the benzylated phosphate monitored by +ESI-MS and *H N M R (87 % yield, M S (+ESI): m/z 1357 [L+Na] +). However, the reaction was completed on a microscale (10 mg) and the desired product was lost upon purification attempts. It is hoped that the phosphate groups can be easily debenzylated by hydrogenolysis using 10 % Pd/C, while the o-methoxy group wi l l be readily removed upon addition of B B r 3 , to afford the desired bone-seeking B F C complex. 8 Scheme 4.7. Synthesis of a B F C using protected phosphate directing groups: (a) N a H , D M F , N 2 , 0°C RT . These B F C , with carefully selected directing groups, suggest the possibility of selectively distributing Ln(III) ions to bone tissue. Along with the pyrone and pyridinone complexes, these species or their successor molecules might have the potential to treat millions of people worldwide who are suffering from bone density disorders. 196 4.5 References 1. Costes, J.-P.; Dahan, F.; Dupuis, A . ; Laurent, J.-P. Chem. Eur. J. 1998, 4, 1616. 2. X u , Z . ; Read, P. W. ; Hibbs, D. E . ; Hursthouse, M . B . ; Mal ik , K . M . A . ; Patrick, B . O.; Rettig, S. J.; Seid, M . ; Summers, D. A . ; Pink, M . ; Thompson, R. C ; Orvig, C. Inorg. Chem. 2000, 39, 508. 3. Kahn, M . L . ; Mathoniere, C ; Kahn, O. Inorg. Chem. 1999, 38, 3692. 4. Murphy, B . P.; Quinti, L . ; Ke l ly , D . G. ; Martin, W. J.; Perotti, A . ; Hursthouse, M . B . ; Gelbrich, T. Inorg. Chem. Commun. 2002, 5, 577. 5. Bismondo, A . ; Bernardo, P. D . ; Portanova, R.; Tolazzi, M . ; Zanonato, R. L . Polyhedron 2002, 21, 1393. 6. Hay, B . P.; Dixon, D . A . ; Vargas, R.; Garza, J.; Raymond, K . N . Inorg. Chem. 2001, 40, 3922. 7. Rodgers, S. J.; Lee, C. W. ; N g , C. Y . ; Raymond, K . N . Inorg. Chem. 1987, 26, 1622. 8. Ryu, J. C ; Shin, H . N . ; K i m , D. H . ; Lee, S. H . Bull. Korean. Chem. Soc. 2001, 22, 1293. 9. Agrawal, Y . K . ; Patel, D . R. J. Pharm. Sci. 1986, 75, 190. 10. Kalu , D . N . J. Bone Miner. Res. 1991,15, 175. 11. Wallach, S.; Rousseau, G. ; Martin, L . ; Azria , M . J. Bone Miner. Res. 1999, 25, 509. 12. Mosekilde, L . Bone 1995,17, 343S. 13. Yeh, J. K . ; Chen, M . M . ; A lo ia , J. F. Bone 1997, 20, 413. 14. Bassily, N . S.; Thabet, S. M . ; Shehata, A . M . J. Drug Res. 2005, 26, 12. 15. Xiong , L . ; Kishida, T.; Ebihara, K . J. Food Sci. 2006, 71, S379. 16. Mitamura, R.; Hara, H . Brit. J. Nutr. 2005, 94, 268. 17. L i u , S.; Wong, E . ; Karunaratne, V . ; Rettig, S. J.; Orvig, C. Inorg. Chem. 1993, 32, 1756. 18. Marques, F.; Guerra, K . P.; Gano, L . ; Costa, J.; Campello, M . P.; Lima, L . M . P.; Delgado, R.; Santos, I. J. Biol. Inorg. Chem. 2004, 9, 859. 19. Gossman, W. L . ; Wilson, S. R.; Oldfield, E. Acta Cyrst. 2003, C59, m33. 197 20. Zeevaart, J. R.; Jarvis, N . V . ; Louw, W. K . A . ; Jackson, G . E . ; Cukrowski, I.; Mouton, C. J. J. Inorg. Biochem. 1 9 9 9 , 73, 265. 21. Neves, M . ; Gano, L . ; Pereira, N . ; Costa, M . C ; Costa, M . R.; Chandia, M . ; Rosado, M ; Fausto, R. Nucl. Med. Biol. 2002, 29, 329. 22. George, A . ; Bannon, L . ; Sabsay, B . ; Di l lon, J. W. ; Malone, J.; Veis , A . ; Jenkins, N . A . ; Gilbert, D . J.; Copeland, N . G. J. Biol. Chem. 1 9 9 6 , 271, 32869. 23. Wang, D. ; Mi l le r , S.; Sima, M . ; Kopeckova, P.; Kopecek, J. Bioconjugate Chem. 2003,14, 853. 24. Kemp, D . S.; Petrakis, K . S. J. Org. Chem. 1 9 8 1 , 46, 5140. 25. Menger, F. M . ; Bian, J.; Azov , V . A . Angew. Chem. Int. Ed. 2002, 41, 2581. 26. Zang, X . - X . ; Lippard, S. J. J. Org. Chem. 2000, 65, 5298. 27. Rump, E . T.; Rijkers, D . T. S.; Hilbers, H . W.; Groot, P. G . d.; Liskamp, R. M . J. Chem. Eur. J. 2002, 5,4613. 28. J. Rebek, J.; Askew, B . ; Kil loran, M . ; Nemeth, D . ; L i n , F.-T. J. Am. Chem. Soc. 1 9 8 7 , 109, 2426. 29. Lujan-Upton, H . ; Okamoto, Y . J. Fluorescence 1 9 9 8 , 5, 355. 30. Takuji, H . ; Kaxuyuki , K . ; Tadafumi, U . ; Seiji, T. Kemp's Triacid Imide Derivative, It's Production Thereof and Selective Metallic Ion Separation Reagent. 1996. 31. Santos, M . A . ; Gama, S.; Gano, L . ; Cantinho, G . ; Farkas, E . Dalton Trans. 2004, 3772. 32. Regino, C. A . S.; Torti, S. V . ; M a , R.; Yap, G . P. A . ; Kreisel, K . A . ; Torti, F. M . ; Planalp, R. P.; Brechbiel, M . W. J. Med. Chem. 48, 7993. 33. Pryor, K . E . ; G . W. Shipps, J.; Skyler, D. A . ; J. Rebek, J. Tetrahedron 1 9 9 8 , 54, 4107. 34. Varma, R. S.; Naicker, K . P. Tetrahedron Letters 1 9 9 8 , 39, 2915. 35. Gardner, R. A . ; Kinkade, R.; Wang, C ; IV, O. P. J. Org. Chem. 2004, 69, 3530. 36. La i , K . ; Ghosh, S.; Salomon, R. G. J. Org. Chem. 1 9 8 7 , 52, 1072. 37. Weinstock, J.; Gaitanopoulos, D . E . ; Oh, H.-J . ; Pfeiffer, F. R.; Karash, C. B . ; Venslavsky, J. W. ; Sarau, H . M . ; Flaim, K . E. ; Hieble, J. P.; Kaiser, C. J. Med. Chem. 1 9 8 6 , 2 9 , 1 6 1 5 . 198 38. Bulls, A . R.; Pippin, C. G . ; Hahn, R. E. ; Raymond, K . N . J. Am. Chem. Soc. 1990, 112, 2627. 199 APPENDIX c o m 3 u 3 o u OO O i n cn oo CN ™ t-- oj O vo -a o ^H ^ ^ CN CN - H OH — o o' CN vo' o "=t r - H i n i n CN c n 00 o O N ' — i »-* O N O N 2^ oo vo oo o r--O N CN r--oo CN CN + 1 c n cn r-o i n i—i oo i n CN O N i n oo" oo cn " S t CN > 0 V O m O N V O V O O © o vo" r-" oo CN «n r-- oo r~-© © © © ^ c u x^  I o d % O N <N ,—I o cn NO rN "3 3 2 CN OH S PH _ wo T—H v- _ ^ i n cn c n r-O N oo i—H © i n O O c n © ,—< O ,—4 i—i CN CN O N O N ^ t , oT o i n 00 m o o O N m i n c n o o ON © © O N + 1 o V O 4382, I-H O N T—< i n 4382, o o r—< c n OO r-- 4382, 1—1 c n 1—H © © i n © © CN I*, CN + C o O o oo N n x © ' 3 V O •*t i n CN o vo PH CN vo CN ^—v oo vo O O i n o r-- c n H^ H^ oo r~- ^ t O O O N i n O N H^ CN o O o "=t 00 V O ^ H N O oo vo oo - o —' CN c n + 1 o oo cn : r 0 0 c n r- — i —' o o "3 -V O V O T t o o o c n O N c n •*t o •*t o c n T t o oo o vo O CN O o SH 1 PH CN CN c n o © + N 1% N 03 5 O r—I NO m vo u 2 CN O oo U 3 ° § 8 1 c n CN c n vo oo O O Nt rH vq r t i n c n O N r n (N rn c n CN' c n cn CN CN O CN O N ^H 00 ^ ^1- ^ t i n O N r-H o CN + 1 O N V O O V O ^93, 1 o CN o c n oo c n r-i-H r- ^93, 1 O H^ i n c n i n O N — 1 CN o <—* »—i o 1—1 T t o O • — i c n O N W c3 T3 3 c2 OH S £ <U <+H s O SH 60 3 'S 3 _ <u u op DX) 00^ & Q ~cT o" s" col ^ HX" K] Q H ^ 1 ^ ' oO O a. o O " <U =L >H o 5 PH O O -o -O on CN PH s <3 200 Table A l . (cont.) Crystal Data [GdNi 2 (bcn) 2 ] + [YbNi 2 (bcn) 2 ] + [GdZn 2 (bcn) 2 ] + empirical formula C 6 o H 6 9 C l G d N 9 N i 2 0 1 0 - . C8i.25H91.25Cl2Yb1.5Nn.25Ni3O14 .75 C 6 oH 6 9ClGdN9Zn 2 0 fw 1386.36 1967.99 1399.72 crystal color purple purple • colorless crystal system monoclinic monoclinic monoclinic space group P2,/n P21/n P21/n a, A 18.4011(17) 11.8772(11) 18.4481(13) b,A 13.0671(12) 20.455(2) 13.0422(7) c, A 25.285(2) 37.828(4) 25.5898(18) a, deg 90 90 90 P, deg 107.266(2) 89.978(4) 107.301(2) y, deg 90 90 90 V, A3 5805.8(9) 9190.1(17) 5878.4(7) Z 4 4 DCaic, g cm" 3 1.586 1.618 1.637 T, (K) 173 ± 2 173 ± 2 180 ± 1 LA 0.71073 0.71073 0.71073 Pcaic, mm" 1 1.885 2.591 1.964 ref. collected, unique 56394, 13836 178040,16478 52254, 12873 Rl f l (a l l ,obsd) 0.0643, 0.0363 0.1123,0.0618 0.0787, 0.0602 w R 2 a ' 6 ( a l l , obsd) ' 0.0752,0.0661 0.1674,0.1693 0.1356, 0.1234 G O F 1.003 1.034 1.101 "RI and wR2 as defined in S H E L X - 9 3 . V = \l[c?(F02) + 0.0322?) 2], where P = (F02 + 2F2)/3 Table A 2 . Selective bond lengths (A), interatomic distances (A) and angles (°) for [LaZn 2 ( tam)2(CH 3 OH)(H 2 0)]C104. Zn(l)-0(1) 2.135(5) Zn(2)-N(5; 2. 124(6) Zn(l)-0(2) 2.093(5) Zn(2)-N(6; 2.155(6) Zn(l)-0(3) 2.157(5) La(l) -0(1) 2.455(6) Zn( l ) -N( l ) 2.154(6) La(l) -0(2) 2.493(5) Zn(l)-N(2) 2.114(6) La(l)-0(3) 2.402(5) Zn(l)-N(3) 2.146(6) La(l)-0(4) 2.459(5) Zn(2)-0(4) 2.188(5) La(l) -0(5) 2.511(5) Zn(2)-0(5) 2.056(5) La(l)-0(6) 2.454(5) Zn(2)-0(6) 2.124(5) Zn( l ) -La ( l ) 3.1913(10) Zn(2)-N(4) 2.133(6) Zn(2)-La(l ) 3.2478(10) 0(1)-Zn(l) -0(2) 82.5(2) 0(5)-Zn(2; )-N(6) 100.6(2) 0(1)-Zn(l) -0(3) 83.9(2) 0(6)-Zn(2; )-N(4) 96.7(2) 0(1)-Zn(l) - N ( l ) 90.0(2) 0(6)-Zn(2; )-N(5) 176.1(2) 0(1)-Zn(l) -N(2) 173.9(2) 0(6)-Zn(2; )-N(6) 90.0(2) 0(1)-Zn(l) -N(3) 97.2(2) N(4)-Zn(2; )-N(5) 87.0(2) 0(2)-Zn(l) -0(3) 83.5(2) N(4)-Zn(2; )-N(6) 87.7(2) 0(2)-Zn(l) - N ( l ) 98.9(3) N(5)-Zn(2; )-N(6) 91.3(2) 0(2)-Zn(l) -N(2) 91.5(2) 0(1)-La(l) -0(2) 68.57(19) . 0(2)-Zn(l) -N(3) 172.8(2) 0(1)-La(l) -0(3) 72.42(17) 0(3)-Zn(l) - N ( l ) 97.2(2) 0(1)-La(l) -0(4) 86.64(17) 0(3)-Zn(l) -N(2) 94.3(2) 0(1)-La(l) -0(5) 81.62(18) 0(3)-Zn(l) -N(3) 89.3(2) 0(1)-La(l) -0(6) 148.19(18) N( l ) -Zn( l ) -N(2) 92.2(2) 0(2)-La(l) -0(3) 70.58(17) N( l ) -Zn ( l ) -N(3) 88.3(2) 0(2)-La(l) -0(4) 126.58(18) N(2)-Zn(l) -N(3) 88.6(2) 0(2)-La(l] -0(5) 145.66(17) 0(4)-Zn(2) -0(5) 82.65(19) 0(2)-La(l] -0(6) 142.99(18) 0(4)-Zn(2) -0(6) 81.8(2) 0(3)-La(l] -0(4) 146.04(18) 0(4)-Zn(2) -N(4) 89.1(2) 0(3)-La(l] -0(5) 82.34(16) 0(4)-Zn(2) -N(5) 97.1(2) 0(3)-La(l] -0(6) 115.66(17) 0(4)-Zn(2) -N(6) 170.8(2) 0(4)-La(l) -0(5) 68.68(17) 0(5)-Zn(2) -0(6) 83.3(2) 0(4)-La(l) -0(6) 70.16(17) 0(5)-Zn(2) -N(4) 171.7(2) 0(5)-La(l) -0(6) 68.03(17) 0(5)-Zn(2) -N(5) 92.9(2) Zn( l ) -La ( l )-Zn(II) 141.89(3) 202 Table A3. Crystallographic parameters, bond lengths (A) and angles (°) for [H 4bcn]C104. C(l)-C(2) 1.268(3) C(14)-N(2) 1.741(3) C ( l ) - N ( l ) 1.547(3) C(15)-C(16) 1.267(3) C(2)-N(2) 1.555(3) C(15)-C(20) 1.481(3) C(3)-N(2) 1.410(3) C(16)-C(17) 1.282(3) C(3)-C(4) 1.756(4) C(17)-C(18) 1.461(4) C(4)-N(3) 1.335(3) C(18)-C(19) 1.258(3) C(5)-N(3) 1.449(3) C(19)-C(20) 1.293(3) C(5)-C(6) 1.571(4) C(20)-O(2) 1.237(2) C(6)-N(l) 1.762(3) C(21)-C(22) 1.539(4) C(7)-N(l) 1.272(2) C(21)-N(3) 1.553(3) C(7)-C(8) 1.758(4) C(22)-C(23) 1.259(3) C(8)-C(9) 1.386(4) C(22)-C(27) 1.629(4) C(8)-C(13) 1.531(4) C(23)-C(24) 1.438(4) C(9)-C(10) 1.624(4) C(24)-C(25) 1.603(5) C(10 ) -C( l l ) 1.495(4) C(25)-C(26) 1.260(4) C( l l ) -C(12 ) 1.367(4) C(26)-C(27) 1.431(4) C(12)-C(13) 1.611(4) C(27)-0(3) 1.257(3) C(13)-0( l ) 1.364(3) C(28)-0(4) 1.336(4) C(14)-C(15) 1.410(3) C(2) -C( l ) -N( l ) 103.37(19) C(18)-C(19)-C(20) 108.9(2) C(l)-C(2)-N(2) 108.32(19) O(2)-C(20)-C(19) 110.0(2) N(2)-C(3)-C(4) 124.23(17) O(2)-C(20)-C(15) 122.9(2) N(3)-C(4)-C(3) 106.89(19) C(19)-C(20)-C(15) 127.1(2) N(3)-C(5)-C(6) 110.2(2) C(22)-C(21)-N(3) 106.31(19) C(5)-C(6)-N(l) 111.01(18) C(23)-C(22)-C(21) 108.6(3) N(l)-C(7)-C(8) 112.77(19) C(23)-C(22)-C(27) 120.3(2) C(9)-C(8)-C(13) 111.0(2) C(21)-C(22)-C(27) 131.15(18) C(9)-C(8)-C(7) 121.4(2) C(22)-C(23)-C(24) 107.3(3) C(13)-C(8)-C(7) 127.7(2) C(23)-C(24)-C(25) 131.7(2) C(8)-C(9)-C(10) 120.6(2) C(26)-C(25)-C(24) 122.2(3) C( l l ) -C(10)-C(9) 127.0(2) C(25)-C(26)-C(27) 105.9(3) C(12)-C(l l ) -C(10) 113.8(3) 0(3)-C(27)-C(26) 105.0(3) C(l l ) -C(12)-C(13) 119.3(2) 0(3)-C(27)-C(22) 122.4(2) 0(1)-C(13)-C(8) 108.0(2) C(26)-C(27)-C(22) 132.6(2) 0(1)-C(13)-C(12) 123.6(2) C(7) -N( l ) -C( l ) 108.0(2) C(8)-C(13)-C(12) 128.3(2) C(7)-N(l)-C(6) 111.70(19) C(15)-C(14)-N(2) 122.91(18) C( l ) -N( l ) -C(6) 117.07(18) C(16)-C(15)-C(14) 110.1(2) C(3)-N(2)-C(2) 109.28(19) C(16)-C(15)-C(20) 121.8(2) C(3)-N(2)-C(14) 122.04(17) C(14)-C(15)-C(20) 128.06(19) C(2)-N(2)-C(14) 117.13(18) C(15)-C(16)-C(17) 111.4(2) C(4)-N(3)-C(5) 102.0(2) C(16)-C(17)-C(18) 125.5(2) C(4)-N(3)-C(21) 118.98(16) C(19)-C(18)-C(17) 125.2(2) C(5)-N(3)-C(21) 103.96(18) 203 Table A 4 . Crystallographic parameters, selective bond lengths (A), interatomic distances (A) and angles (°) for H 3 [ N i ( b c n ) ] 2 C l C v 3 C H 3 C N . N ( l ) - N i ( l ) 2.133(3) 0(3) -Ni ( l ) 1.987(4) N(2)-Ni( l ) 2.125(3) 0(4)-Ni(2) 2.028(3) N(3)-Ni( l ) 2.181(3) 0(5)-Ni(2) 1.991(3) N(4)-Ni(2) 2.086(3) 0(6)-Ni(2) 2.104(3) N(5)-Ni(2) 2.210(3) 0(7) -Cl ( l ) 1.433(3) N(6)-Ni(2) 2.138(3) 0(8) -Cl ( l ) 1.423(4) 0(1) -Ni( l ) 2.034(4) 0(9) -Cl ( l ) 1.420(3) 0(2) -Ni( l ) 2.075(4) O(10)-Cl(l) 1.378(3) 0 ( 3 ) - N i ( l ) - 0 ( l ) 87.28(17) 0(5)-Ni(2)-0(4) 86.44(17) 0(3)-Ni( l ) -0(2) 84.15(17) 0(5)-Ni(2)-N(4) 168.43(16) 0 ( l ) - N i ( l ) - 0 ( 2 ) 85.29(17) 0(4)-Ni(2)-N(4) 89.13(12) 0(3)-Ni( l)-N(2) 100.25(17) 0(5)-Ni(2)-0(6) 87.77(15) 0(1)-Ni( l)-N(2) 170.93(15) 0(4)-Ni(2)-0(6) 85.96(14) 0(2)-Ni( l)-N(2) 90.42(13) N(4)-Ni(2)-0(6) 102.58(15) 0 (3 ) -Ni ( l ) -N( l ) 171.12(15) 0(5)-Ni(2)-N(6) 102.17(17) 0 (1 ) -Ni ( l ) -N( l ) 89.64(13) 0(4)-Ni(2)-N(6) 169.49(14) 0 (2 ) -Ni ( l ) -N( l ) 103.90(17) N(4)-Ni(2)-N(6) 83.50(11) N(2) -Ni ( l ) -N( l ) 83.63(13) 0(6)-Ni(2)-N(6) 88.34(11) 0(3)-Ni(l)-N(3) 93.31(13) 0(5)-Ni(2)-N(5) 88.08(12) 0(1)-Ni(l)-N(3) 102.33(17) 0(4)-Ni(2)-N(5) 102.73(15) 0(2)-Ni(l)-N(3) 171.87(15) N(4)-Ni(2)-N(5) 82.47(11) N(2)-Ni(l)-N(3) 82.42(12) 0(6)-Ni(2)-N(5) 170.12(12) N( l ) -Ni ( l ) -N(3) 79.21(14) N(6)-Ni(2)-N(5) 83.77(11) Table A 5 . Crystallographic parameters, selective bond lengths (A), interatomic distances (A) and angles (°) for [Cu 3(Hbcn) 2](C10 4)2. Cu(2)-0(1) 1.939(3) Cu( l ) -0(3) 1.939(3) Cu(2)-0(3) 1.945(3) Cu(l)-N(3) 2.007(4) Cu(2)-Cu(l) 2.9641(5) C u ( l ) - N ( l ) 2.035(4) Cu(l ) -0(1) 1.931(3) Cu(l)-N(2) 2.274(4) ' 0 ( l ) -Cu( l ) -0 (3 ) 77.56(11) N(3)-Cu(l)-N(2) 83.46(14) N(3) -Cu( l ) -N( l ) 88.37(16) N( l ) -Cu( l ) -N(2) 83.20(13) 0(1)-Cu(l)-N(3) 167.31(14) Cu(l)-0(1)-Cu(2) 100.00(12) 0(3)-Cu(l)-N(3) 95.71(13) Cu(l)-0(3)-Cu(2) 99.50(12) 0 (1) -Cu( l ) -N( l ) 94.43(15) 0 ( l ) -Cu(2) -0( la ) 102.77(11) 0 (3) -Cu( l ) -N( l ) 159.84(14) 0(l)-Cu(2)-0(3) 77.23(11) N(2)-Cu(l)-0(1) 109.14(13) 204 Table A6. Crystallographic parameters, selective bond lengths (A), interatomic distances (A) and angles (°) for [GdNi 2 (bcn) 2 (CH 3 CN) 2 ]C104*CH 3 CN. N ( l ) - N i ( l ) 2.083(3) 0(2)-Gd(l ) 2.384(2) N(2)-Ni( l ) 2.069(3) 0 ( 3 ) - N i ( l ) 2.064(2) N ( 3 ) - N i ( l ) 2.074(3) 0 ( 3 ) - G d ( l ) 2.351(2) N(4)-Ni(2) 2.068(3) 0(4)-Ni(2) 2.094(2) N(5)-Ni(2) 2.079(3) 0(4)-Gd(l ) 2.354(2) N(6)-Ni(2) 2.067(3) 0(5)-Ni(2) 2.065(2) N(7)-Gd(l) 2.573(3) 0(5)-Gd(l ) 2.390(2) N(8)-Gd(l) 2.626(3) 0(6)-Ni(2) 2.039(2) 0(1) -Ni( l ) 2.082(2) N i ( l ) - G d ( l ) 3.1191(5) 0(1)-Gd(l) 2.342(2) Ni(2)-Gd(l ) 3.1182(5) 0(2) -Ni( l ) 2.058(2) Ni ( l ) -0 (1 ) -Gd( l ) 89.48(8) N(6)-Ni(2)-N(4) 85.96(11) Ni ( l ) -0 (2 ) -Gd( l ) 88.90(8) 0(6)-Ni(2)-N(5) 99.42(10) Ni ( l ) -0 (3 ) -Gd( l ) 89.65(8) 0(5)-Ni(2)-N(5) 94.07(10) Ni(2)-0(4)-Gd(l) 88.84(8) N(6)-Ni(2)-N(5) 84.67(11) Ni(2)-0(5)-Gd(l) 88.54(8) N(4)-Ni(2)-N(5) 85.07(11) Ni(2)-0(6)-Gd(l) 89.45(8) 0(6)-Ni(2)-0(4) 83.24(9) 0(2)-Ni( l ) -0(3) 78.79(9) 0(5)-Ni(2)-0(4) 83.37(9) 0(2)-Ni( l)-N(2) 94.26(10) N(6)-Ni(2)-0(4) 98.19(10) 0(3)-Ni(l)-N(2) 100.53(10) N(4)-Ni(2)-0(4) 92.32(10) 0(2)-Ni(l)-N(3) 172.23(10) N(5)-Ni(2)-0(4) 175.99(10) 0(3)-Ni(l)-N(3) 93.86(9) 0 ( l ) -Gd( l ) -0 (3 ) 71.55(7) N(2)-Ni(l)-N(3) 84.56(11) 0 ( l ) -Gd( l ) -0 (4 ) 153.30(8) 0 (2 ) -N i ( l ) -0 ( l ) 83.64(9) 0(3)-Gd(l)-0(4) 92.94(7) 0 (3 ) -N i ( l ) -0 ( l ) 82.85(8) 0 ( l ) -Gd( l ) -0 (6 ) 84.62(8) N(2)-Ni( l ) -0(1) 175.63(10) 0(3)-Gd(l)-0(6) 79.21(8) N(3)-Ni( l ) -0(1) 98.04(10) 0(4)-Gd(l)-0(6) 70.91(7) 0 (2 ) -Ni ( l ) -N( l ) 102.01(10) 0 ( l ) -Gd( l ) -0 (2 ) . 71.48(8) 0 (3 ) -Ni ( l ) -N( l ) 174.48(10) 0(3)-Gd(l)-0(2) 67.07(7) N(2) -Ni ( l ) -N( l ) 84.89(10) 0(4)-Gd(l)-0(2) 123.25(8) N(3) -Ni ( l ) -N( l ) 85.56(11) 0(6)-Gd(l)-0(2) 143.23(7) 0 (1 ) -Ni ( l ) -N( l ) 91.79(9) 0 ( l ) -Gd( l ) -0 (5 ) 109.66(8) 0(6)-Ni(2)-0(5) 80.74(9) 0(3)-Gd(l)-0(5) 146.46(8) 0(6)-Ni(2)-N(6) 93.52(10) 0(4)-Gd(l)-0(5) 71.34(8) 0(5)-Ni(2)-N(6) 173.86(10) 0(6)-Gd(l)-0(5) 67.77(7) 0(6)-Ni(2)-N(4) 175.42(10) 1 N i (2 ) -Gd( l ) -Ni ( l ) 142.417(13) 0(5)-Ni(2)-N(4) 99.93(10) 205 Table Al. Crystallographic parameters, selective bond lengths (A), interatomic distances (A) and angles (°) for [ Y b N i 2 ( b c n ) 2 ] C 1 0 4 ' C H 3 C N . N ( l ) - N i ( l ) 2.069(8) 0 (3) -Yb( l ) 2.190(7) N(2)-Ni( l ) 2.053(7) 0(4)-Ni(2) 2.100(6) N(3)-Ni( l ) 2.064(8) 0 (4) -Yb( l ) 2.183(6) N(4)-Ni(2) 2.062(7) 0(5)-Ni(2) . 2.102(6) N(5)-Ni(2) 2.050(8) 0 (5) -Yb( l ) 2.195(6) N(6)-Ni(2) 2.048(8) 0(6)-Ni(2) 2.084(7) 0(1) -Ni( l ) 2.083(6) 0 (6 ) -Yb( l ) 2.202(6) 0(1) -Yb( l ) 2.203(6) N i ( l ) - Y b ( l ) 2.8990(13) 0(2) -Ni( l ) 2.115(6) Ni (2 ) -Yb( l ) 2.8910(13) 0(2) -Yb( l ) 2.193(6) 0(3) -Ni( l ) 2.076(6) N i ( l ) - 0 ( 1 ) - Y b ( l ) 85.1(2) N(4)-Ni(2)-0(6) 98.8(3) N i ( l ) - 0 ( 2 ) - Y b ( l ) 84.6(2) N(6)-Ni(2)-0(4) 100.7(3) N i ( l ) - 0 (3 ) -Yb( l ) 85.6(2) N(5)-Ni(2)-0(4) 93.4(3) Ni(2)-0(4)-Yb(l ) 84.9(2) N(4)-Ni(2)-0(4) 171.8(3) Ni(2)-0(5)-Yb(l ) 84.5(2) 0(6)-Ni(2)-0(4) 81.6(2) Ni(2)-0(6)-Yb(l ) 84.8(2) N(6)-Ni(2)-0(5) 174.2(3) N(2)-Ni(l)-N(3) 86.3(3) N(5)-Ni(2)-0(5) 98.6(3) N(2) -Ni ( l ) -N( l ) 86.2(3) N(4)-Ni(2)-0(5) 91.0(3) N(3) -Ni ( l ) -N( l ) 86.0(3) 0(6)-Ni(2)-0(5) 82.6(3) N(2)-Ni( l ) -0(3) 173.1(3) 0(4)-Ni(2)-0(5) 80.9(2) N(3)-Ni( l ) -0(3) 100.5(3) 0(4)-Yb( l ) -0(3) 176.0(2) N( l ) -N i ( l ) -0 (3 ) 93.0(3) 0(4)-Yb( l ) -0(2) 106.9(2) N(2)-Ni( l ) -0(1) 98.9(3) 0(3)-Yb( l ) -0(2) 77.1(2) N(3)-Ni( l ) -0(1) 92.7(3) 0(4)-Yb( l ) -0(5) 77.1(2) N( l ) -N i ( l ) -0 (1 ) 174.6(3) 0(3)-Yb( l ) -0(5) 98.9(2) 0 (3 ) -N i ( l ) -0 ( l ) 82.1(2) 0(2)-Yb( l ) -0(5) 175.9(2) N(2)-Ni( l ) -0(2) 92.0(3) 0(4)-Yb( l ) -0(6) 77.2(2) N(3)-Ni( l ) -0(2) 173.6(3) 0(3)-Yb( l ) -0(6) 102.8(2) N( l ) -N i ( l ) -0 (2 ) 100.1(3) 0(2)-Yb( l ) -0(6) 101.9(2) 0(3)-Ni( l ) -0(2) 81.3(2) 0(5)-Yb( l ) -0(6) 77.9(2) 0 ( l ) -N i ( l ) -0 (2 ) 81.4(2) 0 ( 4 ) - Y b ( l ) - 0 ( l ) 103.3(2) N(6)-Ni(2)-N(5) 86.9(3) 0 ( 3 ) - Y b ( l ) - 0 ( l ) 76.9(2) N(6)-Ni(2)-N(4) . 87.4(3) 0 ( 2 ) - Y b ( l ) - 0 ( l ) 77.0(2) N(5)-Ni(2)-N(4) 86.4(3) 0 ( 5 ) - Y b ( l ) - 0 ( l ) 103.2(2) N(6)-Ni(2)-0(6) 92.1(3) 0 ( 6 ) - Y b ( l ) - 0 ( l ) 178.9(2) N(5)-Ni(2)-0(6) 174.7(3) N i (2 ) -Yb( l ) -N i ( l ) 176.39(4) 206 Table A8. Crystallographic parameters, selective bond lengths (A), interatomic distances (A) and angles (°) for [ G d Z n ^ b c n M C H s C N ^ C l O r C F f j C N . Gd(l )-0(4; 2.341(4) Zn(2)-0( 5) 2.086(4) Gd(l )-o(3: 2.350(3) Zn(2)-0( 4) 2.108(4) Gd(l )-0(6] 2.357(3) Zn(2)-N( 5) 2.129(4) Gd(l )-0(2) 2.382(4) Zn(2)-N( 6) 2.136(4) Gd(l ) - o ( i ; 2.385(4) Zn(2)-N( 4) 2.152(4) Gd( 0-0(5: 2.394(4) Zn( l ) -0( 2) 2.047(4) Gd( )-N(8; 2.572(5) Zn( l ) -0( 1) 2.093(4) Gd(l )-N(7; 2.619(5) Zn( l ) -0( 3) 2.122(4) Gd(l l)-Zn(2) 3.1254(6) Zn(l)-N( 2) ' 2.134(4) Gd(l l)-Zn(l ) 3.1259(7) Zn(l)-N( 3) 2.137(4) Zn(2)-0(6) 2.070(4) Zn(l ) -N( 1) 2.150(5) 0(4) - G d ( i ; 1-0(3) 155.36(13) N(6)-Zn(2; )-N(4) 83.81(17) 0(4) - G d ( i ; 1-0(6) 71.86(13) 0(6)-Zn(2; )-Gd(l) 48.94(10) 0(3) - G d ( i ; 1-0(6) 93.86(12) 0(5)-Zn(2; )-Gd(l) 49.97(10) 0(4) - G d ( i ; 1-0(2) 85.56(13) 0(4)-Zn(2] )-Gd(l) 48.51(10) 0(3) - G d ( i : 1-0(2) 71.33(13) N(5)-Zn(2; )-Gd(l) 134.22(12) 0(6) - G d ( i ; (-0(2) 76.31(13) N(6)-Zn(2; )-Gd(l) 127.33(12) 0(4) - G d ( i : )-0(l) 107.11(13) N(4)-Zn(2 )-Gd(l) 127.85(12) 0(3) - G d ( i ; 1-0(1) 72.57(13) 0(2)-Zn(l ) -0 ( l ) 80.15(15) 0(6) - G d ( i ; 1-0(1) 144.18(13) 0(2)-Zn(l )-0(3) 82.86(14) 0(2) - G d ( i ; 1-0(1) 67.99(13) 0(1)-Zn(l )-0(3) 83.35(14) 0(4) - G d ( i ; 1-0(5) 72.66(13) 0(2)-Zn(l )-N(2) 93.15(16) 0(3) - G d ( i ; )-0(5) 121.16(12) 0(1)-Zn(l )-N(2) 171.89(17) 0(6) - G d ( i ; 1-0(5) 67.42(12) 0(3)-Zn(l )-N(2) 100.46(16) 0(2) - G d ( i ; )-0(5) 141.88(13) 0(2)-Zn(l )-N(3) 173.54(17) 0(1) - G d ( f )-0(5) 147.90(13) 0(1)-Zn(l )-N(3) 102.71(17) 0(6) -Zn(2) -0(5) 78.75(14) 0(3)-Zn(l )-N(3) 91.68(16) 0(6) -Zn(2) -0(4) 82.56(14) N(2)-Zn(l )-N(3) 84.42(18) 0(5) -Zn(2) -0(4) 83.96(14) 0(2)-Zn(l )-N(l) 102.17(17) 0(6) -Zn(2) -N(5) 104.12(16) 0(1)-Zn(l )-N(l) 93.56(17) 0(5) -Zn(2) -N(5) 94.09(15) 0(3)-Zn(l )-N(l) 173.60(17) 0(4) -Zn(2) -N(5) 172.58(16) N(2)-Zn(l )-N(l) 83.32(18) 0(6) -Zn(2) -N(6) 93.16(15) N(3)-Zn(l )-N(l) 83.52(18) 0(5) -Zn(2) -N(6)- 170.47(16) Zn( l ) -0(3 )-Gd(l) 88.56(12) 0(4) -Zn(2) -N(6) 100.11(16) Zn(2)-0(4 )-Gd(l) 89.07(13) N(5) -Zn(2) -N(6) 82.92(17) Zn(2)-0(6 )-Gd(l) 89.59(13) 0(6) -Zn(2) -N(4) 171.62(16) Zn(l ) -0(1 )-Gd(l) 88.28(13) 0(5) -Zn(2) -N(4) 104.89(16) Zn( l ) -0 (2 )-Gd(l) 89.45(14) 0(4) -Zn(2) -N(4) 90.27(15) Zn(2)-0(5 )-Gd(l) 88.18(13) • N(5) -Zn(2) -N(4) 83.31(17) 207 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059743/manifest

Comment

Related Items